Selected Data Centre Articles — 1 Jan 2025 - 31 Jul 2026

This is an XPath based search through the daily links archive. But, note, this is a string search and not a keyword search. Thus entries with the letters "data centre" or variants thereof will be fetched, regardless of what the article is actually about:

		//li[h5/a[contains(text(), "data center")
			or contains(text(), "Data Centre")
			or contains(text(), "datacenter")
			or contains(text(), "DataCentre")]
		or blockquote[p[contains(text(), "data center")
			or contains(text(), "data centre")
			or contains(text(), "datacenter")
			or contains(text(), "datacentre")]

• PF states limit reached – on three different host at nearly the same time

  What are the chances that three different hosts, in thee different datacenters all display these messages within seconds of each other?

• TikTok parent ByteDance plans to spend $7B on cloud-based GPUs this year to fuel its AI ambitions

  The company is said to be planning to spend a whopping $7 billion on Nvidia Corp.’s most powerful graphics processing units to fuel the development of cutting edge artificial intelligence models. It doesn’t seek to buy any chips, but rather just rent access to them via data centers located outside of mainland China.

• ByteDance plans to sidestep U.S. sanctions by renting Nvidia GPUs in the cloud — report says it has set aside $7 billion budget

  The report says ByteDance plans to invest over $20 billion in AI infrastructure, including $7 billion in accessing advanced Nvidia GPUs in the cloud, data centers, and even submarine cables. The U.S. prohibits ByteDance from purchasing Nvidia GPUs and using American cloud services. However, it cannot block ByteDance’s access to cloud services elsewhere, for example, in the Middle East or Asian countries. As a result, ByteDance can access American processors while technically adhering to U.S. sanctions against China’s AI and HPC sectors.

• Schneider report looks at how AI energy demand may play out

  Policymakers need to carefully guide the future consumption of electricity by AI datacenters, according to a report that considers four potential scenarios and suggests a number of guiding principles to prevent it from spiraling out of control.

  The research published by energy infrastructure biz Schneider Electric follows the IEA Global Conference on Energy & AI last month. Titled Artificial Intelligence and Electricity: A System Dynamics Approach, it looks at the emerging schools of thought relating to AI and the associated impact on electricity consumption.

• Microsoft pauses datacenter construction in Wisconsin

  Whether the Wisconsin delays are tied to the closed-loop design, or some other element of datacenter design, is unknown. Microsoft didn't answer questions to that end.

  Redmond did tell us that it still plans to invest $3.3 billion in the project by the end of 2026. It also expects to complete one hyperscale data center in Mount Pleasant as originally planned, which it tells us will be one of its five largest in the world.

• US government goes atomic as datacenter power crisis looms

  Constellation Energy has won contracts worth more than $1 billion from the US government to supply nuclear power to over 13 federal agencies, validating efforts by datacenter operators to secure their own atomic sources.

• Kicking datacenters' drinking habit is nearly impossible

  But it's not just power we have to worry about. These facilities are capable of sucking down prodigious quantities of water.

  In the US, datacenters can consume anywhere between 300,000 and four million gallons of water a day to keep the compute housed within them cool, Austin Shelnutt of Texas-based Strategic Thermal Labs explained in a presentation at SC24 in Atlanta this fall.

• Investigation: in Marseille as elsewhere, digital infrastructures are taking over the territory

  Since the end of 2023, the Marseille collective Le Nuage était sous nos pieds (The cloud was beneath our feet) has been investigating, analysing and fighting against the social, ecological and political impacts of digital infrastructures in Marseilles, in particular international undersea cables and data centres. The collective is made up of Marseille residents affiliated to at least three organisations: the collectif des Gammares a Marseille collective for popular education on water issues, Technopolice Marseille which analyses and fights against police surveillance technologies, and La Quadrature du Net, an association for the defence of fundamental freedoms and rights in the digital environment. In this article, we present part of the collective’s investigation into digital infrastructures in Marseille, their socio-environmental impact and the noxious world they represent, an investigation that we are extending beyond the Marseilles area, inspired in particular by the debates at the Le nuage était sous nos pieds festival that took place on 8, 9 and 10 November 2024 in Marseille.

• Building Ultra Long Range TOSLINK

  Optical data transmission is now everywhere in the datacenter, and my own stuff for bgp.tools is no exception. All connectivity in my world now comes out of Small Form-factor Pluggable (or, SFP) ports.

  The cool thing about SFP’s is that they are for the most part (when they don’t contain a FPGA and ARM CPU to run Linux) very basic inside.

• Trump announces Emirati businessman will invest $20B in US data centers

  Sajwani is the founder of DAMAC Properties, a major property development company based in the United Arab Emirates. In 2017, DAMAC built a golf course called Trump International Golf Club Dubai. The company has reportedly paid a multimillion-dollar licensing fee to the Trump Organization, while Sajwani contributed between $1 million and $5 million to Trump’s 2016 presidential campaign.

• OpenAI’s Sam Altman urges Trump to ease power plant, data center construction

  On Tuesday, Trump announced that Emirati billionaire Hussain Sajwani plans to invest $20 billion “over a very short period of time” to build data centers across the U.S.

• Trump touts $20B datacenter deal by Dubai property developer

  The impact of datacenters on the environment has been a matter of ongoing concern, at least until now. Google, for example, recently said that its datacenter emissions are up 48 percent from 2019, and up 13 percent from 2023. This is largely the result of an expanded datacenter footprint to support the intensive computing required to train and operate AI models. AI search queries have been estimated to consume ten times as much energy as traditional search queries.

• Amazon invests $11B in AI datacenters in Georgia

  The online megacorp forked out around $75 billion on capital expenditure in 2024, according to CEO Andy Jassy, who told financial analysts in October that the business will dig even deeper during this calendar year.

• Biden Administration Ignites Firestorm With Rules Governing A.I.’s Global Spread

  The rules have touched off an intense fight between tech companies and the government, as well as among administration officials.

  The regulations, which could be issued as early as Friday, would dictate where American-made chips that are critical for A.I. could be shipped. Those rules would then help determine where the data centers that create A.I. would be built, with a preference for the United States and its allies.

• How Nvidia is creating a $1.4T data center market in a decade of AI

  We are witnessing the rise of a completely new computing era. Within the next decade, a trillion-dollar-plus data center business is poised for transformation, powered by what we refer to as extreme parallel computing, or EPC — or as some prefer to call it, accelerated computing. Though artificial intelligence is the primary accelerant, the effects ripple across the entire technology stack.

• Politicians Blame Los Angeles Fires on Explosion of AI Data Centers

  Meanwhile, as Politico reports, politicians have started pointing fingers at the burgeoning AI industry, which has garnered a reputation for sucking up staggering amounts of water to cool down the quickly-growing number of data centers powering it.

  As such, three California lawmakers introduced bills last week aimed at encouraging AI data processing centers to work toward sustainable water use standards, and holding them accountable to new transparency rules.

• Biden Is Ignoring Environmental Concerns to Boost AI

  Joe Biden’s own Energy Department is warning that data centers’ energy consumption, water use, and emissions are already skyrocketing amid droughts and climate disasters. Biden just signed an executive order to accelerate an AI build-out anyway.

• Amazon Says All It Needs to Do Before Releasing an AI-Powered Alexa Is to Solve the Giant Engineering Problem That Nobody Else on Earth Has Been Able to Solve

  The issue? That's far easier said than done. Despite billions of dollars of investment and the construction of massive data centers to power increasingly complex AI models, even the most advanced chatbots still have a strong tendency to "hallucinate" false claims.

  Some experts have long argued that the issue might be intrinsic to the tech itself. In other words, hallucinations may always be a part of the equation — an unfortunate reality that tech companies are unlikely to admit, especially in the face of all of generative AI's buzz right now.

• Lights, Camera, AI: The Tech Behind Your Favorite Pixar Movies

  The training set up consists of a PyTorch development environment, a Linux instance with two 24GB NVIDIA Quadro P6000 GPUs. Pixar also has a renderer to synthesise the pairs of images, high-quality scenes with diverse shaders, geometry, and lighting conditions, and a data center fine-tuned to render at tremendous scale.

  The researchers claim that they have trained and deployed a production-quality super-resolution model that consistently produces high-quality, artefact-free upscaled images even on scenes with the depth of field or motion blur. Further, the paper states that their latest trained model shows promise towards a pipeline where one can render at 1K and upscale to 2K, which would save 50-75% of the studio’s render farm footprint if used for all intermediate renders.

• Real datacenter emissions are a dirty secret

  At issue, is that emissions from datacenters are likely to be much higher than currently estimated, perhaps more than seven times higher, according to Nightingale, who cited a report last year. This is due to the emissions accounting practices used by the hyperscale operators like AWS, Google, and Microsoft, such as the use of renewable energy certificates to offset emissions in their calculations.

• Bill Gates’ nuclear energy startup inks new data center deal

  TerraPower, a nuclear energy startup founded by Bill Gates, struck a deal this week with one of the largest data center developers in the US to deploy advanced nuclear reactors. TerraPower and Sabey Data Centers (SDC) are working together on a plan to run existing and future facilities on nuclear energy from small reactors.

• Open AI, Oracle, and SoftBank to invest $500 billion in 'Stargate' AI project

  The announcement featured top executives: OpenAI's Sam Altman, Oracle;s Larry Ellison, and SoftBank’s Masayoshi Son. They credited Trump's leadership for facilitating the project. In particular, Trump's decision to roll back a prior executive order by Joe Biden, aimed at curbing AI risks, was framed as pivotal. Also, the current administration promised to simplify power production processes to meet the high electricity demands of AI datacenters that tend to employ hundreds of thousands of power-hungry GPUs. Altman noted that achieving artificial general intelligence domestically would not have been possible without the new administration's support.

• Report: ByteDance sets aside about $20B for AI spending this year

  The report cites two anonymous people familiar with the matter as saying that around half of the amount will be spent on AI-related infrastructure outside China, such as data center resources and networking equipment.

• Exclusive: ByteDance plans $20 billion capex in 2025, mostly on AI, sources say

  The privately held technology giant plans to spend about half of the amount abroad on AI-related infrastructure, primarily data centres and networking equipment, they said.

• OpenAI, SoftBank each commit $19 billion to Stargate AI data center: Report

  OpenAI and Japanese conglomerate SoftBank will each commit $19 billion to fund Stargate, a joint venture to develop data centers for artificial intelligence in the U.S., the Information reported on Wednesday. The ChatGPT maker will hold a 40% interest in Stargate, and would act as an extension of OpenAI, the report said, citing OpenAI CEO Sam Altman speaking to colleagues. His comments imply SoftBank would also have a 40% interest, the report added.

• Mega UK datacenter greenlit, but no word on who's moving in

  Approval was last night granted for a mega datacenter in Hertfordshire, close to London's M25 orbital motorway, clearing the way for construction to begin. The identity of the eventual occupier, said to be a hyperscale operator, has yet to be disclosed.

  The planning application for the site, filed last year by a company called DC01 UK Ltd, was approved by the local authority, Hertsmere Borough Council. If all goes according to plan, facilities at the location are expected to become operational by 2030.

• What happens when we can’t build bigger datacenters anymore?

  Short of a breakthrough in machine learning and with power becoming a limiting factor, AI's continued growth may ultimately hinge on a new kind of supercomputer, one that spans entire countries and potentially even continents.

  The idea here is rather straightforward. If it's no longer practical to build bigger datacenters, start stitching together the ones you already have.

• Meta to Increase Spending to $65 Billion This Year in A.I. Push

  Much of the capital investment, a big jump from 2024, will fund expansion of Meta’s data centers, which provide the computing power needed by A.I. products and algorithms.

• We Aren't Being Told the Real Extent of AI Datacenter Emissions

  The already formidable environmental toll of generative AI is even worse than it seems because businesses aren't being transparent about the emissions given off by their enormous new AI datacenters.

  Take Amazon Web Services, a subsidiary of Amazon and one of the largest cloud computing platforms in the world, which doesn't provide location-based or data center-specific statistics in its emissions reports. That means it's impossible to tell how much greenhouse gases the massive facilities — and specifically, ones that host or train AI models — are producing, right at the most important moment for the public and policymakers to be grappling with that information.

  "The environmental impact of AWS is not broken out specifically, and you can only find the emissions of Amazon overall," Benn Caddy at the IT analysis firm Canalys, told the Register. "In fairness, Google and Microsoft similarly publish annual sustainability reports for their wider respective companies, reflecting emissions beyond their cloud businesses."

• Big Tech wants data centers plugged into power plants; utilities balk

  Looking for a quick fix for their fast-growing electricity diets, tech giants are increasingly looking to strike deals with power plant owners to plug in directly, avoiding a potentially longer and more expensive process of hooking into a fraying electric grid that serves everyone else.

  It's raising questions over whether diverting power to higher-paying customers will leave enough for others and whether it's fair to excuse big power users from paying for the grid. Federal regulators are trying to figure out what to do about it, and quickly.

• Energy Companies Stocks Plummet as DeepSeek Shows AI Doesn't Need Entire Coal Plants to Cheat on Homework

  That's rattled tech stocks in AI-adjacent industries, with the semiconductor company Broadcom down over 17.5 percent over two days, and Nvidia continuing to fluctuate after shedding nearly $600 billion in one day, dragging the NASDAQ down by an astonishing 3 percent.

  Now joining them are energy and utility corporations, as investors reel from DeepSeek's more efficient model seeming to use considerably less power than OpenAI to achieve similar results. Essentially, fossil fuel outfits had been banking on huge new datacenters needing tons of energy.

• Chevron Wants to Tap Into A.I. Boom by Selling Electricity to Data Centers

  The artificial intelligence boom has turbocharged demand for electricity, and everyone who is anyone in the U.S. energy industry wants a piece of the action.

  The latest entrant is Chevron, the country’s second-largest oil and gas company, which sees opportunity in building natural gas-fueled power plants that will feed energy directly to data centers.

• AI's energy obsession gets a reality check

  A reminder about the first piece: OpenAI, Oracle, SoftBank, and an Abu Dhabi–based investment fund called MGX plan to spend up to $500 billion opening massive data centers around the US to build better AI. Much of the groundwork for this project was laid in 2024, when OpenAI increased its lobbying spending sevenfold (which we were first to report last week) and AI companies started pushing for policies that were less about controlling problems like deepfakes and misinformation, and more about securing more energy.

• Linux kernel tweak promises big savings for datacenters

  That's exactly what researchers at Cheriton School of Computer Science at the University of Waterloo have managed to do. Just around 30 lines of code — to the Linux kernel's network stack, and they say you could curb datacenter power consumption by up to 30 percent.

  These changes have now been published as part of the 6.13 kernel release, making their way to the public.

• Team Says They've Recreated DeepSeek's OpenAI Killer for Literally $30

  Jiayi Pan, a PhD candidate at the University of California, Berkeley, claims that he and his AI research team have recreated core functions of DeepSeek's R1-Zero for just $30 — a comically more limited budget than DeepSeek, which rattled the tech industry this week with its extremely thrifty model that it says cost just a few million to train.

  Take it with a grain of salt until other experts weigh in and test it for themselves. But the assertion — and particularly its bargain basement price tag — is yet another illustration that the discourse in AI research is rapidly shifting from a paradigm of ultra-intensive computation powered by huge datacenters, to efficient solutions that call the financial model of major players like OpenAI into question.

• What does DeepSeek mean for AI's environmental impact?

  That could have big environment and climate implications, as training and running current AI models requires vast amounts of energy

  . The long-held assumption was that the next AI wave would require massive data center expansion to satisfy increasing demand.

  Today's more than 8,000 data centers already consume about 1 to 2% of global electricity, according to the International Energy Agency.

• U.S. Sues to Block Tech Deal in First Antitrust Action of Trump Term

  Hewlett Packard Enterprise, or HPE, a business software and services company, announced the $14 billion takeover of Juniper last year, with hopes of combining its data centers with Juniper’s networking business to take on giants like Cisco.

  Regulators in Britain and the European Union cleared the deal this summer. But the Biden administration’s antitrust enforcers had issued what is known as a “second request” for more information, indicating scrutiny of the transaction.

• Intel has missed the boat for AI in the datacenter

  Any hope Intel may have had of challenging rivals Nvidia and AMD for a slice of the AI accelerator market dissolved on Thursday as yet another GPU architecture was scrapped.

• US datacenters in for shock as Canada mulls cutting the juice over Trump tariffs

  According to Heatmap News, the US imports thousands of gigawatt hours of electricity from Canada every year. While this makes up less than 1 percent of total American power consumption nationwide, "it's a significant and growing source of low-cost, low-carbon power for some regions," it states.

• Self-hosting my emails again

  After three years with Purelymail, I’m back at self-hosting my mail server. Not because it’s cheaper (it’s actually much pricier to pay for a VPS), but because my mails are now hosted in Europe (who knows what happens next in the USA), I have more control to configure things how I want, and I can comply with GDPR.

  This time, I decided to use docker-mailserver (boring setup with Postfix + Dovecot + Rspamd with Abusix Guardian Mail + ClamAV) running on a Fedora CoreOS VPS at Hetzner in their Finland data center.

• ZFS Orchestration Tools – Part 1: Snapshots

  Snapshots have a variety of uses, including forming the underpinnings of FreeBSD boot environments feature. With regular snapshots of data files, it can be easy to recover files that were accidentally deleted or modified. While some desktop environments provide a “Trash” folder as a protection against accidental deletions, it can be hard to resist the compulsion to empty it. Moreover, a trash folder only contains deleted files, which is no help if you want to go back to the previous content of a file that has been overwritten.

  With regular snapshots in place, it is tempting to think of them as backups. While they do provide an easy means to recover individual files, they still reside on the same physical storage media and use the same operating system and filesystem as the live data. Backups are meant for disaster recovery, and ZFS snapshots won’t be of any help if your data center burns down.

• The mirage of EU techno-solutionism to the climate crisis

  Technology, as shaped by today’s market-driven priorities, carries immense hidden environmental and social costs. The tech sector is one of the fastest-growing contributors to waste and energy consumption. In 2021, it was responsible for two-to-three percent of global carbon emissions — on par with aviation.

  For instance, Microsoft’s emissions of the poisonous carbon monoxide gas have surged by nearly 30 percent since 2020, largely due to the expansion of its data centres. Similarly, Google’s greenhouse gas emissions in 2023 were almost 50 percent higher than in 2019.

• Datacenter energy use to more than double by 2030 due to AI

  If accurate, this will mean that power consumed by all those data facilities is set to increase by 50 percent in just a couple of years, and Goldman Sachs expects the trend to continue, with about 122 GW of total datacenter capacity online by the end of 2030.

• Call to make tech firms report data centre energy use as AI booms

  Google and Microsoft have reported year on year increases in data centre water consumption since 2020, and many of these water withdrawals come from potable water sources. In its annual environmental sustainability report in 2023, Microsoft said it consumed 6.4m cubic metres of water in 2022, primarily for its cloud data centres – a 34% increase on the year before.

  Google said its data centres consumed 19.5m cubic metres of water in 2022, up 20%.

  There is no reliable data on the quantity of resources used by data centres. In order to make effective policy to reduce their energy consumption, the government needed to collect information at scale, the report said.

• UAE, France to funnel up to €50B into a 1GW AI datacenter

  At one gigawatt in capacity, the facility will be enormous, dwarfing more traditional cloud and hyperscale campuses, which typically top out in the tens of megawatts. With that said, facilities on this scale aren't unheard of, given the insatiable appetite of AI workloads for power-hungry GPUs and high-end networking kits. In the US, social media giant Meta recently began construction of a 2.3-gigawatt facility in Richland Parish, Louisiana.

• London has 400 GW of grid requests for datacenter builds

  While the UK government wants to turbocharge datacenter construction, a newly published report says there are already 400 GW worth of outstanding requests for connection to the power grid around London, and regulator Ofgem estimates 60-70 percent of these will never happen.

• The global race to build AI computing infrastructure

  France is "back in the AI race", President Emmanuel Macron declared Monday evening at Paris's artificial intelligence summit. His comments came just days after the Élysée and the UAE announced their agreement to develop a vast AI campus and data centre in France. But experts warn that the race to dominate the sector with massive investments in AI infrastructure comes with a unique set of challenges.

• [Old] Intel appoints former HPE exec Justin Hotard to lead data center unit - DCD

  Hotard joins from an eight-and-a-half-year stint at HPE, where he most recently served as executive vice president and general manager of High-Performance Computing, AI, and Labs. Prior to that, he held roles at NCR Corp and Symbol Technologies (acquired by Motorola).

  “I am excited about my next chapter to lead the Data Center and AI Business Group at Intel Corporation. I look forward to joining Pat Gelsinger and his talented team to contribute to its vision to create world-changing technologies and deliver on Intel’s mission to bring AI everywhere,” Hotard said on LinkedIn.

• Gov.uk insiders: AI datacenters may be pricey white elephant

  While bit barns are a crucial part of the modern technological economy, building more in haste may not deliver the desired results.

  "Creating more datacenters will not necessarily lead to higher AI usage in the UK. Certainly, when it comes to the use of AI in public services, there are no clear timelines on how quickly the technology can be implemented, and there is a risk that the impact of AI may be far less than expected," TechMarketView Principal Analyst Simon Baxter previously said.

  This point was raised in a panel discussion at the AI Action Summit in Paris this week, where Alexander Vollert, CEO of AXA Group Operations, claimed AI solutions that have worked well are based on task-specific "classic" AI models rather than the large language models (LLMs), on which all the recent hype and investment has centered.

• The UK’s ‘turbocharged’ AI initiative — bringing data center noise to a village near you

  Bitcoin mining in the US is notorious for inflicting ridiculously noisy data centers on small towns across the country.

  We can bring these deafening benefits to towns and villages across the UK — as long as the data centers are labeled “AI.” The government is now accepting applications for AI enterprise zones — immune to local planning permission and green belt rules!

• Nuclear Energy for AI Data Centers

  According to the US Department of Energy, the energy requirement for a large data center is 100 MWe, which is equivalent to the energy consumed by approximately 80,000 households in the United States. In a data center, the process of cooling the servers and Central Processing Units (CPUs) consumes the largest share of electricity. McKinsey and Company, a global consulting firm, estimates that the process of cooling accounts for nearly 40 percent of the total energy consumed by AI data centers. In addition, computing accounts for 40 percent of the electricity while 20 percent is used for IT-related equipment in a data center.

  A report titled “Electricity 2024-Analysis and Forecast to 2026” by International Energy Agency (IEA) highlights that global energy demand from data centers will double by 2026. In 2022, data centers consumed an estimated 460 terawatt-hours (TWh) globally and this figure could increase to more than 1000 TWh in 2026. A measure of how much electricity is consumed by AI data centers can be gauged by the fact that this increased demand is roughly equivalent to the annual electricity consumption of Japan. Another indication of AI’s energy needs is the fact that OpenAI’s ChatGPT requires around 10 times as much electricity as compared to a Google search.

• Electricity demand surges for the world’s two biggest polluters

  That growth reflects broader changes across the world when it comes to how people consume information and what powers their lives. More vehicles and home appliances run on electricity these days. And new AI tools have led to a boom in energy-hungry data centers. That makes it all the more urgent to deploy new sources of energy that can make sure homes and businesses have enough electricity without creating a lot more pollution in the process.

• The Wheels Are Falling off Alberta’s Gas-Powered AI Data Centre Proposal

  Stay up to date with DeSmog news and alerts

  Stay up to date with DeSmog news and alerts

• So you want to build your own data centre

  So in response, we kicked off a ‘Railway Metal’ project last year. Nine months later we were live with the first site in California, having designed, spec-ed, and installed everything from the fibre optic cables in the cage to the various contracts with ISPs. We’re lighting up three more data centre regions as we speak.

  To deliver an ‘infra-less’ cloud experience to our customers, we’ve needed to get good at building out our own physical infrastructure fast. That’s the topic of our blog post today.

• What's driving electricity demand? It isn't just AI and data centers.

  If that sounds familiar, it may be because there’s been a constant stream of headlines about energy demand recently, largely because of the influx of data centers—especially those needed to power the AI that’s spreading seemingly everywhere. These technologies are sucking up more power from the grid, but they’re just a small part of a much larger story.

  What’s actually behind this demand growth is complicated. Much of the increase comes from China, India, and Southeast Asia. Air-conditioning, electric vehicles, and factories all play a role. And of course, we can’t entirely discount the data centers. Here are a few key things to know about global electricity in 2025, and where things are going next.

• [Old] Using GPT-4 to generate 100 words consumes up to 3 bottles of water — AI data centers also raise power and water bills for nearby residents

  The research noted that the exact water usage varies depending on state and proximity to data center, with lower water use corresponding to cheaper electricity and higher electricity use. Texas had the lowest water usage at an estimated 235 milliliters needed to generate one 100-word email, while Washington demanded a whopping 1,408 milliliters per email — which is about three 16.9oz water bottles.

  This may not sound like a lot, but remember that these figures add up fairly quickly, especially when users are using GPT-4 multiple times a week (or multiple times a day) — and this is just for plain text.

• Expert warns Finland’s foreign data centres offer little economic benefit

  Manner singled out Ireland, where data centre energy use surpassed urban household consumption a few years ago. Data centres now account for around one-fifth of Ireland’s electricity use, and experts warn that the rise of AI could push this to one-third within years.

  Despite efforts to transition to renewables, much of the energy consumed by data centres still comes from fossil fuels.

• Microsoft plays down AI bubble, cancels data center leases

  By this measure, AI has yet to actually do a thing for the world economy.

  On that supply side, Nadella talked up “hyperscalers” — where you build AI data centers as fast as possible so you can rent GPUs. He’s a big fan! AI can surely fill them all!

  There’s one problem there — Microsoft is cutting back on data centers.

• World’s largest data center gets go-ahead from Korean govt — facility to require 3 GW of power

  Investment firm Stock Farm Road (SFR) just signed a Memorandum of Understanding (MoU) with South Jeolla Province Governor Kom Young-rok to build the largest data center in the world. The facility will be located in the southwest corner of South Korea with a projected cost of about $35 billion and a capacity of 3 GW, according to a report by Interesting Engineering. This would dwarf other data centers being built around the world, like Portugal’s Sines Data Center (1.2 GW) and Oracle’s 1 GW AI data center. This could potentially even be bigger than Elon Musk’s plan to put a million GPUs in his Colossus supercluster.

• Microsoft scraps some data center leases as Apple, Alibaba double down on AI

  Microsoft’s lease cancellations were detailed on Friday in a note from TD Cowen analysts. According to Bloomberg, the investment bank believes that Microsoft has voided contracts for data centers with about 200 megawatts in combined capacity. One megawatt corresponds to the electricity usage of several hundred households.

  The software and cloud giant reportedly told suppliers that it’s canceling the contracts on account of delays in facility construction and power delivery. According to Bloomberg, Meta Platforms Inc. once used the same arguments as part of a push to reduce capital expenses.

• Meta in talks for $200 billion AI data center project: Report

  Meta executives have informed data center developers that the company is considering building the campus in states including Louisiana, Wyoming or Texas, with senior leaders having visited potential sites this month, the report said.

• Microsoft data centre leases slowing, analysts say

  TD Cowen analysts in a note Friday said the tech giant had scrapped leases for sizeable data centre capacity in the US, suggesting potential oversupply as it builds out artificial intelligence infrastructure.

• The Zero-sum Game of Digitalization: Data Centre Expansion and Sustainability Dilemma

  Europe’s focus on green energy and sustainable infrastructure makes it a crucial region in the global conversation on data centre energy consumption. In 2023, the EU regulated an Energy Efficiency Directive (EED) which included data centre regulations. Starting in May 2024, data centres with an IT power demand exceeding 500 kW will be required to publicly disclose key information on their energy performance, including metrics such as energy consumption, water usage, and waste heat recovery. By May 2025, the European Commission will review the data collected and may introduce additional measures, including mandatory minimum performance standards and a framework for transitioning toward net-zero emissions. Additionally, the European Code of Conduct for Data Centre Energy Efficiency encourages data centres with an IT power demand of 1 MW to adopt best practices. These facilities must also utilize waste heat for applications like district heating or hot water supply unless they can demonstrate that such reuse is technically or economically impractical.

• Sam Altman Says OpenAI Has Run Out of GPUs

  It's a notable admission, highlighting just how hardware-reliant the technology is. AI industry leaders are racing to build out data centers to keep their increasingly unwieldy AI models running — and are ready to put up hundreds of billions of dollars for the cause.

• Minnesota denies Amazon attempt to exempt data center from energy rules

  The decision against Amazon by Minnesota’s Public Utilities Commission comes in a state that, like many others, is grappling with a recent spike in data center construction, with projects also planned by Meta and Microsoft.

• Should we be moving data centers to space?

  After all, energy-guzzling data centers are springing up like mushrooms all over the world, devouring precious land, straining our power grids, consuming water, and emitting noise. Building facilities in orbit or on or near the moon might help ameliorate many of these issues.

  For Steve Eisele, Lonestar’s president and chief revenue officer, a big appeal of putting data storage on the moon is security. “Ultimately, the moon can be the safest option where you can have a backup for your data,” Eisele says. “It’s harder to hack; it’s way harder to penetrate; it’s above any issues on Earth, from natural disasters to power outages to war.”

• HPE to lay off 2,500 staff as stock craters on mixed earnings results and tariffs

  The company, which primarily sells servers, storage arrays and software for enterprise data centers, revealed it’s embarking on a cost-cutting exercise that will result in about 2,500 workers, or 5% of its employee base, being let go. The layoffs will add charges to its second-quarter and full-year results, the company added, causing it to lower its guidance.

• What's next after the AI bubble bursts?

  In other words, bubbles come and go, but capital investment sticks around. Quite recently, the burst of the cryptocurrency bubble paved the way for the AI boom - suddenly there were a lot of cheap GPUs, just sitting in datacenters, waiting to be used for something. Very few people made that connection. Suppose the current AI bubble bursts1. What physical infrastructure would be left behind, and how would it be put to use?

• Get faster OpenVPN on FreeBSD by enabling DCO – easily done

  I’ve been configuring a new gateway server for use in my basement data center (home lab). I had recently read about DCO and FreeBSD’s opvn device.

  DCO (Data Channel Offload) let’s OpenVPN use the encryption features available in many CPUs. In my previous post, I checked; my OpenVPN server and most of the clients are DCO-capable.

  I decided to try it.

• Why Do Researchers Care About Small Language Models?

  Small models are not used as general-purpose tools like their larger cousins. But they can excel on specific, more narrowly defined tasks, such as summarizing conversations, answering patient questions as a health care chatbot and gathering data in smart devices. “For a lot of tasks, an 8 billion parameter model is actually pretty good,” said Zico Kolter, a computer scientist at Carnegie Mellon University. They can also run on a laptop or cellphone, instead of a huge data center. (There’s no consensus on the exact definition of “small,” but the new models all max out around 10 billion parameters.)

• Apple's response to AI

  Ever since Apple revealed their AI strategy to lean into on-device models, there has been a sort of tension with the approach from other companies like OpenAI, Google, and Amazon. Was Apple Intelligence going to work? There are advantages: for user privacy because more data stays on your phone, and for scaling because the load is distributed across millions of phones instead of only running in data centers.

• US trio wants to pipe coal mine gas to fuel datacenters

  But this expansion isn't without its challenges, and powering those multiplying bit barns is one of them. A study last year predicted that US power consumption from datacenters is set to more than double by 2030, while one developer on the other side of the pond proclaimed that access to power is the single biggest constraint on new builds, and that he would be investing "hundreds of millions and more" if it weren't for issues in getting the projects wired up to the national grid.

• Data Centre Energy Use Force Rethink on Grid Planning

  Demand for electricity will grow 15.8% over the next five years compared to earlier predictions of 4.7%, mostly driven by the surge in data centres for artificial intelligence (AI), predicts RMI, formerly known as the Rocky Mountain Institute.

  To handle this surge, new data centres should be built at existing gas interconnection points—such as infrequently used peaker plant sites—where renewables can be added “in ways that don’t affect the rest of the grid, that don’t force you to do upgrades,” Uday Varadarajan, senior principal with RMI, told NBC News

• OpenAI asks the US government for the moon on a stick

  OpenAI would like “AI Economic Zones” where they can do stuff that’s currently just illegal and abuse workers. OpenAI will help the environment with special new laws to build out gas pipelines to new power plants. This is fine, because the artificial general intelligence will save us all. In due course.

  As well as ignoring copyright, OpenAI wants all the classified government data to train on. They need to “test and experiment with real data.” OpenAI even wants the government to invoke the Defense Production Act to prioritize them getting more data center equipment.

• Elon Musk’s Starlink Expands Across White House Complex

  It was not immediately clear when the White House complex was fitted with Starlink after President Trump took office for a second term.

  Starlink terminals, rectangular panels that receive internet signals beamed from SpaceX satellites in low-Earth orbit, can be placed on physical structures. But instead of being physically placed at the White House, the Starlink system is now said to be routed through a White House data center, with existing fiber cables, miles from the complex.

• Majority of AI Researchers Say Tech Industry Is Pouring Billions Into a Dead End

  Published in a new report, the findings of the survey, which queried 475 AI researchers and was conducted by scientists at the Association for the Advancement of Artificial Intelligence, offer a resounding rebuff to the tech industry's long-preferred method of achieving AI gains — by furnishing generative models, and the data centers that are used to train and run them, with more hardware. Given that AGI is what AI developers all claim to be their end game, it's safe to say that scaling is widely seen as a dead end.

• AI scientists are sceptical that modern models will lead to AGI

  “The vast investments in scaling, unaccompanied by any comparable efforts to understand what was going on, always seemed to me to be misplaced,” says Stuart Russell at the University of California, Berkeley, a member of the panel that organised the report. “I think that, about a year ago, it started to become obvious to everyone that the benefits of scaling in the conventional sense had plateaued.”

  Nonetheless, tech companies plan to collectively spend an estimated $1 trillion on data centres and chips in the next few years to support their AI ambitions.

• UK council says freak IT outage didn’t lead to data loss

  "The power outage subsequently impacted the council's central datacenter, which houses computing and networking equipment for the whole council, " the statement went on to say.

  "The serious nature of the unprecedented outage meant that all systems went offline, leading to disruptions to our phone lines and systems while we work to test systems before bringing them safely back online."

• Google to buy cybersecurity firm Wiz for $32B US in company's biggest ever deal

  Wiz, a five-year-old startup founded by four longtime friends who met in the Israeli army when they were still teenagers, is on track for an estimated $1 billion US in revenue this year. After getting its start in Israel in 2020, Wiz now oversees an operation that makes security tools protecting the information stored in data centres from its current headquarters in New York.

• SoftBank to acquire Arm CPUs for datacenter firm Ampere in $6.5 billion cash deal

  Ampere was established in 2018 in Silicon Valley with the backing from Carlyle Group and Oracle with an aim to develop CPUs tailored for the needs of cloud service providers (CSPs), which the company calls 'cloud-native' processors. Since then, the company released two product families — the Ampere Altra with up to 128 cores made on TSMC's N7 node and the AmpereOne with up to 192 cores made on TSMC's N5 manufacturing technology — that gained limited traction in the CSP.

• « Simplification » Bill: a denial of democracy to impose giant data centers in France

  The National Assembly has begun examining the bill on the simplification of economic life. Through its Article 15, this “simplification” bill (or PLS) aims to speed up the construction of huge data centers in France, by allowing the government to impose them on the territories concerned and by increasing exemptions to municipal planning regulations, environmental laws, and to the principle of public participation. Against this new denial of democracy imposed to serve the interest of the tech industry, La Quadrature du Net and the collective “Le Nuage était sous nos pieds”, together with the other members of the“Hiatus” coalition, are calling for the deletion of Article 15 and a two-year moratorium on the construction of large data centers, so as to provide the time to establish the conditions for the democratic control of these digital infrastructures.

• AI data centers: Microsoft steps on the gas

  Microsoft said last year how its data center buildout already meant it was just blowing through all its carbon commitments and that Microsoft increased its CO2 output by about 30% in the last five years.

  But Microsoft wants to use even more natural gas. Bobby Hollis, Microsoft VP for Energy, talked up using more gas and mumbled something about carbon capture and storage at the CERAWeek energy conference this month.

• The AI Boom and Its Data Center Imperative

  According to McKinsey, global demand for data center capacity is projected to grow annually by 19-22% from 2023 to 2030, potentially reaching 171 to 219 gigawatts. This surge is driven by the escalating requirements of AI workloads, particularly in terms of power, cooling, and scalability. For India, the AI revolution presents a unique opportunity to establish itself as a global AI powerhouse by building sovereign, high-performance, and sustainable AI-ready data center infrastructure.

• Reflecting on TikTok’s Role in Society as New Ban Deadline Approaches

  Last month, Mr. Trump told reporters that he could extend the deadline again. And while ByteDance hasn’t confirmed any plans to sell, Oracle, the data center company, and others have emerged as potential suitors.

• Microsoft dropping datacenter leases isn't AI Armageddon

  However, we'd argue there's a far simpler explanation as to why Microsoft might be pulling back on its datacenter leases: many datacenters can't handle the power and cooling demands created by high-end AI hardware.

  Nvidia's massive NVL72 rack-scale systems promise 30x or better performance for inference workloads, and four times better training performance compared to its Hopper GPUs. That performance comes at the expense of denser racks, higher power consumption and more waste heat production. Racks housing Nvidia’s latest are rated for 120kW of power, three times that of the typical Hopper rack. Liquid cooling is no longer an option but an inescapable requirement.

• China built hundreds of AI data centers to catch the AI boom. Now many stand unused.

  Just months ago, a boom in data center construction was at its height, fueled by both government and private investors. However, many newly built facilities are now sitting empty. According to people on the ground who spoke to MIT Technology Review—including contractors, an executive at a GPU server company, and project managers—most of the companies running these data centers are struggling to stay afloat. The local Chinese outlets Jiazi Guangnian and 36Kr report that up to 80% of China’s newly built computing resources remain unused.

• IT Leaders Brace for Tariff Fallout on Infrastructure and Cloud Costs

  It’s hard to know what’s next and how this will play out exactly, but economists and industry experts predict that most or all these costs will be passed on directly to buyers. This is troubling news for U.S. companies that rely on electronics and materials from across the globe to source technology products, build facilities (like data centers), and arm their workforces with the best productivity tools to compete.

• Alibaba Head Warns AI Industry Is Showing Signs of Bubble

  So far, companies have continued to pour tens of billions of dollars into building out massive data centers to meet the demands of increasingly power-hungry AI models.

• Climate Change Contributes to Surge in Global Energy Demand

  Electricity demand rose nearly twice as fast in 2024 as the average over the past decade, as heatwaves drove up air conditioning use, industries ramped up production, and the world’s fleet of electric vehicles, data centres, and artificial intelligence systems grew, finds [pdf] the IEA’s annual Global Energy Review. Nearly all of the additional electricity was supplied by low-emissions sources, with solar and wind seeing record expansion, along with an 8% rise in nuclear power generation.

• Arm targets 50 percent of datacenter CPUs this year

  The Brit chip designer has long touted the AI credentials of chips based on its architecture and is gradually gaining ground in the server market, but Arm's infrastructure chief, Mohamed Awad, now claims that he expects its share of the global market for datacenter processors to surge to 50 percent by the end of the year.

  In an interview with Reuters, Awad claimed that Arm's technology typically offers lower power consumption than processors made by rivals such as Intel and AMD. As a result, it was argued, Arm-based products have become increasingly popular even with the big cloud computing companies, which are concerned about the power drain of their massive bit barns.

• Nuclear power no panacea for AI's growing thirst for energy

  Atomic energy is becoming the preferred solution to address the projected bump in megawatts needed to charge AI in the future, but it simply won't come soon enough in many cases.

  Much has been written about infrastructure and the electricity required to keep feeding the AI beast. Recent research from Goldman Sachs estimated that datacenter energy use will more than double by the end of the decade – just five years away.

• Red Hat OpenStack Services on OpenShift at the edge with Ceph

  Red Hat OpenStack Platform 13 introduced an edge solution called distributed compute nodes (DCN). This allows customers to deploy computing resources (compute nodes) close to consumer devices, while centralising the control plane in a more traditional datacenter, such as a national or regional site.

• Wikimedia Foundation bemoans AI bot bandwidth burden

  That's due to the Wikimedia Foundation's caching scheme which distributes popular content to regional data centers around the globe for better performance. Bots visit pages without respect to their popularity, and their requests for less popular content means that material has to be fetched from the core data center, which consumes more computing resources.

  The heedlessness of ill-behaved bots has been a common complaint over the past year or so among those operating computing infrastructure for open source projects, as the Wikimedians themselves noted by pointing to our recent report on the matter.

• US electricity prices are surging - These Louisiana companies want out

  “Horrifying.” That’s what one consumer advocate calls a projected 90% increase in electricity prices for customers of electric utility, Entergy Louisiana, between 2018 and 2030.

  Similar spikes in the cost of electricity are happening across the United States driven by the need to run data centers and power-hungry industries, replace old power plants and repair damaged infrastructure from more frequent storms, floods and fires supercharged by climate change.

  “Louisiana is a microcosm of what’s happening nationally, but it is happening at a breakneck speed in a way that residents … are not protected. And that is not OK,” said Logan Atkinson Burke, executive director of the nonprofit Alliance for Affordable Energy.

• Arm aims to capture 50% of data center CPU market in 2025

  Arm Holdings hopes to increase its share of the global data center CPU market from 15% to 50% by the end of 2025. Mohamed Awad, senior vice president of infrastructure at Arm, made the claim in an interview with Reuters. The company pins its hopes primarily on AI servers, so consider offerings like Nvidia's GB200 and GB300 machines, custom silicon from large cloud service providers, and Ampere Computing-based systems.

• Something's Gone Wrong With Microsoft's Huge AI Data Center Investments

  That's on top of last week's news that Microsoft had walked away from two data center projects in the US and Europe, piling on to a February announcement that it was cancelling data center leases.

• Trump's Tariffs Are a Bruising Defeat for the AI Industry

  The vast majority of chips that power these data centers come from hard-hit countries like Taiwan, Thailand, Japan, and Vietnam, while the US makes just 11 percent of its chips at home. Trump's tariffs will force countries from those companies to hike their chip prices, and US companies will no doubt hike their prices to compensate, which will ultimately run off to consumers.

• Trump’s tariffs will deliver a big blow to climate tech

  The Pitchbook report, for instance, noted that the surge in development of AI data centers is fueling demand for “dispatchable energy sources.” That means the type that can run around the clock, such as nuclear fission, fusion, and geothermal (though in practice, the data center boom has often meant commissioning or relying on natural-gas plants that produce planet-warming emissions).

• China massively overbuilds empty AI data centres

  Chinese companies still want data centres — even buying capacity in Malaysia to serve the broader Asian market. They’re just not so interested in the cheap local ones.

• US DoE wants devs to fast-track domestic AI datacenters

  Ironically, Trump's new reciprocal tariffs could disrupt efforts to expand AI infrastructure, potentially undermining the administration's objectives. Analysts have reportedly warned that tech giants that have committed to significant AI investments may face challenges as the tariffs target major equipment suppliers from countries such as China, Taiwan, and South Korea. The steep duties, including a 10 percent baseline tariff on all imports, will increase the cost of essential electronics like smartphones, PCs, and datacenter equipment – a category accounting for nearly $486 billion in imports last year.

• Powering innovation with Red Hat Enterprise Linux on Google Cloud

  The modern CIO faces many challenges. They are faced with the constant priority of maintaining consistent performance across cloud and datacenter environments alike to address critical needs around security, scalability and reliability – all while keeping an eye on the future with artificial intelligence (AI). Now more than ever, organizations must prioritize standardizing cloud-based infrastructure on a consistent and flexible operating system. Running RHEL on Google Cloud provides organizations with a powerful, adaptable platform for innovation to address the business needs of today and into the future. 

• Simplification’ law: stop the data center boom!

  This is all the more concerning that the principle of a moratorium — a way of laying the foundations for democratic control of data centers, and of countering the government’s desire to accelerate ever further in defiance of rights and democracy — is supported by a wide range of actors.

• Microsoft puts $1B DC builds on hold amid AI, tariff worries

  The Windows maker confirmed it put on hold $1 billion plans for three bit barn sites in Licking County – at New Albany, Heath and Hebron – citing a strategic investment review for the decision, telling The Register it will "not be moving forward at this time with our plans to build datacenters at the Licking County sites. We will continue to evaluate these sites in line with our investment strategy."

• Amazon buys up all electricity from huge wind farms in western Finland

  Swedish energy company OX2 announced last week that it would invest 700 million euros in the Rajamäenkylä and Halsua wind farms. They will have a capacity of 472 megawatts (MW).

  Amazon said that it will use its wind power investment as part of efforts to offset its total emissions. According to the company, this does not mean that it is planning to build a data centre in Finland.

• Trump’s Abuse of Emergency Declaration to Force Ratepayers to Prop Up Inefficient Coal Power Plants Is Breathlessly Stupid

  “Reviving or extending coal to power data centers would force working families to subsidize polluting coal on behalf of Big Tech billionaires and despoil our nation’s public lands. States planning to move to cleaner, cheaper energy sources could be forced to keep old coal plants up and running for years, forcing nearby residents to breathe dirty air and harming the climate. Trump’s expected use of the threat of power demand growth from AI data centers to ramp up domestic coal mining and consumption is unjustifiable, as Public Citizen recently pointed out to Congress. Trump and his team of incompetents continue to demonstrate their lack of understanding of how energy markets work. Public Citizen is more than happy to meet with Administration officials and walk them through why forcing American families to pay for uneconomic coal power plants is dull-witted and will result in a massive ratepayer-funded subsidy for Big Tech billionaires.

• SONiC: The open source network operating system for modern data centers

  Canonical has identified several key strategies to help network equipment vendors and users of SONiC in the enterprise address some of SONiC’s limitations and enhance reliability and performance for their version of SONiC. This article begins by exploring SONiC’s technical architecture, its origins and current limitations. We then cover Canonical’s efforts to improve it, and its potential impact on enterprise data centers.

• This is Europe's shot to emerge from Silicon Valley's shadow

  Never let a good crisis go to waste is an adage Europe could use now. As it stares down the barrel of a loaded gun from Donald Trump, the region could seize an opportunity if lawmakers and regulators move quickly to bolster its tech sector, a critical growth driver.

  With key services from America like cloud computing and artificial intelligence potentially becoming more expensive, Europe’s homegrown tech industry can reap an immediate advantage. Trump may complain loudly about how the US has been a victim of global trade, but the biggest beneficiaries of the booming AI business have been American. Now so-called hyperscalers – Amazon.com, Microsoft and Google – could suffer from Trump’s tariff plans as the steel, aluminium and copper they need to build vast data centres will become costlier.

• The Amount Amazon Is Actually Making on AI Will Make Your Eyebrows Raise Dramatically

  That AI is a money pit shouldn't be surprising. It's been clear for some time now that generative AI's road to profitability will be a rocky one, if it even leads to that destination at all. As evidenced by the above figures, it's an enormously expensive technology, both to develop and to run. Vast amounts of data are scraped to train the AI models, and the pricey data centers that run them rack up huge energy bills, water bills, and maintenance costs even after they're constructed.

  The other side of the equation — how the AI services will generate money — is also fraught with uncertainty. OpenAI, which is charging customers an eye-watering $200 per month for full access to ChatGPT, for example, reeks of desperation. Power users may think it worthwhile, but most people might have second thoughts about paying even just a $20 subscription fee.

• Lawmakers fear AI data centers will drive up residents’ power bills

  “We’re going to have tremendous stress from AI,” said New Jersey state Sen. Bob Smith, a Democrat who chairs the Environment and Energy Committee. “We have a crisis coming our way in electric rates. These outrageous increases are going to be put on the citizens. Why should they bear the rate increases?”

  Smith has authored a bill that would require new AI data centers in New Jersey to arrange to supply their power from new, clean energy sources, if other states in the region enact similar measures.

• Datacenter energy usage to more than double in next 5 years

  The IEA this week published a sweeping study on the crucial intersection of AI and the energy it gobbles up, and believes datacenters will drive more than 20 percent of electricity demand growth in advanced economies over the next five years, with AI driving most of that energy consumption.

  The US, which accounted for 45 percent of global datacenter electricity consumption in 2024, according to the agency, is expected to see that share grow significantly by 2030. By then, American datacenters are set to consume more electricity than the country's entire energy-intensive manufacturing sector, including aluminium, steel, cement, and chemicals.

• As demand for AI rises, so do power thirsty data centers

  As more things moved online, and computing hardware and chip technology supported faster processing, AI models became attainable to the general public, Acosta said. Current AI models use thousands of GPUs to operate, and training a single chatbot like ChatGPT uses about the same amount of energy as 100 homes over the course of a year.

  “And then you multiply that times the thousands of models that are being trained,” Acosta said. “It’s pretty intense.”

• As demand for AI rises, so do power thirsty data centers

  “Each one of these AI models has to sit on a server somewhere, and they tend to be very, very big,” he said. “So if your millions or billions of users are talking to the system simultaneously, the computing systems have to really grow and grow and grow.”

• What goes up must come down? AI datacenter spending inflates

  Rather more recently, Synergy Research Group said in January that mergers and acquisitions of datacenter businesses throughout 2024 swelled to record highs of $73 billion, and yet more investors are lining up for a piece of the action.

  Having lived through the early days of the [Internet] frenzy, Fabrice Coquio, senior veep at Digital Realty, which bills itself as the world's largest provider of cloud and carrier-neutral datacenter, colocation and interconnection services, is perhaps better placed than most to venture an opinion. Is there a bubble?

• Midwest emerges as top hub in U.S. data center expansion

  Yes, but: The expansion often happens behind closed doors.

  Local governments frequently sign nondisclosure agreements with tech firms, limiting public knowledge of energy and water usage, says Helena Volzer of the nonprofit Alliance for the Great Lakes.

• Microsoft's Huge Plans for Mass AI Data Centers Now Rapidly Falling Apart

  Investment bank TD Cowen also told the publication in February that Microsoft had canceled some leases for US data center power capacity, totaling a "couple of hundred megawatts."

  While Microsoft's exact reasoning for the scaling back remains unclear, waning demand and fears of a growing AI bubble are easy culprits.

• Microsoft says it's 'slowing or pausing' some AI data center projects, including $1B plan for Ohio - CBS Pittsburgh

  The computing it takes to run AI tools is expensive and requires a large amount of electricity, so much so that Trump this week cited AI needs as part of the justification for using his emergency authorities to boost the declining U.S. coal industry, a reliable but polluting energy source. Tech companies have also sought to tap into nuclear power, including a proposed Microsoft-backed revival of the shuttered Three Mile Island plant in Pennsylvania, which would feed an electricity grid supplying data centers in Ohio as well as Virginia, the nation's biggest data center hub.

• AWS: Customers would flee Azure if licensing costs were fair

  AWS estimates that half of the workloads Microsoft enterprise customers run on Azure would migrate to its own datacenters if only the licensing costs of doing so were not prohibitively high and a competitive deterrent.

  This claim is made in the latest submission by AWS to the UK's Competition and Markets Authority's (CMA) Cloud Services Market Investigation.

• Accelerating towards cable convergence with Red Hat and Intel

  OpenShift is an enterprise-grade platform designed to simplify development, deployment and management across hybrid and multicloud environments. Built on Red Hat Enterprise Linux (RHEL) and Kubernetes, OpenShift provides a scalable and consistent experience across on-premise data centers, public clouds and edge environments across the broadband network.

• AI Electricity Demand Could Mean Grid Strain or Gain

  A new report this week from the International Energy Agency (IEA) confirms that the rapid spread of data centres, driven by soaring demand for artificial intelligence technologies, is one of several factors putting pressure on electricity systems around the world.

  “A typical AI-focused data centre consumes as much electricity as 100,000 households, but the largest ones under construction today will consume 20 times as much,” the IEA writes. Poor planning could worsen their impact, but AI also has the potential to improve how power grids function.

• Amazon also slows down its data centre commitments

  It’s not just Microsoft slowing down on data centres — it’s Amazon too.

• Chinese data centers refurbing and selling Nvidia RTX 4090D GPUs due to overcapacity — 48GB models sell for up to $5,500

  One surprising aspect is that AI data centers are doing this despite the uncertainty surrounding AI chip supply from the U.S. The White House has recently blocked China-compliant Nvidia H20 and AMD MI308 chips for export to China, and there are also some rumors that the 5090D might also be affected by the ban. In fact, it's rumored that Nvidia has mentioned suspending supplying 5090D chips to its board partners, although it did not mention stopping sales entirely.

• iTWire - Nokia selected by Australian cloud provider ResetData for networking backbone

  Telecoms equipment vendor Nokia has been selected by Australian cloud provider ResetData to supply a networking backbone for its rollout of sovereign liquid immersion-cooled ‘AI Factory’ data centres across the country.

• Adyen Cyberattack Disrupts Payment Services Across Europe

  Global payment platform Adyen has confirmed it was the target of a Distributed-Denial-of-Service (DDoS) attack on April 21, 2025, which disrupted services for several customers across Europe. The Adyen cyberattack caused significant delays and failures in processing transactions, highlighting the growing cyber threat of digital attacks on critical financial infrastructure.

  The cyberattack on Adyen began at 18:51 CEST on April 21, when Adyen’s internal monitoring systems flagged unusual levels of errors and slow responses across several of its payment services hosted in European data centers. The company’s engineering team swiftly launched an investigation and identified the disruption as a DDoS attack.

• Nvidia aims to solve AI's water consumption problems with direct-to-chip cooling — claims 300X improvement with closed-loop systems

  Modern cloud data centers not only consume an immense amount of power for computing and cooling, but also a significant amount of water, as most use evaporative liquid cooling.

• Google admits depreciation costs soaring amid bit barn build

  Google owns and operates 135 datacenters across the world and uses colocation providers as part of its cloud interconnect services. Yet the CFO admitted it is still not able to entirely meet customer demand.

• TikTok owner weighs data center project in Brazil, sources say

  ByteDance, the Chinese parent company of TikTok, is weighing a major investment in a data center in Brazil, three people familiar with the matter told Reuters, tapping abundant wind energy on the country's northeast coast.

• Trump tariffs taking out AI data centres too

  At least the data centres are paid for by private companies and not bank loans. This means the bubble pop won’t cause a banking crisis too.

• Cyber: Forever a top priority for government IT

  Where data is located deeply affects this trustworthiness. In the case of government agencies, the majority of data — vast amounts of structured, unstructured and semi-structured data points that could be used to train models — is siloed. This type of data structure makes it very difficult to ensure consistency, security and governance due to the lack of cross-validation. Additionally, because public sector programs are funded individually, some departments have migrated workloads to the public cloud already, while others remain hosted in on-premises data centers.

• UK nabs 3 in £3M bribery probe tied to Dutch Microsoft DC

  UK authorities on Wednesday arrested three individuals in connection with a multi-million-pound bribery probe tied to the construction of a Microsoft datacenter in the Netherlands.

• Small Towns Are Rising Up Against AI Data Centers

  "Hyper scale data centers bring few benefits to communities," claims Peaceful Peculiar, a grassroots campaign that successfully repelled a Diode Ventures data center in Peculiar, Missouri. "Dozens of communities around the nation severely struggle from the presence of data centers. It's only about maximizing their profits, so they choose properties as close as possible to a large power supply."

• TikTok to build 1-billion-euro data centre in Finland

  TikTok said Tuesday it was investing 1 billion euros ($1.1 billion) to build a data centre in Finland, days after the EU slapped it with a massive fine accusing it of sending Europeans' personal data to China. The European Union last week hit TikTok with a 530-million-euro ($600 million) fine, accusing it of sending Europeans' personal data to China and failing to guarantee it was shielded from access by Chinese authorities.

• How Is Elon Musk Powering His Supercomputer?

  All that processing takes power to run, and so the xAI team moved about thirty-five mobile methane-gas-powered generators onto the site to support the data center. These are truck-mounted units, many of them designed by Caterpillar, which give off some of the same brew of pollutants as other gas-combustion device—including nitrogen oxides and formaldehyde—and which are currently operating without a permit. “xAI has essentially built a power plant in South Memphis with no oversight, no permitting, and no regard for families living in nearby communities,” the Southern Environmental Law Center said, in a report released in April. (Full disclosure: I volunteer every year to judge the S.E.L.C.’s Phil Reed prize for best environmental writing about the South). The S.E.L.C. has called for an “emergency order” from the city to require xAI to cease the use of these generators, with a twenty-five-thousand-dollar daily fine if the company refuses. The mayor of Memphis, Paul Young, a supporter of the project, addressed the concerns at a meeting with community members in March. “I want to figure out how we can exploit this project for us,” he said. “I know you all feel like it’s us getting exploited, but we need to speak from a place of strength.” After the S.E.L.C. issued its report, Young explained that the company has a permit application pending with the Shelby County Health Department to run fifteen generators. “There are thirty-five, but there are only fifteen that are on,” he said. “The other ones are stored on the site.”

• Pakistan turns to bitcoin miners, AI data centers to use surplus power

  Bilal Bin Saqib, chief executive officer of the council, told Reuters the location of the mining centre will be finalised based on the availability of excess power in specific regions. Documents seen by Reuters outline the role of Changpeng Zhao, founder of Binance, who will serve as a strategic adviser to the Pakistan Crypto Council.

• What's the carbon footprint of using ChatGPT?

  The inevitable counter-argument to the idea that the impact of ChatGPT usage by an individual is negligible is that aggregate user demand is still the thing that drives these enormous investments in huge data centers and new energy sources to power them. Hannah acknowledges that: [...]

• [Old] Microsoft pulls back from more data center leases in US and Europe, analysts say

  The tech giant's withdrawal from new capacity leasing was largely led by the decision not to support additional training workloads from ChatGPT maker OpenAI, the analysts led by Michael Elias said in a note.

• Microsoft Abandons Data Center Projects, TD Cowen Says

  Microsoft Corp. has walked away from new data center projects in the US and Europe that would have amounted to a capacity of about 2 gigawatts of electricity, according to TD Cowen analysts, who attributed the pullback to an oversupply of the clusters of computers that power artificial intelligence.

• A strategic approach to migrating virtualized workloads to Red Hat OpenShift Virtualization

  There are plenty of reasons why teams decide to move their virtual machines (VMs). Sometimes it’s about switching vendors to cut costs, other times it’s about upgrading to infrastructure that’s a better fit for the tools and services they plan to use. In many cases, it’s part of an effort to simplify, consolidate systems and reduce the sprawl across multiple data centers.

• Minister: TikTok data centre could anger the United States

  A government minister suggests that a planned data centre may be a way for China to bypass curbs on AI chip exports.

• Why are so many data centres popping up in Finland?

  The money generated by data centres in Finland mostly ends up outside of the country. The majority of profits are usually funnelled back into the business to acquire more processors and servers.

  Even though data centre deals are worth billions of euros, only about 20 to 30 percent of the funds remain in the country, according to an estimate by Jouni Salonen, a senior advisor at Business Finland.

  However, that is a rough estimate, due to the competitive nature of the industry, companies behind the centres tend to carry out business as quietly as possible.

• Meta faces Democratic probe into plans to power a giant data center with gas

  Sen. Sheldon Whitehouse (D-RI), ranking member of the Senate Committee on Environment and Public Works, shot off a letter to Meta CEO Mark Zuckerberg on Wednesday demanding answers about how much energy the data center would use and the greenhouse gas emissions that would be generated. Powering the new data center with gas “flies in the face of Meta’s climate commitments,” the letter says.

  Tech companies are rushing to build out data centers to train and run new AI tools, driving up electricity demand. In this case, power utility Entergy wants to meet that demand with new gas infrastructure, raising concerns about the impact Meta’s data center will have on the environment and local residents.

• China begins assembling its supercomputer in space

  The benefits of having a space-based supercomputer go beyond saving communications time, according to South China Morning Post. The outlet notes that traditional satellite transmissions are slow, and that “less than 10 per cent” of satellite data makes it to Earth, due to things like limited bandwidth and ground station availability. And Jonathan McDowell, a space historian and astronomer at Harvard University, told the outlet, “Orbital data centres can use solar power and radiate their heat to space, reducing the energy needs and carbon footprint.” He said both the US and Europe could carry out similar projects in the future, writes SCMP.

• Website Carbon Calculator v3 | What's your site's carbon footprint?

  The [Internet] consumes a lot of electricity. 1021 TWh per year to be precise. To give you some perspective, that’s more than the entire United Kingdom.

  From data centres to transmission networks to the billions of connected devices that we hold in our hands, it is all consuming electricity, and in turn producing carbon emissions equal to or greater than the global aviation industry. Yikes!

• The data center boom in the desert

  Much of Nevada has suffered through severe drought conditions for years, farmers and communities are drawing down many of the state’s groundwater reservoirs faster than they can be refilled, and global warming is sucking more and more moisture out of the region’s streams, shrubs, and soils.

  “Telling entities that they can come in and stick more straws in the ground for data centers is raising a lot of questions about sound management,” says Kyle Roerink, executive director of the Great Basin Water Network, a nonprofit that works to protect water resources throughout Nevada and Utah.

• AI could keep us dependent on natural gas for decades to come

  To power the data center, Entergy aims to spend $3.2 billion to build three large natural-gas power plants with a total capacity of 2.3 gigawatts and upgrade the grid to accommodate the huge jump in anticipated demand. In its filing to the state’s power regulatory agency, Entergy acknowledged that natural-gas plants “emit significant amounts of CO2” but said the energy source was the only affordable choice given the need to quickly meet the 24-7 electricity demand from the huge data center.

• AI’s energy impact is still small—but how we handle it is huge

  The rising energy cost of data centers is a vital test case for how we deal with the broader electrification of the economy.

• TikTok explains why it's establishing a data centre in Finland

  The other Project Clover data centres are located in Norway and Ireland and serve the same purpose — storing TikTok users' data.

• My hunch is that Microsoft's Copilot is as much a dead end as the Metaverse

  What's telling is that even Microsoft seems to be cooling on its AI investments. Reports suggest they've cancelled plans for significant data center expansions; a sign they're not seeing the AI demand they expected. Across industries, AI adoption rates have slowed dramatically according to recent studies.

• Unleashing intelligent operations at the edge

  For too long, the edge has been viewed as just another spot to place workloads with tactical proximity or convenience. But at Red Hat, we see something fundamentally different. This isn't about extending the data center; it's about transforming how businesses operate and how organizations can make a difference in their bottom line. Think about it: bringing compute power and, crucially, intelligence right to where the data bursts into existence. That's where real value is unlocked.

• AI power and water use is through the roof, and 80–90% is per query

  But the bottom line is: each individual query you make to ChatGPT really does pump out more carbon.

  Another claim is that AI data centres don’t use lots of water. Which they very much do: [...]

• How AI Demand Is Draining Local Water Supplies

  Each time you ask an AI chatbot to summarize a lengthy legal document or conjure up a cartoon squirrel wearing glasses, it sends a request to a data center and strains an increasingly scarce resource: water.

  The data centers that power artificial intelligence consume immense amounts of water to cool hot servers and, indirectly, from the electricity needed to run these facilities.

• ‘It will destroy this place:’ West Virginia residents fight for future against proposed data center

  A Virginia-based company, Fundamental Data, was applying for an air permit from the West Virginia Department of Environmental Protection for what it called the “Ridgeline Facility.” The company’s heavily redacted application showed plans to build an off-the-grid natural gas power plant between Thomas and Davis. That power plant will likely be designed to power an enormous data center just a mile out from Tucker County’s most populous and tourist-attracting areas.

  Earlier this month, representatives for Fundamental Data — who did not respond to requests for comment on this article — told the Wall Street Journal that the facility could be “among the largest data center campuses in the world,” spanning 10,000 acres across Tucker and Grant counties if fully realized.

• Trump trade policy to cost Nvidia $10.5B in lost revenues

  US export controls blocking the sale of Nvidia’s H20 GPUs to will cost the company $10.5 billion in lost revenues in the first half of the 2026 fiscal year, executives revealed on Wednesday's Q1 earnings call.

  The export rules, which went into effect in April, effectively cut Nvidia off from the Chinese datacenter market, but not before it managed to deliver roughly $4.6 billion of planned $7.1 billion worth of H20 shipments expected for Q1.

• AI could consume more power than Bitcoin by the end of 2025

  AI already accounts for up to a fifth of the electricity that data centers use, according to de Vries-Gao. It’s a tricky number to pin down without big tech companies sharing data specifically on how much energy their AI models consume. De Vries-Gao had to make projections based on the supply chain for specialized computer chips used for AI. He and other researchers trying to understand AI’s energy consumption have found, however, that its appetite is growing despite efficiency gains — and at a fast enough clip to warrant more scrutiny.

• Schneider Electric says US grid will be less stable by 2030

  Schneider used forecast data and metrics supplied annually to the North American Electric Reliability Corporation (NERC) by US Regional Transmission Operators (RTOs) and Independent System Operators (ISOs) to gauge the level of grid capacity and safety margins. It then analysed how well it might accommodate all projected demand, including that from datacenters, between 2025 and 2030.

• A Network Status Panel The Way It Should Be

  Sometimes a project forms itself around a component rather than an idea, and thus it was that [Maximilien] found himself building a data rate monitor for the connection between two data centers. Some MD0657C2-R LED dot matrix displays for not a lot needed a project.

• AI Angst

  The rent is too damn high · I promise I’ll talk about genAI applications but let’s start with money. Lots of money, big numbers! For example, venture-cap startup money pouring into AI, which as of now apparently adds up to $306 billion. And that’s just startups; Among the giants, Google alone apparently plans $75B in capital expenditure on AI infrastructure, and they represent maybe a quarter at most of cloud capex. You think those are big numbers? McKinsey offers The cost of compute: A $7 trillion race to scale data centers.

  Obviously, lots of people are wondering when and where the revenue will be to pay for it all. There’s one thing we know for sure: The pro-genAI voices are fueled by hundreds of billions of dollars worth of fear and desire; fear that it’ll never pay off and desire for a piece of the money. Can you begin to imagine the pressure for revenue that investors and executives and middle managers are under?

• Europe's cloud datacenter ambition 'crazy' says SAP CEO

  However, he warned about attempts to replicate the services of the large US cloud providers, which claim to offer data sovereignty in Europe. In recent weeks, Microsoft, Google and most recently AWS have all made similar noises about being safe data havens for European customers worried about the policies and rhetoric coming out of the White House.

• A Political Battle Is Brewing Over Data Centers

  The rapid growth in the number of data centers across the US has seen a corresponding rise in local pushback against them. While the projects bring in tax dollars, they often use massive amounts of electricity and water. A recent BloombergNEF analysis found that AI’s electricity demand in the US is expected to triple by 2035, while in Virginia data centers currently use as much electricity as 60 percent of the households in the state.

• Ex-Google employee must face charges of stealing AI secrets for Chinese tech companies

  The federal government claims 38-year-old Linwei Ding, aka Leon Ding, stole sensitive Google trade secrets and other confidential information while secretly affiliating himself with companies in the AI industry. He is accused of transferring files which included the foundational data for Google’s advanced supercomputing data centers, designed to support machine learning workloads and train and host large AI models, according to an indictment from last year.

• This Relay Computer Has Magnetic Tape Storage

  Magnetic tape storage is something many of us will associate with 8-bit microcomputers or 1960s mainframe computers, but it still has a place in the modern data center for long-term backups. It’s likely not to be the first storage tech that would spring to mind when considering a relay computer, but that’s just what [DiPDoT] has done by giving his machine tape storage.

• The Back Of The AI Envelope

  Lets try to figure out the return on this investment. We will assume that the $80B is split two ways, $40B to Nvidia for hardware and $40B on building data centers to put it in. Depreciating the $40B of hardware over five years is very optimistic, it is likely to be uneconomic to run after 2-3 years. But that's what we'll do. So that is minus $8B/year on the bottom line over the next five years. Similarly, depreciating the data centers over 20 years is likely optimistic, given the rate at which AI power demand is increasing. But that's what we'll do, giving another minus $2B/year on the bottom line.

• xAI Data Center Emits Plumes of Pollution, New Video Shows

  xAI quickly built an enormous data center and supercomputer at an empty factory site in Memphis last year, aimed at providing computational power for xAI’s chatbot, called Grok. Having difficulty securing enough grid power to fuel the energy-hungry data center, xAI brought in 35 portable gas turbines, and assembled them without environmental permits or pollution controls.

  A new video, recorded by Oil Field Witness, an environmental group, shows vast plumes of pollution coming from those gas turbines.

  “It’s a power plant. That’s what it is,” Sharon Wilson, director of Oil Field Witness, told Gas Outlook.

• Chinese AI outfits smuggling suitcases full of hard drives to evade U.S. chip restrictions — training AI models in Malaysia using rented servers

  This was a meticulously planned operation and took several months of preparation. Sources say that the engineers chose to fly in the data on hard drives, because it would take a lot of time to transfer the data online without attracting attention. They then divvied up the hard drives between four passengers to avoid raising alarm bells with Malaysian customs and immigration officers. The Chinese personnel then proceeded to a Malaysian data center, where their company rented 300 Nvidia AI servers to process the data and build the AI model.

• Effective infrastructure automation to reduce data center costs

• 'A Black Hole of Energy Use': Meta's Massive AI Data Center Is Stressing Out a Louisiana Community

  Details about how Meta's nearly Manhattan-sized data center will impact consumers' power bills are still secret.

• NREL Maps Out US Data Infrastructure

  Spending time as wee hackers perusing the family atlas taught us an appreciation for a good map, and [Billy Roberts], a cartographer at NREL, has served up a doozy with a map of the data center infrastructure in the United States. [via LinkedIn]

• AI Is Turbocharging Global Inequality

  That warning is based on new data from researchers at Oxford University showing the distribution of the world's most powerful data centers. It's a rare glimpse: historically, companies and nations have tried to keep information about these facilities under close wraps.

• AI's power needs could short-circuit US infrastructure

  Power required by AI datacenters in the US may be more than 30 times greater in a decade, with 5 GW facilities already in the pipeline..

• Cut data center energy costs with bare metal automation

• A trip through vintage datacenter networking

  Donald Davies and his team at National Physical Laboratory (NPL) came up with packet switching in 1965 (although Paul Baran had independently devised a similar concept a few years earlier as distributed adaptive message block switching).

• OpenAI turns to Google's AI chips to power its products: The Information

  The move, which marks the first time OpenAI has used non-Nvidia chips in a meaningful way, shows the Sam Altman-led company's shift away from relying on backer Microsoft's data centres, potentially boosting Google's tensor processing units (TPUs) as a cheaper alternative to Nvidia's graphics processing units (GPUs), the report said.

• Arm-based servers to grow 70% this year

  Arm-based servers are rapidly gaining traction in the market with shipments tipped to jump 70 percent in 2025, however, this remains well short of the chip designer's ambitions to make up half of datacenter CPU sales worldwide by the end of the year.

• Are we already paying for "Free" AI photography? The hidden costs of generative AI laid bare | Digital Camera World

  Generative AI, like AI processing in general, takes a lot of processing power. So, like, whatever… who cares? Tech gets better all the time, computers get faster, nothing bad has happened, it’s a big tech problem not mine, right?

  Well, hold on. Current estimates suggest that an AI processing search query takes about ten times the electricity of a regular query, and that it’s just the energy requirements we need to think about but the land, water and materials needed for the construction of semiconductor plants and data centers. This is according to a June 2024 analysis from Barclays Research, which suggests annual US data center power demands could grow from 14%-21% every year to 2030. US data-centre demand could triple by 2030, from 150-175 terawatt hours (TWh) in 2023 to as much as 560 TWh – equivalent to 13% of the current US electricity demand.

• Chasing A Raspberry Pi Bottleneck

  The Raspberry Pi has been used for many things over its lifetime, and we’re guessing that many of you will have one in perhaps its most common configuration, as a small server. [Thibault] has a Pi 4 in this role, and it’s used to back up the data from his VPS in a data centre. The Pi 4 may be small and relatively affordable, but it’s no slouch in computing terms, so he was extremely surprised to see it showing a transfer speed in bytes per second rather than kilobytes or megabytes. What was up? He set out to find the bottleneck.

• Google’s electricity demand is skyrocketing

  We got two big pieces of energy news from Google this week. The company announced that it’s signed an agreement to purchase electricity from a fusion company’s forthcoming first power plant. Google also released its latest environmental report, which shows that its energy use from data centers has doubled since 2020.

• Tariffs, trade turmoil drive up bitbarn cost and build times

  A survey by datacenter specialist Onnec finds that AI workloads have increased by an average of 42 percent over the past twelve months, with nearly two-thirds of operators saying that requirements to support AI were higher than expected, stretching the capacity of both existing and planned sites.

• AI Might Kill Us All (With Carbon Emissions)

  So-called artificial intelligence (AI) is all the rage right now between your grandma asking ChatGPT how to code in Python or influencers making videos without having to hire extras, but one growing concern is where the power is going to come from for the data centers. The MIT Technology Review team did a deep dive on what the current situation is and whether AI is going to kill us all (with carbon emissions).

• Elon Musk confirms xAI is buying an overseas power plant and shipping the whole thing to the U.S. to power its new data center — 1 million AI GPUs and up to 2 Gigawatts of power under one roof, equivalent to powering 1.9 million homes

  The challenges only become more intense as the company moves forward — Musk faces a monumental challenge with powering his next AI data center, one that is predicted to house one million AI GPUs, thus potentially consuming the same amount of power as 1.9 million households. Here's how the data center could consume that much power, and how Musk plans to deliver it.

• Elon Musk Obtains Permit to Spew Pollution

  In Boxtown, the historically Black neighborhood in South Memphis where xAI's data center is situated, Musk's unfettered pollution has ripped the band-aid off a wound that had barely begun to heal. As Capital B News reported earlier this year, the neighborhood was once home to the Allen Fossil Plant, an electrical facility that left pits of noxious coal ash and a lengthy legacy of environmental racism behind when it was forced to close in 2018.

  In the year since the data center opened and Colossus went online, the smog from Musk's gas turbines has been veritably choking out local residents in a district already struggling with heightened asthma rates due to its proximity to industrial pollution.

• AMD CPU Transient Scheduler Attacks security flaw revealed

  Made public today is a fresh round of security issues, this time for AMD CPUs with Transient Scheduler Attacks. It affects quite a lot of processors including desktop, mobile and data centre.

• Increasingly Paranoid OpenAI Has Installed Fingerprint Scanners and Airgapped Systems to Prevent Secrets Escaping

  Per the Financial Times, the company has gone as far as installing fingerprint "biometric access controls" around its offices, as well as electronically-dependent security airlocks, similar to the kind found in industrial cleanrooms. It likewise beefed up "physical security" — presumably meaning security guards — in its datacenters, and shored up its cybersecurity team for good measure.

• Microsoft says regulations are cramping its Euro expansion

  Microsoft intends to more than double its European datacenter capacity by 2027, but suspects this won't be easy because of all the red tape and environmental safeguards it faces.

• Doing a bit of stress work on a new HDD

  As foreshadowed in x8dtu – drive problems, I will be visiting a data center soon to replace a 4TB HDD. The replacement HDD arrived last night. It was unceremoniously tossed onto the front porch by the courier. However, it was properly packaged. I’m sure it’s fine.

• x8dtu: adding in the smaller drive

  I was up at 5:30 AM today. I packed the car and headed out. I arrived within the datacenter at about 8:15 or so. By 8:50, I was on IRC and the photos of the FreeBSD racks were uploading. Since I was going there anyway, I did some inventory and disposal work (a decommissioned server, about 25 old HDD, and various bits and pieces).

  I must say though, I’m not liking this option. Right now, I have two copies of my data, one in each of the zpools you’ll see listed later. Soon, I’ll destroy one of them, partition the larger drive to match the smaller drive, and add that larger drive to the smaller zpool. Destroying data causes me to go all heebie-jeebie.

• Why Big Tech is threatened by a global push for data sovereignty

  This has always been true, but was perhaps less of an issue when the US government wasn’t both unpredictable and antagonistic. Now, it isn’t just Europe that is seeking to own its own technology future: Global Majority countries are too. And they’re taking steps to make sure it happens.

  Africa is building new data centers. While some more of the wealth and investment will flow into the continent as a result, it doesn’t appear that many of them are actually owned by Africans, which feels like a necessary next step. Otherwise they’re essentially being colonized by tech companies from places like the US and China.

• Meta Built a Data Center Next Door. The Neighbors’ Water Taps Went Dry.

  The Morrises’ experience is one of a growing number of water-related issues around Newton County, which is a one-and-a-half-hour drive east of Atlanta and has a population of about 120,000 people. As tech giants like Meta build data centers in the area, local wells have been damaged, the cost of municipal water has soared and the county’s water commission may face a shortage of the vital resource.

• Meta to invest ‘hundreds of billions’ in new multi-gigawatt AI data centers

  The first data center cluster that the company is building as part of the project is called Prometheus. According to Zuckerberg, it’s set to launch next year and will consume multiple gigawatts of power. One gigawatt corresponds to the electricity use of about one million households.

• Meta's Zuckerberg pledges hundreds of billions for AI data centers in superintelligence push

  Its first multi-gigawatt data center, dubbed Prometheus, is expected to come online in 2026, while another, called Hyperion, will be able to scale up to 5 gigawatts over the coming years, Zuckerberg said in a post on his Threads social media platform.

• Trump announces billions in investments to make Pennsylvania an AI hub

  Pennsylvania is a leading gas-producing state and an epicenter of the US fracking boom; some of the largest investments linked to the initiative were from energy companies, including “a $25 billion investment in data center and energy infrastructure development in Northeast Pennsylvania” from Blackstone, and $15 billion in expansions announced by First Energy.

• Meta plans multi-GW data center that's nearly the size of Manhattan — Zuckerberg promises enormous AI splash as company uses 'tents' to try and keep up with rate of expansion

  They will also leverage pre-fabricated power and cooling modules and will draw power from on-site meta substations, with smart workload management tools helping to maximize the power available to each datacenter. That can also include on-site natural [sic] gas generation, too, with Meta already building two 200MW natural gas plants in Ohio.

• Meta reveals plan for several multi-GW datacenter clusters

  “We're actually building several multi-GW clusters,” he wrote. “We're calling the first one Prometheus and it's coming online in '26. We're also building Hyperion, which will be able to scale up to 5GW over several years. We're building multiple more titan clusters as well. Just one of these covers a significant part of the footprint of Manhattan.”

• Trump and the Energy Industry Are Eager to Power AI With Fossil Fuels

  Natural [sic] gas from Pennsylvania and the Appalachian region, in particular, has faced market challenges both from ultra-cheap natural [sic] gas from the Permian Basin in Texas and New Mexico as well as a lack of infrastructure to carry supply out of the region. These economic headwinds are “why the industry is doing their best to sort of create this drumbeat or this narrative around the need for AI data centers,” says Clark Williams-Derry, an energy finance analyst at the Institute for Energy Economics and Financial Analysis. It appears to be working. Pipeline companies are already pitching new projects to truck gas from the northeast—responding, they say, to data center demand.

• Meta Building Data Centers in Tents, Part of Bid to Catch up in AI - Business Insider

  Zuck confirmed some of the details of SemiAnalysis's report via a Facebook post on Monday. The CEO said Meta will build several new AI data centers that use more than 1 gigawatt of power each. This is how data center capacity is measured, and anything in the 1 gigawatt range is absolutely massive (or at least it was until now).

• Gates unfazed as TerraPower credits dodge Trump cuts

  TerraPower is working to develop a sodium-cooled small modular nuclear reactor. Its design has neither been tested in the real world, nor certified by the Nuclear Regulatory Commission. Gates claimed in the interview that virtually all of TerraPower's current demand in the US is related to datacenters, and predicted that the AI industry would "fill up our order book for the first 10 to 20 units."

• AI data centers are using more power. Regular customers are footing the bill • Michigan Advance

  Between 2024 and 2025, data center power usage accounted for $9 billion, or 174%, of increased power costs, a June report by Monitoring Analytics, an external market monitor for PJM Interconnection, found. PJM manages the electrical power grid and wholesale electric market for 13 states and Washington, D.C., and this spring, customers were told to expect roughly a $25 increase on their monthly electric bill starting June 1.

• Their water taps ran dry when Meta built next door

  Meta's data center in Newton County, Georgia, is linked to water problems. Beverly and Jeff Morris faced dry taps and sediment issues. Local wells are damaged, and water costs are soaring. Newton County may face a water deficit by 2030. Data centers consume vast amounts of water. The county is upgrading recycling facilities.

• The Emerging Challenges of AI Energy Growth

  AI growth, the advent of “hyperscalers”, and plans for new power-hungry data centers dotting the country from coast to coast have overturned previous assumptions of a stable US energy demand growth outlook. One state in particular is at the epicenter of America’s AI revolution: In 2024 alone, Virginia connected fifteen new data centers and anticipates adding another fifteen by the end of 2025. These are not isolated occurrences: already an established hub for US data centers, a recent WoodMackenzie report showed that Virginia lags only Texas as the top destination for newly announced data centers since January 2023 (boasting over 23,000 MW of capacity in the pipeline). Much of this development has been driven by Northern Virginia’s long-standing “Data Center Alley” concentrated around Washington, DC. Meanwhile, the state’s primary utility company, Dominion Energy, has suggested that the average Virginia ratepayer could see their power bills increase by 50 percent over the next fifteen years driven largely by power-hungry new data centers coming online.

• The Epochalypse: It’s Y2K, But 38 Years Later

  Picture this: it’s January 19th, 2038, at exactly 03:14:07 UTC. Somewhere in a data center, a Unix system quietly ticks over its internal clock counter one more time. But instead of moving forward to 03:14:08, something strange happens. The system suddenly thinks it’s December 13th, 1901. Chaos ensues.

• Nvidia AI chips worth $1 billion entered China despite US curbs: FT

  These Nvidia AI chips were reportedly smuggled to China in the three months following tighter US export controls, including high-end B200 processors banned in China, Financial Times reported. Nvidia stated that using smuggled products available on the Chinese black market for data centres is technically and financially inefficient.

• China is developing nation-spanning network to sell surplus data center compute power — latency, disparate hardware are key hurdles

  Many data centers were built on the assumption that state-owned companies and government agencies would purchase computing power. However, demand has fallen short of expectations, leaving many data centers idle or operating at a 20% - 30% load, way below their capacity.

• China's AI data center boom goes bust: Rush leaves billions of dollars in idle infrastructure

  Insiders believe Chinese officials will not abandon these projects, viewing them as growing pains rather than failures. The government is expected to take over floundering centers and assign them to more capable operators. However, for those operators that cannot rent their capacity to clients that can pay, the bubble has clearly gone bust.

• How much water do data centers use and are we prepared?

  If the projections of data center load growth from the region's major utilities become reality, these new facilities in Arizona, Colorado, Nevada, New Mexico and Utah could have annual water use of 21,600 acre-feet (7 billion gallons) by 2035, according to the data. This amount of water serves the annual needs of up to 194,000 people.

• Data Center Impacts in the West: WRA's Policy Solutions

  Left unaddressed, data centers’ loads threaten Western states’ air quality, climate, and water supplies, while risking utilities’ ability to meet statutory, corporate, and science-based emissions reductions goals. With the right policies in place, however, data centers can catalyze investments in innovative clean energy resources while driving broader system transformation.

• How could a proposed hyperscale data center affect Bessemer?

  Instead, public officials, many bound by non-disclosure agreements with the developer, have relied on the assertions of the company proposing the project—a company that has so far refused to answer media questions about the impact of their plans on the community and the environment.

• Big Tech Asked for Looser Clean Water Act Permitting. Trump Wants to Give It to Them

  Environmental lawyers aren’t so sure that a nationwide permit for data centers, regardless of their size, would follow the intent of the Clean Water Act. “What makes [a blanket data center exemption] a little bit tricky is that the impacts are gonna differ quite a bit depending on where these are,” McElfish says. While one data center may impact just a “fraction of an acre,” he says, by rebuilding a stream crossing or filling in a wetland, other data centers in different areas of the country may have much larger impacts to local waterways during their construction.

• AI Datacenters Are Raising Nearby Residents' Electric Bills

  Serving 67 million customers, the PJM region covers just over a dozen states, including Indiana, Maryland, Michigan, Pennsylvania, Virginia, as well as DC. Some of these states will see their energy bills surge by more than 20 percent this summer, Reuters reported; in Philadelphia, according to WaPo, the typical bill rose by about $17. And in Columbus, Ohio, prices spiked by $27.

  "We are seeing every region of the country experience really significant data center load growth," Abe Silverman, a researcher of energy markets at Johns Hopkins University, told WaPo. "It's putting enormous upward pressure on prices, both for transmission and for generation."

• Electricity rates in Ohio and elsewhere rise due to AI and cloud computing

  This time around, though, it is possible to trace the price hikes in these cities to a specific source: the boom in data centers, those large warehouses of technology that support artificial intelligence, cloud computing and other Big Tech wonders. They consume huge amounts of electricity, and, as they proliferate, the surging demand for electricity has driven up prices for millions of people, including residential customers who may not ever use AI or cloud computing.

• Power usage in Wyoming AI data center could eclipse consumption of the state's human residents by 5x — tenant of colossal investment remains a mystery

  The upcoming huge AI data center is so big that it will need its own dedicated energy production using a mix of gas and renewables. Despite this, and the positive spin from local politicians, some experts still think regular households could see their bills increase due to the AI industry's unquenchable thirst for energy.

• Cheyenne to host massive AI data center using more electricity than all Wyoming homes combined

  The latest data center, a joint effort between regional energy infrastructure company Tallgrass and AI data center developer Crusoe, would begin at 1.8 gigawatts of electricity and be scalable to 10 gigawatts, according to a joint company statement.

  A gigawatt can power as many as 1 million homes. But that’s more homes than Wyoming has people. The least populated state, Wyoming, has about 590,000 people.

• Datacenter lobby blows a fuse over EU efficiency proposals

  The Climate Neutral Data Centre Pact (CNDCP or the Pact) is made up of more than 100 bit barn operators and trade associations, and claims to account for more than 85 percent of Europe's datacenter capacity. It lists AWS, Microsoft, and Google among its signatories, as well as firms like Digital Realty, NorthC, and Vantage Data Centers.

• EU plans $30 billion investment in gigawatt AI data centers — multiple sites to host 100,000 AI GPUs each as bloc plays catch-up to US and China

  The EU's effort, if realized, is probably one of the world's largest publicly funded initiatives in artificial intelligence, probably well below what Chinese authorities (both federal and local) have invested in AI data centers, but well ahead of what other big economies invest in their AI efforts.

  Henna Virkkunen, European Commission executive vice president for technology policy, told CNBC that while Europe has a strong talent base — reportedly 30% more AI researchers per capita than the U.S. — their limited access to computing has held back development. Building massive AI data centers is designed to solve this problem and kick-start the AI sector across the EU.

• UK needs bit barns to lead AI deployment, says Blair group

  In its report, Blair's think tank says Brtain lacks the resources to keep up with the US, China and the Gulf States, which are all pumping hundreds of billions of dollars into vast, energy-hungry datacenters for AI training. The UK doesn't have the money, the available land or the energy resources to follow suit, it adds.

• SoftBank buys Foxconn's Ohio plant to jumpstart Stargate AI push: Report

  SoftBank Group is reportedly acquiring Foxconn’s Ohio EV plant to kickstart its $500 billion Stargate data center project with OpenAI and Oracle. The move signals SoftBank’s push into AI infrastructure.

• Tesla Shutting Down Its AI Supercomputer As Staff Leaves in Droves to Join Competitor

  As Bloomberg reports, the head of the project, dubbed Dojo and which CEO Elon Musk used to hype up immensely, is leaving the company. The team has already lost around 20 workers to a separate data center firm that's been poaching former Tesla executives.

  The rest of the team will reportedly be reassigned to other data center projects.

• Russian data centre in Helsinki sparks concerns over legal loopholes

  News emerged over the weekend that Beeline, one of Russia's leading mobile operators, has opened a telecommunications data centre in Helsinki. Now Finnish authorities say they were unaware of the data centre's existence.

• DEFRA says: delete your emails to … save water?!

  You might think the issue involves the government-advocated data centres that use vast amounts of fresh water each day for cooling. Or perhaps the water companies that leak vast amounts of water and just don’t do maintenance.

  No, it’s you. Personally. You’re the problem. Not them.

  [...]

  But either way, the amount of power does not depend how much data’s stored on the disk. It just doesn’t. It has no relation.

• The AI Industry Just Got Some Horrible News

  For years now, big tech and its financial backers have been blowing past the warning signs that human-level AI — the theoretical technology with hundreds of billions of dollars riding on it — could turn out to be a pipe-dream, at best.

  Now, however, Wall Street appears to be yanking its money out of one of the largest datacenter providers in the game, possibly an early indicator that reality could be finally catching up to the massively hype-dependent AI industry.

• Delete old emails and pictures as data centres require vast amounts of water to cool their systems.

  As David Gerard pointed out, the document makes no mention of the inefficiencies of (privatized) water companies nor the thirsty data centers focused on artificial intelligence. (The great irony is that the suggestion to delete emails to save water may have come from generative AI. Is this how AI will end humanity?)

• AI experts warn that China is miles ahead of the US in electricity generation — lack of supply and infrastructure threatens the US’s long-term AI plans

  A U.S. analyst of Chinese technology said that the country has already solved its energy problem — at least in terms of power for its AI infrastructure. Rui Ma, founder of Tech Buzz China, posted on X that the country’s massive investments in advanced hydropower and nuclear technologies meant that its “electricity supply is secure and inexpensive.” This is in contrast to the U.S., where many AI data centers are disrupting its electricity grid and supply, resulting in a lack of supply and price increases for every user.

• World's largest data center campus could be coming to central Utah

  The first domino fell when Orem-based Fibernet MercuryDelta LLC in May filed a request to rezone nearly 1,200 acres of property — located southeast of Delta — from agricultural land to heavy industrial land for its potential 20-million-square-foot data center campus called Delta Gigasite.

• ChatGPT 5 power consumption could be as much as eight times higher than GPT 4 — research institute estimates medium-sized GPT-5 response can consume up to 40 watt-hours of electricity

  The University of Rhode Island's AI lab estimates that GPT-5 averages just over 18 Wh per query, so putting all of ChatGPT's reported 2.5 billion requests a day through the model could see energy usage as high as 45 GWh.

  A daily energy use of 45 GWh is enormous. A typical modern nuclear power plant produces between 1 and 1.6 GW of electricity per reactor per hour, so data centers running OpenAI's GPT-5 at 18 Wh per query could require the power equivalent of two to three nuclear power reactors, an amount that could be enough to power a small country.

• OpenAI’s GPT-5: The Great Energy Mystery

  This increase in power usage is not a technical aside. With ChatGPT's projected 2.5 billion daily requests, GPT-5's electricity appetite for one day may match that of 1.5 million US homes. The power and related carbon footprint of high-end AI models are quickly eclipsing much other consumer electronics, and data centres containing these models are stretching local energy resources.

• Electricity prices are climbing more than twice as fast as inflation

  Power-hungry data centers have been popping up all over, to serve the boom in artificial intelligence. The Energy Department projects data centers and other commercial customers will use more electricity than households for the first time ever next year. That's a challenge for policymakers, who have to decide how to accommodate that extra demand and who should foot the bill.

• Data Center Lobbyists Clear the Way for Mega-Projects in Rural Georgia

  A DeSmog review of public records suggests that industry lobbyists and company representatives prevailed upon Coweta County officials to dilute earlier versions of the proposed planning rules for data centers.

  Specifically, residents opposing Project Sail are concerned that the latest draft of the rules — known as an ordinance — no longer includes certain provisions designed to limit adverse environmental impacts and require special public hearings for proposed data centers.

• Google datacenters to be powered by molten salt reactors

  The small modular reactor startup, along with the Tennessee Valley Authority (TVA), one of the largest utilities in the country, aim to bring 50 megawatts of nuclear power to Tennessee.

• Reflecting on 10 years in open source network security: Lessons from FastNetMon

  Ten years ago, I wrote the first lines of FastNetMon in my home lab, trying to detect volumetric Distributed Denial-of-Service DDoS attacks in a small data centre environment I was managing. What started as a proof-of-concept tool with hardcoded thresholds and no config file has since grown into a globally deployed open source DDoS detection system, shaped entirely by the needs and input of the network operator community.

  In this post, I want to reflect on two parallel journeys: How DDoS attacks have evolved over the last decade, and how the FastNetMon Community project has grown in response — from a homegrown script to a battle-tested detection platform trusted by ISPs, hosting providers, and enterprises.

• OpenAI’s Sam Altman points to AI ‘bubble’

  A number of experts agree: The Apollo investment group’s chief economist noted that tech firms’ price-earnings ratios are unusually elevated, reminiscent of the late-1990s dotcom boom and subsequent crash, and also that spending on data center infrastructure accounts for a worryingly high percentage of US GDP growth.

• Disaster Recovery: Achieving Instantaneous Hot-Hot with OpenShift

  If you're moving from a colo that someone manages to VMware, the cloud, or a differentVMware data center, they all look and feel different.

• Kilowatt Madness

  And this is only the start. Many analysts expect further large increases in electricity prices over the next year or more, largely because of surging demand from AI data centers. The electricity outlook is sufficiently scary that Texas — Texas! — has passed a new law giving the grid operator the right to cut off data centers during periods of power shortage.

  An aside: In the America I grew up in, people who made big boasts about what they would achieve then completely failed to deliver were considered unserious blowhards. What happened to that country?

• Data Center Boom Straining U.S. Electric Grid

  Large data center developments that are connecting to PJM, one of largest regional transmission organizations (RTO) in the U.S., should provide their own power generation, according to U.S. electricity market researcher, Monitoring Analytics, MobileWorldLive reported. The research firm warns that the existing electrical grid infrastructure is unable to handle surging demand.

• Anubis.

  If that sounds familiar, it’s because it’s similar to how bitcoin mining works. Anubis is not literally mining cryptocurrency, but it is similar in concept to other projects that do exactly that, perhaps most famously Coinhive and JSECoin.

  So how do some useless SHA-256 operations prove you’re not a bot? The argument goes that this simply makes it too expensive to crawl your website. The typical datacenter used by an AI crawler

  This… makes no sense to me. Almost by definition, an AI vendor will have a datacenter full of compute capacity. It feels like this solution has the problem backwards, effectively only limiting access to those without resources or trying to conserve them.

• Data centers are drying up the Port of Marseille: ‘They consume enormous amounts of electricity’

  City Hall has a plan to make the presence of cruise ships more bearable: electrify them and connect them to the grid, so they can shut down their engines and still continue operating. The idea is that shipyard repairs — as well as loading and unloading operations at the dock — will also use this energy. But it’s unclear whether there will be enough electricity to make this feasible. This is because data centers in the port area, active since 2021, have priority when it comes to the energy that’s available.

• UK bit barn builds turn to gas rather than wait for grid pow

  This is a situation that has already played out across the Atlantic, with pipeline companies in the US disclosing last year that they were getting lots of interest from bit barn operators to supply them with large quantities of natural gas because of delays in getting wired up to the grid.

  Typically, the end result is the campus gaining a small on-site power station to supply electricity for the individual data halls. This is most likely generated by gas turbines, although in some cases solid-oxide fuel cells might be used, as is the case in some sites managed by global datacenter giant Equinix.

• Google quietly removes net-zero carbon goal from website amid rapid power-hungry AI data center buildout — industry-first sustainability pledge moved to background amidst AI energy crisis

  It appears that Google is starting to hide its internal climate advocacy goals from the public eye. The [Internet] titan has silently removed its goal to "pursue net-zero emissions" across all operations by 2030 from its Sustainability webpage, as first spotted by Canada's National Observer.

• Red Sea submarine cable outage slows Microsoft cloud

  Junior cloud Linode has also warned of “network congestion and latency due to multiple faults in the undersea cables that are part of the optimal routes out of the data centers.”

• Oracle shares soar on AI contracts

  The US software maker said it expected to get $144 billion per year by the end of this decade from AI companies using its cloud computing capabilities, much of it from OpenAI, which The Wall Street Journal reported will buy 4.5 gigawatts’ worth of data center capacity from Oracle — enough energy to power millions of homes.

• Proxmox delivers datacenter manager beta

  Proxmox’s main offering is the Virtual Environment (PVE), a hyperconverged infrastructure suite that allows users to run and manage virtual machines and containers. Proxmox Server Solutions GmbH, the Austrian company behind the product, says users employ its wares on over 1.6 million hosts.

• Intel loses chief architect behind its Xeon CPUs

  His innovations aren't limited to the datacenter either, with his architectural contributions playing a significant role in the success of Intel's Core and Atom processor families as well.

• Malaysia reins in data centre growth, complicating China's AI chip access

  The Southeast Asian country has drawn in data centre investments from US technology giants like Microsoft, Amazon, Alphabet's Google and their Chinese counterparts Tencent, Huawei and Alibaba in recent years, spurred by cheap land and electricity costs and robust local AI demand prospects.

• OpenAI’s staggering numbers

  It reportedly centers on new data centers announced in July, designed to deliver 4.5 gigawatts of power — enough to supply a city of four million people. The deal could also partially replace the stalled $500 billion Stargate project, unveiled in January but left in limbo for lack of funding.

• Google lands £400M MoD contract for secure UK cloud services

  The arrangement promises Google AI, data analytics, and cybersecurity, all in sovereign datacenters built in the UK. Defense intelligence and national security specialists will also use the systems to share secure information between UK partners, the MoD statement said. The aim is to strengthen secure communication links between the UK and US, adding to the security partnership the two nations share.

• OpenAI signs contract to buy $300 billion worth of Oracle computing power over the next five years — company needs 4.5 gigawatts of power, enough to power four million homes

  In one of the largest cloud contracts ever signed, OpenAI has inked a deal with Oracle to purchase $300 billion worth of computing power over the next five years, according to WSJ. That's close to 100 times OpenAI's 2024 total revenue, but follows earlier gigantic partnership commitments between the two companies, including Oracle signing on to the $500 billion Stargate data center plan, and plans for another joint data center project announced in July.

• Navigating complexity, delivering value: our take on why Forrester named Red Hat OpenShift a Leader

  Today's IT leaders face intense pressure to accelerate innovation while managing a complex mix of technologies. They are faced with the challenge of integrating AI, modernizing legacy systems and controlling costs within a complex hybrid cloud that spans datacenters, public clouds and the edge. The core question for every technology leader is how to build a platform strategy that tames this complexity, strengthens security and reliability, and creates a foundation for future growth.

• Who is Larry Ellison, who might soon become the world’s richest person?

  It traces back to 1977, when Ellison, Bob Miner and Ed Oates founded Software Development Laboratories in Santa Clara, California. They were contracted by the CIA to build a database program code-named “Oracle.”

  Five years later, the company changed its name to Oracle and went public in 1986. Throughout the four decades, the company has found various levels of success in technology, but most recently in demand for its data center capacity from AI customers.

  [...]

  In 2010, he signed the Giving Pledge, a Bill Gates-backed initiative that promises to give away at least 95% of one’s wealth to charity. Ellison recently announced on X he was “amending” that effort, funneling some of the money toward a technical institute he founded with the University of Oxford.

• US electric grids under pressure from energy-hungry data centers are changing strategy

  Now the concept is emerging in the 13-state mid-Atlantic grid and elsewhere as massive data centers are coming online faster than power plants can be built and connected to grids. That has elicited pushback from data centers and Big Tech, for whom a steady power supply is vital.

  Like many other states, Texas wants to attract data centers as an economic boon, but it faces the challenge of meeting the huge volumes of electricity the centers demand. Lawmakers there passed a bill in June that, among other things, orders up standards for power emergencies when utilities must disconnect big electric users.

• Microsoft boosts Wisconsin data center spending to $7 billion

  Microsoft on Thursday said it plans to build a second massive artificial intelligence data center in Wisconsin, bringing its spending in the state to more than $7 billion. The new $4 billion project will join a $3.3 billion data center in Mount Pleasant in the southeastern corner of the state, announced last year.

• Nebraska Faces Backlash Over Plan to Use Only $44M of $405M Federal Broadband Funds

  “If there were a trophy for squandering opportunity, Nebraska would already have it on the shelf,” said Gage County Board member Emily Haxby, a longtime advocate for better rural service. John Nelson, CEO of Hamilton Telecommunications, added that “a lot of locations will be left behind.”

  Inside Towers reported the debate now centers on whether unspent funds will be clawed back by the Trump administration or repurposed for “non-deployment” broadband projects like digital training and data centers. Other states, including Louisiana, argue federal law allows them to retain excess funds for broadband purposes.

• NVIDIA are investing $5 billion in Intel to develop new chips together

  A huge announcement for the entire computing industry came today, with NVIDIA and Intel teaming up to make new chips for data centre and consumers too. This is of course, largely about AI, and they want to ensure they stay ahead of AMD and Arm chip makers.

• Intel and Nvidia are teaming up to make a new chipset. Should Apple be worried?

  For the server market, Intel will build custom x86 CPUs that Nvidia will use in its AI datacenter products. Nvidia already rules the AI datacenter business and if it can eke out a little more performance or efficiency with a custom CPU, that’s good, but it hardly matters to those who use Apple products.

• Big Tech Dreams of Putting Data Centers in Space

  The companies frantically building and leasing data centers are well aware that they’re straining grids, driving emissions, and guzzling water. The electricity demand of AI data centers in particular could increase as much as 165 percent by 2030. Over half of the energy powering these sprawling facilities comes from fossil fuels, threatening to reverse progress toward addressing the climate crisis.

• How an SA Data Centre and Subsea Cables Power Zim’s Internet

  Over the past nine years, around $2 billion a year has gone into new subsea cable construction. But between 2025 and 2027, more than $13 billion worth of new systems are expected to come online.

  Unlike the early 2000s, when state-owned enterprises and telecom companies joined forces and led in this, the investment map has shifted.

• AI hype train may jump the tracks over $2T bill, warns Bain

  Management consultant biz Bain calculates the industry is heading toward creating an extra 100 gigawatts of capacity in the US by 2030 to meet demand, but it also estimates spending of $500 billion per annum on building datacenters will be needed to get there.

• 7 Top Google Cloud Platform (GCP) Courses on Udemy in 2025

  It consists of a set of physical assets, i.e., computers, hard disks, and virtual resources that are contained in Google’s data centres in each region around the globe.

• Underground data fortresses: the nuclear bunkers, mines and mountains being transformed to protect our ‘new gold’ from attack

  This subterranean data centre is located in a former nuclear bunker that was constructed in the early 1950s as a command-and-control centre for the Royal Air Force’s radar network. You can still see the decaying concrete plinths that the radar dish once sat upon. Personnel stationed in the bunker would have closely watched their screens for signs of nuclear missile-carrying aircraft.

• Wind and solar will power datacenters more cheaply than SMRs

  But analysis from the Centre for Net Zero (CNZ) says it would cost 43 percent less to power a 120 MW data facility with renewables and a small amount of gas-generated energy, when compared with an SMR.

  It claims that a microgrid comprising offshore wind, solar, battery storage, and backed up by gas generation, would be significantly cheaper to run annually than procuring power sourced from a nuclear SMR.

• Datacenter fire downs 647 South Korean government services

  The fire struck on Friday night at a datacenter operated by South Korea’s National Information Resources Service. Korean media report that technicians replacing a lithium-ion battery inadvertently sparked a blaze. As is often the case with battery fires, firefighters struggled to control the blaze, which reached 234 batteries.

• Residents Shut Down Google Data Center Before It Can Be Built

  Google nixed plans for a gigantic $1 billion data center on more than 460 acres of land in Indiana after residents hotly protested the proposal due to concerns that the complex would jack up electricity prices for neighbors and suck away untold gallons of water in an area already plagued with drought.

• Jerome and Norwich township officials concerned with data center surge

  Ohio townships are concerned that the rapid growth of data centers is straining local emergency services. Firefighters report frequent emergency calls and delays in accessing high-security Amazon data center sites.

• First Responders Are Being Overwhelmed by Data Center Fires

  For one thing, the facilities are incredibly noisy, their cooling systems likened to a “distant jet that never leaves.” Their collective energy demand is so high that they’re driving up utility prices throughout the country, in some cases prompting insufferable CEOs to resort to methane gas generators, choking local communities with noxious fumes. Their water intake is so enormous that taps in nearby homes can stop running, while their excessive pollution has been linked to a growing cancer risk.

  That’s before we even get into the massive financial problem with the AI data center gold rush, which in many ways is now holding up the stagnant US economy.

  As if all that weren’t enough, one community is now learning about another obnoxious data center trait: their tendency to burst into flames.

• A.I. Is on the Rise, and So Is the Environmental Impact of the Data Centers That Drive It

  To power the digital world—from day-to-day digital communications, websites and data storage—data centers require energy to power the hundreds of servers within them. With the advent of more hyperscale data centers being built to support A.I., data center energy use has increased.

  Benjamin Lee, a computer scientist at the University of Pennsylvania, breaks the high energy consumption of A.I. into two categories.

• Data centers to account for 9% of electricity demand in the U.S by 2035 — Nuclear power could help sate AI demand

  But with individual graphics cards requiring upwards of 500W of power, there are frank discussions taking place over where all the extra electricity for these projects is going to come from. In June this year, EpochAI published a study which showed that the most capable AI supercomputing systems were pulling more power with each iteration, and doubling those needs every couple of years. Elon Musk's xAI Colossus supercomputer requires over 280 Megawatts - 20 times the power needed for the top AI data center in 2019, according to the study.

• AI data center boom sends some wholesale electricity prices soaring up to 267% in five years, says report — as global rollout of AI factories continues apace

  In its analysis of Grid Status' energy data analytics, Bloomberg found that for areas located near new data centers built to run and train AI, wholesale electricity prices had risen by up to 267%. Although that cost doesn't necessarily have to be passed on to the consumer, some of its cited examples clearly show that happening.

• Google Announces $4 Billion Investment in AI Data Center in West Memphis, Arkansas

  West Memphis — and its future AI datacenter — is right across the Mississippi River from downtown Memphis, and just a couple dozen miles from xAI’s first site in Memphis. That obviously leads to the obvious question about power deliver to the datacenter.

• AI’s Unquenchable Thirst for Water Could Rival That of Some Nations

  Luccioni describes a December 2024 conference hosted by the International Energy Agency, which brought big tech companies into the same room as energy grid operators from across the world. To do their jobs, the latter, Luccioni says, need to know what electrical demand will be. “Data centres will use energy 24/7,” she says, but electricity providers “don’t get any information, so they don’t know how many data centres are going up.” It’s kind of just forced upon them. They can’t plan ahead. “And so they were panicking a little, being like, ‘We need numbers, we need at least some projections, we need to know where we’re going. We need to know how we have to build out our capacity, because otherwise we can’t respond to demand.’”

  The reason why all this information is so critical was underscored early this summer, when a heat wave struck the entire eastern North American continent, driving humidex temperatures into the mid-forties. Electricity providers struggled to keep up with demand from air conditioning, which, during bouts of increasingly common extreme summer heat, keeps many people alive. As utilities warned of impending rolling blackouts, tech companies were racing to plan a mass expansion of energy-sucking data centres.

• ‘Media Need to Report on the Real Cost of Data Centers’: CounterSpin interview with Jai Dulani and Vivek Bharathan on data center opposition

  Janine Jackson interviewed MediaJustice’s Jai Dulani and Vivek Bharathan of the No Desert Data Center Coalition about data center opposition for the September 26, 2025, episode of CounterSpin. This is a lightly edited transcript.

• AI data centers are swallowing the world's memory and storage supply, setting the stage for a pricing apocalypse that could last a decade

  Once-cheap SSDs, DRAM, and HDD prices are climbing fast as AI demand and constrained supply converge to create the tightest market in years.

• China to launch commercial underwater data center — facility expected to consume 90% less power for cooling

  Chinese company Highlander is building a server pod near Shanghai, where it will be submerged underwater to reduce the power consumption used by the facility for cooling. According to the South China Morning Post, Highlander is expected to sink the underwater pods in October. The servers will operate commercially, with state-owned institutions like China Telecom being among the first customers.

• China to sink servers off Shanghai in underwater data center trial

  “Underwater facilities can save around 90 per cent of energy consumption for cooling,” said Yang, Highlander’s vice-president.

• OpenAI reportedly closes new $6.6B secondary sale at $500B valuation

  The ChatGPT developer’s newly closed secondary sale apparently also included contributions from SoftBank Group Corp. and Abu Dhabi’s MGX. Both companies are backing OpenAI’s Stargate initiative, a $500 billion push to build a network of artificial intelligence data centers in the U.S. The facilities’ combined compute capacity is expected to exceed 10 gigawatts.

• Microsoft inks $33 billion in deals with 'neoclouds' like Nebius, CoreWeave — Nebius deal alone secures 100,000 Nvidia GB300 chips for internal use

  Microsoft is renting external GPU data centers for internal use so it can, in turn, rent out its own facilities to other customers.

• You Are Not Prepared for Merriam-Webster's Announcement of Its New AI Model

  “It never hallucinates. It does not require data centers, and uses no electricity,” the booming voice continues. “It’s a powerful tool that will change how you communicate — forever.”

  Then the video cuts to the reveal. It’s Merriam-Webster’s newest “Collegiate Dictionary,” the 12th edition, which was announced last week. A glorious, actually physical book, slowly rotating. Not a prompt box in sight. A red tome, as opposed to a black box.

• AI Data Centers Are Skyrocketing Regular People's Energy Bills

  A very granular Bloomberg analysis of wholesale electricity prices across the country paints a picture of doubled wholesale prices and soaring consumer costs over the past five years. It’s complex, but one thing is for certain: those within 50 miles of data centers are bearing the brunt of it.

  The issue is that about two thirds of the country’s power consumption runs on either a state or regional grid. This means the costs necessary for network expansion and maintenance is trickling down to residents and businesses’ electricity bills.

• Bezos dreams of orbital datacenters powered by the sun

  However, even if we can make putting bit barns in orbit economical, there is no shortage of challenges to overcome. Datacenters are massively complex facilities. Even with an abundant supply of power, an orbital datacenter would still need a way to reject a gigawatt of thermal energy through radiation.

  For reference, the ISS's radiators are capable of rejecting about 70 kilowatts of thermal energy. Bit barn operators will also have to contend with bit-flips on a fairly regular basis unless the hull can be sufficiently hardened against charged particles from the sun and cosmic rays from the outer reaches of space — standard ECC probably isn't going to cut it.

• [Old] 'Shark Tank' Star Kevin O'Leary Says Bitcoin Mining, AI Data Centers Are Locked In A 'Power Struggle' Amid Scarce Electricity In US

  He added that setting AI data centers and Bitcoin mining operations risks raising local power rates, which regulators avoid.

• OpenAI Agrees to Use Computer Chips From AMD

  On Monday, OpenAI said it would begin using AMD chips in the second half of next year as it built computer data centers. The new facilities would be separate from the data centers that OpenAI has committed to building in Texas, New Mexico, Ohio and an unnamed site in the Midwest.

• Google to invest $10 billion in Visakhapatnam for Asia's largest data centre cluster

  The technology giant plans to invest $10 billion (around Rs 88,730 crore) to develop a one-gigawatt data centre cluster in the city.

• Intel rethinking how it contributes to open source community

  Speaking to press and analysts at Intel's Tech Tour in Arizona last week, Kevork Kechichian, who now leads Intel's datacenter biz, believes it's time to rethink what Chipzilla contributes to the open source community.

• Softbank-owned chipmaker Graphcore will invest $1.3 billion in India over 10 years: CEO Nigel Toon

  Graphcore, a UK chip firm, will invest $1.3 billion in India over the next decade. The company is opening its first office in Bengaluru, hiring 100 AI engineers initially. This move taps into India's talent pool for chip design. Graphcore's smartchips power AI data centers. The investment signals a significant expansion into the Indian market.

• Brazil to begin construction on TikTok data center in six months, minister says

  Plans for the project, which is expected to be based in the northeastern state of Ceara, were disclosed by Reuters in April, when three people familiar with the matter had said TikTok's Chinese parent company ByteDance was weighing the move.

• Coal making a comeback as US datacenters demand power

  This spike in demand is driving an unexpected resurgence in coal generation, which has increased nearly 20 percent year-to-date. The research note, seen by The Register, states: [...]

• Michigan township sued by AI data center builder and disgruntled residents over opposition to the site — mounting concerns about rising power bills and water usage fuel growing skepticism

  In the latter case, CRG wanted to build a 440-acre data center near multiple wells, backed by an unnamed tech company. The township reportedly didn't get straight answers from the developers about the water usage and noise pollution, with one constituent remarking the companies seemed to want to "shove [the project] down the people’s throats". After drawing strong criticism, CRG withdrew its application, and the city council called a special meeting to ban data centers for one year, passing it unanimously.

• Cali governor derails bill to monitor datacenter water use

  Assembly Bill 93 would have mandated that data center operators provide water suppliers with estimates of expected usage before obtaining business licenses, followed by annual reports of actual consumption when renewing permits. However, in an October 11 letter [PDF] to the California State Assembly, Newsom declined to sign the bill.

• AI data centers won't have to disclose water use in California

  New data centers have been rapidly proliferating in California and other western states as the rise of artificial intelligence and growing investments in cloud computing drive a construction boom. The centers, full of equipment, generate lots of heat and can use large quantities of water to cool their servers and interiors. Many companies don’t reveal how much they use.

  Assembly Bill 93, introduced by Assemblymember Diane Papan (D-San Mateo), would have required new data centers to disclose their expected water use when they apply for a business license and would have required all to report their water consumption annually.

• As electricity bills rise, candidates in both parties blame data centers

  As electricity bills rise, a growing number of US candidates in both parties are pointing to the high energy costs of data centers — booming thanks to tech companies’ AI investments — as the culprit. While the issue isn’t yet a flashpoint in statewide races, it’s already an overwhelming source of debate in local ones, especially in Virginia.

• We're all going to be paying AI's Godzilla-sized power bills

  Today, a large, AI-dedicated datacenter typically requires 100 megawatts (MW). That's roughly equivalent to the energy used by 100,000 homes. That's a lot of power, but it's not that much more than your typical hyperscaler datacenter. However, there are already a lot of AI datacenters. By last count, we're up to 746 AI datacenters.

  Think that's a lot? That's nothing compared to where we're going.

• Google to invest $15 billion in India AI data center

  The data center and AI hub will be built in the port city of Vishakhapatman in the southeastern state of Andhra Pradesh over the next five years.

  The data center campus will have a capacity of 1-Gigawatt initially, but will be scaled to "multiple gigawatts," Kurian said at the event, referring to the electricity demand required by data centers powering AI.

• OpenAI losing 3x what it makes. Will porn save it?

  OpenAI is valued at $500bn and has committed to spending more than $1tn on datacenters and GPUs and what-have-you in pursuit of all the things AI is good for. But it currently tallies only $13bn in annual recurring revenue and is losing about three times what it already spends, writes Thomas Claburn: "ChatGPT is so popular that almost no one will pay for it."

• Old-school material could power quantum computing and cut data center energy use

  Barium titanate, first discovered in 1941, is known for its powerful electro-optic properties in bulk, or three-dimensional, crystals. Electro-optic materials like barium titanate act as bridges between electricity and light, converting signals carried by electrons into signals carried by photons, or particles of light.

• AI Is Going to Kill Everyone You Love. The Surprise Is How.

  This vision benefits fossil-fuel companies because in the world of data centers, speed and reliability are key. Solar and wind would require massive battery capacity to supply enough nonstop energy—which can be expensive—and a lot of land area. Gas plants are relatively quick to build, and enjoy subsidies from the Trump administration. As a result, gas, not renewables, has become the go-to power source for AI slop-bots and Grok’s white supremacy posts. And that means the electricity used by data centers is dirty—about 48 percent more carbon-intensive than the US average. Even if promises of future efficiency come through, it’ll be too late: Once these plants are built, there’s no going back. The three gas plants that’ll power Meta’s AI data center in Louisiana are built to operate for 30 years. By the time the plants are retired, most of Miami-Dade County will already be submerged by rising seas.

  Each new piece of infrastructure locks us in for decades of emissions, at a time when—the science is overwhelmingly clear—we cannot go on building fossil-fuel infrastructure and have a livable planet. The threat, in Louisiana, is not abstract. It is immediate.

• AI-related data centres use vast amounts of water. But gauging how much is a murky business

  Her main concern, shared by other local opponents, is the amount of municipal drinking water the 200,000-square foot data centre would need for its cooling system. In a region beset by drought, Barnwell says similar-sized facilities can churn through 70,000 litres of potable water a day.

• OpenAI Inks Deal With Broadcom to Design Its Own Chips for A.I.

  After signing deals to use computer chips from Nvidia and AMD, OpenAI plans to design and deploy its own chips as it spends hundreds of billions of dollars building new computer data centers.

• Microsoft denies Mexico data center linked to water shortages, local illnesses, and power outages — stomach bugs and even hepatitis reported in region as 1.5 Gigawatt AI data center buildout looms

  Though causation does not imply causality and the individual incidents are isolated and not necessarily linked, the reports of problems have escalated in the last year since the data center complex in Queretaro was opened. The NYT report cites Dulce María Nicolás, a resident and mother of two from the nearby town of Las Cenizas, who claims water outages have lasted weeks and led to school cancellations. Víctor Bárcenas, who runs a local health clinic, claims that because of power outages he attributes to the data centers, one of his patients couldn't be treated locally and had to be rushed to a hospital over an hour away.

  Bárcenas also claimed that a recent lack of clean water even led to a bout of hepatitis spreading rapidly over the summer of 2025.

• AI Data Centers Create Fury From Mexico to Ireland

  Directly linking any data center to local power and water shortages is difficult. Yet building in areas with unstable grids and existing water strains has pressured already frail systems, according to experts, increasing the potential for cascading effects.

  In country after country, activists, residents and environmental organizations have banded together to oppose data centers. Some have tried blocking the projects, while others have pushed for more oversight and transparency.

• The ratepayer backlash to data center expansion is here

  Copious data centers have sprung up across the United States, nearly doubling in number between 2021 and 2024, with no end in sight to their rapid spread. According to consulting firm McKinsey & Company, companies are projected to spend $1.6 trillion on data center hardware in the US by 2030.

  It’s not just the existing facilities that are creating heftier bills; even data centers that have yet to be built are driving up power prices today.

• Will the AI Boom Lead to Water and Electricity Shortages?

  Data centers house the computers, servers, and other hardware used to process and store digital information. With their enhanced processing capabilities, large “hyperscale” complexes are the preferred data centers for the computation-heavy training and use of AI models. They can cover an area of over 1 million square feet, roughly equal to 17 football fields. Water is used to maintain humidity and as a coolant for the heat-generating machines, and as American data centers have grown in size and number, so has their water consumption, from 5.6 billion gallons in 2014 to 17.4 billion in 2023.

  Communities are already feeling the squeeze. In Newton County, Georgia, a single Meta-owned center accounts for about 10 percent of all water use. Homeowners near the facility told The New York Times they believed the data center’s construction had damaged their wells: Their taps yield brown water, or no water at all. The community is also facing water shortages, and prices have skyrocketed; in the future, locals may have to resort to rationing. That’s a steep price to pay so that Mark Zuckerberg can sell AI-enabled spy glasses.

• Grounded jet engines take off again as datacenter generators

  The Financial Times says two-thirds of gas turbines for electricity generation come from three manufacturers: Japan's MHI, Germany's Siemens, and GE Vernova, the latter formed from a spin-out of General Electric's energy businesses.

  It quotes an MHI executive as saying: "There's so much demand right now that we can't meet it all," and that there is a lot of demand out of North America, meaning customers are now facing a three-year waiting list for delivery of turbine generators.

• The Path to Digital Sovereignty: Why an Open Ecosystem is the Key for Europe

  This isn’t a move towards isolation, with true sovereignty requiring more than just local data centers. Businesses are re-evaluating their provider dependencies and are increasingly looking for a flexible approach that allows them to define their own digital destiny through a collaborative and open ecosystem of partners.

• Why your electric bill is so high now: Blame AI data centers

  Electricity prices have been going up pretty dramatically over the past year. In some places, they’re rising by double-digit percentages, and they’re projected to rise even further. We’re talking about prices that are paid by consumers, so this is actually showing up on people’s power bills, which is why it’s getting a lot of attention.

• What Americans Think About the Environmental Impact of AI, According to a New Poll

  The results of the poll, conducted in September, suggest that as AI reshapes work, communication and culture, it’s also sparking anxieties about how the growing energy demands could further harm the environment.

  It takes massive amounts of electricity to power AI. Electricity consumption from data centers is set to more than double globally by 2030, according to the International Energy Agency. The United States accounts for by far the largest share of the projected increase, followed by China. In many places, the electricity for data centers will come from power plants that burn coal, oil and natural gas. Burning these fossil fuels for electricity emits carbon dioxide, trapping heat in the atmosphere and warming the planet.

• How Data Centers Actually Work

  Tech giants have been investing hundreds of billions of dollars into AI data centers just this year alone. But as the deals pile up, so have the concerns around their viability and sustainability. Michael Calore and senior correspondent Lauren Goode sit down with senior writer Molly Taft to discuss how these energy hungry facilities actually work, the different industry interests at stake, and whether it’ll all come crumbling down.

• What Tech Insiders Actually Think of AI Is Extremely Revealing

  Tech companies are committing hundreds of billions of dollars to enormous data center buildouts, while industry leaders are promising an inevitable future in which superintelligent machines will usher in an era of unprecedented prosperity.

  But many tech insiders are quietly singing a very different tune, expressing a growing conviction that’s been circulating for a while now: that AI is immensely overhyped.

• AWS outage: Myths vs reality

  A single AWS region is a single point of failure. With a lot of dedicated work, you can add another cloud provider or AWS region or datacenter into the mix until finally, at tremendous effort and expense, you have added a second single point of failure. Good work; now you're subject to your existing issues with us-east-1, but you've added in your workload having problems whenever Azure hiccups too.

• Nokia shares hit near 10-year high after Nvidia invests $1 billion in AI push

  Since joining Nokia earlier this year, new CEO Justin Hotard, who earlier led Intel's (INTC.O), opens new tab data centres and AI group, has focused on growing the Finnish firm's data centre business, betting on AI growth.

• Amazon cuts 14,000 corporate jobs as spending on artificial intelligence accelerates

  Since 2024 started, Amazon has committed to about $10 billion apiece to data center projects in Mississippi, Indiana, Ohio and North Carolina as it builds up its infrastructure to tries to keep up with other tech giants making leaps in AI. Amazon is competing with OpenAI, Google, Microsoft, Meta and others. In a conference call with industry analysts in May, Jassy said the potential for growth in the company's AWS business is massive.

• Meta's stock slides on AI bubble fears, as Zuckerberg calls for increased spending

  In recent months, concerns have been growing among some critics that the extravagant spending by Meta and its big technology peers is fueling a bubble in the AI industry that may one day burst, with potentially devastating consequences for the stock market. The company has invested billions of dollars into AI, not only on the data center infrastructure it needs but also in acquiring talented developers, such as Scale AI Inc. founder Alexandr Wang, who now heads up the company’s Superintelligence Labs unit.

• Report: AI Data Center Boom Threatens U.S. Climate Goals

  A massive expansion of data centers, set to be powered primarily by fracked gas, could account for 10% of the economy-wide emissions and 44% of the power sector emissions allowable to meet the U.S. 2035 climate target, known as the nationally determined contribution, or NDC. Data centers are projected to account for more than 12% of U.S. electricity consumption by 2030.

• Amazon strategised about keeping its datacentres’ full water use secret, leaked document shows

  Amazon strategised about keeping the public in the dark over the true extent of its datacentres’ water use, a leaked internal document reveals.

  The biggest owner of datacentres in the world, Amazon dwarfs competitors Microsoft and Google and is planning a huge increase in capacity as part of a push into artificial intelligence. The Seattle firm operates hundreds of active facilities, with many more in development despite concerns over how much water is being used to cool their vast arrays of circuitry.

• Leaked Document Shows Amazon Scheming to Keep AI Data Center Water Use Secret

  If your data centers are guzzling more water than a major US city, you face a choice: come clean to the public and risk bad press — or hide the evidence to keep key figures in the green.

  For Amazon executives, evidently, the answer is to bury the data down far beneath the water table.

• Will A.I. Destroy the Planet?

  A single data center can use as much power as the city of Philadelphia. And they’re popping up everywhere. These sprawling buildings, filled with rows of computing equipment, are the factories of the A.I. economy; they power all those mundane chatbot searches, sucking up tons of energy in the process. As the OpenAI C.E.O. Sam Altman put it, “I do guess that a lot of the world gets covered in data centers over time.”

  For our latest issue, the reporter Stephen Witt was invited (“after what felt like two hundred phone calls”) inside a Microsoft facility, still under construction. I caught up with Witt to discuss what he saw there—and what A.I.’s massive energy consumption means for our planet.

  This conversation has been edited and condensed.

• Nvidia's market capitalization hits $5.12 trillion — AI powerhouse is the first company in history to hit seismic milestone

  The market capitalization of Nvidia passed the $5 trillion mark on Wednesday, following a host of potentially lucrative deals and projects the company announced at its GPU Technology Conference event on Tuesday. The company's stock price continues to be driven by the broad interest towards artificial intelligence and Nvidia's unique position in the AI supply chain that now spans from humble chips for inference to gigawatt-scale data centers for massive AI training.

• Datacenter biz and nuke firm join forces for Texas AI ranch

  The 1,600-acre (about 6.5 square kilometers) site in Victoria County was chosen for its location, existing infrastructure, and available energy resources to support a gigawatt-scale datacenter campus, the pair say, close to electric transmission lines, fiber networks, and natural gas pipelines.

• Data centers

  Where are my data? Partly in a data center; probably with your data too… So, where are they?

• Azure stumbles in Europe, Microsoft blames 'thermal event'

  The West Europe region is in the Netherlands.

  The Windows giant blamed the problems on “a thermal event affecting datacenter cooling systems, which led to a subset of storage scale units going offline in a single availability zone.”

• China bans foreign AI chips from state-funded data centers, report claims — crackdown would include removing Nvidia, AMD, and Intel chips from builds in early stages

  The report, citing unnamed sources with direct knowledge, claims regulators have instructed state-funded projects to use only Chinese-made silicon going forward. That includes domestic chips from Huawei, Cambricon, and Enflame, among others.

  While the official directive has not been published, the guidance would mark a formal shift from earlier policies that merely discouraged foreign chip purchases. Under the new dictat, even parts like Nvidia’s H20 — a model specifically designed to comply with U.S. export controls — will now be off the table. This follows Chinese port crackdowns on all Nvidia imports in early October.

• The AI Data Center Boom Is Warping the US Economy

  Some tech companies have spent so much on AI recently that they have been forced to look for new sources of funding. One noteworthy example is Meta, which recently announced a $27 billion deal to develop a cluster of data centers in Louisiana. The project was created through a special purpose vehicle (SPV), an increasingly common organizational structure that allows firms to avoid putting large amounts of debt on their balance sheets. Last week, Meta said it also raised another $30 billion in debt through a more conventional channel: selling corporate bonds.

• Study: Big growth in data centres could raise electricity prices by 10% by 2030

  The consulting firm's projections considered a sharp growth in the number of data centres, with a combined electricity demand of 2,500 megawatts (MW). That is roughly one and a half times more than the Olkiluoto 3 nuclear reactor, which is Europe's largest.

• Google to test building AI data centres in space

  Google plans to launch AI data centers into space by 2027. These solar-powered satellites will use proprietary AI chips. The project aims to leverage constant sunlight and vacuum cooling for efficiency. This initiative pushes Google's deeptech research boundaries. Other companies like SpaceX are also exploring similar space-based computing.

• Google wants to build AI data centers in space using satellites

  In the blog post, Google mentions that there are several major challenges with the project, including that the satellites in the network will need to move in a much tighter formation than existing satellite networks, and that the Tensor chips will also need to cope with cosmic radiation while in orbit around the Earth.

• Rideshare giant dumps 200 cloudy Macs, saves $2.4 million

  After crunching the numbers, the company decided to build its own racks full of Mac Minis and house them in a Malaysian datacenter.

• The futile future of the gigawatt datacenter — by Nicholas Weaver

  The scale of the investment is so large it is distorting the economy. These massive expenditures, however, are shortly going to prove to be at least a half-trillion-dollar money pit: these massive data centers are simply not fit for purpose.

  But to understand why this is a waste of money, it is critical to discuss two technologies: machine learning (ML) in general and large language models (LLMs) in particular.

• Power crunch threatens to derail AI datacenter construction

  Nearly half of respondents (48 percent) cite power access as the biggest scheduling constraint, with grid connection wait times stretching years.

• Meta plans $600 billion US spend as AI data centers expand

  Last month, the social media giant sealed a $27 billion financing deal with Blue Owl Capital to fund its Louisiana data center, its biggest project globally.

  Meta said in October it would invest $1.5 billion in a data center in Texas, breaking ground on its 29th such facility globally.

• India's data centre capacity set to grow 5x to 8GW by 2030: Jefferies

  The report highlights that India's colocation data centre capacity, where enterprises lease infrastructure from operators, has already increased fivefold to 1.7GW, with occupancy levels at 97%, underscoring strong demand.

• Voters’ Anger at High Electricity Bills and Data Centers Looms Over 2026 Midterms

  Voter anger over the cost of living is hurtling forward into next year's midterm elections, when pivotal contests will be decided by communities that are home to fast-rising electric bills or fights over who's footing the bill to power Big Tech's energy-hungry data centers.

• Sam Altman backs away from OpenAI's statements about possible U.S. gov't AI industry bailouts — company continues to lobby for financial support from the industry

  The original confusion stemmed from remarks made by CFO Sarah Friar, who said at a Bloomberg event that a federal “backstop” could help unlock investment for future data center expansion. Those comments, widely interpreted as a request for taxpayer guarantees, drew scrutiny given the scale of AI server costs and energy requirements. While Altman has said OpenAI is investing heavily in future compute, it is not seeking a bailout for its facilities. “We plan to be a wildly successful company, but if we get it wrong, that’s on us,” he said.

• The State of AI: Energy is king, and the US is falling behind

  In the age of AI, the biggest barrier to progress isn’t money but energy. That should be particularly worrying here in the US, where massive data centers are waiting to come online, and it doesn’t look as if the country will build the steady power supply or infrastructure needed to serve them all.

• If the US Has to Build Data Centers, Here’s Where They Should Go

  A new analysis tries to calculate the coming environmental footprint of AI in the US and finds that the ideal sites for data centers aren’t where they’re being built.

• Proving Safety at Scale: SPARK, RISC-V, and NVIDIA’s Security Strategy

  Modern systems, from data centers to embedded systems, are built from an ever-growing assortment of embedded controllers, accelerators, and firmware components. At scale, traditional “test more” approaches struggle to keep pace: tests can show the presence of bugs but cannot guarantee the absence of whole classes of failures that attackers routinely exploit. That is why the conversation is shifting from securing software after it is written to engineering security into the software from the start.

  In a presentation by Adam Zabrocki and Marko Mitic of NVIDIA at DEF CON, they discuss the challenge of securing a proprietary, billion-core architecture.

• US taxpayers kept in the dark over datacenter subsidies

  Hundreds of billions is being lavished on datacenter construction across the country, however, a report published today claims many of these projects are being funded at the expense of taxpayers, with few states confirming the names of the beneficiaries.

• iTWire - Cloud Outages Are Inevitable, But Lock-In Shouldn’t Be

  True resilience looks completely different. It means your critical data workloads are natively portable. Imagine being able to move your core applications not just between regions, but also between entirely different cloud providers, and back to your on-prem data centre, without the cost and disruption of a massive, months-long refactoring project.

  That’s the only way to ensure that when one environment goes down, another can take over quickly, securely, and without major disruption.

• The AI Boom Is Fueling a Need for Speed in Chip Networking

  Chip giants like Nvidia, Broadcom, and Marvell already have well-established networking bona fides. But in the AI boom, some companies are seeking new networking approaches that help them speed up the massive amounts of digital information flowing through data centers. This is where deep-tech startups like Lightmatter, Celestial AI, and PsiQuantum, which use optical technology to accelerate high-speed computing, come in.

• Ypsilanti, not Ann Arbor, will bear the cost of the U-M LANL data center

  The University should not delegate the responsibility of elevated prices and environmental degradation on people that will never benefit from research resources the data center project will provide. And given that the main sustainability goal in Campus Plan 2050 is a net-zero endowment, its construction is completely hypocritical. Siphoning money into massive construction data center projects without feasible sustainable infrastructure plans while claiming to pursue carbon neutral investments is contradictory in nature and flat out wrong in practice.

• The Data Center Resistance Has Arrived

  “The number one issue was affordability,” he says. “But a very close second was data centers and the concern around them just sucking up the water, the electricity, the land—and not really paying any taxes.”

• Most states don’t disclose which companies get data center incentives, report finds

  Despite data centers’ significant energy requirements, states frequently compete heavily to land the projects, which invest millions or even billions into new construction. But the study noted those projects often employ nondisclosure agreements, project code names and subsidiary names that hide the firms behind the new server farms.

  “Only when governments disclose information on which companies get public money and what they do with it can there be meaningful analysis, greater public participation, and wiser use of public financial resources,” the report says.

• Cloudy Data, Costly Deals: How Poorly States Disclose Data Center Subsidies [PDF]

  At least 36 states have economic development subsidies specifically crafted for data center projects, but only 11 disclose which companies receive them: Arizona, Connecticut, Illinois, Indiana, Minnesota, Nevada, Ohio, Pennsylvania, Texas, Washington, and Wisconsin.

  Even where there is some disclosure, it is often fragmentary and confusing. Consider: [...]

• Data Center Watch worries that anti-AI activism is working

  Even the data centre trade press doesn’t buy 10a’s scaremongering. Data Center Dynamics notes that when 10a says a data centre was “blocked,” that also includes when a company submits a very similar proposal nearby. But that’s normal — data center plans typically have a pile of similar proposals ready to deploy.

• GPU depreciation could be the next big crisis coming for AI hyperscalers — after spending billions on buildouts, next-gen upgrades may amplify cashflow quirks

  Most corporations operate with an understanding that their servers will remain relevant for between three and five years, but in the world of "AI factories," where the speed and efficiency of your data center may equate to how much you can earn, even falling one generation behind could be terminal. What happens when the pace of innovation accelerates beyond the potential profitability of the hardware you spent so much time and money investing in?

• Data centers are the desert’s thirsty new neighbor

  But in a few short weeks, the deal has generated intense controversy. Like the residents of dozens of other U.S. communities facing the arrival of a data center, many in Doña Ana County are wary. A large data center could use millions of gallons of drinking water a day to keep its equipment cool, and the industry already accounts for more than 4% of total U.S. electricity consumption in a given year.

• AI datacenters can't run indefinitely on fossil fuel

  Gartner warns that fossil fuel dominance in on-site power generation is not sustainable, given the rapid rise in datacenter energy consumption due to AI servers.

  The research firm forecasts that electricity required by datacenters worldwide is set to expand 16 percent this year alone, and likely to double to 980 terawatt hours (TWh) by 2030.

• Cloudflare Outage Impacts Millions of Internet Users

  Cloudflare acts as an intermediary between users and servers, helping millions of websites stay online even during high traffic or cyberattacks. It’s used by about 20 percent of all websites on the [Internet], according to Inside Towers Intelligence. Based in San Francisco, Cloudflare connects to over 330 data centers in more than 125 countries. When it experiences an outage, it can disrupt access to these sites globally.

• Saudi Arabia backs Elon Musk's xAI with data center deal

  The goal is to use the country's ample energy reserves, deep pockets, abundance of land and links to global fiber optic networks to provide cheap AI computing power, including to companies based elsewhere. Humain has set a target of powering 6% of the global AI workload in the coming years.

• Dramatic Surge in Water Demand Predicted by 2040 Puts Ohio Farmers and Industry on Collision Course

  The competing demands of agriculture and industry – particularly the 130 data centers in central Ohio already consuming millions of gallons of water a day to cool computer equipment – would require billions of gallons of water daily, according to a 15-county Central Ohio Regional Water Study released this year by the Ohio Environmental Protection Agency.

• Environmental groups join call to open DTE data center contracts to public scrutiny

  The Sierra Club, Michigan Environmental Council, Natural Resources Defense Council and Citizens Utility Board of Michigan are calling on the Michigan Public Service Commission to initiate a contested case proceeding in reviewing DTE Energy’s contracts, arguing that this type of proceeding is needed for the groups to properly evaluate the company’s proposal and advocate for Michiganders who will be impacted by the project.

• RAM is so expensive that stores are selling it at market prices

  Generative “AI” data centers are gobbling up trillions of dollars in capital, not to mention heating up the planet like a microwave. As a result there’s a capacity crunch on memory production, shooting the prices for RAM sky high, over 100 percent in the last few months alone. Multiple stores are tired of adjusting the prices day to day, and won’t even display them. You find out how much it costs at checkout.

• World's Most Stable Raspberry Pi? 81% Better NTP with Thermal Management - Austin's Nerdy Things

  This post details how I achieved an 81% reduction in frequency variability and 77% reduction in frequency standard deviation through a combination of CPU core pinning and thermal stabilization. Welcome to Austin’s Nerdy Things, where we solve problems that 99.999% of people (and 99% of datacenters) don’t have.

• Amazon commits up to $50 billion to boost AI, supercomputing infrastructure for agencies

  The company announced the up to $50 billion investment Monday, stating it is estimated to add nearly 1.3 gigawatts of AI and supercomputing capacity through new data centers. The project is expected to break ground in 2026, Amazon said, describing it as the “first-ever AI and high-performance computing (HPC) purpose-built infrastructure for the U.S. government.”

• Meta’s Accounting of Its Louisiana Data Centre ‘Strains Credibility’

  The phrase “marvel of financial engineering” does not seem like a compliment. In addition to the evidence from Weil’s article, Meta is taking advantage of a tax exemption created by Louisiana’s state legislature. But, in its argument, it is merely a user of this data centre.

• Microsoft partners with Powertrust to develop 270MW of solar projects in Mexico, Brazil

  The partnership represents a significant investment in Mexico’s renewable energy infrastructure at a time when the country is positioning itself as a data center destination for major tech companies.

• Asahi admits ransomware may have spilled data on 2M people

  In its latest update, Asahi said attackers entered via compromised network equipment at a Group datacenter facility in Japan and deployed ransomware on the same day, encrypting data on multiple live servers and some connected PCs. This forced a broad operational suspension – order processing systems were shut down, shipments paused, and customer service lines silenced. The company isolated the datacenter within hours, but ransomware gangs don't need much time when the door is already open.

• Uh-oh, accounting concerns circle the AI boom

  In his Breaking Analysis this week, Dave Vellante makes the case that GPUs have plenty of legs, quelling most concerns that lengthening depreciation times are a red flag. That said, Meta is particularly, shall we say, creative, when it comes to accounting for, or not accounting for, its multibillion-dollar spending on data centers.

  Sharpening the concern over AI’s payoff is that the current generative approaches may soon reach their limit, as another AI expert opined this week.

• Amazon Data Center Linked to Cluster of Rare Cancers

  This meant that the data center itself began taking on the toxic sludge as it drew on groundwater to cool its electronics. When it did, evaporation only further concentrated the wastewater, which occasionally contained nitrate levels eight times higher than Oregon’s safe limit. The super concentrated data center water then made its way back into the waste system, where it ostensibly piled up all over again.

• Datacenters in space are a terrible, horrible, no good idea.

  The short version: this is an absolutely terrible idea, and really makes zero sense whatsoever. There are multiple reasons for this, but they all amount to saying that the kind of electronics needed to make a datacenter work, particularly a datacenter deploying AI capacity in the form of GPUs and TPUs, is exactly the opposite of what works in space. If you've not worked specifically in this area before, I'll caution against making gut assumptions, because the reality of making space hardware actually function in space is not necessarily intuitively obvious.

• Unpacking India’s data centre boom

  Demand for digital infrastructure fuelled by AI and regional data protection laws have unleashed a new wave of investments in India’s data center industry. Global hyperscalers and Indian conglomerates are set to invest more than $50 billion in the next five to seven years in the domestic data centre industry, taking total capacity to about 9 gigawatts (GW) from 1 GW now.

• OpenAI partners shoulder nearly $100 billion debt to fund its AI ambitions

  OpenAI’s rapid expansion is being financed largely through heavy borrowing by its partners rather than by the company itself. According to the Financial Times, suppliers and data centre operators linked to OpenAI now carry close to $100 billion in debt, even as OpenAI remains mostly debt-free.

• OpenAI data center planned for farmland in Saline

  Less than a half-hour away from the University of Michigan in Saline Township, OpenAI — parent company of ChatGPT — and tech giant Oracle have plans to develop a $7 billion data center. In an announcement Oct. 30, the companies said the development is set to supply 4.5 gigawatts to Stargate to help advance artificial intelligence infrastructure.

• Sam Altman Is Suddenly Terrified

  To once again leave its competition behind, OpenAI has committed to spending well over $1 trillion on data center buildouts to support new users and its latest AI models. That’s despite burning through billions of dollars each quarter.

• Smartphone prices set to jump as memory crunch hits consumer tech

  The squeeze spans almost every type of memory, from flash chips used in USB drives and smartphones to advanced high-bandwidth memory (HBM) that feeds AI chips in data centres. Prices in some segments have more than doubled since February, according to market research firm TrendForce, drawing in traders betting that the rally has further to run.

• The Consequences of the National AI and Data Centre Plan

  I think that misses the mark.

  For one, the plan is very clear that this is only the beginning – the start of a phased, adaptive approach to governing AI and the infrastructure that underpins it. It explicitly positions itself as a framework that will evolve, including future updates to “national data centre principles”. In many ways, it should be seen as a statement of intent: directionally strong, deliberately flexible, and designed to be refined in consultation with industry and state governments.

  Secondly, no government–in Australia, Europe or across Asia Pacific–can be expected to have a perfectly formed AI rulebook at this stage. We’re at the front end of one of the biggest technology shifts since the industrial revolution. Every national plan today is, by necessity, a first attempt. Innovation invariably moves faster than regulation, and AI is no different.

• AWS CEO Matt Garman Wants to Reassert Amazon’s Cloud Dominance in the AI Era

  You might think Amazon’s biggest swing in the AI race was its $8 billion investment in Anthropic. But AWS has also been building in-house foundation models, new chips, massive data centers, and agents meant to keep enterprise customers locked inside its ecosystem. The company believes these offerings will give it an edge as businesses of all shapes and sizes deploy AI in the real world.

• The Anacé Indigenous people are protesting TikTok’s construction of the largest data center in Brazil

  This Indigenous community has launched a peaceful battle with the support of lawyers, NGOs and the Public Prosecutor’s Office against the Chinese company. They fear that the mega-data center it plans to build on land which they consider to be their own will negatively impact them. They’re also concerned because no “free, prior and informed consultation” took place, as mandated by the Indigenous and Tribal Peoples Convention of 1989 (an International Labour Organization convention, often referred to as Convention 169). Around the world, this convention is as frequently ignored by investors as it is invoked by Indigenous peoples.

  ByteDance, the company that owns the social media network that has captivated hundreds of millions of internet users, has partnered with a Brazilian wind energy company, Casa dos Ventos, to build a 300-megawatt data center that will be the most powerful one in Brazil.

• As data centers flock to Michigan communities, what protections exist for residents? • Michigan Advance

  While proponents of these facilities – which house servers, storage devices and other pieces of online infrastructure – point to the role data centers play in the development of technologies like artificial intelligence and the potential economic boons that could come with them, environmental advocates and nearby residents are pointing to the burden the projects could put on their communities.

  Chief among the concerns with the projects making their way to the state is the large amount of energy needed to power these facilities, with an analysis from Bloomberg News finding a 267% increase in energy prices in communities located near data centers.

• AI Data Centers Are Making RAM Crushingly Expensive, Which Is Going to Skyrocket the Cost of Laptops, Tablets, and Gaming PCs

  For the past few months, RAM prices have been shooting through the roof, as the rapid buildout of AI datacenters causes an ever tightening shortage of memory chips. What was once considered to be one of the most affordable components of building a gaming rig has now doubled or even tripled in price. Prices are so volatile, in fact, that some retailers are now selling RAM kits at market prices that go up and down — or usually up and up — by the day, instead of having a fixed price tag. Or as The Verge put it: they’re being sold “like lobster.”

  And now, a new omen is casting a somehow even more ominous shadow over the future to come.

• Racing is not winning for Indiana data centers

  But the biggest pressure of all comes from the explosive growth of artificial intelligence and the data centers needed to run it. By 2035, AI-driven data centers are expected to add the electricity demand equivalent or greater than building a second Indianapolis, a second Fort Wayne, a second Gary, and a second Evansville combined.

  Take the planned data center campus in New Carlisle for example. It alone is projected to use 2.2 gigawatts of power — almost 10% of Indiana’s entire current electricity demand.

• TikTok’s first data center in Latin America will be in Brazil and will run entirely on wind power

  TikTok’s data center in Brazil will be powered by wind energy from wind farms that are also under construction, according to the company. The company explains that it will use a closed-loop water cooling system for the supercomputers and will not draw power from the electrical grid, “avoiding impacts on the local supply” and “reinforcing its commitment to the energy transformation and decarbonization of the global economy.”

• Why copper markets are feeling the pinch due to aggressive AI data center expansion — expansive buildout to demand 1.1 million tonnes of copper annually by 2030, close to 3% of global demand

  Alongside energy, water and computer chips, the supply squeeze is being felt further up the chain, too.

• Hackaday Links: December 7, 2025

  We stumbled upon a story this week that really raised our eyebrows and made us wonder if we were missing something. The gist of the story is that U.S. Secretary of Energy Chris Wright, who has degrees in both electrical and mechanical engineering, has floated the idea of using the nation’s fleet of emergency backup generators to reduce the need to build the dozens of new power plants needed to fuel the AI data center building binge. The full story looks to be a Bloomberg exclusive and thus behind a paywall — hey, you don’t get to be a centibillionaire by giving stuff away, you know — so we might be missing some vital details, but this sounds pretty stupid to us.

• National Data Center Moratorium Now!

  We, the undersigned organizations collectively representing millions of people in all 50 states, call on you to support a national moratorium on the approval and construction of new data centers.

  The rapid expansion of data centers across the United States, driven by the generative artificial intelligence (AI) and crypto boom, presents one of the biggest environmental and social threats of our generation. This expansion is rapidly increasing demand for energy, driving more fossil fuel pollution, straining water resources and raising electricity prices across the country. All this compounds the significant and concerning impacts AI is having on society, including lost jobs, social instability and economic concentration.

• Activist groups urge Congress to pause datacenter buildouts

  Organizations endorsing the letter collectively represent millions of citizens across all 50 US states, the group claims. It urges members of Congress to suspend the construction of new facilities until adequate regulations are enacted to protect communities and the environment from the damage it claims is already being inflicted.

• Datacenters accused of hoarding grid capacity

  The problem stems from how grids operate: Grid operators can't expand networks without lengthy regulatory approval, and capacity is allocated strictly on a first come first served basis. This prompts developers to reserve power for projects that may never be built, while datacenter operators apply for more than they need to accommodate future growth and planning uncertainty.

• How Red Hat OpenShift AI simplifies trust and compliance

  These standards set the rules for encryption, access control, auditing, and data handling. They also introduce operational constraints that limit where and how AI runs.Red Hat OpenShift AI helps to bridge that divide, allowing organizations to build and deploy protected AI where the data lives, across datacenters, public clouds, and edge environments.

• Wind and Solar Power Frozen Out of Trump Permitting Push

  "It is difficult to understand how these actions align with the President’s stated goals of unleashing American energy and ensuring energy affordability," three Democratic senators, Michael Bennet, John Hickenlooper and Ben Ray Lujan, said in a letter to Burgum last month.

  U.S. electricity demand is expected to increase 32% by 2030, with data centers accounting for more than half that growth, according to an analysis by power sector consulting firm Grid Strategies.

• The Data Center Boom Could Trigger Blackouts

  Last week, the monitor overseeing PJM filed a complaint with the nation’s top electric utility regulator, warning of unreliable service for its sixty-five million customers across the Midwest and mid-Atlantic. Monitoring Analytics asked the Federal Energy Regulatory Commission to pause connecting larger artificial intelligence–powering data centers to its grid until PJM can ensure “reliable, economic, and environmentally acceptable” service and guarantee that the growing data center energy burden won’t produce unnecessary blackouts.

• Dirty data centers will drive up our utility bills 

  According to a MIT study, a single data center can use as much energy as an entire city, and frighteningly, big tech could transform the Reno area into the largest data center market in the world if we don’t take action. Greedy businesses are making opportunistic moves on data center expansion, and Nevada is where they come for cheap land, cheap energy, and tax breaks. As a result, multi-billion-dollar tech companies whose massive data centers power AI growth are consistently proposing new data centers. These giant warehouse-like structures house hundreds of computer servers and other IT infrastructure to store, process, and manage data. That number of computers and servers requires enormous amounts of energy and also generates a lot of heat. To keep the computers cool, data centers use large amounts of water and place a massive burden on the energy grid, which drives up energy costs for everyone else to line the pockets of big tech’s AI companies.

• Fear of Missing Out on data center boom leads Arkansas officials to ignore hard questions | Arkansas Advocate

  Is it appropriate or wise for state and local governments to provide corporate welfare subsidies to some of the most profitable companies on the planet — and to negotiate those deals in secret to avoid appropriate public scrutiny? Is opening up the candy store really necessary to compete for these projects, or are state and local leaders — driven by their fear of missing out — getting played?

  As we try to move away from our reliance on fossil fuels to combat climate change, should we really be encouraging environmentally unfriendly projects that undermine our renewable energy goals by driving up electricity demand beyond what renewable sources can possibly supply?

• How UMich is forcing a data center on Ypsilanti

  Over the past year, residents have frequented Ypsilanti Township Council meetings, distraught over one thing: the University of Michigan’s new data center. Officials have fielded questions about environmental, noise, electricity and infrastructure concerns that they cannot answer. Not because the council doesn’t care, but because the University has refused to engage transparently with them.

• Nvidia weighs expanding H200 production as new China orders rush in, report claims — ByteDance and Alibaba listed as suitors for 'large orders' in the wake of sanctions lift

  Nvidia is evaluating whether to increase production of its H200 data center GPUs after demand from Chinese customers is expected to exceed current supply, according to people familiar with the matter, as reported by Reuters. The discussions follow a policy shift this week in which the U.S. government said Nvidia would be allowed to export the H200 to China under a framework that includes a 25% fee on sales.

• AI datacenter boom could end badly, Goldman Sachs warns

  But doubts have been raised about just how much return on investment (ROI) will come from the current AI frenzy, with IT firm Lenovo finding earlier this year that many business leaders were unconvinced that AI is worth the expense, while another report claimed that forecasts of future demand are based on little more than guesswork.

• Google Data Centers Are Returning Nuclear Power to Tornado Country

  A destructive storm in 2020 prematurely shut down Iowa’s only nuclear plant. With Google’s plans to reopen it to power nearby data centers, will extreme weather threaten the reactor’s safety?

• Fox News Instructs Viewers to Buy Plastic Christmas Trees So the Real Ones Can Be Sacrificed to Feed the Insatiable Hunger of AI Data Centers

  Maryland is one of many battlegrounds across the country that have sprung up over the rapid construction of AI data centers. The sprawling complexes can bring soaring energy bills, and headache-inducing noise pollution. Some have also been accused of hoovering up their area’s water, leaving behind a desiccated water table. The actual construction of these colossal projects puts a massive strain on communities, too: a once-quiet Louisiana town where a Meta data center is now being built has seen a 600 percent surge in automobile crashes as massive construction trucks keep barreling down its tiny roads and getting into horrific wrecks, including near a school playground.

• How Tech’s Biggest Companies Are Offloading the Risks of the A.I. Boom

  Those deals had one thing in common: They allowed companies that make massive quarterly profits to reduce their financial exposure to the frenetic, global buildup of data centers.

• What Does a Database for SSDs Look Like?

  But first, the bigger picture. It’s not only SSDs that have come along since databases like Postgres were first designed. We also have the cloud, with deployments to excellent datacenter infrastructure, including multiple independent datacenters with great network connectivity between them, available to all. Datacenter networks offer 1000x (or more) increased throughput, along with latency in the microseconds. Servers with hundreds of cores and thousands of gigabytes of RAM are mainstream.

  Applications have changed too. Companies are global, businesses are 24/7. Down time is expensive, and that expense can be measured. The security and compliance environment is much more demanding. Builders want to deploy in seconds, not days.

• Global Internet Traffic Rose 19 Percent in 2025, Cloudflare Data Shows

  The company operates one of the world’s largest content delivery, DNS, and security infrastructures, with data centers in hundreds of cities and visibility into traffic for millions of websites, applications, and APIs. Here is a summary of the most important points.

• Data Centers In Spaaaace!

  The AI bubble has been kept inflated by journalists uncritically reporting whatever CEOs say as they frantically pump the stock. Right now, you can observe a wonderful example of this by searching the Web for "orbital data centers". My recent search turned up pages of links, including SpaceX’s Lofty IPO Valuation Hinges on Big Bet on Outsize Growth from Bloomberg's Bailey Lipschultz, Sana Pashankar, and Loren Grush: [...]

• When People Realize How Good The Latest Chinese Open Source Models Are (And Free), The GenAI Bubble Could Finally Pop

  Although the field of artificial intelligence (AI) goes back more than half century, its latest incarnation — generative AI — is still very new: ChatGPT was launched just three years ago. During that time a wide variety of issues have been raised, ranging from concerns about the impact of AI on copyright, people’s ability to learn or even think, job losses, the flood of AI slop on the Internet, the environmental harms of massive data centers, and whether the creation of a super-intelligent AI will lead to the demise of humanity. Recently, a more mundane worry is that the current superheated generative AI market is a bubble about to pop. In the last few days, Google’s CEO, Sundar Pichai, has admitted that there is some “irrationality” in the current AI boom, while the Bank of England has warned about the risk of a “sharp correction” in the value of major players in the sector.

• Senate trio questions DC operators over rising energy costs

  Senators Elizabeth Warren (D-MA), Chris Van Hollen (D-MD), and Richard Blumenthal (D-CT) sent letters to Google, Microsoft, Amazon, Meta, CoreWeave, Digital Realty, and Equinix on Monday, demanding answers on their contributions to rising utility rates, which the lawmakers blame firmly on the explosion of datacenter construction projects.

  Because the massive scale of AI datacenters requires more power than most grids can supply, the Senators explain in their letters, utility companies have to build out their grids to the tune of billions of dollars.

• New calculator helps evaluate the economics of datacenters in space — running the numbers on orbital computing reveals a brutal reality

  Even though McCalip's analysis already puts the kibosh on the financial side, Andrew Côté's exceedingly deep dive into the engineering difficulties faced by such an enterprise delivers the killing blow. That's also an excellent read , as the list of things that can go wrong with setting up a datacenter in space can be reasonably summed up with "yes." Did you know, for example, that AI accelerators tend to up and die a lot when used at full tilt, and can't exactly be replaced in 5 minutes in space?

• The data center cooling state of play (2025) — Liquid cooling is on the rise, thermal density demands skyrocket in AI data centers, and TSMC leads with direct-to-silicon solutions

  Just 10 years ago, in 2015, data center liquid cooling was primarily limited to specialized applications such as supercomputers and mining farms; its global adoption rate at the time was at best around 5%. Air cooling dominated due to lower upfront costs and simpler infrastructure. But by 2020, the adoption of liquid cooling began to rise and reached around 10% as cloud hyperscalers started to strive for efficiency . However, with power densities of approximately 5 –10kW per rack on average, air cooling was sufficient for most.

  Fast forward a few years, and power density per rack increased to 15kW, 20kW, 30kW, or even 40kW (in the case of Nvidia H100-based racks). These soaring power densities led to the steady adoption of liquid, as the technology moved from specialized setups to enterprise and cloud data centers.

• Bernie Sanders proposes AI datacenter moratorium

  US Senator Bernie Sanders (I-VT) is calling for a nationwide moratorium on datacenter construction, saying it would "give democracy a chance to catch up … and make sure that the benefits of technology work for all of us, not just the wealthiest people on Earth."

• Bills would target water use, utility costs tied to Michigan data center boom

  Senate Bills 761, 762 and 763 each propose amending existing state statutes.

  SB 761 bans those withdrawing more than an average of 2 million gallons of water a day for consumptive use ― water that evaporates or does not return to the water system ― from obtaining a permit in the state.

  Meanwhile, SB 762 requires the Michigan Public Service Commission to publish annual reports regarding the total energy spending and water usage of data centers, and SB 763 prohibits passing infrastructure improvement costs on to residential ratepayers.

• Oracle stock falls over US data center funding troubles

  The Financial Times reported that Oracle’s primary backer for its largest data center projects won’t finance a $10 billion, 1-gigawatt facility in Michigan over concerns about the software giant’s rising debt and AI spending. Cloud-computing firms including Oracle, Microsoft, and Meta have amassed a combined $500 billion in obligations toward data center leases, Bloomberg reported.

• RAM Shortage 2025: How AI Demand is Raising DRAM Prices

  Executive Summary: In late 2025 the global memory industry is grappling with an unprecedented RAM (DRAM) shortage. Exponentially rising demand – especially for high-bandwidth memory (HBM) and next-generation server memory – is colliding with manufacturing constraints, leading to record-high prices and supply rationing. Major industry players like Samsung, SK hynix, and Micron have shifted investment and production capacity toward AI-oriented memory (HBM, advanced LPDDR), deliberately holding back commodity DRAM expansions ([1]) ([2]). AI data centers (from giants like Nvidia, Google, and Microsoft) are voraciously consuming HBM and LPDDR, effectively outbidding consumer PC, gaming, and smartphone markets for available DRAM. ([3]) ([4]). The result is that even traditional memory segments face shortages: contract DRAM prices are surging (e.g. 16Gb DDR5 chip prices rose from ~$6.84 to ~$27.20 in Q4 2025 ([4])), and bargain DDR4 is scarce as production lines convert to DDR5. Leading analysts now warn that high memory prices and tight supply could persist into 2027–2028 when new fabrication plants come online ([2]) ([5]). This report provides a comprehensive analysis of the December 2025 RAM shortage: its technical and economic causes, industry responses, effects on various sectors, and foreseeable future developments, drawing on extensive data and expert commentary.

• [Old] NVIDIA Transitions Fully Towards Open-Source GPU Kernel Modules

  With the R515 driver, NVIDIA released a set of Linux GPU kernel modules in May 2022 as open source with dual GPL and MIT licensing. The initial release targeted datacenter compute GPUs, with GeForce and Workstation GPUs in an alpha state.

  At the time, we announced that more robust and fully-featured GeForce and Workstation Linux support would follow in subsequent releases and the NVIDIA Open Kernel Modules would eventually supplant the closed-source driver.

• LLMs are a Cognitohazard

  Let me start, though. I think each and every one of these things ignores the externalities of running an LLM: The wholesale scraping of data frequently without consent (and the associated pushing of costs onto the site being scraped), the massive build-out of data centers that will never be used, a global chip shortage because the AI industry can just do that, the proliferation of slop and disinformation, and the trend of people experiencing psychosis when interacting with chatbots.

  Okay, with that out of the way, let's dig in.

• AI has pumped hyperscale – but how long can it last?

  As a result, quarterly capex reported by the hyperscale operators has inflated, growing by almost 180 percent and reaching $142 billion in the third quarter of this year, according to the firm's research. The volume of datacenter capacity added each quarter has also risen by 170 percent.

• Google Buys Data Center Company for $4.75 Billion

  The all-cash [sic] deal is set to help Google bolster the supply of power to its data centers, the giant computing facilities that power A.I. Intersect, a privately held company that operates data centers and energy plants and is based in San Francisco, had already been working with Google to build the computing sites, and Google had invested in the company last December.

• Advancing digital connectivity and innovation in China

  China’s 14th Five-Year Plan (2021 – 2025) for the Information and Communications Industry emphasizes 5G rollout, next-generation network expansion, and the development of national data centres to support broad economic transformation. Looking ahead, recommendations for the 15th Five-Year Plan (2026 – 2030) include building a modern industrial system, strengthening technological self-reliance, expanding digital trade and services, and advancing the Digital China initiative to support economic modernization and international collaboration.

• PC prices could rise by 8% in 2026 due to memory shortages

  Research firm IDC predicts that the average price of computers could rise by up to 8% in 2026 due to a global shortage of memory chips (RAM and NAND).

  The background is the sharp increase in demand for HBM memory for AI data centers, which is prioritized over the production of consumer memory, which is less profitable for manufacturers.

• AI data centers are forcing obsolete 'peaker' power plants back into service

  The Fisk power plant is among a growing number of ⁠so-called "peaker" electric generating units being pressed into duty across the U.S. as the nation's electrical grids struggle to keep up with growing demand from data centers powering Big Tech's investments in artificial intelligence.

• Surviving The RAM Apocalypse With Software Optimizations

  To the surprise of almost nobody, the unprecedented build-out of datacenters and the equipping of them with servers for so-called ‘AI’ has led to a massive shortage of certain components. With random access memory (RAM) being so far the most heavily affected and with storage in the form of HDDs and SSDs not far behind, this has led many to ask the question of how we will survive the coming months, years, decades, or however-long the current AI bubble will last.

• AI data centers may soon be powered by retired Navy nuclear reactors from aircraft carriers and submarines — firm asks U.S. DOE for a loan guarantee to start the project

  Texas-based HGP Intelligent Energy sent a proposal to the U.S. Department of Energy (DOE) to repurpose two retired U.S. Navy nuclear reactors for use in an AI data center project at Oak Ridge National Laboratory in Tennessee, under President Donald Trump’s Genesis Mission. Bloomberg says that the company aims to use two old reactors to deliver 450 to 520 megawatts of power, but did not confirm where they would come from.

• Billion-Dollar Data Centers Are Taking Over the World

  Now the tech industry is in the fever-dream days of generative AI, which requires new levels of computing resources. Big Data is tired; big data centers are here, and wired—for AI. Faster, more efficient chips are needed to power AI data centers, and chipmakers like Nvidia and AMD have been jumping up and down on the proverbial couch, proclaiming their love for AI. The industry has entered an unprecedented era of capital investments in AI infrastructure, tilting the US into positive GDP territory. These are massive, swirling deals that might as well be cocktail party handshakes, greased with gigawatts and exuberance, while the rest of us try to track real contracts and dollars.

• APT Down and the mystery of the burning data centers

  The dump sparked a government investigation, and big corporations like LG, Lotte and Korea Telecom were asked to explain themselves. The government also mandated an on-site audit in the data center where the hacks had taken place. On the day of the audit, some li-ion batteries in the data center mysteriously caught fire. The blaze destroyed close to 100 servers (which had no backup) and plunged public service in South Korea into disarray.

  Shortly after, the Lotte data center burned as well - the corporation had been victim of a breach recently, albeit by a different threat actor. In the beginning of October, one of the officers examining the government data center fire tragically died by his own hand.

• SoftBank stakes $4B on securing AI data center power and capacity — DigitalBridge purchase indicative of AI industry's increasing investments in energy supply

  SoftBank's deal comes days after Alphabet announced that it would acquire Intersect — another provider of data center and energy infrastructure solutions — for $4.75 billion in cash [sic]. Taken together, the two deals demonstrate how AI infrastructure planning is moving upstream into power origination and grid access, rather than continuing to focus solely on server procurement.

• SoftBank reportedly finalizes OpenAI investment with $22.5B cash infusion

  OpenAI will use some of the funds to finance its Starlink data center initiative. The project, which the company launched at the start of the year, aims to deploy 10 gigawatts of computing capacity in the U.S. by 2029. The effort is supported by Oracle Corp. and SB Energy, a SoftBank unit that develops renewable energy installations.

• Elon Musk reveals plan to expand xAI's 'Colossus' data center to 2 gigawatts

  The facility, known as “Colossus,” already houses one massive data center building and there is currently a second one under construction at a nearby site that has been dubbed as “Colossus 2.” The third site is adjacent to Colossus 2, located near the town of Southaven, Mississippi, according to property records seen by The Information.

• Researchers Are Hunting America for Hidden Datacenters

  A team of researchers at Epoch AI, a non-profit research institute, are using open-source intelligence to map the growth of America’s datacenters. The team pores over satellite imagery, building permits, and other local legal documents to build a map of the massive computer filled buildings springing up across the United States. They take that data and turn it into an interactive map that lists their costs, power output, and owners.

• As Google, Musk, Bezos, Altman back space data centres, can sky really be the next frontier for AI

  The notion has gained traction as the AI race hits a fever pitch, fueling fears of a potential bubble. Meta, OpenAI, Microsoft, Amazon and other big tech companies are investing hundreds of billions in data centres worldwide, with OpenAI alone committing $1.4 trillion to such projects. Saudi Arabia and other nations are also pouring money into these efforts, while smaller companies pile up debt and take on financial risks to join the frenzy.

• The Great Unracking: Saying goodbye to the servers at our physical datacenter

  For almost our entire 16-year existence, the SRE team has managed all datacenter operations, including the physical servers, cabling, racking, replacing failed disks and everything else in between. This work required someone to physically show up at the datacenter and poke the machines.

  We’ve since moved all our sites to the cloud. Our servers are now cattle, not pets. Nobody is going to have to drive to our New Jersey data center and replace or reboot hardware. Not after last week.

• U.S. electricity grid stretches thin as data centers rush to turn on onsite generators — Meta, xAI, and other tech giants race to solve AI's insatiable power appetite

  Elon Musk was the first to use gas turbine generators to power a data center at this scale and duration at xAI’s Memphis Supercluster. In 2024, the AI startup signed a 50-MW deal with the Tennessee Valley Authority (TVA), which only became operational months after.

  Aside from the delay, this is nowhere near enough of the 155MW that the 100,000 H100 GPUs running on the site require. There was also a 150-MW substation under construction on the site, which also required additional setup. Waiting for these power sources to come online would have negated Musk’s historic 19-day setup of an AI data center, so he turned to VoltaGrid to deliver the power he needed to run the Colossus site.

• Data center ‘gold rush’ pits local officials’ hunt for new revenue against residents’ concerns

  Twiggs County commissioners recently approved a massive data center project, sparking a legal battle and deepening distrust between citizens and their elected officials in the rural middle Georgia community.

  The decision was made despite vocal opposition aired at a contentious public hearing and accusations that the process was improperly rushed to bypass a state review that was at the time paused.

• Most venture capital deals are now AI

  The US stock market rests on a few huge companies that are all-in on AI, and the economy is only technically not in recession because of big dollar numbers attached to huge data centre deals that will never happen. The whole US economy is a bet on AI working out and the bubble never popping.

  Venture capital turns out to work the same way now. They’re just putting their whole bag on red and spinning the wheel: [...]

• Mississippi Lands $20B xAI Data Centre: Jobs Surge but Pollution Fears Grow

  Governor Tate Reeves revealed on Thursday that Southaven will become home to a $20 billion xAI data hub, a move he described as the most significant private capital injection Mississippi has ever witnessed. This massive project by Elon Musk's artificial intelligence firm represents a historic milestone for the state's industrial landscape.

• Texans Are Fighting a 6,000 Acre Nuclear-Powered Datacenter

  But it’s unclear if Palantir wants any part of Project Matador. One unnamed client—rumored to be Amazon—dropped out of the project in December and cancelled a $150 million contract with Fermi America. The news hit the company’s stock hard, sending its value into a tailspin and triggering a class action lawsuit from investors.

  Yet construction continues. The plan says it’ll take 11 years to build out the massive datacenter, which will first be powered by a series of natural gas generators before the planned nuclear reactors come online.

• Meta inks deals to supply a staggering 6 gigawatts in nuclear power for data center ambitions — enough wattage to supply 5 million homes

  The power that Meta acquired will come from three different providers — Vistra Corp. with its existing nuclear plants and a couple of startups: OpenAI-backed Oklo, Inc., and TerraPower LLC, which is supported by Bill Gates and Nvidia. Vistra will deliver 2,176MW of power to Meta from its Perry and Davis-Besse nuclear plants in Ohio, with a further increase of 433MW across the two sites and the Beaver Valley plant in Pennsylvania. On the other hand, both Oklo and TerraPower are still working on their small modular reactor prototypes, which we don’t expect to be deployed until the 2030s.

• Stop aruging each other over LLMs, you stupid fucks // crescentro.se

  You want change? Pick up the phone and start calling your representatives.

  Call them until Tim Cook is dragged by his ear to the same courtroom that let his company off the hook for gatekeeping the App Store, and until he is forced to explain in painful detail how is enabling distribution of CSAM on his App Store exactly keeping iPhone users safe.

  Call them and demand Sundar and Satya be dragged onto the Congress floor and repeatedly slapped with a slipper until they stop forcing the taxpayers to finance their power and water use, for data centres they did not want and that they do not profit from.

  Call your MEP and get them to lock Sam Altman in a room with 50 of European Union’s finest bureaucrats until he elaborates exactly how he’s complying with European privacy, safety, and competition regulations2.

  Because what you’re doing now is exactly what the ones who started this mess are hoping you will do. It’s not making anyone’s life better. And in a decade, the only thing people will remember is judgement and hurt.

• AI datacenter boom reshapes US clean energy bets

  The preference among hyperscale datacenter operators for baseload power, typically supplied by gas turbine generation, will likely hold back the growth of wind and solar power purchase agreements (PPAs), it says, but forecasts 2026 will be the year traditional wind and/or solar companies step out of their core competencies and "de-silo" into adjacent and complementary markets.

• This post was almost angry, now it's just confused

  The core of Bezos's argument, as I understood it, was creating a parallel between electricity distribution and data center distribution, specifically in the context of AWS and why he thought that worked for data centers. At no point did he mention personal computers, and he certainly didn't express any desire to move all customer data to his cloud. The explicit text of his words was that AWS solved the problem of every business needing its own physical server farm. AWS offered a different solution, one that has been quite successful and convenient on the whole.

• EPA rules that xAI’s natural gas generators were illegally used

  The power plants drew the ire of local communities and legal organizations. The company was facing a lawsuit for contributing more ozone and particulate emissions in an already polluted region. The company was operating as many as 35 turbines, and only 15 were ultimately permitted. Today, xAI has 12 turbines providing power to its data centers there.

• ‘Every Data Centre Is a U.S. Military Base’ – Pixel Envy

• Memphis Press Turning Blind Eye to Grok's Creation of Sexual Deepfakes of Adult and Children - 512 Pixels

  At the end of December, stories began to break about Grok being used by X users to generate non-consensual sexual imagery of people, including children.

  I’ve been writing about xAI for over a year, as the company has built two massive data centers here in my hometown of Memphis, Tennessee. There have been concerns about the natural gas turbines powering the data centers and the amount of water xAI was using to keep its hardware cool.

• Microsoft hiring energy strategists for Asian datacenters

  The software giant last week advertised for three “Senior Energy Program Managers” – two in Australia and another in Singapore – who will be “responsible for strategizing and executing “end-to-end” plans for Microsoft’s energy requirement for a specific geographical market within the APAC region.”

• EPA: Elon's methane-powered xAI datacenter illegally generating electricity

  The turbines harm air quality and increase the risk of cancer, asthma and respiratory diseases. The neighborhood is historically poor and black; targeting such communities is a longtime strategy for heavy industry and the energy sector looking for places to site polluting or otherwise risky facilities without facing significant local opposition. But in this case they found just that, and then some. Here's Abre' Conner, the director of environmental and climate justice for the NAACP, which filed a lawsuit against xAI last July claiming its unpermitted turbines violated the Clean Air Act.

• Data center developers pull back from real estate

  The business of selling power-equipped real estate to data center developers is getting more challenging, one of the top global real estate investment firms told Semafor.

• Power scarcity drives datacenters to Texas

  The demand for energy from these massive server farms is growing rapidly, with the report claiming that total IT infrastructure load could roughly double over the next few years, from about 80 GW in 2025 to 150 GW or more by 2028.

• Elon Musk's xAI Colossus 2 is nowhere near 1 gigawatt capacity, satellite imagery suggests — despite claims, site only has 350 megawatts of cooling capacity

  Based on 550,000 Nvidia Blackwell AI accelerators, xAI's Colossus 2 is advertised as the industry's first AI facility that consumes one gigawatt of power for AI inference and training. But for now, the data center codenamed 'Macrohard' purportedly only has 350 MW of cooling capacity — not nearly enough to cool down 550,000 Blackwell GPUs at full power, even in winter. As a result, Musk's Jan. 19 announcement may have been premature, to put it mildly. Epoch AI expects the supercomputer to reach 1 GW by May.

• 'We can't completely vacate the client market' says Intel amid wafer supply shortages — Nova Lake still on-track for late 2026 release, 14A in 2028

  Intel reconfirmed its commitment to the consumer market during its Q4 2025 earnings, despite a heavy focus on wafer supply shortages and increased demand from the data center and AI markets. The company is shifting its internal wafer supply to the DCAI segment while relying on external wafer supply for CCG (Client Compute Group).

• Microsoft CEO says AI needs to have a wider impact or else it risks quickly losing ‘social permission’ — also says that the technology should benefit more people to avoid a bubble

  The rush to build AI infrastructure is putting a strain on many different resources. For example, we’re in the middle of a memory chip shortage because of the massive demand for HBM that AI GPUs require. It’s estimated that data centers will consume 70% of memory chips made this year, with the shortage going beyond RAM modules and SSDs and starting to affect other components and products like GPUs and smartphones.

• The AI Boom Will Increase US Carbon Emissions—but It Doesn’t Have To

  That’s the message of a new analysis from the Union of Concerned Scientists released Wednesday, which models a variety of scenarios for how to fuel the coming AI boom. The US is poised to see a 60 to 80 percent increase in electricity demand through 2050, with data centers alone making up more than half of the increase by the end of this decade, the analysis finds. If policies stay the same as they currently are—with attacks on renewable energy being embedded into regulatory regimes and few significant national policies restricting carbon emissions from power plants—we could see between a 19 and 29 percent increase in CO2 emissions from US power plants tied just to the energy needs of data centers over the next 10 years.

• Kushner Reveals Dystopic Plan to Build Data Centers on Ruins of Gaza Genocide

  Tellingly, numerous parts of the map located next to residential areas are dedicated to industry and “data centers.” Ruinous technology like AI, reports have said, are slated to be a major part of the White House’s plan for Gaza, with other slides in the pitch deck reported by The Wall Street Journal showing a transformation of the Strip into a “smart city” with “tech driven governance.”

• Ex [cryptocurrency] miners building 430 MW datacenter in secret place

  [Cryptocurrency] miner turned AI infrastructure provider Applied Digital announced it has broken ground on a 430 MW data center somewhere in the southern US, but it isn’t yet ready to reveal the location of its new facility.

  The company has faced negative reactions to past projects. In Harwood, North Dakota, locals opposed the company’s plans to build a data center campus. Another of the company’s projects, in Ellendale, Colorado, also attracted attention – not all of it positive.

• Congress wants veto power over Trump administration for AI chip exports — new proposed AI Overwatch Act would shift ultimate control of high-performance chip exports

  Legislators from the U.S. House of Representatives Foreign Affairs Committee this week advanced the AI Overwatch Act, originally introduced in December, which would give ultimate control over the exports of high-performance data center-grade AI processors to adversary nations to Congress. As reported by Reuters, the bill advanced after the White House introduced its new export rules for fairly advanced AI GPUs from AMD and Nvidia to China, along with a mechanism to get a 25% fee from the exporters.

• TikTok’s US joint venture off to a rocky start

  The TikTok USDS (US Data Security) Joint Venture - the entity backed by Trump ally Larry Ellison's Oracle, private equity firm Silver Lake, and Emirati-backed investment firm MGX, with ByteDance retaining a minority stake - took to X on Monday to explain the issues as resulting from a power outage at a US datacenter.

• Microsoft plans more server farms, despite water worries

  The technology giant has been given approval to build 15 more server farms at Mount Pleasant in Wisconsin, reports say, near its existing AI datacenter campus, which Microsoft has said is on track to come online in 2026 and has billed as "the world's most powerful AI datacenter."

• Data Centers Are Driving a US Gas Boom

  The findings from Global Energy Monitor, a San Francisco–based nonprofit that tracks oil and gas developments, come as the Trump administration is both encouraging data center build-out and doing away with pollution regulations on power plants and oil and gas extraction. They will also almost certainly mean an increase in US greenhouse gas emissions, even if some of the projects tracked by Global Energy Monitor never get built.

• DMs Mentioning Jeffrey Epstein Were Blocked on TikTok, Worrying Users

  Besides the data center troubles and the still unexplained blocking of the Epstein DMs, TikTok USDS has faced scrutiny for the terms of its user agreement. After the U.S. takeover, American users were prompted to sign amended policy guidelines that reflected the venture’s broadened power to collect data from customers, including “precise” GPS location and data from user interactions with TikTok’s AI tools. This language raised alarm bells from activists who warned that the app could theoretically be used to track the movements of protesters, though the GPS feature has reportedly not yet been enabled and is expected to be opt-in only.

• US gas power projects tripled in 2025 on AI demand

  In other words, the US is looking at adding nearly 50 percent to its existing world-leading fleet of gas-fired power plants. More than a third of those new plants are directly linked to datacenter projects, GEM said.

  If all the new projects in development end up being fired up, that'll equate to around 53.2 billion tonnes of lifetime CO₂ emissions, GEM global oil and gas plant tracking project manager Jenny Martos told The Register. 12.1 billion tonnes of that addition will come from US projects alone. For reference, the existing global fleet of gas and oil-fired power plants is estimated to account for 54.3 billion tonnes of lifetime emissions.

• One-way heat diode could help batteries run cooler and last longer

  The technique, known as thermal rectification, could extend battery life in cell phones, electric vehicles, satellites, and even improve AI data center efficiency.

• Tech CEOs Say AI Is Ushering in an Age of Abundance, But Instead the Evidence Shows That It's Pushing Down Wages

  “That is equal to 3.0 percent of GDP,” Baker writes, “which is certainly enough to give us a recession, especially when added to the collapse of spending on data centers and other AI investments.”

  Making matters worse, that trend line doesn’t look like it’s going to reverse back to the old average anytime soon. It all supports Baker’s long-running thesis that the AI bubble bursting would actually be a phenomenal economic development for working class Americans.

  “The sooner the AI bubble bursts,” he explains, “the better it will be for almost all of us, except the AI whizzes.”

• You Will Laugh Out Loud When You Hear What the Tech Industry Is Spending a Swimming Pool's Worth of Money to Convince the Public

  According to the Financial Times, data center operators are “planning to go on the offensive with a lobbying blitz” as well, trying to get ahead of the growing public backlash. One data center executive told the FT that lobbying spending is a flash in the pan compared to the tens of billions being spent on infrastructure.

• Energy Bills Have Soared Recently. How Can States Bring Costs Down?

  Electricity prices have soared recently because of fast rising demand for energy from residents, businesses and data centers used for artificial intelligence. The national average retail rate was 5 percent higher in November from a year ago, nearly double the overall inflation rate. The average household spent $178 a month for 1,000 kilowatt-hours of electricity, according to the Energy Information Administration.

• Rising memory costs bump Raspberry prices by up to $60

  The RAM shortage is caused by increased datacenter demand, which has encouraged memory makers to focus more on selling pricey HBM (high-bandwidth memory) to hyperscalers. It has also taken some of the regular DRAM out of the market as rack systems require it as well. Nvidia’s Vera Rubin NVL72, for example, requires 54 TB of LPDDR5X memory.

• SpaceX data centers

  I thought there was consensus that it was bad to burn through GPUs in space where it’s nearly impossible to maintain them? It’s gotta be more practical to build huge solar farms on earth. Here we can use the new energy for more than just AI training.

• PC vendor warns of upcoming price hikes due to SSD and memory volatility — PowerGPU to pass costs to customers once existing inventory depletes

  PowerGPU, a system integrator known for its high-performance gaming PCs, just issued a devastating price warning on social media. The company said on its X account that costs of SSDs and other parts have gone up, and that it expects to increase prices soon. This is catastrophic news for gamers and enthusiasts who have been suffering from continuous prices increases on memory and storage since late last year, which has been brought by the incessant demand for memory and storage chips for AI data centers.

• As tech companies race to build data centers, communities are pushing back

  It’s a fight flaring across the country, in red and blue states, from Oklahoma to Indiana to Pennsylvania, pitting big tech companies and their partners against local activists up in arms about the environmental and community impacts of data centers, as well as potential disruptions from the artificial intelligence technology they make possible. Power-hungry data centers are also being blamed for rising electricity prices. That issue was central to November’s gubernatorial elections in New Jersey and Virginia, the latter of which has the largest concentration of data centers in the country.

• The OpenAI and Nvidia $100b not-a-deal is off

  The press went wild! Stock prices got quite the boost! Line went up!

  It turns out this impossible data centre deal wasn’t possible. The Wall Street Journal broke the news that “The $100 Billion Megadeal Between OpenAI and Nvidia Is on Ice.” Jensen Huang of Nvidia was not impressed with OpenAI: [...]

• Memory and CPU shortages to push up PC prices

  Constraints on specific configurations, alongside emerging CPU availability issues, are expected to limit choice and increase pricing pressure from Q2 onward, according to Context.

  "Manufacturers are prioritizing production for AI datacenter infrastructure, redirecting capacity away from consumer-grade memory and storage towards high-bandwidth memory (HBM) and advanced storage required for large-scale AI workloads," said senior retail analyst James Bates.

• Coding Agents and Software Development

  I hate the fact that all these LLMs are trained on stolen data.

  I hate how much energy is used to train these models.

  I hate that computer hardware like RAM is getting so much more expensive because it’s being bought in bulk for data centres to power LLMs.

  Morally, I think this is terrible technology, and I don’t want it to be a part of my career, or even the world in general, frankly.

  And there are other reasons to be skeptical about LLMs. The business model is a big question for all of these companies. They don’t make a profit, even when they charge people £200 a month. But I can’t see that being anything other than a short term problem as economies of scale, and the progress of time, bring costs down.

  Another problem often raised is: what happens to the next generation of software developers? If having an agent is like having a junior developer, and you’re the senior developer, where do the new senior developers come from? I suspect the answer to that is: where did the new assembly developers come from?

• The Generative AI Race

  Over the past couple of days Amazon, Google, Microsoft and Meta have seen their stocks plummet when they released generally good earnings, but gigantic plans to each spend about $200 billion over the next year building out new data centres and compute capacity to power their AI ambitions. All these companies already possess a huge number of data centres and gigantic compute resources. Why do they need so much more? And will these huge investments ever pay off? Let’s look at what these companies are doing with all these billions of dollars of servers.

• Michigan Sierra Club presents petition calling for moratorium on data centers

  Sierra Club Michigan on Wednesday issued a petition asking the Legislature to ensure Michigan’s laws can keep pace with the significant energy and water use demanded by hyperscale data centers being proposed across the state while providing cost protections for residents. Without improved protections the group warned that these facilities could drive up energy bills, strain water resources and undermine the state’s effort to meet a 100% clean energy standard by 2040.

• Data center pause

  Wired reporting on a proposed bill to pause building data centers in New York for three years: [...]

• New York Is the Latest State to Consider a Data Center Pause

  Red and blue states alike have introduced legislation in recent weeks that would halt data center development, citing concerns from climate to high energy prices.

• Why I joined OpenAI

  There's so many interesting things to work on, things I have done before and things I haven't. I'm already using Codex for more than just coding. Will I be doing more eBPF, Ftrace, PMCs? I'm starting with OpenAI's needs and seeing where that takes me; but given those technologies are proven for finding datacenter performance wins, it seems likely -- I can lead the way. (And if everything I've described here sounds interesting to you, OpenAI is hiring.)

  I was at Linux Plumber's Conference in Toyko in December, just after I announced leaving Intel, and dozens of people wanted to know where I was going next and why. I thought I'd write this blog post to answer everyone at once. I also need to finish part 2 of hiring a performance engineering team (it was already drafted before I joined OpenAI). I haven't forgotten.

• Anthropic says its datacenters won't spike your leccy bill

  These steps are welcome, but won't make much of a dent as they apply only to Anthropic's own datacenters, which make up a small part of the US bit barn armada. A growing body of evidence suggests those facilities are driving up the price of electricity across America.

• Details on long-expected Meta data center campus unveiled

  Meta, which owns technologies like Facebook, Instagram and WhatsApp, will construct 13 total buildings spanning four million square feet on 1,500 acres. That includes 10 data center buildings, as well as facilities for logistics, warehousing and administrative support functions.

• Data Centers Are Scrambling to Power AI Boom With [Fossil] Gas

  Many of these data centers want the kind of flexible, around-the-clock energy associated with combined-cycle natural gas turbines. These heavy-duty machines burn gas to spin turbines and generate electricity, then capture the associated heat and use it to spin the turbines some more. As far as fossil fuel generation goes, they are among the most efficient options for so-called dispatchable baseload power. But with demand for these turbines surging and supply increasingly tight, developers are turning to creative alternatives.

  The upshot of all this creativity is clear: Much of the data center build-out is poised to be powered by natural gas — and the climate consequences that come with it.

• The U-M data center meeting in Ypsilanti was never about community input

  The University’s proposed data center in Ypsilanti Township has been controversial because U-M officials have consistently provided few details and have left township representatives out of the process. In partnership with LANL, the development is meant to provide high-functioning research capacity for the University. Currently, project leaders are deciding between two potential sites for the project: a property on Bridge Road and Textile Road and one at the American Center for Mobility. They will not pay taxes on the property — which one Ypsilanti resident and accountant had calculated would amount to tens of millions of dollars for the township. They also will not be subject to standard zoning laws since this development will be built by a public university.

• 'We have the investors back': World copper rush promises new riches for Zambia

  Surging demand from the artificial intelligence, green energy and defence sectors has exponentially boosted demand for the workhorse metal that underpins power grids, data centres and electric vehicles.

  The scramble for copper exposes geopolitical rivalries as industrial heavyweights -- including China, the United States, Canada, Europe, India and Gulf states -- compete to secure supplies.

• The RAM crunch could kill products and even entire companies, memory exec admits

  It’s genuinely possible that some companies won’t be able to secure enough RAM. AI data centers are gobbling up the vast majority of the world’s memory supply as part of a global buildout, creating an unprecedented imbalance in supply and demand that’s seen RAM prices triple, quadruple, or even sextuple over the past handful of months. Even Nvidia might skip shipping a gaming GPU for the first time in 30 years. Even Apple may have trouble securing enough RAM now, not to mention memory chips for SSDs, and other vital components.

• Wi-Fi routers are expected to explode in price due to RAM shortage

  With AI data centers buying up all the memory on the market, memory prices have risen by 90 percent since the fourth quarter of 2025. Year over year, we’re looking at a price increase of 600 percent. Market analysts expect memory prices to rise further in 2026, peaking in the first half of the year: [...]

• The A.I.pocalypse Hits Video Game Hardware

  Orders that haven’t yet been fulfilled, for data centers that haven’t been built yet, for AI services which don’t generate money, to perform services most people don’t even want.

  I feel like this is going to lead to a really strong resurgence in the Right to Repair movement, as we’re all going to be holding onto our tech longer.

• Why Self-host?

  I recently shared my current Homelab setup with a colleague and was asked a pretty simple question I just took for granted... why?

  Why go through the hassle of configuring servers, installing applications, setting up containers and spending quite a substantial amount of money on hardware that will not even run under optimal data center conditions (consumer-grade internet connection, no failover, no auto migrations)?

  I will also give some specific recommendations on what you could and maybe should self-host.

• Farmer Hailed as Hero for Rejecting Huge Payment to Turn His Land Into a Giant Data Center

  The farmer was offered $60,000 per acre to build a data center on his property. But giving up his family legacy wasn’t in the cards for him.

  “I was not interested in destroying my farms,” he told WPMT. “That was the bottom line. It really wasn’t so much the economic end of it. I just didn’t want to see these two farms destroyed.”

  Instead, he sold the development rights in December for just under $2 million to a conservation trust, taking a significant loss but guaranteeing that it would stay farmland in perpetuity.

• I’m Tired Of These Useless Jackasses Making The Computer Expensive

  There is a historic data center buildup going on to fuel this bubble, and what used to be reliable, cheap and boring parts have shot through the roof price wise. If you bought a stick of RAM in September and flipped it you could make more than most stocks.

  The other component of this is that there are functionally three companies that produce this stuff at scale and which dominate the market–SK Hynix, Micron, and Samsung–and they kind of don’t give a shit if they sell to Sam Altman or you. But before you get mad, they want you to know that it isn’t their fault and that this was inevitable as their stock shot straight up. Micron shuttered Crucial, their consumer-facing RAM brand, in December to focus on AI. Factories are being built, but scaling up takes years, although there’s some hope that Chinese manufacturers could scale up. What’s worse, this has also come for flash memory as well, and as a guy who had to go to Microcenter and pay twice as much as he would have in August for an M.2 SSD, it’s rough out there. Guess what else uses RAM? Graphics cards! And while prices did briefly crash, they are now shooting back up just in time for, to quote Steve Burke of Gamer’s Nexus, an “inevitable opportunity to screw consumers”.

• Community activist arrested at data center meeting in Claremore

  "One or two guys spoke in favor of it, they let one of those guys go over several seconds, probably 20, 30 seconds, really said nothing about him going over but when Darren Blanchard did it, they brought the police in and said get him out of here and he said ok, he willingly put down the mic, went up front and said I'm done and they arrested him."

• Man Opposing Data Center Arrested for Speaking Slightly Too Long

  An Oklahoma man tried to talk about a data center coming to his community. Police arrested him when he went a few seconds over his time limit.

• Barrage of Emails From AI Politics Platform Defeats Clean Air Initiative

  So far, reality has demonstrated the exact opposite. Not only are electrical demands of data centers supercharging our carbon emissions, but the technology itself is now being used to actively resist climate regulations.

• Performance Max Preserving Attribution

  Now that other Big Tech companies have joined the Meta/Mozilla effort and are working on getting “Attribution” (formerly “Privacy-Preserving Attribution”) through W3C, it’s pretty clear that competing standards is the least of the problems here. Besides user harms like the alarming fraud-friendliness, and the fact that it’s based on cranking through extra garbage data (“for privacy”) on data centers in the USA—right at the time that lots of people are concerned about the impact of data centers on the environment and energy markets and the need to stop depending on IT services in the USA (read the room, people) a big question for advertisers needs to be: can this thing ever give an honest answer?

• As data centers look to rural New England, Maine considers a moratorium

  When local officials in rural Wiscasset, Maine, voted on November 4, 2025, to pause a data center discussion in the community of around 4,000 year-round residents, it may have been a sign of what’s to come. As the nationwide expansion of data centers arrives in New England, questions about electricity prices, grid reliability, and impact on water resources are forcing elected leaders to pump the brakes.

• Data center tax breaks are on the chopping block in some states

  “Who is actually benefiting from these massive data centers that, in many cases, are the size of one or two shopping malls combined?” asked Michigan Democratic state Rep. Erin Byrnes, who introduced a proposal to repeal the state’s data center tax exemptions. “They have a large footprint in terms of land and energy usage. And by and large, it’s not going to be the average resident who lives near a data center who’s going to benefit.”

• Data center bill stalls after last-minute change opposed by industry finds support

  The Georgia Senate avoided voting on Senate Bill 34, which moved out of committee earlier this week after it had been changed to reflect language from House Bill 1063, which Georgia Power did not oppose and the data center industry preferred.

• Residents Say Elon Musk's AI Facility Is Like Living Next Door to Mordor

  The $20 billion facility, run by Musk’s AI firm xAI, is powered by 27 methane gas turbines that run day and night, belching fumes and emitting a constant noise like jet engines, NBC News reports. The turbines were trucked in because Southaven, the unsuspecting rural community that Musk chose to build the data center in, can’t provide the electricity the facility needs.

• Musk's AI power plant generates sound and fury in Mississippi

  Nine more temporary turbines arrived in December, bringing the total to 27. The company has since applied for a permit to put in 41 permanent turbines at the 114-acre site to generate electricity for AI data centers.

  Tech giants like Microsoft, OpenAI and Google are making billions in AI investments across the country to build up the computing and electric power the technology demands. That requires vast land and resources and, for communities like Southaven, a tradeoff between the promise of new jobs and tax revenue vs. environmental and energy costs.

• Big Tech Signs White House Data Center Pledge With Good Optics and Little Substance

  He was flanked by representatives from Microsoft, Meta, OpenAI, xAI, Google/Alphabet, Oracle, and Amazon.

  Bipartisan anger about data centers and their potential impact on consumers’ electric bills has exploded over the past year. As the White House goes all in on AI, the pledge marks a significant salvo by the Trump administration to assure voters that they will not be affected by rising costs.

  But electricity experts and industry insiders threw doubt on how much power the White House actually has to create meaningful consumer protections.

  “This is theater,” says Ari Peskoe, the director of the Electricity Law Initiative at the Harvard Law School Environmental and Energy Law Program. “This is a press release designed to make it seem like they are addressing this issue. But this issue can only really be addressed by utility regulators or Congress. The White House doesn’t really have a lot of moves here, and I don’t think the tech companies themselves are the most important parties on cost issues.”

• PC graphics cards are now nearly 100 percent Nvidia

  The JPR report complements a second report issued earlier this week that included integrated PC graphics, which also highlighted the rise of graphics cores flowing into the workstation and data center. Overall, total add-in card shipments increased by 36 percent year-over-year. But the new data, from the fourth quarter of 2025, also showed a sequential drop in card shipments of 11.5 percent from the third quarter.

• Scaling the future of Open RAN: Red Hat joins the OCUDU Ecosystem Foundation

  Launched under the stewardship of the Linux Foundation, OCUDU represents an important shift in how cellular networks are built. By creating a carrier-grade, open source software stack for 5G, 5G-Advanced, and 6G, the project aims to do for the Radio Access Network (RAN) what Linux did for the data center: Breaking down segregated technology areas and fostering an environment where RAN innovation moves at the pace of open source innovation.

• Demand for AI Data Centers Sends Prospectors Hunting for Land and Power

  As the A.I. boom enters its fourth year, Mr. Janous and his team have become modern-day land men. They work at the intersection of utility companies and tech giants, securing the power and sites necessary for the hundreds of billions of dollars of data centers being built across the country. Their product — powered land, they call it — has become one of the nation’s most valuable commodities.

• Temporarily banning data centers draws more interest from state, local officials

  Lawmakers in at least 11 states — Georgia, Maryland, Michigan, New Hampshire, New York, Oklahoma, South Carolina, South Dakota, Vermont, Virginia and Wisconsin — have introduced legislation this session that would temporarily ban data centers, according to Good Jobs First, a watchdog group that focuses on economic development incentives.

• Southaven Residents Fight xAI’s Plan for 41 Gas Turbines

  Directly in front of her property along the Mississippi-Tennessee border sits a tangle of massive power lines, an extension of the electrical grid operated by Memphis Light, Gas and Water.

  At the end of her street, a few hundred yards away from her bedroom, sprawls Colossus 2, the second data center built by Elon Musk’s company xAI to power his controversial AI chatbot, Grok. And less than two miles across the state line in Southaven, Mississippi, is an xAI-owned energy plant with over two-dozen towering gas turbines.

• Temporarily banning data centers draws more interest from state, local officials

  While no states have successfully enacted a moratorium, several cities and counties have banned new data center projects over the past year. In August, St. Charles, Missouri, became one of the first cities in the nation to ban data center construction for a year.

• Voters being misled about the cause of Michigan’s economic decline

  A lot of voters might not even be reading or hearing about Michigan’s decline if one question in the chamber’s poll is any indication. Opposition to the explosion of large-scale data center proposals in the state has been a huge issue. But the poll found that more than 40% of voters haven’t seen or heard anything about them.

• AI datacenters may gulp NYC's daily water supply at peak

  Without new water efficiencies, datacenters across America may require 697 million to 1.45 billion gallons of extra peak water capacity per day by 2030, the study estimates. This compares with New York City's daily water supply of about a billion gallons.

• [2603.02705] Small Bottle, Big Pipe: Quantifying and Addressing the Impact of Data Centers on Public Water Systems

  Water is a critical resource for data centers and an efficient means of cooling. However, meeting the growing water demand of data centers requires substantial peak water withdrawals, which many communities in the United States cannot supply, especially during the hottest days of the year. This largely overlooked water capacity constraint is emerging as a bottleneck for data centers and can force operators to rely on less efficient dry cooling, further stressing the power grid during summer peaks. In this paper, we focus on the direct water withdrawal of U.S. data centers for cooling and examine their impacts on public water systems. Our analysis indicates that, if the 2024 water use intensity persists, U.S. data centers could collectively require 697-1,451 million gallons per day (MGD) of new water capacity through 2030, comparable to New York City's average daily supply of roughly 1,000 MGD. Under an optimistic scenario with a compound annual water use intensity reduction by 10%, the water capacity demand decreases to 227-604 MGD, although high-growth IT loads could still require enough capacity to hypothetically supply about half of New York City for most of the year. The total valuation of the new water capacity is on the order of \$10 billion, reaching up to \$58 billion in the high-growth case. These impacts are highly concentrated on communities hosting data centers. Finally, we provide recommendations to address the growing water capacity demand of U.S. data centers, including reporting peak water use, developing corporate-community partnerships, adopting a Water Capacity Neutral approach (colloquially "Pipe Neutral") to allow host communities to retain limited water capacity resources, and implementing coordinated water-power planning to responsibly leverage water for peak power reduction and opportunistically utilize surplus power to mitigate impacts on public water systems.

• How the spiraling Iran conflict could affect data centers and electricity costs

  It’s a week later and the conflict has only escalated since the US and Israel launched strikes against Iran, killing Supreme Leader Ayatollah ​Ali Khamenei. Energy infrastructure has become a key leverage point in the unfolding war, with Israel hitting Iranian fuel depots and Iran targeting Gulf neighbors’ oil and gas infrastructure in its own strikes. Iran’s paramilitary Revolutionary Guard threatened on Tuesday not to “not allow the export of even a single liter of oil from the region to the hostile side and its partners until further notice.” Iran has reportedly also started to lay mines in the strategic Strait of Hormuz, through which one-fifth of global petroleum consumption and liquefied natural gas (LNG) trade used to move.

  I talked to Blakemore again today about what Iran’s continued chokehold on the Strait of Hormuz means for energy costs and US tech companies’ rush to build out energy-hungry AI data centers.

• China's ByteDance to access 36,000 Blackwell GPU cluster through Malaysia cloud operator — Nvidia confirms no objections, deal is in line with US export controls

  The cluster, worth around $2.5 billion and consisting of 500 NVL72 GB200 rack-scale systems, will be formally owned and operated by Aolani Cloud in Malaysia. The hardware will be supplied through Aivres, a company that builds servers based on Nvidia GPUs, according to the WSJ, which cites people familiar with the arrangements. An Aolani spokesperson told WSJ that the company currently operates with roughly $100 million worth of hardware, so the scale of the proposed expansion is vast, but it is not completely clear who is funding it. ByteDance is also reportedly considering additional deployments, such as a cluster containing over 7,000 B200 GPUs at a data center in Indonesia.

• Subscription watch: Managing your hybrid cloud estate

  Managing a hybrid cloud environment spanning on-premise data centers, edge deployments, and multiple public clouds often results in subscription sprawl. Even in simpler environments, it can be challenging to maintain clear visibility into subscription use. Organizations frequently struggle to answer a basic question: “Exactly how much of our purchased Red Hat capacity are we actually using right now?”

• Americans' Anger Against AI Data Centers Is Boiling Over

  As flagged by 404 Media, the majority of people surveyed said they think data centers are bad for the environment, home energy costs, and quality of life for people living nearby.

  Compared to the amount of hate data centers got, there wasn’t a lot of love going around. A paltry 4 percent of Americans said data centers were “mostly good” for the environment, compared to 39 percent who said the opposite. Meanwhile, only 6 percent believed data centers have positive effects on their neighbors’ quality of life and home energy cost, versus 30 percent and 38 percent who said they were “mostly bad,” respectively.

• People Hate Datacenters, Survey Finds

  Datacenters are terrible neighbors. The buildings drive up the cost of energy for people who live nearby, consume massive amounts of water, and can produce noises and fumes that hurt locals. In Mississippi, locals are concerned about the pollution and noise caused by an xAI datacenter powered by gas turbines. A proposed datacenter project near Amarillo, Texas would be powered by four massive nuclear generators and pull water from an aquifer with dwindling reserves. In an effort to quell fears about power consumption, Trump made Amazon, Google, Meta, Microsoft, OpenAI, Oracle, and xAI sign a pledge to keep energy costs down. But a pledge isn’t a law. It’s not even an executive order.

• Bernie Sanders on the billionaire tax, data centers, and Iran

  Vermont Sen. Bernie Sanders (I-VT) is no stranger to singling out the richest of the rich. Along with Rep. Ro Khanna (D-CA), Sanders recently introduced the Make Billionaires Pay Their Fair Share Act, a 5 percent annual wealth tax on anyone in the US worth over a billion dollars.

  The act would affect 930 people — the very tippy-top of the 0.01 percent. Elon Musk would owe roughly $42 billion per year. Mark Zuckerberg would owe $11 billion.

• Meet Canonical at NVIDIA GTC 2026: NVIDIA CUDA and NVIDIA Vera Rubin NVL72 support in Ubuntu 26.04 LTS

  With Ubuntu 26.04 LTS, Canonical is continuing its collaboration with NVIDIA by directly distributing NVIDIA CUDA in Ubuntu, and by delivering day-one platform readiness for NVIDIA Vera Rubin NVL72 rack-scale architecture. Together, these advancements provide developers and enterprises with a stable, enterprise-grade foundation for building, scaling, and operating AI workloads across edge, workstation, and data center environments.

• Alabama House committee OKs Senate bill regulating data centers

  SB 270, sponsored by Sen. Lance Bell, R-Pell City, allows the state agency that regulates public utilities to take into the account that data centers pay for the additional cost to operate their facilities and that data centers will provide a benefit to the committee when it reviews a contract between a public utility and a large-load data center.

  Bell did not provide any additional comments on the legislation aside from introducing the bill.

• UMich finalizes purchase of Textile Road parcel for data center

  The data center has met opposition from township residents and members of the University community due to concerns over its planned use in the development of nuclear weapons and its potential impact on the electrical grid and environment. Local officials say the University has not properly communicated with the township throughout the site selection process.

• Tiny Township Fears Iran Drone Strikes Because of New Nuclear Weapons Datacenter

  Winters delivered a report to the town’s Board of Trustees about the proposed datacenter during a public meeting on Tuesday. “Los Alamos, which produces the nuclear weapons, is a high value target,” he said. He pointed to America’s war in Iran as proof that the datacenter would be a target, noting that Iran’s drones had disabled AWS servers in the Middle East. “This is not a commercial datacenter. A Los Alamos datacenter is going to be the brains of the operation for nuclear modeling, nuclear weaponry.”

• SoftBank and AEP plan massive Ohio gas-fired AI data center campus

  The arrangement, which would involve building a 10-gigawatt artificial intelligence server warehouse, is part of ‌a wider Washington-Tokyo trade ⁠deal. SB ⁠Energy, a unit of SoftBank, plans to connect a 9.2-gigawatt gas plant to the local grid and power the new data center ​at the federally owned Portsmouth site in Pike County, Ohio.

• A data center opened next door. Then came the high-pitched whine.

  Shaw’s experience is a possible cautionary tale for President Donald Trump, who is pitching a new wave of data-center-specific power plants as the solution to easing voters’ anxieties about the economic costs of the artificial intelligence boom. A pledge he signed last Wednesday with leaders of Amazon, Google, OpenAI and other tech giants calls for data companies to “build, bring, or buy the new generation resources and electricity needed to satisfy their new energy demands,” while “paying the full cost of those resources.”

• Australia to datacenter operators: BYO energy or stay home

  The expectations also call on datacenter operators to prioritise Australia’s national interest, use water sustainably and responsibly, invest in local skills and jobs, and do all that while strengthening the nation’s “research, innovation and local capability.”

• The AI Race Is Pressuring Utilities to Squeeze More From Europe’s Power Grids

  National Grid, which operates the transmission network in England and Wales, says that proposed data centers representing more than 30 gigawatts of power demand are awaiting connection to its grid, equal to two-thirds the peak demand of Great Britain. Even accounting for the likelihood that some of those data centers will never be built, there is currently not enough room to accommodate them.

  The wait for permission to plug in is causing some data center projects to collapse, undermining European ambitions to capture a share of the hundreds of billions of dollars that AI labs are spending on compute. “Across Europe, projects are being canceled because there’s no access to the grid,” claims Taco Engelaar, managing director at grid optimization company Neara.

• Arm Holdings, in Break From Past, Will Sell Its Own Computer Chips

  The company, a British unit of Japan’s SoftBank, on Tuesday announced plans for the first silicon product that Arm will design and sell since its founding in 1990. It is a microprocessor aimed at data centers running artificial intelligence tasks.

• Arm moves beyond IP with AGI CPU silicon — 136-core data center chip targets AI infrastructure with Meta as lead partner

  The chip packs up to 136 Neoverse V3 cores running at up to 3.2 GHz all-core and 3.7 GHz boost across two dies, all within a 300-watt TDP. It supports 12 channels of DDR5 memory at up to 8800 MT/s, delivering more than 800 GB/s of aggregate memory bandwidth or 6GB/s per core with a target of sub-100ns latency. I/O includes 96 PCIe Gen6 lanes and native CXL 3.0 support for memory expansion and pooling.

• Microsoft and Nvidia launch AI partnership to speed up nuclear power plant permitting and construction — simulation tools and generative models could hasten historically lengthy processes

  The idea of letting generative AI anywhere near safety-critical nuclear infrastructure might give the average reader pause, but it's already happening in the real world. Aalo Atomics, an Austin-based startup building modular nuclear reactors for data centers, has said that it reduced its permitting process workload by 92% using Microsoft's Generative AI for Permitting solution, saving an estimated $80 million annually.

• Terafab, Elon Musk’s unrealistic project

  Elon Musk no longer speaks in gigawatts — the unit that has become the benchmark for measuring computing power dedicated to generative AI models. The Tesla CEO is now thinking in terawatts, or one thousand gigawatts. On Saturday, he unveiled Terafab, a project for a giant factory designed to produce chips for his self-driving cars, his robots, and the data centers he ultimately hopes to deploy in space. His goal: to reach a capacity of one terawatt per year. An extraordinary figure, roughly fifty times current global production.

• Panasonic's already sold future datacenter battery output

  Shifting production from automotive to compute and working on supercapacitors as another way to protect workloads

• Senators Demand to Know How Much Energy Data Centers Use

  In a joint letter sent to the Energy Information Administration Thursday morning, seen by WIRED, Hawley and Warren press the agency to publicly collect “comprehensive, annual energy-use disclosures” on data centers. This information, they write, is “essential for accurate grid planning and will support policymaking to prevent large companies from increasing electricity costs for American families.”

• Data center boom raises water, power questions

  Data centers, which support everything from cloud storage to artificial intelligence systems, require large amounts of electricity to operate and significant volumes of water to keep equipment cool. According to estimates cited in the legislation, a single 100-megawatt data center can use about as much water as 2,600 households. Some newer facilities are planned at more than ten times that size, increasing both energy and water demand in the areas where they are built.

  The number of facilities is also rising quickly. There are roughly 4,000 active data centers across the country, with about 3,000 more either planned or already under constructionOpens in a new tab.. This expansion is being driven in part by the growing use of artificial intelligence, which requires large-scale computing power.

• Crusoe to build 900-megawatt AI data center for Microsoft in Texas

  Billions of dollars are being spent on data centers to meet the massive computing ‌requirements for ⁠GenAI. Microsoft ⁠alone reported $37.5 billion in capital expenditure for the quarter ended December 31, ​with about two-thirds going toward computing chips for services like Copilot and ChatGPT.

  Located next to Crusoe's existing AI facilities in Abilene, Texas, the new campus will bring the site's total projected capacity ​to 2.1 gigawatts. On average, one ⁠gigawatt is enough ‌to power 750,000 homes.

• Senators want datacenters to come clean on power consumption

  The letter references the Ratepayer Protection Pledge that President Trump urged the biggest AI and cloud companies to sign earlier this month. The agreement intends to shield American consumers in the event that bills rise due to AI's growing energy appetite. However, as The Register noted at the time, there is no enforcement mechanism.

  Senators Warren and Hawley say both Congress and the public currently lack the data needed to hold these companies to this voluntary pledge. Standardized reporting is essential to delivering the oversight necessary to combat rising utility costs.

• Why the real bottleneck in enterprise AI isn't GPUs - it's data

  For the past two years, the conversation around enterprise AI has been dominated by GPUs: who has them; who can buy them; who is waiting months for them. But a new wave of announcements at NVIDIA's recent annual AI conference suggests something more subtle is happening inside enterprise data centres. While the GPU shortage grabbed headlines, the real constraint on AI adoption is increasingly data infrastructure. And that’s the problem companies like Everpure are now trying to solve.

• Scaling clean energy and reliability in Michigan

  Google’s data center operations will be served by 2.7 gigawatts (GW) of new resources for the grid, including solar power, advanced storage technologies and demand flexibility. This Clean Capacity Acceleration Agreement with DTE (the same structure as the Clean Transition Tariff) will bring new, clean resources online, while supporting the state’s transition away from coal-fired power. As part of our standard approach to building new data centers, Google will fully cover its electricity costs and infrastructure needs, helping to ensure that its growth protects local ratepayers and actively bolsters the long-term resilience of the state’s electricity grid.

• Two wildly different data centers reveal a ‘Fork in the Road’ on how to meet electricity demand

  A proposed Google data center in Michigan is notable for its use of renewable energy and its ability to ramp down power use during periods of high demand.

  The plan, announced last week, is a partnership with the utility DTE Energy and could serve as a model for cleanly and efficiently providing electricity to a data center.

  But there are many examples of data center developers going in the opposite direction, with proposals to build giant gas-fired power plants to go along with giant data centers. The largest of these, also announced last week, is a 9,200-megawatt gas plant in southern Ohio being developed by SoftBank Group Corp. of Japan and its subsidiary SB Energy.

• Seminole Nation of Oklahoma Passes Moratorium on Data Centers

  In a landmark unanimous vote, the Seminole Nation of Oklahoma has become one of the first Native American Tribes to enact a complete moratorium on data center development within its jurisdiction.

  On March 7, the Nation’s Tribal Council approved a resolution by a 24-0 vote to “implement a moratorium on the advancement of generative Artificial Intelligence (AI) technology and hyperscale data center development within the Seminole Nation and within tribal lands and territories.” The resolution bars “any inquiries, discussions and/or developments concerning any entity seeking to develop a data center of any size within the Seminole Nation during the moratorium period.”

• Seminole Nation Becomes First Indigenous Group to Ban Planet-Cooking Data Centers From Its Land

  The Seminole Nation of Oklahoma just became the first Indigenous nation to officially ban data center construction from lands under its jurisdiction.

• You Can’t Escape the AI Tax

  RAM is suddenly so expensive because memory is powering the AI boom. Data centers require huge amounts to run the models that underlie AI tools such as ChatGPT and Claude—especially as they become capable of handling more complicated tasks. This year, a group of tech giants—Amazon, Alphabet, Meta, Microsoft, and Oracle—is set to collectively spend half a trillion dollars on the AI build-out. Roughly a third of that money is being spent on memory alone, according to Dylan Patel, the founder of SemiAnalysis, a popular semiconductor-research firm.

  The insatiable demand has “cannibalized our conventional consumer-electronics supply,” Yang Wang, an analyst at Counterpoint Research, a market-research firm, told me. Every major RAM manufacturer has shifted production lines to service AI data centers. This year, 70 percent of memory-chip products made globally will be destined for them. In South Korea, where two of the biggest RAM manufacturers are based, Silicon Valley executives are reportedly booking hotels in the country’s tech districts, frantically hoping to secure inventory. A Korean newspaper has given them a name: RAM beggars.

• Data Centers Causing Huge Temperature Spikes for Miles Around Them, Study Suggests

  The data centers at the heart of the AI boom are producing so much heat that they’re spiking land temperatures for miles around them by up to 16 degrees Fahrenheit, new research suggests. The effect is so pronounced that the researchers say they’re creating entire “heat islands.”

• New Mexico delays decision on data center air permits

  At least one construction project,Opens in a new tab. a proposed 17-mile, $60 million pipeline to supply gas to power the data center, relies on the state’s approval of the air quality permits to move forward.

  Rep. Angelica Rubio (D-Las Cruces), one of the lawmakers who requested a public hearing, told Source NM she hopes the people most impacted by this project will have “a real seat at the table.”

• Data centre segment may attract investment commitment of $54 billion this year: CBRE

  Real estate consultant CBRE released a report 'India Alternate Sectors Outlook 2026', stating that India's total data centre stock across major cities is projected to increase by 30% year-on-year in 2026 with an estimated 500 MW of new supply addition.

  In 2025, the total data centre capacity in the country reached the 1,700 MW mark, driven by a record 440 MW of new supply addition.

• Oracle believed to have cut 10,000 positions across multiple divisions as mass layoffs begin to fuel AI investments — company is reportedly reducing headcount to fund data centers

  Oracle has reportedly begun mass layoffs based on reports by various people who worked at the company. According to the BBC, one employee noted that around 10,000 positions have been affected based on the number of active staff accounts in the firm’s Slack messaging system so far. Other people from across the company have also made posts about the situation on social media, like LinkedIn.

• Why Mistral AI is taking on debt

  This is not Mistral’s first use of debt. In 2024, the company had already borrowed €132 million alongside a €468 million equity round. More such operations are likely as it moves to finance additional data centers. Taking on debt allows the company to limit dilution for its founders and shareholders at a time of massive capital needs. But the strategy carries risks: it ultimately requires generating enough revenue to repay creditors.

• Data Centres Are on Track to Wreck the Planet. Can We Stop Them?

  Across Canada and around the world, enormous data centres are popping up in the midst of suburban developments and rural topography alike. These giant buildings are what feed the digital appetites of our modern society: long steel racks with servers stacked up to the ceiling and trillions of bytes of data being processed every second, information running through fibre optic cables across towns, countries, and oceans. And now, with the explosion of artificial intelligence and our ever-deepening reliance on it, data centres have become a major part of our modern life as well as a key concern in how we manage our relationship with the world around us. Consider this: due to the greater computational power required by the AI-processing architecture, the energy draw of these facilities doubled just between 2017 and 2023.

  But what are data centres? How and where are they built, and how do they impact our lives, both digital and physical? What are the benefits and the tripwires associated with their existence, and will this very rapid expansion be a footnote in the history of the future or a cautionary tale about development without limits or meaningful regulation?

• There's a Blinking Warning Sign for the Data Centers in Space Industry

  Flash forward to the present, however, and the data centers that are popping up everywhere are amid the AI boom are most decidedly not being built in the ocean. Sources told Reuters that the project was figuratively sunk by lack of client demand and unviable economics for reasons that could also plague Musk’s orbital facilities.

  “These problems are likely to be more ​severe in space than under the sea,” Roy Chua, founder of industry research firm AvidThink, told Reuters.

• A New Google-Funded Data Center Will Be Powered by a Massive Gas Plant

  Michael Thomas, the founder of Cleanview and author of a new report on Google’s power strategy for its data centers, says that Google’s focus on and continued commitment to renewables is often held up by environmental groups as an example of Big Tech doing things right. But the plans for this campus, he alleges, illustrate how even big tech companies with stated climate goals and a public commitment to renewable energy are exploring fossil fuel investments as the AI race heats up.

• Half of planned US data center builds have been delayed or canceled, growth limited by shortages of power infrastructure and parts from China — the AI build-out flips the breakers

  The trade-war between the U.S. and China has forced server makers out of the People's Republic, greatly reducing reliance of American companies on producers from Tianxia. However, China remains the world's largest producer of electrical equipment that is required to build power infrastructure inside and outside of AI data centers. To that end, shortages of power delivery equipment, including devices from China and other countries, are slowing project timelines, Bloomberg reports.

• DC In The Data Center For A More Efficient Future

  If you own a computer that’s not mobile, it’s almost certain that it will receive its power in some form from a mains wall outlet. Whether it’s 230 V at 50 Hz or 120 V at 60 Hz, where once there might have been a transformer and a rectifier there’s now a switch-mode power supply that delivers low voltage DC to your machine. It’s a system that’s efficient and works well on the desktop, but in the data center even its efficiency is starting to be insufficient. IEEE Spectrum has a look at newer data centers that are moving towards DC power distribution, raising some interesting points which bear a closer look.

• Sony Shuts Down Nearly Its Entire Memory Card Business Due to Flash Shortage

  The aforementioned SSD shortage is due entirely to the seemingly endless hunger of AI datacenters for memory. The cost of memory components is going up because the supply for the actual dies that make up all types of SSDs is finite and being largely consumed by enterprise companies looking to expand AI data centers. But it isn’t stopping at price increases. Because supply is finite, that also means that at some point, there simply isn’t enough hardware to even make certain products anymore.

• [Old] FTC vs Microsoft: The Broadest Antitrust Probe Since the 1990s

  The formal investigation began in November 2024, when Khan signed off on a sprawling civil investigative demand spanning hundreds of pages. The document, obtained and reviewed by Bloomberg, compels Microsoft to turn over nearly a decade of data about its operations, from 2016 through 2025.

  Federal investigators are seeking detailed information about Microsoft's AI operations, including the costs associated with training models and acquiring training data. They're looking into the firm's data centres, its struggles to meet customer demand for computing power, and its software licensing practices. Most significantly, they're scrutinising Microsoft's relationship with OpenAI, particularly whether the partnership was structured to avoid merger review requirements.

• Iranian missile blitz takes down AWS data centers in Bahrain and Dubai — Amazon reportedly declares “hard down” status for multiple zones

  The AWS sites in the Middle East each have three compute zones, with both data centers reporting “hard down” and “impaired but functioning” zones. More importantly, the company said in its internal communications, “We do not have a timeline for when DXB and BAH will return to normal operations.”

• Investors push Amazon, Microsoft, and Google to disclose data center water and power consumption — more than a dozen resolutions ask for transparency ahead of annual investor meetings

  More than a dozen investors are pressuring Amazon, Microsoft, and Alphabet's Google to provide detailed data on water and energy consumption at their U.S. data centers, Reuters reported today. The pressure comes as all three companies have recently scrapped multibillion-dollar data center projects following community opposition, and as North American data centers consumed nearly 1 trillion liters of water in 2025, according to market research firm Mordor Intelligence.

• Shots fired over proposed datacenter in Indianapolis

  Ron Gibson, a city-county representative for Indianapolis's 8th District, was woken in the early hours of Monday by gunfire. He found that 13 shots were fired at his front door, and a note left on the doorstep reading "No data centers."

• Maine House advances data center moratorium

  The bill, LD 307, places a moratorium on data centers with a load of 20 megawatts or more. House lawmakers passed the bill by a vote of 82-62 Monday, advancing it to the Senate.

• Maine Is Close to Passing a Moratorium on New Datacenters

  Anger against datacenters is mounting across the country. The massive complexes aren’t good neighbors. They use public land, increase the electricity rates of everyone near them, and have negative effects on water quality and noise levels. The deals to construct them are sometimes cut in secret and local communities have little to no say in what’s being built near them. In Texas, a 6,000 acre datacenter plans to consume water from a dwindling aquifer to power nuclear power plants in the desert. In Michigan, a township is pushing back against a $1.2 billion AI datacenter meant to service America’s nuclear weapons scientists.

• Maine Could be First State to Pause Data Center Construction

  The pause, which would last until November 2027, is meant to give regulators the chance to assess the impact those centers have on the environment and the electric grid, according to reason.com. The Maine House of Representatives recently passed House Bill 307 to prohibit “approval for the development, construction or operation of a data center with a load of 20 megawatts or more” anywhere in the state.

• Data centers are straining the grid. Can they be forced to pay for it?

  This boom coincides with a dramatic rise in U.S. electricity prices, driven by inflation and the rising cost of adapting to wildfires, hurricanes, and other extreme weather. But these massive facilities have also strained the grid — and in some cases — contributed to rising prices. For instance, last year, an independent monitor for PJM, the grid operator that serves 13 northeastern states and Washington, D.C., projected that powering data centers would result in higher electricity generation costs, which would ultimately be passed on to consumers. And in cases where the buildout hasn’t yet led to price hikes, utilities and grid operators expect that it’s just a matter of time if tech companies follow through on their plans. Indeed, the Federal Reserve Bank of Dallas estimates that with data center electricity demand expected to double in the next five years, wholesale power prices could rise by as much as 50 percent.

• Bain Capital's data center unit removes disgraced tenant suspected of smuggling Nvidia GPUs to China — Megaspeed previously alleged to have spent roughly $2 billion on AI processors for illicit distribution

  Bain Capital's Bridge Data Centers (BDC) has terminated its contract with Megaspeed International at its Malaysian site after the U.S. government began a probe into whether Megaspeed smuggled restricted Nvidia AI accelerators to China. The company reallocated 68.4 MW of its power capacity from Megaspeed to cloud provider Zenplayer, reports Bloomberg.

• Texas’s Tax Breaks for Data Centers May Be Largest for Any State in US

  Lawmakers approved the tax break more than a decade ago, when data centers were smaller and required fewer resources. From 2014 to 2022, the exemption amounted to between $5 million and $30 million in lost state revenue per year. By 2023, that skyrocketed to more than $150 million, and this year Texas is forgoing at least $1.3 billion — a number that is rapidly increasing every year, based on state projections.

• xAI's Memphis Water Treatment Plant 'on an Indefinite Pause'

  Part of xAI’s deal with the city of Memphis included an $80 million greywater facility designed to reduce the amount of water the company needs for cooling its data centers. The Mid-South sits atop the Memphis Sand Aquifer that provides us with some of the very best water in the country. It is our best natural resource by far, and one that many people have worked hard to protect.

• Hybrid chip delivers 4x more current for AI data center power

  Engineers at the University of California San Diego have developed a new chip architecture that could reshape how data centers handle power.

  The design focuses on improving how graphics processing units convert high-voltage electricity into the lower levels required for computation.

• Investors are going nuclear to keep UK's AI datacenters fed

  It points to the growth in power-hungry AI infrastructure in the UK, driven partly by the government's AI Opportunities Action Plan, published last year, in combination with "geopolitical energy volatility" as factors driving this trend.

• Intel inks multiyear data center chip partnership with Google

  Google’s C4 instances use a specific variant of Xeon 6 called Granite Rapids. It’s based on a core design called P-core that includes multiple AI-focused optimizations. One of those optimizations is AMX, a set of extensions to the machine language in which Intel chips express computations. AMX speeds up a calculation called multiply-accumulate that AI models run frequently during inference.

• As Sam Altman pushes for IPO, the wheels are coming off OpenAI

  As a product. OpenAI is trapped between a far cheaper Gemini offering from Google, with its rising popularity, and Anthropic's Claude, which delivers results people are willing to pay for; it seems OpenAI is stuck. The company's bet on Nvidia hardware and massive datacenter presence has trapped them in an insanely expensive model that cannot compete with Google, and its own failings have ensured it cannot compete with Anthropic.

• Why are communities pushing back against data centers?

  Tech, data policy expert says concerns legitimate over rising power rates, water use, environmental issues amid mushrooming growth

• Reputation grows for Oklahoma data center opponent - Oklahoma Energy Today

  Blanchard agrees with those in Montana who raised concerns about the extensive amounts of water that would be needed. He explained many towns like Broadview are so-called “tank towns” where water is trucked in to the community.

• Europe To Big Tech: We Don’t Want You, Either

  Last week, in Claremore, Oklahoma, a man named Darren Blanchard was arrested during a heated public hearing over a proposed data center. Blanchard, who was speaking against the project, reportedly exceeded his allotted time in front of the Claremore City Council by about 30 seconds. For that infraction, he was arrested by Claremore police and charged with trespassing. Blanchard and other Claremore residents are fighting a data center being pushed by Beale Infrastructure, a company that is backed by Blue Owl Capital, an asset management firm that recently financed a Meta data center project in Louisiana. (I wrote about the Blue Owl/Meta project last November because the financing deal reeks of Enron. Blue Owl recently restricted redemptions in one of its private funds, and its stock price has plummeted.)

  There’s more.

• The “Moderate” Think Tank Pushing Dems to Loosen AI Rules

  Ascendant DC think tank Searchlight Institute, pushing Democrats to the center, has ties to megadonor Simone Coxe, whose Nvidia-linked money could boost AI-backed efforts to defend data center build-outs and limit AI regulation.

• Amazon rejects AWS climate disclosure proposal

  All of that extra infrastructure needs power, and the proposal notes that utilities in states such as Virginia – the datacenter capital of the world – now have to build new gas-powered generator plants to meet the growing demand, or even keep coal-fired facilities online. All of this is pumping millions of tons of extra greenhouse gases into the atmosphere.

  As a result, Amazon faces questions over how it intends to deliver on its climate promises. The company relies heavily on renewable energy credits (RECs), according to the proposal, which asks whether the volume purchased will increase and whether enough will be available. Amazon's investors would benefit from analysis that explains how the company will tackle those concerns, it states.

• Amazon Is Expanding Data Centers in Central Mississippi, Governor Announces

  Amazon already has an artificial intelligence data processing center in Ridgeland and another data center campus in nearby Canton. With the additional $12 billion, it’ll be expanding its operations in Madison County and creating 700 jobs, Reeves announced Thursday. The company is also investing $1 billion into a project in Clinton that is projected to add 100 jobs to the community.

• Army Corps reviews Google data center proposal, seeks public input

  A public notice published April 6 by the U.S. Army Corps of Engineers’ Little Rock District said the project would include five industrial buildings totaling about 1.43 million square feet. The development would also include two office buildings, an electrical substation and supporting infrastructure such as transmission line corridors, parking lots, access roads and stormwater facilities.

  The project is expected to use more than 100 megawatts of power — enough to supply electricity to roughly 80,000 to 100,000 homes, based on average U.S. household energy use. There were approximately 88,000 households in Little Rock in 2024, meaning this Google data center could use as much power as all the households in Little Rock combined.

  According to the Corps, the development would result in the filling of nearly 17 acres of wetlands and more than 6,000 feet of streams.

• OpenAI pulls out of Stargate UK data centre

  This exciting facility was announced in September last year. It would be built in an “AI growth zone” in northeastern England with local partner Nscale.

  Nscale are best known for the vaporware data centre outside London that turned out to still be a scaffolding yard. Nscale hadn’t even applied for planning permission.

• AI Rejected: Tracking The Great Data Center Revolt

  Since January 1, there have been more rejections or restrictions of data centers in the US than in all of 2025. Furthermore, the number of US data center rejections in 2026 is already nearly equal to the total number of wind projects rejected worldwide in 2025! And more rejections are coming. A recent poll found that 65% of Americans surveyed oppose data centers in their communities. (Details on that in a minute.)

• The Origins of GPU Computing

  Government-funded academic research on parallel computing, stream processing, real-time shading languages, and programmable graphics processing units (GPUs) directly led to the development of GPU computing. GPUs are used in modern datacenters and have enabled the current revolution in artificial intelligence (AI). Nvidia, which makes GPUs, is now the most valuable company in the world. This transformation of computing and resulting economic value created was enabled by more than 30 years of government-funded research. Government funding not only helped develop many of the key technical innovations; it also enabled the training of large numbers of students who have conveyed this technology to industry.

  This article traces the origins of GPU computing. We start by describing the development of the technologies on which GPU computing is built (parallel computing, parallel graphics systems, programmable shaders, and stream processing), and then we detail how this technology was transferred to Nvidia and other companies and came to be applied to modern machine learning.

• Google Plans its First Data Center in France

  Google is apparently pressing ahead with plans for its first data center in France, with a preliminary consultation set to be held to study the project’s environmental impact. The French National Commission for Public Debate (CNDP), which organizes inquiries into large infrastructure projects with implications for the environment, has announced that a consultation will be held to study the proposed Google data center near Châteauroux, a town in central France.

  Datacenter Dynamics reported last year that Google was seeking to buy 195 acres of land at the Ozans Business Park, south east of Châteauroux. The site was set to cost about $68 million.

• Missouri town fires half its city council over data center deal

  On the same day as the Festus election, voters in Port Washington, Wisconsin, a Milwaukee suburb, where tech giants Oracle and OpenAI are building a $15 billion data center campus, also registered their disapproval by overwhelmingly passing a first-of-a-kind referendum to restrict future projects. At least three other cities across the country will vote on similar measures this year.

  The rout of half the Festus City Council was fueled by a surge in voter turnout and widespread frustration with the data center approval process.

• Cities Face Resilience Challenges in Data Centre Investment Boom

  Data centres are both real estate and infrastructure, the C40 Centre writes. They are infrastructure, as the sector has high barriers to entry (both financial and technical) and offers essential digital services predominantly via predictable, long-term contracts. They are also inherently real estate, given their need for land in specific locations, the specific nature of their architecture, and their financial management (they’re typically leased to third-party tenants, similar to other warehousing and light industrial uses).

  This dual nature of data centres requires the intervention of cities to ensure their development is beneficial to the public, not just tech companies. Investors have already started questioning the likes of Amazon, Google, and Microsoft on their data centres strategies, pointing out that transparency about the impact of new developments on local communities may help with building trust and support and how the lack of such transparency is making it hard to understand how the tech giants may meet their climate targets.

• Meta doubles down on partnership with Broadcom on custom AI processors

  The company said in an announcement that it is committing to an initial deployment of 1 gigawatt’s worth of its Meta Training and Inference Accelerators, a custom-designed chip for AI workloads that Meta runs in its own data centers. Companies use “gigawatts” to measure the scale of AI hardware orders based on the total electrical power required to run them, as opposed to a simple count of individual chips. Ultimately, the extended partnership envisions Meta deploying multiple gigawatts of the new chips, which are based on Broadcom’s technology.

• Why are 'we' building all these data centres again?

  It is worth asking who will buy goods and services if unemployment increases substantially. If a data centre that replaces 1,000 jobs was built by 100 people, and can be run by 10 people, is that really an improvement from a societal perspective?

• Many states don’t report losses from data center tax breaks, study says • Michigan Advance

  Good Jobs First said in most cases, states are failing to disclose incentives in violation of the Governmental Accounting Standards Board, a private organization that sets financial reporting standards for state and local governments.

  “No form of state spending is more out of control today than data center tax abatements,” Greg LeRoy, executive director of Good Jobs First and primary author of the study, said in a news release. “Hyperscale data centers are not only extractive of electricity, water, and land; they are also undermining public budgets.”

• Despite Soaring Costs, 14 States Fail to Disclose How Much Revenue They Lose to Data Center Tax Abatements - Good Jobs First

  Most of these failures to disclose violate Generally Accepted Accounting Principles (GAAP), as set forth by the Governmental Accounting Standards Board (GASB). Since FY 2017, those principles require most governments to disclose lost revenue when they themselves grant tax abatements. They also require other governments which routinely lose revenue passively — such as local governments that lose a share of the sales and use tax when a state exempts a data center — to also disclose.

  Those are the key findings of “Data Center Tax Abatements: Why States and Localities Must Disclose These Soaring Revenue Losses,” a study just released by Good Jobs First, the national watchdog group on economic development subsidies.

• Landmark data center moratorium passes Maine Legislature

  “What we’re talking about here is an ability for us to absorb and understand the impact of data centers potentially on the State of Maine,” Sen. Mark Lawrence (D-York) said Wednesday ahead of the Senate vote. “The states that have had data centers come in have had tremendous impacts.”

  The bill, LD 307, bans data centers larger than 20 megawatts until November, 2027. It also creates the Maine Data Center Coordination Council, and instructs the council to provide strategic input, facilitate planning considerations and evaluate policy tools to address data center opportunities.

• Maine bans new data centres until November 2027

  Moratorium bills are in progress in 12 other states.

  The reasons are the obvious. The electricity grids can’t handle the sudden new load. Nobody wants to live near the noise. The data centres use all the fresh water — and then go to court to keep their water use secret.

• The US Government Will Ask Data Centers How Much Power They Use

  The letter was sent to senators Elizabeth Warren and Josh Hawley on April 9 by the head of the Energy Information Administration, Tristan Abbey, and comes in response to a previous inquiry from the senators about the EIA’s plans to get more information about data centers. WIRED reported on Hawley and Warren’s letter last month.

• The Future of Everything is Lies, I Guess: Where Do We Go From Here?

  I want you to think about “AI” in this sense.

  Some of our possible futures are grim, but manageable. Others are downright terrifying, in which large numbers of people lose their homes, health, or lives. I don’t have a strong sense of what will happen, but the space of possible futures feels much broader in 2026 than it did in 2022, and most of those futures feel bad.

  Much of the bullshit future is already here, and I am profoundly tired of it. There is slop in my search results, at the gym, at the doctor’s office. Customer service, contractors, and engineers use LLMs to blindly lie to me. The electric company has hiked our rates and says data centers are to blame. LLM scrapers take down the web sites I run and make it harder to access the services I rely on. I watch synthetic videos of suffering animals and stare at generated web pages which lie about police brutality. There is LLM spam in my inbox and synthetic CSAM on my moderation dashboard. I watch people outsource their work, food, travel, art, even relationships to ChatGPT. I read chatbots lining the delusional warrens of mental health crises. [...]

• Many states don’t report losses from data center tax breaks, study says

  The study found that Alabama, Arkansas, Idaho, Iowa, Indiana, Louisiana, Maryland, Missouri, Mississippi, North Carolina, North Dakota, Oklahoma, South Carolina and Utah all failed to report data center incentives, which generally include sales, use and property tax breaks.

  For years, states have used incentives and tax breaks to compete for data centers, sought for their massive investment in construction and equipment. Currently, 38 states offer dedicated tax incentives for data centers, according to the National Conference of State Legislatures.

• Data Center Tax Abatements: Why States and Localities Must Disclose These Soaring Revenue Losses - Good Jobs First

  Fourteen states and scores of localities across the U.S. fail to disclose how much revenue they lose to data center tax abatement programs. Yet such losses are known to be soaring in states that do disclose, with three states already losing $1 billion or more per year.

  Most of these failures to disclose violate Generally Accepted Accounting Principles (GAAP), as set forth by the Governmental Accounting Standards Board (GASB). Since FY 2017, those principles require most governments to disclose lost revenue when they themselves grant tax abatements. They also require other governments which routinely lose revenue passively to also disclose.

• EU denies bowing to tech lobby on data centre green impact

  But in 2024 it took up a submission by tech companies requiring all individual information on data centres be classified, according to an investigation led by Investigate Europe, a journalism cooperative, in partnership with The Guardian, Le Monde and other media.

  As a consequence, information on the precise impact of individual data centres is kept from the public even if demanded through freedom of information requests -- in a possible breach of EU transparency rules, the report said.

• Ypsilanti Township proposes blocking water to UMich data center

  The Ypsilanti Township Board of Trustees approved a resolution Wednesday that could prevent the University of Michigan from drawing the 500,000 gallons of water needed to operate the $1.25 billion data center it plans to construct in partnership with Los Alamos National Laboratory.

• Maine Legislature Passes First-in-Nation Pause on Large Data Centers

  The measure would block the creation of new data centers that draw more than 20 megawatts of power until the fall of 2027 and establish a mechanism to study their impact on the electrical grid, according to The Washington Post. During this moratorium, a coordinating council will convene to produce recommendations and guidelines to shape the future of the state’s policies on large data centers. The council will include government officials, experts and other stakeholders. The bill allocates $95,000 to support the council’s activities, notes The Hill.

• Data Center Bills Led 2026 Virginia General Assembly

  Several bills were introduced to regulate the sector, with some successfully passing, reports the Virginia Mercury. One key measure, Senate Bill 253, focuses on electricity costs. It allows regulators to shift more energy-related expenses, such as power purchased through regional auctions and infrastructure like substations, onto data centers and large industrial users, rather than residents.

• Housing, data center legislation dies on final day of 2026 session

  Another proposal that didn’t make it to the governor’s desk was sponsored by Rep. Josh Bray, R-Mount Vernon, seeking to ensure the costs of infrastructure and service to meet the demands of power-intensive data centers would not be borne by other ratepayers. Bray’s House Bill 593 cleared the House in March but stalled in the Senate.

• Users complain of UK Azure capacity problems

  The Open Sorcerer is not alone. A look at social media has shown plenty of users running into similar problems, with one commenting, "Pushing their datacenters to capacity with no real plan to build out or expand them is just piss poor planning." Another also claimed they'd been "pushed towards other regions such as Sweden."

• HK's data centre boom: Powering innovation or jeopardising climate goals?

  Hong Kong’s data centres consumed 7,131 terajoules of electricity in 2023 – up by more than 75 per cent in just five years.

  If we take the Environment and Ecology Bureau’s data as a reference, greenhouse gas emissions from data centres through electricity use rose by 35.6 per cent, from 680,164 tonnes of CO2-equivalent in 2018 to 922,392 tonnes in 2023. That is equivalent to the annual emissions of roughly 200,000 Hong Kong residents.

• Foo Local political revolts threaten to derail US data center projects — mounting delays are already costing AI hyperscalers billions

  Just this week, a small town in Missouri ousted half of its city council for not doing their due diligence in protecting local communities from the harms of AI data center construction, and they're pushing to remove the rest of them, the mayor included. A resident of Claremore, Oklahoma, was arrested in February for speaking too long during a town hall meeting to discuss a data center project. In Virginia, voter support for data centers has collapsed to just 35% from 69% in 2023, halting efforts to build what would have been one of the largest data centers in the country.

  Suffice to say, while major data center projects are the darling of tech CEOs and many politicians, the general populace is making its voice heard; through collective action, they are putting the brakes on $10's of billions of dollars of investment and derailing the plans of major corporations. As of the time of writing, half of all planned U.S. data center builds have been delayed.

• Many states don’t report losses from data center tax breaks, study says

  “No form of state spending is more out of control today than data center tax abatements,” Greg LeRoy, executive director of Good Jobs First and primary author of the study, said in a news release. “Hyperscale data centers are not only extractive of electricity, water, and land; they are also undermining public budgets.”

• Data Center Tax Abatements: Why States and Localities Must Disclose These Soaring Revenue Losses [PDF]

  Hyperscale data centers are by far the most controversial issue in U.S. economic development today — for their noise, voracious electricity and water consumption, air pollution, and massive footprints. They have also emerged as the nation’s biggest new budget headache for states and localities: Tax-abatement laws written for much smaller data centers, predating massive artificial intelligence (AI) facilities, are now unexpectedly costing governments billions of dollars in lost tax revenue.

  However, Good Jobs First finds that in violation of Generally Accepted Accounting Principles (GAAP), as set forth by the Governmental Accounting Standards Board (GASB), at least 14 states and scores of localities are failing to disclose tax abatement revenue losses they are suffering to data centers.

• My solution for copying backups around the homelab

  I have database servers outside the homelab, as in not in my basement. They are in datacenters. I don’t let them push the backups into the basement. Instead, I let them call home asking for the backups to be picked up. I prefer it that way. As I describe it, it may seem complex to do multiple steps when one step will do. However, this solution promises that the backups are ready and you’re not pulling back a half-completed backup.The process is: [...]

• Want to Resist a Data Center? These Organizers Share How They Did It.

  The prolific construction of massive data centers that house the physical computing power for artificial intelligence (AI) is galvanizing resistance in localities across the United States. Communities are fighting back against the billionaire tech and financial power behind these projects and their numerous harms, from their noise and pollution to their hyperconsumption of water and electricity.

• Everything Is Changing Fast: A Brisk Tour Through Shifting Views

  As I mentioned here recently, one way you can see the new attitude in action is as local campaigns, often successful, to prevent tech corporations from building data centers in their area and ICE from buying warehouses to convert to prisons. There's a quite exhilarating wave of opposition to AI right now. Elites, including a lot of mainstream media as well as the creators and profiteers of the technology, often insist that it's inevitable, all-powerful, and we should just lie down and let the tanks of big tech run us over. Again. The public is not having it, and whether they're seeing the slop and sludge AI spreads, the threats to jobs, the corrosive effect of chatbots on vulnerable users, the potentials for profound dangers, or the reckless amorality of those in charge of the technology, they are skeptical about the promises. There were three attacks recently, including two targeting the home of OpenAI CEO Sam Altman and one the home of an Indianapolis city councilman who voted for a data center there.

• North Dakota has the cheapest electricity in the country, data shows

  There are growing concerns in North Dakota and elsewhere that large power users, like data centers, could drive up the cost of electricity for residential consumers. MDU and Otter Tail representatives said they hope the new industries will actually benefit existing power customers over time.

  Otter Tail’s Hoff said the utility’s customers are not going to have to pay more for electricity to compensate for new infrastructure needed by a single large customer, like a data center.

• AI backlash is coming for elections

  More than 60 percent of both Republicans and Democrats polled by Ipsos earlier this year agree that the government should regulate AI for economic stability and public safety, and that the technology’s development should slow down. Still, “when you just ask folks, ‘what’s on your mind?’ AI and data centers aren’t rising to the top of the list — at least not yet,” says lead pollster for Ipsos Public Affairs Alec Tyson.

• A Data Center Is Getting a $77 Million Tax Break. It Promises to Create 1 Job.

  On a February morning in 2024, a little-known agency in Rockland County held a public hearing on a proposed subsidy for the expansion of a JPMorganChase data center in Orangeburg, near the New Jersey border. In return for nearly $77 million in tax breaks, the project promised to create exactly one permanent job.

  No one showed up. After 20 minutes of silence, an agency official called the meeting to a close. Two weeks later, the subsidy deal was approved.

• 0

  Is this the end of the PC gaming handheld dream? With shortages everywhere, we're now seeing the Lenovo Legion Go S price skyrocket. With the Steam Deck out of stock pretty much everywhere, and the prices of other handhelds much higher anyway - the AIpocalypse continues to claim victims as AI data centres buy up everything.

• Now generally available: Red Hat Confirmed Sovereign Support drives digital autonomy for global enterprise

  For organizations in highly regulated industries, achieving true digital autonomy requires more than just localizing physical data centers. It demands developing a strong sovereign support model to help build, deploy and run critical workloads in-jurisdiction. This requires verifiable control over data flows, encryption, software supply chains, and operational transparency. 

• When Your Digital Life Vanishes

  Technological progress may be increasing our exposure. A.I. agents are becoming notorious for accidental deletions, while the proliferation of data centers has wildly inflated the cost of storage. And, despite exponential growth in capacity, the average hard drive’s life span remains just under seven years. Considering the hundreds of zettabytes of data estimated to exist in the world, it’s as though a million Libraries of Alexandria were saved from annihilation solely by hamsters on wheels.

• Luddites and AI datacenters

  Like many people, I have heard a lot about Luddism and Luddites, but only in the context of it being a general term for someone who is anti-technology. I was interested in learning more about the actual historical movement: what kind of people participated, what it was, and what it accomplished. I read Merchant’s and Mueller’s books, plus others2, to try and figure all of this out. Who were the actual, historical Luddites? What can we learn from them about burning down datacenters?

• Meta breaks ground on over $1 billion data center in Oklahoma's Tulsa

  This ‌will be Meta's first data centre in Oklahoma and it will be located at the Fair ⁠Oaks Innovation Park, spanning more than 2 million square feet, according to the company.

• Data center growth threatens air quality and progress toward cleaner energy grid

  The research, building on new government data, tracks a dramatic slowdown in the previously accelerating pace of coal plant retirements as utilities respond to rising electricity demand from data centers and other sources. Compared to 2022, when the country was retiring power plants at a pace that would lead to a coal-free grid by 2040, the current pace of plant retirements would keep some of our dirtiest ones online until 2065. Power generation from coal-fired power plants increased by 13% in 2025 as U.S. electricity generation hit an all-time record.

• Why cloud computing still runs on coal and gas

  The United States has more data centers than anywhere else in the world and the extra energy demand is straining transmission grids and driving up the cost of electricity. Struggling grid operators are turning to polluting fossil fuels to quickly supply the power. In some cases, they are considering nuclear energy, or even bringing back oil, gas and coal power plants that had been slated for closure.

  According to recent analysis by Reuters, the country's biggest power grid postponed or canceled the planned closure of 60% of its fossil fuel plants last year. PJM Interconnection covers 13 eastern states, including Virgina, the "data center capital of the world." Eleven of the plants due to close were so-called "peakers," which are brought online during periods of spiking electricity demand.

• Minnesota's energy system faces headwinds

  Maybe we’ve grown too complacent. After years in the doldrums, electricity demand is set to grow here in Minnesota and across the U.S. amid booming demand for AI “compute” and the slow but steady electrification of transportation, industry and heating.

  Google insists it’ll pay for the electric infrastructure needed to power the massive data centers it plans to build near Duluth and Rochester. But utility bills are rising faster than inflation already, so it’s understandable that many Minnesotans are in “believe it when we see it” mode. Xcel Energy electric customers saw rates jump 5.2% last year, with more increases on the way.

• Indigenous Activists Decry “Data Colonialism” of AI Boom in Their Communities

  The artificial intelligence industry’s data center boom is the latest chapter in a long history of environmental racism and resource exploitation in vulnerable Native communities, says Oglala Lakota and Northern Cheyenne activist Krystal Two Bulls, the executive director of Honor the Earth, an Indigenous-led environmental justice organization that is tracking over 100 proposed data center projects on tribal and rural lands. We speak to Two Bulls about the myriad impacts of what she calls a “modern-day iteration” of “settler colonialism,” including noise pollution, cancers and respiratory illnesses, water depletion, energy grid overload and even “ecological collapse.” As tech companies set their sights on Indigenous lands, Two Bulls says, “We’re always the one that ends up having to sacrifice our relationship to land, air, water, our communities and our nonhuman relatives.”

• Europe risks falling behind US, China on AI data centre build-up, Nokia CEO says

  Nokia's CEO raises an alarm about Europe's lagging status in the development of AI data centers, warning that inadequate infrastructure and low investments could push businesses towards the more adaptive environments of the US and China. He also points out the ongoing regulatory and energy supply challenges.

• Michigan’s gubernatorial hopefuls share approaches to AI data center regulations at press forum

  Michigan Advance asked the candidates about the rising tide of opposition to data centers, which is becoming more organized, more forceful and uniting rural conservative Republicans and urban progressive Democrats. In what ways would they regulate and rein in data centers, and would they support an outright ban on new facilities?

  Here’s how the candidates responded. Their responses have been edited for clarity and brevity: [...]

• Community Votes to Deny Water to Nuclear Weapons Data Center

  America’s nuclear scientists plan to break ground on an AI data center next week, but the Township where it’s being constructed just put a 365 day hold on providing it with water.

• New Gas-Powered Data Centers Could Emit More Greenhouse Gases Than Entire Nations

  New gas projects linked to just 11 data center campuses around the US have the potential to create more greenhouse gases than the country of Morocco emitted in 2024. Emissions estimates from air permit documents examined by WIRED show that these natural gas projects—which are being built to power data centers to serve some of the US’s most powerful AI companies, including OpenAI, Meta, Microsoft, and xAI—have the potential to emit more than 129 million tons of greenhouse gases per year.

• Meta cuts 10% of staff, cancels 6,000 open roles in AI efficiency push

  The reduction in headcount is Meta’s largest since the 21,000-plus job cuts the company pushed through in 2022 and 2023, a period Zuckerberg called a “year of efficiency.”

  Meta’s capital expenditure guidance for 2026 sits as high as $135 billion, nearly double 2025’s roughly $72 billion, with the bulk earmarked for data centers, custom silicon and other AI infrastructure. Meta still trails OpenAI Group PBC, Google LLC and Anthropic PBC in generative AI.

• Investigation: RAM prices are falling. Don’t fall for it

  AI data centers are gobbling up memory manufacturing capacity. The trend is expected to continue for quite a while, and there’s almost no capacity left for consumer-level RAM and storage fabrication. As a result, prices for RAM, finished computers, storage, game consoles, phones, even products as innocuous as SD cards are rising higher and higher. It’s a terrifying development for PC enthusiasts on the hunt for more performance without spending a fortune.

• Frederik Braun: Multiple things can be true at the same time

  AI is also bad for the environment, due to their excessive electrical power requirements. AI is making hardware expensive. Memory and storage are no longer affordable, most of it going straight to people who are buying it to fill a datacenter that has not yet been constructed and with money they don't have.

  AI has plagiarized all the works that are (publicly) available on the internet. It destroys and devalues creative work. AI is also unpredictable. Given the stochastic nature of these systems, there is no way to make them reliable.

• Google's $15 billion data centre in Vizag to host 5-GW capacity

  Google Cloud is building a massive AI-ready data centre in Vishakhapatnam. It will have a capacity of 5 gigawatts. This investment is Google's largest in India. The project will be completed over five years. The new campus will be part of Google's global network. It will enhance data transfer capabilities.

• Tech Companies Are Using Insidious Tactics to Build Data Centers on Indigenous Lands, Activists Say

  “We were there in Tulsa to pass the moratorium last month, and we were there this morning in Oklahoma City to show that the community does not want these data centers,” Ash Leitka, director of Honor the Earth’s Department of Sovereignty and Self Determination told Futurism. “We want clean water. We want privacy and security. We want jobs and economic stability, all of which hyperscale data centers threaten and negatively impact. The falsely generated demand for AI and AI infrastructure truly is a death cycle.”

• Just 11 AI Data Centers Could Belch More Fumes Than Entire Countries

  The magazine examined emissions estimates provided by gas power projects that are being built to supply energy to the data centers. Construction of these sprawling facilities has surged to meet the demands of the AI industry, and to get them online as soon as possible, many of the newly built data centers are relying on gas power. This strategy means data centers don’t have to wait to plug into local power grids and be saddled with the controversy that invites, such as surging energy bills. Gas turbines can be trucked in on-site and start providing power almost immediately.

• In wartime, megascale data centers may make way for distributed architectures

  Among the targets in the war between Iran and the US have been data centers. AWS was hit by drones, and Iran has threatened to target US tech. This piece makes the point that these buildings don’t just store vast amounts of civilian customer data: increasingly, they store military data, too. That make them an even more attractive target and makes the security consequences of an attack that much worse.

  Meanwhile, data centers — including here in Pennsylvania, where I live, as well as Chile, India, and many other places around the world — have been the cause of significant objections from local populations. They push energy costs up, have a serious environmental footprint, and can even change the local climate.

  So why have these giant megascale data centers at all?

• Janet Mills vetoes Maine data center ban

  Gov. Janet Mills vetoed a bill Friday that would have made Maine the first state to impose a moratorium on large new data centers, even as local ‌opposition to the electricity-hungry facilities grows.

• US data centers to get power from Hyundai's 684MW energy system deal

  The contract involves delivering equipment capable of producing a total of 684 megawatts of electricity.

  With a value exceeding $400 million, this agreement stands as the largest engine supply deal ever achieved by the company. It also marks a key step in Hyundai’s efforts to expand its presence in global energy infrastructure, particularly within the rapidly growing data center sector.

• Finland to keep election data at home after all

  The Ministry of Justice says a new electoral information system will be hosted in a data centre in Finland. Following a competitive bidding process, the Canadian IT firm CGI was selected to supply the centre.

  The ministry said the move is driven by the need to keep elections secure and running reliably.

• Why Maine Governor Vetoed Landmark Data Center Pause

  Maine Conservation Voters Executive Director Maureen Drouin stated that Mills’ veto sides with developers. “Across the country, the development of large-scale data centers has far outpaced the ability of policy and lawmakers to properly regulate them and establish sensible protections,” Drouin said. “Maine had a chance to push pause and establish the right regulatory framework to protect its people, their wallets and the environment from polluting, resource-hungry data centers. This veto flies in the face of that responsibility and the bipartisan will of the Maine Legislature, passing the buck to the next Governor to rein in large-scale data centers after they’ve arrived.”

• AI Race Pressure: Microsoft, Meta & 98 Firms Cut Jobs | Tech Layoffs

  According to Layoff Dot FYI, which tracks employee layoffs in the technology industry, 98 companies have announced their intention to reduce more than 93,000 employees this year. OpenAI has changed its plan for a very large expensive infrastructure project. The company has abandoned its plan to build and operate its own data centers. Instead, the company will rent servers from existing cloud companies. This will have some impact on its balance sheet. Still, it may have to spend ₹56 lakh crore.

• Ubuntu Plans to Land AI Features Throughout 2026

  Ubuntu powers the majority of the world's AI workloads today. As artificial intelligence moves from big data centers to our personal laptops, many users feel nervous. They worry about privacy, system bloat, and losing control over their machines.

  To address this, Canonical developers are now planing to make AI development safe and fast without forcing it on people who don't want it. They are taking a "principled and focused" approach that prioritises local privacy and user choice.

• Oracle plans to power its New Mexico DC with fuel cell farm

  This week, the database giant announced an expanded collaboration with Bloom Energy to deploy 2.45 gigawatts — the equivalent of two or three nuclear reactors — of fuel cell generators alongside the datacenter complex. Previously, Oracle had committed to purchasing 1.2 gigawatts of Bloom fuel cells with the option to expand to 2.8 gigawatts.

  Fuel cells are one of several energy sources hyperscalers have adopted to supplement grid power in areas where local utilities either can't or won't meet demand on their own.

• WI Vote Signals New Data Center Regulatory Model

  The Public Service Commission (PSC) of Wisconsin has approved significant modifications to a proposed data center tariff from We Energies (Wisconsin Electrical Power Company, part of WEC Energy Group). It requires large-load customers to cover the full cost of generation and grid infrastructure serving them.

• Utah TRIGA microreactor to power AI data center for first time

  The University of Utah is preparing an unusual nuclear experiment this summer. Its TRIGA research reactor will generate electricity for the first time in its 50-year history. That electricity will power a small AI data center on-site.

• Stop the AI Build-Out, Start the Fight

  The grassroots resistance to artificial intelligence data centers that is springing up in communities across the country outlines the kind of working-class coalition many of us on the Left have always dreamed of — a diverse, nonpartisan, top-bottom movement against Big Tech billionaires that has the potential to reshape American politics in incredibly positive ways.

• CA Soon May Have Over 300 Data Centers. Locals Worry About Water Supply Threats.

  What’s going on is the second-largest new data center being considered statewide, which would be less than half a mile from Padilla’s stucco home in the center of Imperial Valley. If finished by 2028, as the developer expects, the at least 950,000-square-foot, two-story data center could be the largest operating statewide, taking up 17 football fields’ worth of land.

  The roughly $10 billion, 330-megawatt data center would require 750,000 gallons of water a day to operate, said developer Sebastian Rucci, who insists electricity and water costs won’t rise due to the data center.

• US university's research reactor to generate power in a first

  The 2-3 kW output from the reactor will be used to power a high-performance GPU node to that will execute a live artificial intelligence (AI) workload.

  The recent surge in AI applications has increased demand for data centers in the future. With high-performance GPUs also consuming significant energy, electricity demand is expected to double by the end of this decade.

• The Impossible Things We Have to Believe

  In a time of energy shortage and, as noted, climate challenges, we have to accept that the world is going ALL IN on massive AI data centers that slurp up our water and emit vast amounts of CO₂ and gigawatts of heat. Apparently THIS is a priority!

• When looking at data centers, polling says transparency is top of mind for Michiganders

  In a poll of 653 Michigan residents, commissioned by the progressive advocacy group Progress Michigan and conducted by Public Policy Polling, 69% of respondents said they would strongly support additional transparency laws after being informed that data centers receiving state tax breaks are not required to publicly disclose details related to their water and energy use, infrastructure costs or commitments to job creation.

• City says Project Blue must stop using Tucson Water for data center construction

  City Manager Tim Thomure told Beale Infrastructure officials in an letter on Monday that Tucson does not support the construction of the data center and the project is not eligible for any city resources, including water.

• Study finds AI server farms creating their own local heat islands

  The same AI bubble promising to reshape the future may also be quietly reshaping the climate. A new study suggests massive data centers are generating localized "heat islands," raising temperatures around them and potentially compounding the effects of global warming in already vulnerable communities.

• AI data centers appear to be creating their own microclimates

  Researchers from multiple institutions, including the University of Cambridge and Nanyang Technological University, used satellite data from that time to assess rising land surface temperatures at AI data centers worldwide. After conducting an analysis, they estimated that surrounding surface areas typically increase by an average of 2 degrees Celsius — or 3.6 degrees Fahrenheit — once AI centers start operating, suggesting that the data center heat island effect “is real and significant, especially in the context of global warming and climate transformation.” Overall, “our results show that the data heat island effect could have a remarkable influence on communities and regional welfare in the future,” researchers said in the study.

  The implications also suggest that building AI data centers in heat-stricken areas of California could have dire consequences on local communities.

• (PDF) The data heat island effect: quantifying the impact of AI data centers in a warming world

  The strong and continuous increase of AI-based services leads to the steady proliferation of AI data centres worldwide with the unavoidable escalation of their power consumption. It is unknown how this energy demand for computational purposes will impact the surrounding environment. Here, we focus our attention on the heat dissipation of AI hyperscalers. Taking advantage of land surface temperature measurements acquired by remote sensing platforms over the last decades, we are able to obtain a robust assessment of the temperature increase recorded in the areas surrounding AI data centres globally. We estimate that the land surface temperature increases by 2°C on average after the start of operations of an AI data centre, inducing local microclimate zones, which we call the data heat island effect. We assess the impact on the communities, quantifying that more than 340 million people could be affected by this temperature increase. Our results show that the data heat island effect could have a remarkable influence on communities and regional welfare in the future, hence becoming part of the conversation around environmentally sustainable AI worldwide.

• Rage Against The Data Center

  It’s short because the videos below are powerful examples of the raging backlash against Big Tech, an issue I began writing about last December, in “The Data Center Backlash Is Global.”

  Since then, the opposition to Big Tech, AI, and data centers has grown faster than I ever imagined. I have been covering land-use conflicts over alt-energy projects for 16 years. In that time, I’ve interviewed dozens of people all over the world about their opposition to solar, wind, and battery projects. I’ve documented hundreds of examples of rejections or restrictions on alt-energy projects in the Renewable Rejection Database.

  But the rage against data centers is different.

• The Environmental Costs of The AI Boom

  The computing power required to train and run AI systems is a major source of this environmental burden. Thousands of GPUs (Graphic Processing Units) power many advanced models, and Yale Climate Connections reported that some systems rely on 100,000 GPUs, drawing as much electricity as a small city. The outlet also reported that AI data centers now consume 4.4 percent of all US energy, with projections that AI alone could use as much electricity as 22 percent of US households by 2028. A February 27, 2026, report in the San Francisco Examiner linked the AI data-center boom to rising electricity demand and new power plant development, much of it fueled by natural gas rather than renewables. Yale Climate Connections similarly noted that, although AI may support climate research and environmental monitoring, the technology depends on data centers that consume enormous amounts of water and energy.

• Denmark presses pause on new data center grid connections as total requests hit 60 GW — Nordic nation is the latest to put the brakes on AI buildouts

  For reference, Denmark's current power peak power consumption is estimated around 7 GW, so it's easy enough to understand why the action was necessary. The pause is meant to last three months, though there's speculation that that period might be extended given the need to reprioritize many of the requests.

• Denmark faces data center reckoning amid power grid strains

  In March, Denmark's state-owned grid operator Energinet introduced a temporary pause on new grid connection agreements due to an "explosion" in capacity requests, a spokesperson told CNBC. Around 60 GW of projects are waiting for connections. That far exceeds Denmark's peak electricity demand of around 7 GW. Data centers account for nearly a quarter (14 GW) of the 60 GW potential new grid connection projects, the spokesperson said.

• University Claims Withholding Water From Nuclear Weapons Data Center Is 'Unlawfully Discriminatory' to Data Centers

  The University of Michigan did not return 404 Media’s request for comment.

  Ypsilanti Township has been fighting the proposed datacenter for more than a year now. Data centers are wildly unpopular in the United States. They often cause noise pollution, affect water quality, and drive up utility bills for their neighbors. Local opposition to the Ypsilanti Township data center has been compounded by its connection to America’s nuclear weapons industry.

• Utah data center water fight intensifies

  Nearly 3,900 protests were filed over water concerns.

  Residents fear impacts on the Great Salt Lake and local water supplies.

• Maryland citizens slapped with $2 billion power grid upgrade bill for out-of-state AI data centers — state complains to federal energy regulators, says additional cost breaks ‘ratepayer protection pledge’ promises

  The Maryland Office of People’s Counsel (OPC), a state agency that represents its utility consumers, filed a complaint before the Federal Energy Regulatory Commission (FERC) regarding PJM Interconnection, LLC’s plans to charge it $2 billion of the $22 billion it spent to upgrade its grid to accommodate increasing demand from data centers. According to the OPC’s press release, this $2 billion bill will cost the state’s consumers an extra $1.6 billion in the next ten years alone — that means an extra $823 million for residential (approx. $345 per customer), $146 million for commercial (approx. $673 per customer), and $629 million for industrial customers (approx. $15,074 per customer).

• AI data center firm IREN's stock soars after it strikes $2.1B deal with Nvidia

  The partners plan to deploy up to 5 gigawatts of Nvidia’s DSX-branded infrastructure in the neocloud company’s global network of data centers. Investors reacted positively to the news, with IREN’s stock gaining more than 27% at one point today, before settling back to an 8% gain in late trading.

• Senator at center of Utah AI data center debate gets physical, slaps phone out of reporter’s hand — reporter covering cases of harassment against his business

  A witness who saw the interaction called the authorities, resulting in the filing of a police report. Once the situation had cooled down, law enforcement officers told Wang that the Senator was apologizing for the incident. However, the crew was also served with a trespassing notice, warning them that they are prohibited from going on the property or premises of the business for a year.

• Residents Furious After Their Town Board Rejected an OpenAI Data Center, But a Billionaire Developer Forced It Through Anyway

  Residents in Saline Township, Michigan, however, thought they had avoided the drama after their township board and planning commission both voted to decline a 21 million square foot data center in their backyard. It was exactly what Saline’s 2,883-some residents wanted. Unfortunately for them, the data center developer soon sued the tiny township, Fortune reported, which was ultimately bullied into accepting the $16 billion development.

  Back in September, Saline’s planning commission rejected the request to rezone 575 acres of farmland for the data center, proposed by the company Related Digital, a subsidiary of a real estate conglomerate owned by billionaire Steven Roth (who’s been in the news lately for other reasons.)

• Data centers: Tech boom with downsides

  This is why data centers in Germany are classified as part of the country's critical infrastructure and afforded special protection. In March 2026, the federal government published a new national Data Center Strategy, illustrating how significant they are. It plans to double Germany's data center capacity by 2030 and will also aim to reduce its dependence on non-European providers.

• Setback for Microsoft's Africa cloud ambitions

  In May 2024, Microsoft partnered with UAE-based AI firm G42 to invest US$1-billion in a data centre in Kenya as part of its efforts to expand cloud computing services in East Africa. The project was announced during Kenyan President William Ruto’s state visit to Washington under the Joe Biden administration.

• A data center drained 30M gallons of water unnoticed — until residents complained about low water pressure

  The neighbors of a data center in Georgia are steaming after they discovered the facility had sucked up nearly 30 million gallons of water — without initially paying for it.

  Outrage started bubbling up last year when residents of an affluent subdivision named Annelise Park in Fayetteville, Georgia, noticed their water pressure was unusually low.

  When the county utility investigated, officials discovered two industrial-scale water hookups feeding a data center campus located 20 miles south of downtown Atlanta. One water connection had been installed without the utility’s knowledge, and the other was not linked to the company’s account and therefore wasn’t being billed.

• AI data center project secretly sucked 29 million gallons of water over 15 months before detected by residents complaining about low water pressure — officials refuse to fine builders of massive 6.2 million-square-foot facility over unauthorized water use

  QTS told Politico the 29 million gallons were consumed during temporary construction activities, including concrete work, dust control, and site preparation. The company markets a "closed-loop" cooling system for its data centers, which recirculates the same water rather than drawing from the municipal supply. Once operational, QTS said its facilities would only require water for domestic needs like bathrooms and kitchens.

  However, the discrepancy between QTS’s stated and actual water usage remained undetected for months, with Politico reporting that the county’s water system director, Vanessa Tigert, attributed the oversight to a procedural error during the county's transition to a cloud-based metering system.

• Those damn data centers | Factor This Brief

  Meanwhile, online rhetoric on the topic is growing increasingly violent. Deteriorating sentiment towards AI has been accompanied by a rise in direct threats against individuals perceived as driving the technology forward, warns one nonprofit intelligence firm, as well as threats against policymakers and corporations involved in developing new data centers. Physical security, it turns out, might be just as important to computing projects as cybersecurity.

  So how does this all shake out? Your guess is as good as mine, but one thing is for sure: the temperature is rising, and, as we all know, cooling things down can get very expensive for data centers.

• AI data centers face increasing complaints about inaudible but 'felt' infrasound — citizens complain high- and low-frequency sounds do not register on decibel meters but cause adverse health effects

  Data center projects have faced resistance from residents and communities over their impact on power prices, but another complaint is being raised more frequently — noise pollution. One form of sound pollution is called infrasound, which is inaudible to humans but can be felt, and some claim it causes headaches, insomnia, nausea, and anxiety. Then there's the normal garden-variety sound pollution. The Environmental and Energy Study Institute (EESI), a non-profit organization, said that high- and low-frequency sounds emitted by these industrial sites can be heard and felt for hundreds of feet in surrounding areas, with noise levels reaching as high as 96dB for 24 hours a day and seven days a week.

• AI data center developers target rural territory to bypass city construction bans and regulations — rural locations allow sites to bypass city council approvals, rezoning votes, land-use reviews, and reduce public scrutiny

  Developers are showing increased interest in unincorporated county land for putting up data centers, reducing regulatory friction, and allowing construction to proceed at a much quicker pace. According to a SemiAnalysis post on X, since these parcels of land are outside city limits, they do not have to go through city or town council approvals, rezoning votes, and land-use reviews. Many projects are experiencing delays because of pushback from the communities surrounding potential projects, with many jurisdictions approving moratoriums on data center construction, some of them permanent bans.

• Workers Have a Secret Weapon Against the AI Build-Out

  Left to their own devices, the likes of Altman and the other titans of the tech billionaire elite would like nothing more than to remake society into one where all else is subservient to their profit margins. Amazon founder Jeff Bezos is currently on a fundraising spree to bring in $100 billion to Project Prometheus, of which he is the co-CEO. Describing itself as a “manufacturing transformation vehicle,” the fund aims to buy up, hollow out, and implement AI and other forms of automation across manufacturing firms in various sectors.

  This future is being actively contested on multiple fronts. Communities across the country have banded together to fight the construction of data centers in their towns and have won moratoriums blocking future construction in others. Unions are beginning to bargain around AI and workers are developing tactics to fend off the impacts of AI in an increasing number of workplaces. In 2024, automation, though not directly related to AI, was a central issue of the three-day strike by the International Longshoremen’s Association (ILA) that shut down ports across the East and Gulf Coasts.

• Banks turn data centre loans into 2008-style financial time bombs

  An asset bubble is an opportunity. By late 2024, financiers were already constructing collateralised GPU obligations, lending against video cards.

  CoreWeave borrows against its Nvidia GPUs — which have a useful life of maybe two years. CoreWeave then uses the loan to buy more Nvidia GPUs. This is fine.

  Investment banks have been hard at work creating “novel types of debt structures” since then. What a great phrase that is.

  Now banks and private lenders find themselves stuck with a large pile of horrifyingly rickety loans for data centres.

• How a data center derailed $240,000 for affordable housing in Wiscasset

  Yet the land captured public attention in full force when plans to build a $5 billion data center were made public in September of 2025. Davis, who lives along the Back River about a mile and a half from the site, became concerned as soon as he found out about the data center. Protect Wiscasset came together last fall to organize against a development.

  But Davis would soon learn that the data center was just part of the story. As the facilities are planned, often without much public transparency, communities also find themselves dealing with the opportunities lost to the possibility of a sprawling data center.

• Local data center battles in Kentucky are contentious. They’re also inspiring runs for office

  The more than three-hour-long meeting in April, archived on social media, heard from people with concerns ranging from worries about the use of “prime farmland” for the project and whether the costs of electricity infrastructure upgrades needed to serve data centers would be burdened by other ratepayers.

• Charlotte Weighs Pause on Data Centers Amid Community Pushback

  The Charlotte City Council is moving to temporarily halt new data center projects as residents raise concerns about noise, water use and rising electricity demand, according to Axios Charlotte. Council member Dimple Ajmera is pushing for a 90-day pause, warning that data centers can currently be built without rezoning hearings in several districts. Construction is already underway on five large facilities in University City.

• Should data centers pay for grid upgrades? Oregon regulators think so

  The Oregon Public Utilities Commission recently approved several elements of Portland General Electric’s proposals for charging customers based on their contribution to growth – which is meant to ensure that data centers will pay for new infrastructure that supports their growth.

  “The decision reflects an important step toward balancing growth, reliability and affordability for Oregon customers,” said John McFarland, Chief Customer Officer. “As energy demand grows, it is critical that the costs of new infrastructure are allocated fairly and transparently. Our focus remains on maintaining reliable service, supporting economic development and protecting residential and small business customers from unnecessary cost impacts.”

• Microsoft's massive Kenya AI data center would require switching off 'half the country' to meet power requirements, government says — $1 billion project stalls over capacity disagreements and lack of infrastructure

  The project, announced in May 2024 during Ruto’s visit to Washington, was supposed to bring a geothermal-powered data center to the Olkaria region in Kenya's Rift Valley. G42 was to lead construction, with the facility running Microsoft Azure in a new East Africa cloud region. The first phase targeted 100 megawatts of capacity and was expected to be operational by this year, with a long-term goal of scaling to 1 gigawatt.

  President Ruto isn’t exaggerating about shutting off half the country’s power. Kenya’s total installed electricity capacity sits between 3,000 and 3,200 megawatts, and peak demand reached a record 2,444 megawatts in January, according to data from KenGen, the country’s government-owned electricity producer.

• SoftBank to manufacture its own batteries with water-based tech to power AI data centers — targets gigawatt-hour-scale production by 2028

  SoftBank has announced that it will begin manufacturing battery cells and energy storage systems at its facility in Sakai, Osaka, targeting gigawatt-hour-scale production by the fiscal year ending March 2028.

• AI datacenters in space do not have a cooling problem

  Of course, you don’t just need radiators to put a datacenter in space. You need a similar quantity of solar panels, the GPUs themselves, and all kinds of other supporting equipment. If a GPU dies in an Earth datacenter, you can go in and swap it out; if it dies in space, you just have to leave it dead and keep going with less capacity.

  It’s still wildly impractical to build AI datacenters in space. But it’s not impossible, and it’s certainly not impossible because of the cooling, which is a relatively minor component of the total mass that would have to be launched into space.

• Silicon Valley Needs Your Electricity

  We regret to inform you that nearly 50,000 residents of the Lake Tahoe region are about to lose their power source. NV Energy, the Berkshire Hathaway-owned utility that has supplied wholesale electricity to the area for decades, has informed Liberty Utilities — the small California company that retails the power to those 50,000 households — that it will stop providing service after May 2027. The reason, helpfully provided by NV Energy itself, is that the utility needs the capacity for data centers. Google, Apple, and Microsoft are building or expanding facilities in the Tahoe-Reno Industrial Center east of Reno, and the load growth from these projects is, in NV Energy‘s own filings, projected to constitute 75% of major-project demand in Northern Nevada through 2033. Data centers were 22% of Nevada’s electricity in 2024. They are projected to reach 35% by 2030.

• Nearly 50,000 Lake Tahoe residents face power loss as utility redirects lines to data centers

  NV Energy, the Nevada utility that has supplied the bulk of Lake Tahoe’s electricity for decades, told Liberty Utilities—the small California company that services the region—that it will stop providing power after May 2027. The reason? NV Energy needs the capacity for data centers. As in: the energy supplier for the Lake Tahoe region is telling the utility company that it has less than a year to find another power source.

• Data Center Opponents Claim Victory in VA

  Virginia has become the center of the U.S. data center industry, with more than 660 facilities already operating and hundreds more planned. One of the biggest proposed developments, the 2,100-acre Digital Gateway project in Prince William County, has now stalled after strong public opposition, reports Bloomberg. The project, backed by Compass Datacenters and Brookfield Asset Management, was approved in 2023 despite a divided county vote and would have created around 23 million square feet of data center space, making it one of the largest hubs in the world.

• Kevin O'Leary-Backed AI Data Center Transforms Rural Utah Into a National Flash Point

  This hyperscale data center is set to consume 9 gigawatts of power — more than double the total energy consumption of the entire state of Utah — relying solely on natural gas-fired energy generation. Reports indicate this would increase the state’s emissions by 50%. An estimated 16.6 billion gallons of water per year would also be required to operate the facility’s turbines. The quantity of water isn’t the only concern, but also the source. Stratos would have to tap water sources whose diversion environmental groups warn will impact the Great Salt Lake, the largest saline lake in the Western Hemisphere and a critical ecological hub for millions of birds and other species. Warmer winter temperatures have reduced the desert state’s snowpack, the primary water source for residents and the Great Salt Lake. As Utah’s water crisis deepens, the shrinking lake is approaching record-low water levels.

  But the data center is now facing its first major wave of organized resistance.

• Google reportedly in talks with SpaceX to launch its orbital data centers — partnership could mark a historic turning point and boost upcoming IPO

  This news has the potential to boost the impending SpaceX IPO to infinity and beyond. That offering is expected to be the largest of all time, and was already expected to reach stratospheric levels of $1.5 to $1.7 trillion. As of this writing, neither company has offered any comment on the presumably ongoing negotiations.

• Keeping the doors open | The Tor Project

  We're Unredacted, a US-based 501(c)(3) non-profit. We build and operate Internet infrastructure that helps people reach the open Internet and protect their right to privacy. We do this by operating a network of over 300 servers around the world. We're a way through when the front door is locked, and a place to communicate when the public square isn't safe. Most of the work is invisible: datacenter work, hardware, automation, open source software, bandwidth, abuse queues, monitoring alerts, and the late nights spent keeping all of it online.

• Norway, facing an energy crunch, hunts for new solutions

  Now, a few miles up the river, Google is building a $700 million data center outside the city of Skien. The amount of energy needed to cool its high-performance servers will be immense. One media report raves: “This decision highlights Norway’s growing attractiveness as a hub for data centers, driven by its robust infrastructure and abundant renewable energy resources.”

  The problem is, those abundant renewable energy resources don’t feel all that abundant anymore. Norway’s growing thirst for energy has tipped it from huge surpluses to growing deficits. There are no more hydropower sources left. Many experts suggest wind power is the best solution – with one proposed project above Porsgrunn in Lannerheia. But those efforts have been shut down by voters, deeply unpopular.

• Nobody Wants Data Centers in Their Backyard

  Anew Gallup poll confirms that most Americans hate the prospect of data centers coming to their communities. This revelation has been met with some anger among the tech crowd and its extremely online fans, but the opposition is entirely predictable for a few reasons worth reviewing, even if they are obvious: [...]

• xAI grows data center gas fleet despite legal scrutiny – This Week in Cleantech

  This week’s episode features special guest Molly Taft from WIRED, who discusses Elon Musk’s AI company xAI’s decision to add 19 natural gas turbines to its Colossus 2 data center in Southaven, Mississippi.

• Kevin O’Leary’s giant Utah data centre: will it happen?

  The people who live there? They hate it.

  The planned data centre is supposed to use 9 GW of power by itself. The whole state of Utah currently uses 4 GW.

  The data centre doesn’t have a customer signed up. That’s what did in the President Donald J. Trump Advanced Energy and Intelligence Campus in Texas last month.

  And giant data centre projects don’t have a great track record of getting finished in the past couple of years.

• AI data centers require 36 times more fiber than designs with standard servers — severe glass shortages push cable lead times out to a full year

  The scale of demand from AI infrastructure dwarfs anything the fiber industry could ever have planned for. Data center fiber demand grew roughly 76% year-on-year in 2025, according to CRU data cited by various industry publications, and the segment is projected to account for 30% of total global fiber demand by 2027. In 2024, that figure was below 5%.

• Inside the fight over America’s data centers

  Make no mistake: Americans hate data centers.

  A recent poll from Gallup shows 70 percent of Americans oppose a data center in their local area, including 48 percent who are strongly opposed. That 70 percent number is tied to several concerns, environmental questions and quality of life chief among them, and it’s up 18 percent (!) in just two months, when Gallup asked the same question in March.

  Nonetheless, data centers keep going up at a rate that is nothing short of astonishing.

• Americans Oppose AI Data Centers in Their Area

  The data center question parallels the wording Gallup uses to ask about local nuclear power plant construction. In the same March survey, 53% of Americans say they oppose building a nuclear energy plant in their area, far less than the 71% opposed to data center construction. Since Gallup first asked the nuclear power plant question in 2001, the high point in opposition has been 63%.

• AI data centers trigger massive 'irreversible' 76% electricity price spike in largest US region — federal watchdog demands tech giants pay for their own power infrastructure

  Monitoring Analytics, the federally mandated independent watchdog that keeps an eye on the critical PJM Interconnection that distributes power around the U.S., said in a new report [PDF] that a massive 75.5% increase in power costs in the largest region of the U.S. has been directly caused by data centers, and it also blamed the regional market operator for failing to keep up with the rising demand. The price increases have been steep; wholesale electricity prices went up from $77.78 per MWh in the first quarter of 2025 to $136.53 per MWh in the same period of this year.

• Datacenters slurping juice help drive 75% jump in PJM power prices

  “The price impacts on customers have been very large and are not reversible,” the report states, but the bad news doesn’t stop there. “The price impacts will be even larger in the near term unless the issues associated with data center load are addressed in a timely manner.”

  Based on the rest of the report, a timely resolution to the datacenter load issue shouldn’t be expected, at least not in a way that’ll benefit locals.

• Our servers seem to have surprisingly low power consumption

  One of the reason these numbers surprise me is that many of the idle numbers are lower than my desktops. I have a mental image of servers as not being particularly low power or power efficient, just as they're not particularly quiet, but that seems to be wrong. I suppose it's not too odd that people making 1U servers care about power usage and power density, since that's definitely a concern in general in data centers, it just hadn't really occurred to me before.

• Europe built sovereign clouds to escape US control. Then forgot about the processors

  But most datacenters and qualified cloud operators still rely heavily on Intel or AMD processors. And inside those processors sits a computer beneath the computer: management engines operating at Ring -3, below the operating system, outside the control of host security software, persistent even when the machine appears powered off. Under the US Reforming Intelligence and Securing America Act (RISAA) 2024, hardware manufacturers count as "electronic communications service providers" subject to secret government orders.

• AI Has Broken Containment

  The path here was not the inevitable result of some technological, scientific, or economic law. Nor is continuing down it. To the extent we are already living in the AI future, it is the result of a series of calculated decisions by the biggest tech firms and their investors. Silicon Valley has spent ungodly sums on AI and data centers: Microsoft, Amazon, Meta, and Google alone have already spent more on data centers since the launch of ChatGPT than the federal government spent to build the entire interstate highway system. Those expenditures are set to grow, even as consensus opinions on whether all this spending constitutes an economic bubble fluctuate every few months. Meanwhile, AI companies have been hard at work partnering with local and federal government agencies, major colleges and research universities, Fortune 500 companies, and media organizations to weave their products into everyday life.

  All of this spending and all of these partnerships were set in motion years before the technology was actually capable or reliable enough for widespread usage. Now these same companies are barreling forward to consummate their technological revolution. For everyone else, the AI future is beginning to feel less like something you participate in and more like something that happens to you.

• China says 'world's first' offshore wind-powered underwater data center has entered full operation, houses 2,000 servers — 24 megawatt subsea AI facility uses ocean water for passive cooling and offshore wind for power

  The 24 MW facility houses nearly 2,000 servers (including GPU clusters from China Telecom and LinkWise), and is expected to process artificial intelligence, big data annotation, and 5G infrastructure workloads. Unlike conventional land-based data centers that rely heavily on industrial chillers and large HVAC systems to remove waste heat, the Shanghai UDC uses the surrounding seawater as a massive passive heat sink. The servers are sealed inside pressure-resistant subsea modules deployed roughly 35 meters beneath the surface, where stable ocean temperatures continuously absorb heat generated by the computing hardware.

• Living near a data center in Phoenix? Your neighborhood may be getting hotter

  A new study from Arizona State University is the first to directly measure air temperatures both upwind and downwind of data centers in real time, and what researchers found warrants attention from city planners. Heat discharged by data centers is raising air temperatures in surrounding neighborhoods by an average of 1.3 to 1.6 degrees Fahrenheit — and in some cases, by as much as 4 degrees Fahrenheit. The effect was detectable up to a third of a mile, roughly five city blocks, from the facilities’ perimeters.

• Are data centres moving to DC power?

  We’ve recently explored the societal level trade-offs involved in building data centres to run Artificial Intelligence (AI) workloads. A big part of that discussion related to the power required to run these giant rack farms of computers. The topic was explored in an IEEE article on the use of direct current (DC) in data centres, and robustly discussed on Hacker News.

• Former OpenAI Staffers Warn That xAI’s Poor Safety Record Could Complicate SpaceX’s IPO

  In a letter directed to investors published on Tuesday, the ex-staffers highlighted what they describe as “unpriced risks” related to xAI that could complicate SpaceX’s reported plans to raise up to $75 billion as part of its IPO. The rocket company’s private valuation shot up to over $1 trillion after it acquired xAI last year. Musk claimed his rocket company could launch data centers into space for his AI lab, but the letter's authors argue that xAI’s poor record on safety issues could complicate how investors view the combined company as it gets ready to submit its IPO prospectus filing.

• Utah law narrows data center water protests

  His organization filed an objection to the proposed change in water use for the data center, raising concerns including about groundwater and the feasibility of the project, and condemning the “complete absence of details in the application.”

  Powered by its own natural gas plant, developers say the project would produce 9 gigawatts of energy, which is double the state’s consumption. They’ve laid out a plan to use a closed-loop cooling system, recirculating the same water instead of continually drawing more.

• The power play behind Hyperion, Mark Zuckerberg’s colossal data center being built in rural Louisiana

  But a power imbalance is playing out. The tech companies are able to secure humongous tax breaks from local governments that are loath to pass up an opportunity to turn otherwise underutilized land into potential investment dollars and the promise of good-paying jobs. But many times, the number of local jobs isn’t guaranteed, and massive demands on water and power resources burden the infrastructure at unprecedented levels. Property values near data center sites skyrocket, which makes homes harder to purchase for those looking to move in. For existing residents, it also sends things into a bit of disarray: residents’ home values soar, but that also means tax assessments could rise, leading to higher property tax bills.

  The tax breaks and incentives that Louisiana gave Meta are worth billions. The company will pay no sales tax on the tens of billions of dollars’ worth of expensive GPUs and advanced networking gear that will fill the data center’s racks. The local power authority even agreed to build new power lines and infrastructure to accommodate the astonishing amount of electricity that will be needed to feed the data center, which will be one of the largest in the world.

• Meta Secured Over $3 Billion in Tax Breaks in Louisiana to Build a Data Centre

  This is the very same project where Jonathan Weil, of the Wall Street Journal, found “aggressive accounting” that “strains credibility”. Neither of these advantages would be possible for a less-resourced competitor. Meta is a company so rich it benefits immensely without carrying nearly as much risk as the scale of this project would imply.

• Public officials reconsider NDA process amid Project Blue outrage

  As officials in Pima County and the City of Tucson discussed and planned one of the region’s largest development projects in history, for at least two years not a single hint about the project was released to the public, largely because of non-disclosure agreements, commonly referred to as NDAs.

  Before the proposal was considered by the Pima County Board of Supervisors and Arizona Luminaria reporting revealed the company behind the data centers to be Amazon Web Services, important details had been unavailable to the public.

• Top 10 Reasons Why Communities Are Fighting Data Centers

  Indeed, the look on Eric Schmidt’s face when he got booed on Saturday by students at Arizona State University during his speech, when he talked about AI, was worth 1,000 words. Schmidt (net worth: $42 billion) was flustered and appeared genuinely bewildered that he — one of the titans of Silicon Valley — would get booed over anything. Further, he seemed hurt that the students who were there to hear his wisdom (“build a team”) were not down with his vision of AI as their virtual companion. At one point, he even pleaded with the boo-birds, “If you’d let me make this point, please...” (The full video of his speech is available here.)

  Schmidt’s disastrous speech shows that young Americans, in particular, are pissed off. It also shows how Big Tech, including Google, which withdrew its plans for a big data center project in Indianapolis last year before Indy’s city council could reject it, is drawing more and more public anger.

• Michigan township treasurer resigns, citing death threats over data center

  A Saline Township treasurer announced that she was resigning from her position after she says she received numerous deaths threats over the development of a data center. Other city officials say the threats have come through emails and phone calls.

• Yeah, we know why people don’t want data centers

  My take, which isn't necessarily different or counter to Matt's five points, is people don't want data centers in their backyard because:

  • It doesn't create jobs
  • It increases energy costs
  • It increases energy use
  • They feel weird about the products

• Arkansas Explained: Understanding the data center boom and debate

  More than 1,500 centers are planned across the country.

  Some take issue with the vast amounts of power and water large data centers use, and the potential impact on utility bills. Others chafe against the public’s lack of input into projects and the economic incentives that attract them, brought about by nondisclosure agreements signed by local officials. Some have other environmental fears, such as the potential for ongoing noise or increased flood risk from filled wetlands.

• Critics Warn Utah Megadata Center Could Devastate Great Salt Lake

  The controversy in Utah is a stark illustration of a wider trend. Across the United States, data centers are drawing bipartisan backlash as communities clash with tech giants and developers over strained water supplies and spiking energy costs.

• Data center battles started in the states. Now it’s Congress under siege.

  At the grassroots, intense opposition continues breaking out. In early May, community members screamed and booed when commissioners in Box Elder County, Utah, unanimously approved two resolutions in support of a 40,000-acre data center campus proposed for an unincorporated site in the county.

  Asked if he sees any momentum in Congress to do anything significant about the potential spread of data centers, Sen. Josh Hawley, R-Mo., said in an interview, “Only with voters.”

• Arkansas’ data center fights boil down to trust and transparency

  Pulaski County has become the latest front in the war over data centers, with opponents stepping up pressure for a moratorium on new data centers so local-level restrictions can be considered. It’s a debate that’s playing out across the country, as more communities are pushing back against the facilities, citing concerns about water use, utility rates and other quality-of-life issues.

• Americans overwhelmingly oppose data centers. Women most of all.

  More than two-thirds of adults oppose the construction of the massive and costly complexes used to power artificial intelligence, with a majority saying they’d prefer to have a nuclear power plant in their backyard instead. While women and men overwhelmingly expressed opposition, women did so more intensely. Out of 1,000 adults surveyed, 55 percent of women said they strongly oppose data centers, compared to 43 percent of men. In fact, men were more likely to favor data centers, citing their economic benefits and job opportunities.

• SoftBank hires banks for US IPOs of SB Energy and AI robotics spinoff Roze, sources say

  SoftBank has hired JPMorgan, Goldman Sachs, Morgan Stanley, Citi and Mizuho for SB Energy's IPO, which could come as early as September, the sources said. SB Energy could seek a valuation of more than $50 billion in its market debut, the people said. At the same time, SoftBank has tapped Goldman Sachs, JPMorgan, Mizuho and Morgan Stanley for the IPO of its autonomous robotics company Roze, which will focus on building data centers and using robotics to improve the efficiency of AI infrastructure construction and is targeting a similar time ⁠frame, the ‌sources added.

• Erin Brockovich starts tracking AI data centers, calls on affected communities to submit issues — website shows more than 2,700 reports from across the US raising various concerns

  "The RACE to build AI infrastructures is unfolding town by town across America. In some places, data centers are welcomed. In others, they are delayed, contested or abandoned altogether,” Brockovich wrote on her website. “This MAP captures the real-world footprint of that race — revealing patterns of growth, conflict and uncertainty.” At the moment, the website lists 33 operational (built & running) AI data centers, 44 under-construction projects, and 27 proposed sites. It also shows 2,716 community-reported locations nationwide as of May 24, 2026.

  There are several primary concerns against data centers, the biggest of which is energy consumption, which has been causing utility prices to skyrocket across large swaths of the United States. Water usage is also another issue that communities have been raising, with these massive-scale projects using a lot of water and catching the attention of lawmakers. There are also reports of noise pollution, with some members of the public complaining about the infrasound phenomenon often brought about by large-scale industrial operations. Brockovich also lists location risks, scalability, and e-waste as potential long-term problems that the industry must deal with in the future.

• When the AI boom subsides, the data centers will remain. What we do now matters

  Data centers now account for 6% of US energy use, and their water use is similarly staggering. 13% of the underlying workloads are useless: zombie processes that have been left running by inattentive owners whose priorities now lie elsewhere. Beyond the environmental impacts, which are no joke, data center consumption is pushing up people’s bills and disrupting communities. And beyond that, they push up real estate costs, with real knock-on effects for communities. It’s no surprise, then, that legislation is being written to limit their growth.

  It’s not that we shouldn’t have data centers. But their footprint is enormous, and the effects are sometimes disastrous. We need to consider the effect on people’s quality of life more than the impact to GDP, not least because economic indicators like these don’t actually show if people’s lives are improving.

• The Permission Slip

  Look at where the money is going and you can see the permission slip being cashed. Stargate, half a trillion dollars. The hyperscalers, tens of billions each per year. The Anthropic–Akamai arrangement, nearly two billion more. The collective bet of the wealthiest companies in the world is that you fix intelligence — including its honesty — by buying more of it. The data center operators are happy. The chip vendors are ecstatic. The labs raising money at valuations with too many zeros are happy. Everyone in that chain has the same incentive, which is to believe that the answer is more.

  The customers who will eventually pay for all of it are the ones who should be asking whether any of this is true.

• Unlawful by Design: Exposing the Human Rights Costs of Generative AI [PDF]

  Amnesty International adopts a supply chain lens in its human rights analyses of AI systems, which implies understanding AI products as reliant on complex supply chains that are each comprised of a network of actors responsible for the various aspects of the system’s training, development and deployment.1 These actors, from graphics processing unit (GPU) manufacturers, data centre providers, web crawlers, model developers and data annotators, to end user design and outputs, co-produce the technology’s functionality.2 This briefing focuses on the “data pipeline” aspect of the supply chain of generative AI products, specifically, the stages related to data capture, analysis, and processing. It examines the key parameters and implications of design choices concerning the training data of generative AI models, with a focus on methods and sources of data collection, data processing, model scaling and data outputs.

  This briefing examines how standalone generative AI systems, based on unlawful web scraping, are in conflict with international human rights law (IHRL) and standards through their design, development and deployment. While these technologies promise sophisticated automation and efficiency, they rely on data collection and model training practices that abuse privacy rights, enable discrimination, and threaten freedom of expression and thought.

• Hundreds of Rural Albertans Line Up to Battle Data Centre Goliath

  The Alberta Utilities Commission (AUC) doesn’t officially keep track of the statistics. But Victor Choy, lead application officer at the regulator, told The Energy Mix the project has generated more formal requests to participate in a hearing than any case he’s seen in 18 years on the job. It’s difficult to say whether it’s the most people who’ve ever filed a Statement of Intent to Participate (SIP), said Choy, because SIPs can be submitted by both individuals and groups.

• Utilities report as many as 30 data centers under discussion in Kentucky

  But Kentuckians in communities near where these data centers want to locate have strong concerns: about the potential enormous electricity usage and whether it could exacerbate electricity bills; the source of water and the amount needed to cool data centers; whether property values will go down for homeowners living near a data center; the potential for noise pollution and in some cases the prospect of rural land being transformed for a data center site. What about water use?

• The NVIDIA Tax

  I live in Virginia, which means I have a front-row seat to the strangest tax increase in modern American life. Nobody voted for it. It isn’t on any ballot. But it’s showing up on the electric bills of people who have never typed a prompt into a chatbot and wouldn’t know a GPU from a garden hose.

  In January, Consumer Reports profiled a man in Manassas who had lived in the same house for nearly forty years and opened an electricity bill for $281 — roughly triple what he’d paid the month before. He is not a heavy user. He did not buy a data center. He simply happens to live near “Data Center Alley,” the Northern Virginia corridor where, by Consumer Reports’ accounting, electricity prices in the densest data-center counties have climbed 267 percent over five years. Nearly three-quarters of Virginia voters now blame those buildings for what’s happening to their bills, and they aren’t wrong to. Abigail Spanberger won the governorship partly on a promise to do something about it.

  This is not just a Virginia story. Residential electricity prices nationwide are up more than 36 percent since 2020. Goldman Sachs told clients that power prices rose 6.9 percent in 2025 — more than double headline inflation — and that data centers will account for fully 40 percent of the growth in electricity demand through the end of the decade. The mechanism by which this lands on ordinary people is almost elegantly unfair: in at least 40 states, utilities are allowed to bill customers in advance for grid construction that hasn’t been finished yet. So the retiree in Manassas isn’t just paying for the power the data centers use. He is pre-financing the substations being built to feed them.

  Here is the part nobody says out loud at the ribbon-cuttings. That $281 bill is, in part, a margin payment to a single chip company in Santa Clara.

• I Put a Datacenter GPU in My Gaming PC for £200

  I already had an RTX 4080. 16GB of VRAM. Good enough for gaming, not good enough for the models I wanted to run locally. The next step up in GPU land is either spend a fortune on a card with more VRAM, or find another way.

  I found another way.

  I bought a datacenter GPU that doesn’t even have a normal PCIe connector, stuck it in my gaming PC with an adapter, and now I have 32GB of VRAM across two GPUs running a 27 billion parameter model at 32 tokens per second. The whole thing cost me £200.

• AI and data sovereignty in Postgres: An answer to the datacenter energy crisis

  A typical new hyperscale datacenter can face grid-connection bottlenecks of up to seven years in certain markets, well before the necessary transmission, substations, generation capacity, and transformers are in place. McKinsey, meanwhile, estimates that global datacenter spending could reach $7 trillion by 2030 - a figure comparable to the size of a top-12 global economy.

• Cox orders ‘higher bar’ for data centers, says public input ‘absolutely matters’

  Cox’s remarks were a reversal from the fervent, exasperated tone he used at a news conference in April when asked about opposition to the Stratos project, initially billed as a 9-gigawatt project that Cox recently said in its first phase would top out at about 1 gigawatt of energy generated by a natural gas power plant.

• The data center debates are only a preview of Arkansas’ AI challenges

  The backlash in some Arkansas communities about data centers’ drain on their resources like electricity and water reflects just one element of the public’s worries about this technology.

  The centers are vaulting to one of the top agenda items for the Legislature next year. But lawmakers face a session that needs to be dominated by wider discussions about AI, and how to contain its harmful effects on Arkansans.

• The Pirate Bay Remains Resilient, 20 Years After The Raid

  Twenty years ago today, dozens of Swedish police officers stormed a Stockholm data center, seizing The Pirate Bay's servers. The entertainment industry hoped the raid would finish the site for good. Instead, the police action inadvertently helped to create one of the most resilient and iconic websites on the Internet, one that remains online today.

• Leaving the Tech World Professionally

  I can't even begin to explain how awful AI has ruined the tech world. I used to love hearing about new hardware, new software, open source projects, and ways that people were trying to solve problems we ran into. Now AI is shoved into everything we do, where it's not even needed or make sense to be applied to, and it's ruined the creative world - video, music, photography - and nothing feels "real" anymore. And not many people realize how it's not actually intelligent. It needs more and more data to be trained on, and costs so much to run via datacenters that it's obliterating the planet and people's wallets. Even if we keep LLMs around for fast, computational usages for coding and transactions, where "AI" makes the most sense, hopefully it can be downloaded onto local computers to run there instead of in "the cloud".

• Neighbors Horrified by Data Center Twice the Size of Manhattan

  Given the widespread backlash to data centers across the entire country, it shouldn’t come as a surprise that many of these residents are now rushing to council meetings to forcefully refute the plans. After all, they’ve watched as other areas that welcome the facilities struggle with rising electricity prices, stressed water systems, and noise pollution.

  Worse yet, the Great Salt Lake is already in crisis: it’s rapidly disappearing amid devastating droughts across the state. An extremely resource-intensive data center could place a massive new strain on it, regardless of the many reassurances from developers.

• More cities are pressing pause on data centers as local backlash grows

  Hearing backlash from residents, cities and counties across the country in recent weeks have blocked planned data centers amid concerns over rising electricity prices and environmental harms.

  The local actions come as state lawmakers also are looking to limit or repeal the incentives for the centers, which are sprawling campuses of computer servers that store and transmit the data behind apps and websites.

• SoftBank to spend up to $87 billion on French AI data centers — country offers ample nuclear grid that US sites lack

  There’s a stark contrast between this and SoftBank’s American plans, where, to feed its planned 10GW data center in Ohio, the company has to build its own energy generation infrastructure in the form of a natural gas plant that’ll cost roughly $33 billion to build and generate roughly 9.2GW. France lets it skip that step and plug into an existing low-carbon fleet, sidestepping the strained grids and rising local opposition slowing U.S. buildouts.

• AI Grifters Are Making Anti-Data Center Slop With AI

  People hate data centers. Local and state communities around the country are passing moratoriums on their construction. Data centers are noisy and their neighbors have pressing concerns about water use, increased electricity bills, and the quality of the jobs developers are promising to create. Some areas, like Ypsilanti Township where the University of Michigan is planning a large data center, are even worried about becoming targets in future wars. These anxieties are now the focus of AI spam farms on Facebook.

  This is the same algorithmically boosted “shrimp Jesus” style AI spam scheme we’ve reported on before. There are people, some of them in foreign countries, who churn out hundreds of AI-generated images across multiple pages to engage users and turn a profit on ads and links. It’s impossible to know who, exactly, is putting up all these state-themed anti-AI pages. I reached out to several of the pages through Facebook Messenger but got no response. Many of the pages provide the same contact email but I didn’t receive a response when I contacted it.

• Shift work and booking maintenance on the calendar

  Then I talked to 2nd and 3rd, and they mostly wanted the window to close two hours ahead of time, just like with the datacenters.

  It was pretty clear. You could almost predict what shift a person worked on based on how they responded to my informal queries: they're either the ones generating the shit, or the ones being stuck with it.

  My manager at the time had an excellent point: since 2nd/3rd end up doing the majority of this work, their opinions should matter more. He was a good guy for thoughts like that and others in the same sense, and that's how we dealt with it.

• Data Centers Have Become Shockingly Unpopular, Poll Finds

  In short, the AI backlash has grown immensely over the past year or so — and the latest numbers put the trend in stark relief.

  According to a new Heatmap poll, at least seven in ten Americans would oppose a data center being built near their home. That’s a seismic shift from last September, when a similar poll found only 42 percent of Americans were opposed.

• Lombardo backs natural gas pipeline expansion for data centers

  The pipeline, which has a current daily average capacity of about 200 dekatherms, stretches from Nevada’s northern border to Reno-Sparks, Carson City, and Lake Tahoe. The application calls for 212 miles of new or expanded pipeline, including expansion to the Tahoe Regional Industrial Center in Storey County.

• Data Center Operators Are Trying to Fix Their Water Use Problems

  Lawrence Berkeley National Laboratory predicted in a 2024 report that hyperscale data centers could consume up to 33 billion gallons of water by 2030 if they relied heavily on evaporative cooling. That’s on par or even less than other thirsty industries, like agriculture or oil and gas—a single fracked well can use 1.5 to 16 million gallons of water—but it poses a risk in regions where water is already scarce. The risk is particularly acute in summer, when data center cooling needs tend to skyrocket at the same time as municipal water use.

• EU plans energy standards for data centres amid concerns over soaring power use

  The European Union will develop minimum energy-efficiency standards for data centres, it said on Wednesday, as concerns grow over their rapidly rising power use. EU data centre capacity is expected to more than ‌double in the coming ⁠years, ⁠reaching 28 gigawatts by 2030 from 12 GW last year. That expansion will lift their share of EU electricity consumption beyond the current 2.5%.

• Data centers, redistricting and Great Salt Lake take center stage in Republican CD2 debate

  From a debate stage in Salt Lake City, the two candidates pitched themselves to voters in Utah’s newly drawn 2nd Congressional District, staking out differences in how they see plans for a massive hyperscale data center in the district and on President Donald Trump’s proposed $1 billion in federal funding to help the drying Great Salt Lake.

• Arkansas’ largest city approves new data center ordinance

  The nearly five-hour meeting was packed with local residents who provided feedback on the ordinance and Google’s plans to build a 300,000-square-foot data center in the Port of Little Rock. Mayor Frank Scott Jr. proposed the ordinance amid significant online pushback against the centers last month.

  Many residents implored the directors to delay voting on the regulations in favor of a temporary moratorium on data centers while stronger regulations were developed. Several criticized the process that brought the data center to Little Rock as lacking transparency, and warned of consequences for elected officials at the ballot box.

• NV Energy betting the house on data centers

  The utility is projecting it needs 47% more electricity than it anticipated just two years ago in its last IRP, primarily to satisfy the needs of data centers in the north. That part of the state is “a sought-after location for data centers due to the expanse of developable land, a favorable climate for efficient cooling, and robust fiber connectivity,” according to its 41-volume plan filed in May with the Nevada Public Utilities Commission.

• UN Calculates Nation-Sized Environmental Footprints for AI and Data Centres

  By 2030, data centres will account for nearly 3% of the world’s projected electricity use, with 935 trillion watt-hours. If data centres were a country, the country would be projected to rank sixth-highest in power use in 2030. That would produce nearly 399 million tonnes of carbon dioxide, the report said. The study focused on energy use and didn’t examine the massive amount of water used to cool data centres.

  “If you look at these numbers, we’re seeing scales comparable to nations,” said study co-author Kaveh Madani, a water scientist and director of the United Nations University Institute for Water, Environment and Health in Canada. “The demand is enormous.”

• Beyond tokens per watt – using Ubuntu 26.04 LTS for AI

  But in the pursuit of tokens, it’s important to remember that hardware efficiency isn’t the only factor influencing data center operating costs, or the output of useful, revenue-generating AI work. While TpW is crucial, we also need to consider time-to-value and the impact of human productivity, which are largely determined at the software level.

• ‘Bots have now passed human traffic online,’ Cloudflare boss laments — says agentic traffic wasn’t expected to eclipse real people until next year

  We were also interested in looking at Cloudflare’s breakdown of human/bot traffic by country. The most bot-ridden traffic comes from the tiny island of Gibraltar (92.1%), followed by Singapore (76.4%), then Iran (76.4%). While some of these places have a lot of data centers and hosting infrastructure compared to population size, Iran’s high bot count may rather come from the heavy use of VPNs with automated scraping and bypass tools. Cloudflare has also previously flagged Iran as a hotspot for malicious bot activity.

• Seattle to pass one-year AI data center moratorium next week — will use window to study community impact of AI buildouts

  The city intends to spend the one-year hold on data center development to study its impact on the community surrounding it and what the local government should do to ensure that its constituents benefit from the project. This includes putting up pollution standards, determining energy connection requirements, setting contract terms, setting labor standards, and other regulations that will apply to data centers. Wilson also mentioned that Seattle will ascertain if these projects are a “good use of urban land” during this period.

• Microsoft buys land for data centre on Finland's west coast

  Microsoft has signed a preliminary agreement to buy about 190 hectares in the Vaasa region of western Finland for a possible data centre.

• O'Leary's AI Data Centres Face Backlash on Both Sides of Border

  Utah Senate President Stuart Adams is demanding O’Leary reduce the footprint of the project by 75%, shrinking it from 40,000 to just 10,000 acres (16,200 to 4,050 hectares) and include more environmental protection and transparency measures.

• Rise Up, People! Make AI Optional!

  The first problem: By Google’s own calculations, the AI Overviews are incorrect 28% of the time.

  The bigger problem: AI is an environmental disaster. It’s already a monstrous energy hog, and its appetite is doubling every six months. (You can thank the 4,200 data centers that AI companies have built all over the country so far. 1,500 more are going up as we speak.)

  By 2030, AI will consume 945 terawatt-hours of electricity—enough to power every household in California, Texas, Florida, New York, Ohio, and Pennsylvania combined.

  60% of that will come from polluting power sources.

• Data Centre Will Be One of Earth's Largest Heat Sources, Physicist Warns

  “These are not data centres in any familiar sense of the term,” wrote [pdf] Dr. Robert Davies, a physicist and complex systems scientist, in an assessment for the Sturgeon Lake Cree Nation. “From here on, I’ll refer to these energy-and-heat behemoths—massive compute fused with massive power generation—as Gigascale AI Smelters, smelting data and material strip-mined from people and planet.”

  At a realistic generation efficiency for a gas plant, Davies said supplying 9 GW of electricity would mean “burning fuel at a continuous rate on the order of 16 to 18 GW, day and night, year-round.”

  “Essentially the entire fuel burn ends up as heat released at the site, because the electricity is consumed onsite and degrades, in full, to heat.”

• Five large data centers eyed for Seattle

  City Light recently sent the companies proposing the data centers cost estimates for the grid infrastructure upgrades that are needed to support them. And it plans to send the updated large load contracts to the mayor’s office for review soon.

  Depending on how the companies feel about those terms, they may or may not follow through on their proposed projects. Data center operators are proposing many more projects than they actually build, market analysts say, likely in search of the most favorable terms.

• AI Will Consume as Much Water as a Billion People By 2030, UN Report Estimates

  The report, released this week, found that the environmental cost of AI is being “systematically mismeasured” because current assessments focus on the carbon emissions from training large language models while overlooking the tech’s broader water and land footprint. The water footprint comes from cooling and powering the data centers, and the land footprint comes from the energy infrastructure and supply chains that go into building and running them.

• The Environmental Cost of Artificial Intelligence: Carbon, Water, and Land Footprints | United Nations University

  The report moves beyond a carbon-only lens by quantifying the carbon, water, and land footprints associated with the electricity used to train, deploy, and operate AI systems at scale. Its central finding is that AI’s environmental costs depend not only on how much electricity is used, but also on where that electricity is generated and which energy sources power it. Every kilowatt-hour used by AI carries carbon, water, and land implications, and these footprints do not always move in the same direction: low-carbon electricity is not automatically low-water or low-land. The report also shows that AI’s footprint is shaped by both major infrastructure trends, including the rapid growth of data centers, and everyday use patterns, including model choice, output length, modality, and the growing use of text, image, and video generation.

• Rising Emissions, Depleting Water and Vanishing Land—UN Scientists: AI Is Threatening Natural Resources for Billions | United Nations University

  By 2030, the global data centres powering artificial intelligence are projected to consume 945 terawatt-hours of electricity. This is nearly triple the combined annual electricity use of Pakistan, Bangladesh, and Nigeria—countries collectively home to more than 650 million people. Their associated water footprint will equal the basic annual domestic water needs of all 1.3 billion people in Sub-Saharan Africa, and their land footprint will exceed 14,500 square kilometers, roughly twice the Jakarta metropolitan area, home to more than 32 million people.

• Meta putting up tents across the US to house AI servers, like ‘a scene out of the movie Mad Max’ — structures take three months to build and use jet engines for power

  Another factor that allowed Meta to bring its data centers online at a much faster pace is its use of “behind-the-meter” power, in which the company installed its own turbines to produce power on-site rather than relying on grid power. This is similar to what Musk did with his Memphis Supercluster, which he initially powered with portable power generators. However, Meta’s turbines would be a permanent feature on the Ohio site, as it’s designed to run independently of the power grid.

  [...]

  This combination of seemingly temporary structures and jet engines strapped to the ground is certainly an interesting combination. As one reporter said, “It’s like a scene out of the movie Mad Max.”

• Industry coalition urges Trump administration to take urgent action as AI data centers' extreme memory consumption threatens other industries — AI-driven memory chip shortage could raise prices in automotive, medical, telecommunications sectors

  A coalition of nine US trade associations has urged the Trump administration to take immediate action on what it describes as an emerging memory chip shortage driven by the explosive growth of AI data centers. In a June 3 letter sent to US Commerce Secretary Howard Lutnick and Treasury Secretary Scott Bessent shared with Tom's Hardware, the organizations — representing telecommunications providers, automakers, medical device manufacturers, and major retailers — warned that AI infrastructure deployments are consuming an outsized share of global memory production, creating supply constraints and price increases that could ripple across large segments of the US economy.

• Data centers are booming. Here’s how they’re endangering communities across the US

  In the U.S., Virginia has more data centers than any other state – over 600, two-thirds of which are in the northern Virginia suburbs of Washington, D.C. In 2023, the state’s data centers consumed about 26% of Virginia’s total electricity supply – a higher share than in any other state.

  We study science communication, climate science and public health, so we wanted to understand how data centers in Virginia affect the people who live near them and the broader public.

  We found that the data centers that already exist affect nearby residents and the nation as a whole in five main areas: air quality, water quality, noise levels, land use and energy costs.

• Zoo Officials Horrified by AI Data Center Menacing Their Endangered Animals

  And it’s not just residents in rural America trying to keep out facilities, which have been linked to rising electricity prices, water issues , and copious amounts of noise. The Nashville Zoo in Tennessee has turned into the latest battleground, with officials arguing that a proposed 69,000-square-foot data center next door could be devastating to animals living in captivity a mere 50 yards away.

  As NBC News reports, Southeast Asian clouded leopards, which the zoo is breeding, could be particularly vulnerable. Zookeepers are concerned that the cacophony could stop the creatures from reproducing entirely.

• Some Kentucky counties and cities are hitting pause on data centers

  Residents’ passionate opposition to data centers in his county led the local government to enact in late May a moratorium preventing data center construction for the next 12 months. That opposition includes ongoing speculation by residents and a county commissioner over local land deals connected to a company that works on “energy and data center infrastructure” investments.

• How can wave energy tech support the AI race?

  As part of these efforts, Eco Wave Power U.S. is advancing discussions relating to research and development collaborations with academic institutions, including Florida-based universities, as well as additional technology and infrastructure stakeholders. These initiatives are expected to focus on the integration of AI layers into wave energy infrastructure, including digital twin technologies, predictive maintenance systems, operational optimization, and future applications relating to renewable energy solutions for data center infrastructure.

  The company believes that “proximity-based renewable energy generation” near ports, coastal cities, and infrastructure hubs may become increasingly important as electricity demand from AI computing continues to grow.

• Amazon inks multibillion-dollar optical manufacturing deal with Corning

  Corning is a major supplier of fiber-optic cables for data centers. It makes riser cables, which are used to link together the different floors of a data center, and plenum cables optimized for server rooms. The company’s cables ship with specialized jackets that don’t release toxic materials when they come into contact with a flame or water. That helps reduce safety risks in the event of data center malfunctions such as short circuits.

  The company also makes outdoor cables, which cloud providers can use to link together their data centers. The cables come in armored and dielectric versions. The armored variety is designed to reduce scratches, while dielectric cables can be placed near high-voltage power lines without the risk of electric inference.

• OpenAI: ‘Likely’ Chinese influence operation tried to use ChatGPT to stir debate on data centers

  In both cases, OpenAI said the operators “likely originated” in China. The anti-data center content was traced to an unnamed Chinese technology company that holds multiple contracts with regional Chinese governments, and both clusters used VPNs to evade restrictions, prompted ChatGPT in simplified Chinese and asked for both English and Chinese-language outputs, all while posing as Americans on social media platforms like X and YouTube.

• The future of Siri, or: why private inference isn’t private enough

  Apple has since “expanded” PCC to encompass Google’s hardware as well. I will confess that I find the details of the new “expanded” PCC just a bit vague. It sounds a lot like Apple is primarily going to rely on Google’s existing confidential compute (running in Google datacenters) to process this data, but they’re bolting on a new layer of technical security to control which models are actually running. In any case: security experts can argue about whether this is good enough to keep Cozy Bear away from your data. What I will grant is that it’s probably good enough to keep Google and Apple from accessing your stuff, which is what most people are worried about in the first place.

  So why am I so nervous?

• Texas farmer donated land for children decades ago but City officials quietly sold the land for massive data center construction

  Public records later showed the property passing through multiple nonprofit and municipal entities before eventually reaching the City of Taylor in 2003.

  Five years later, city authorities transferred ownership to the Taylor Economic Development Corporation for approximately $15,000, substantially changing the property’s administrative direction afterward.

  The controversy intensified in 2025 after the Taylor Economic Development Corporation sold the same property to developer Blueprint for $10 million.

• Farmer donates land for a park, city sells it for data center development — $10 gift became $10M for city government, with $30M tax expected over next decade

  To introduce this case, let’s go back to 1999, when the $10 deed was inked. For some background, 404 Media talked to long-time local Pamela Griffin, who used to play on the farmland, and watched her children grow up and enjoy the same freedom. Griffin recalled that old farmer Mr. Bland used to talk to her father from time to time. According to her, Bland once said to her dad, “I see the kids don’t really have nowhere to play.” He continued, “I’m thinking about giving this land for parkland because these kids need somewhere to play.” The original July 1999 deed has since been unearthed, and the farmer did indeed follow through with his words. Now, let’s make the following chain of events simple using a bullet point timeline: [...]

• Datacenter growth may run into a power wall by 2030

  The research giant expects global datacenter electricity consumption to reach 565 terawatt-hours (TWh) in 2026, as power demand rises from 104 GW in 2025 to 132 GW this year.

  This is higher than the 500 TWh per year Gartner estimated two years ago that AI-optimized servers would consume by 2027.

• Goodbye, Leaf-and-Spine Networks?

  What exactly did they do? They rediscovered the way Plexxi tried to build data center fabrics. Instead of spine switches, Plexxi tried to connect leaf switches directly, first with CWDM (they were dreaming about dynamic leaf-to-leaf bandwidth), later with a prewired middlebox (what AWS engineers call ShuffleBox).

  Obviously, you’d waste a lot of bandwidth that way, as there are always some leaf switches that do not exchange traffic even though they have a direct link. Plexxi solved that with unequal-cost multipathing (the traffic also uses longer paths, not just direct links); the AWS blog post calls that Routing through Randomness.

• Brad Paisley joins fight as zoo's dispute with AI data center escalates, petition tops 330,000 signatures — Nashville weighs sweeping hyperscale ban

  The pressure now shifts to the Metro Planning Commission, which is set to hold a public hearing Thursday, June 11th, on legislation that would ban data centers larger than 500,000 square feet across Davidson County and keep smaller ones at least half a mile from homes, schools, and zoos. The 69,220-square-foot facility at 648 Grassmere Park would draw an estimated 50 MW from the local grid, and the zoo argues its noise and light could disturb vulnerable species, including clouded leopards it is working to conserve.

• Officials powerless to stop 8 new data centers that could transform small Texas county

  However, some commissioners and residents remain skeptical, saying the benefits are uneven, and data centers create few permanent jobs after their labor-intensive construction is finished. For example, one Hood County data center proposal shows a peak construction workforce of 2,000 dropping to a permanent workforce of 220, according to the project’s concept plan.

  Hood County Commissioner Dave Eagle said there are “too many unanswered questions” about data centers, and they’re being asked to greenlight plans with incomplete information about their impact on the community.

• Bernie Sanders’ AI Sovereign Wealth Fund Plan

  We should be wary of this possibility because we’ve seen it before. Ownership of substantial stakes in oil companies by the Norwegian sovereign wealth fund, the world’s largest, does not seem to have steered those corporations to pro-environmental policies. Instead, the Norwegian government’s dependence on those companies has inhibited them from taking climate action. Here in the US, public employee pension funds merit the same criticism: the fiduciary duty to generate wealth overwhelms any intention to direct their corporate holdings in the public interest.

  A better answer is to separate the two goals. The standard way to share private rewards with the broader society that made them possible is taxation. Senator Elizabeth Warren has proposed an excise tax on datacenters’ energy use. Others have proposed an AI token tax, which has much the same effect.

• Federal Adviser Says Data Centres ‘Provide Markets’ for Gas

  A list of proposed data centres, which provide the computing power for AI, gives clarity on what kind of energy resources the federal government hopes to grow. Massive data centres proposed for Alberta, including the Kevin O’Leary-backed Wonder Valley project, will have their energy needs “supplied by natural gas,” the Privy Council document reads.

• AI data centers are taking over. These Americans are fighting back

  In nearly every state, hyperscale data centers powering artificial intelligence are being proposed and built. These facilities need much more computing power compared to other data centers and that comes with a much larger footprint, in the literal and carbon sense. A typical hyperscale data center needs 10 acres, or 435,000 square feet of land. However, many of these proposed sites plan to use over 1 million square feet, which could fit a few Eiffel towers or roughly 50 Boeing 737 jets.

  The list of potential drawbacks for communities receiving a data center are plentiful. The huge increased pressure on energy grids translates to higher utility bills, which puts additional strain on gas and coal plants, which of course jacks up toxic air pollution. Furthermore, the centers’ cooling systems need a lot of water, draining local aquifers and water tables, with the more efficient systems still requiring significant amounts. Even noise pollution from constant humming of the servers impacts local quality of life.

• Amazon Data Centers In Mississippi Have Already Raised Electricity Rates for Local Customers, Report Suggests

  A new report written by consultancy Synapse and commissioned by advocacy groups Earthjustice and Environmental Advocates Mississippi attempts to calculate the cost of 3 planned Amazon data centers to Entergy Mississippi customers, who share an energy utility with the centers. These hidden costs may offer a window into the broader burden borne by residents living near data centers around the country. The report estimates that residential customers of Entergy Mississippi, one of the state’s regional energy monopolies, have paid $38 million as of March 2026 for infrastructure and other costs related to data centers and will have paid $74 million by the end of the year.

• The More Data We See About SpaceX's IPO, the More We're Wincing

  The more data we see about the blockbuster IPO, the harder it becomes to rationalize it. Investing in Musk’s rocket company is starting to feel less about trusting it to ever make money and more about signaling confidence in the billionaire’s vast ambitions to send humans to Mars or build out a one million satellites-strong AI data center in orbit. Others are likely looking to go along for the ride, hoping to get rich on the hype alone. That’s an inherently risky gambit, especially given the volatility of Musk’s other publicly traded company, Tesla.

• The FBI built a small town to simulate cyberattacks

  It’s a training facility where the bureau can recreate real-world scenarios for training and research purposes. All of the various buildings and facilities are hooked up the way they would be in a real town. There’s even a small data center with over 200 servers that can be hacked, infected with malware, and studied. But, importantly, all of the systems in the fake town are cut off from the outside world, which means there’s no danger of any malicious code or anything from escaping containment.

• My rural Kansas town has a bad case of data center fever. Boom or bust, we will never be the same.

  Seven in 10 Americans are opposed to local construction of data centers for artificial intelligence, according to a Gallup survey released last month, including nearly half who are strongly opposed. That opposition was tied to concerns about the environment and quality of life.

  Those concerns were echoed by one of the organizers of the opposition to the Emporia data center. Amanda Mendoza, a teacher who lives across the street from one of the tracts that would be used for development, said she and other landowners felt betrayed by the city commission because of the secrecy surrounding the project.

• Multiple small Tennessee counties pass temporary data center bans — Nashville also passed near-unanimous moratorium on first reading

  Tennessee is currently home to 63 data centers, according to the Data Center Map, with Elon Musk’s massive Colossus and Colossus 2 data centers located in Memphis. However, it seems that Tennesseans are pushing back on these developments, as several counties are considering and passing temporary bans. This is especially true for rural areas, where many data center developers take interest because of cheaper land, less regulatory friction, and weaker resistance.

• Trump Admin Admits to Using Grok to Bomb Iran In Federal Lawsuit Backing Musk

  The information came to light when the Department of Justice (DOJ) intervened in a federal environmental lawsuit filed by the NAACP against Musk’s data center in Mississippi. The DOJ argued that the judge should dismiss the lawsuit because the data center is vital to issues of national security — and admitted that xAI’s Grok Gov Model, owned by Musk, has been used to conduct U.S. military operations in Iran.

  In a court filing, the DOJ claimed that the NAACP’s attempt to stop Musk’s company from running natural gas turbines in Mississippi “threatens American national, economic, and energy security” as the AI company “supports the Department of War’s military operations.”

• SpaceX Stock Has Now Started to Fall

  In its SEC filings, SpaceX also admitted that Musk’s plan to place one million enormous data centers in Earth’s orbit may not be economically feasible, let alone physically possible, casting a long shadow over the now-trillionaire‘s infamously overambitious aspirations.

• Finnish spy agency warns of hidden security threats from data centres

  But if a foreign company simply leases a plot of land and builds a new data centre on it, it may fall completely off the radar of security authorities.

  "We don't currently have any legislation that requires notification of such data centre projects," says Kivimäki.

  The same blind spot applies to situations where a foreign entity simply rents space in a ready-made data centre.

• Amazon employees say they’re facing termination for backing data center limits

  On June 10th — one week after the hearing, and one day after the City Council passed a milestone moratorium on data centers — Patrick Schloesser, Darius Irani, and Liesl Wigand were each called into an impromptu meeting with Amazon’s “Employee Relations.” HR representatives told the employees that the company was investigating them and said there could be disciplinary action, up to and including termination. On Thursday, the three filed a legal complaint requesting that the Seattle Office for Civil Rights investigate the matter, alleging that Amazon engaged in prohibited employment discrimination.

• ARM Tries to be Everywhere

  ARM is best known for providing the CPU chips used in smart phones and tablets. Its CPUs are power efficient allowing longer battery life while still providing a lot of processing power. ARM has expanded beyond this market, first into the low power microcontroller market and then more recently into the higher end namely the cloud data centers. When Apple released their M series ARM CPUs for their Mac laptop and desktop computers, Apple showed how to build high end CPU chips with integrated graphics and AI processors. Apple also showed how to leverage the ARM CPU architecture to provide extremely fast on-chip RAM for a significant performance boost, while keeping power usage under control. Now ARM seems to be expanding into every niche of computing and threatening other players at both the low and high ends of computing.

• Donald Trump, Champion of Renewable Energy

  Yes, $2.5 billion to destroy already-approved, cost-effective clean energy projects while Americans are suffering from soaring electricity prices thanks to data centers and high gasoline prices.

• The unlikely role of Operation Epic Fury in a Mississippi AI data center lawsuit

  A Defense Department official disclosed that U.S. forces used Elon Musk’s Grok AI tool to help deploy “over 2,000 munitions to 2,000 distinct targets” during the first four days of combat operations in Iran.

  The disclosure offered a glimpse into how generative AI is being used by the military. It came in a court filing supporting the Justice Department’s effort to dismiss a lawsuit targeting a Mississippi power plant, also owned by Musk, that supplies electricity to data centers used to run the AI system.

• New Mexico judge dismisses Jupiter data center suit

  A government watchdog group, the New Mexico Foundation for Open Government, filed its own lawsuit this week, accusing county officials of withholding public records related to the data center by improperly claiming a narrow exemption designed for terrorist threat response plans.

• A second data center could be coming to North Richland Hills. Many residents don't want it | KERA News

  A majority of residents at the meeting said they oppose the new project. Many brought up issues with large-scale data centers in other parts of the state, including communities south of Tarrant County.

  Following a presentation, Backes answered questions from several residents about the project.

  “Say hypothetically every person in this room like we took a vote right now and everyone said ‘no, I don't want this data center,’” asked Taylor Stevenson. “Would you still continue with that?”

  “I think I'd have a lot of questions about that process and about the due process involved in that,” replied Backes.

  “That sounds like you still would,” Stevenson said.

• Bodycam Shows Moment Cops Arrested a Man for Speaking Too Long at Data Center Meeting

  In February, police in Claremore, Oklahoma arrested farmer Darren Blanchard for speaking a little too long during a community meeting about data centers. The city charged Blanchard with criminal trespass, a crime with a $200 penalty, but he’s vowed to fight the charge. He recently shared video of the bodycam footage for the first time with 404 Media and answered our questions about the moment cops arrested him for going over his time at a February 17 community meeting of the Claremore City Council.

• Why Americans are fighting AI data centers

  Yet grassroots anti-data center activists are hardly wrong to be worried about artificial intelligence — it is one of the most formidable policy problems we face today. AI’s ultra-wealthy makers promise a world of unprecedented progress and prosperity, but also say they might eliminate everyone’s job and possibly annihilate humanity in the process.

• AI data center boom to drive inflation

  The AI data center boom is causing a new wave of inflation, data suggested. As oil prices drop with the Iran war quiescent, US consumers might hope price rises could level off. But demand for chips and energy is driving up costs for everything from smartphones to electricity.

• Apple is hiking the prices of MacBooks and iPads due to the memory chip shortage

  In a statement, Apple said that growing number of AI data centers has "created an extraordinary surge in demand for memory and storage" and that it has "never seen a component price increase this much, this quickly."

• AI can't be trustworthy without data center transparency

  If the infrastructure powering AI is developed through secrecy, exclusion or governance failures, public confidence will become increasingly difficult to keep.

• Why is Apple asking me to pay more for Big Tech’s AI obsession?

  But this isn’t some fluke, or temporary supply chain problem. Companies are choosing data center clients over ordinary buyers because “the same chip earns far more inside an AI server than inside a consumer device,” according to Srikanth Jagabathula, professor of technology, operations, and statistics at the NYU Stern School of Business. Regardless of whether people are clamoring for more AI, and more AI data centers, or not.

• The Anti-Data-Center Movement Is Reshaping Michigan Politics

  Lawrence’s campaign sees data centers as a potent topic to rally voters to his side in the Democratic primary in Michigan’s 7th district, to be held in August. Internal polling conducted by Data for Progress of likely Democratic primary voters in the district shared with WIRED shows that more than 40 percent of respondents were “much more likely” to vote for a candidate who opposed data centers. The message resonated even more with respondents under 45: Almost 80 percent of younger voters said they’d be much more likely or more likely to support an anti-data-center candidate. (The 7th district includes the college county of Ingham.)

• ‘We’re up against forces that have all the money in the world’: Erin Brockovich on her battle against AI datacentres

  Datacentre developers often enter into nondisclosure agreements with local officials, so it is not possible to see why they were approved without environmental-impact assessments or input from residents. “I am getting reports from people where their local leaders are changing zoning laws for this to happen,” she says, incredulous that anyone would bypass democracy to that degree.

  She does not believe this is a story of sudden, nationwide corruption or that local governments are setting out to stonewall residents. “What I’m seeing now is that councils, having heard community responses, are trying to hit pause and they’re getting sued [by the developers] for $100m-plus. They cannot withstand that.” In one case, Hill County in Texas, county commissioners, not expecting the public outcry over a planned datacentre, voted on a year-long moratorium to halt building. The county was then sued by the developers for $100m in damages and, according to the Texas Tribune, has now backed down.

• Texas county rescinds data center moratorium after lawsuit

  A rural North Texas county that appears to be the first in the state to pass a data center moratorium has rescinded the measure after being sued by a developer for $100 million in damages.

• Arizona governor vetoes bill that would weaken referendums and empower data centers

  HB 2873, passed by the state legislature in late April, would have allowed referendum petitions brought forth by Arizona citizens to be withdrawn before they are put to a vote. The Arizona Secretary of State defines referendums as the “method by which voters may veto a law (or part of a law) by gathering signatures from registered voters to place the issue on the ballot.” The bill was first introduced in the Arizona House of Representatives in January 2026 and after passing both chambers, the bill made its way to Hobbs’ desk in June for her rejection.

  According to Natali Fierros, executive director of Rural Arizona Action (RAZA), the organization assumed that Hobbs would most likely sign the bill. However, pressure from the public and advocacy organizations may have pushed her to veto.

• America Is Having MacBook Sticker Shock

  Blame the AI boom. Data centers use a gargantuan amount of memory chips. Amid the frenzied AI build-out, the world doesn’t have enough to go around. In a statement issued the day of the price increases, Apple wrote that “the rapid expansion of AI data centers” had forced its hand: “We have never seen a component price increase this much, this quickly.” Seemingly every laptop maker has already jacked up prices at this point. In October, Microsoft raised Xbox prices by up to $70; last week, the company announced another $100 hike. The effects of the memory shortage extend beyond laptops and gaming consoles. A coalition of trade groups recently wrote to the Trump administration that the memory shortage was jeopardizing the “production and availability of automobiles, medical devices, and other manufactured goods.”

• Musk Tries To Bribe Memphis So They'll Overlook His AI Data Centers Are Killing Black Toddlers

  This synopsis was from a recent lawsuit by civil rights groups pointing out Musk's Colossus 2 AI data center is openly violating the Clean Air Act: [...]

• Massive Data Center Cooks Nearby Residents Alive Amidst Deadly Heatwave

  In the context of a record-breaking summer heatwave, it’s a bitter pill to swallow for a public already extremely wary of data centers appearing in their backyards. The backlash to the trend has grown considerably, with critics citing surging electricity prices, low water pressure, and plenty of noise as reasons for opposition.

  The research surrounding the effects data centers have on their immediate surroundings is still in its infancy. According to one yet-to-be-peer-reviewed paper by researchers at Cambridge, data centers can spike temperatures by as much as 16 degrees Fahrenheit and an average of 3.6 degrees Fahrenheit in their immediate vicinity by creating a “data heat island effect.”

• County With 37 Data Centers Tells Schools to "Turn Off Your Lights" to Save Electricity Ahead of Scorching Heat Wave

  The eyebrow-raising demand sums up why data centers continue to be so controversial with locals, as communities across the country fiercely protest against existing data centers and the construction of new ones. These sprawling facilities devour an enormous amount of power, require heaps of water to keep cool, and generate tons of noise, on top of being an eyesore that requires hundreds of acres of land be flattened. Studies have shown that the electricity bills of ordinary people who live near data centers are soaring, and 404 noted the plight of one woman who used solar panels and a heat pump to keep her energy costs down and still saw her electricity bill double.

• Our focus is data centres because they are big fibre guzzlers: Corning's Sudhir Pillai

  Specialty glass and fibre-optic maker Corning is betting heavily on the rapid expansion of AI data centres to accelerate its growth momentum in India, said Sudhir Pillai, the US-based company’s managing director and president for India, the Middle East, and Africa.

• County with 37 data centers tells schools to turn off lights to save electricity

  Matthew Gault of 404 Media acquired email sent by County Manager John Vithoulkas to thousands of county employees on June 26, warning that the rate Henrico pays for electricity in all county and school facilities was about to rise by 25%, adding an estimated $5 million to costs in the coming fiscal year. He told staff to expect further increases. To offset them, Vithoulkas asked workers to turn off lights and computers at the end of the day, close blinds to block heat, unplug idle chargers and appliances, and avoid space heaters, which he said can cost the county $150 to $300 a year each.

• An Updated Look at the Environmental Impact of A.I. Data Centres

  Google’s power consumption is now greater than the amount of electricity generated by New Zealand.

• SoftBank unveils plans to enter the US neocloud business with SB Neo

  SoftBank Corp. says it has been running a beta operation of its GPU cloud service in Japan since April. It’s powered by a software stack called Infrinia AI Cloud OS, which is a software stack for AI data centers. The company says it wants to apply the expertise it has gained from running this business on a much bigger scale in the U.S.

• AI data center builder Crusoe reportedly raising $3B at $30B valuation

  Crusoe builds data centers for major tech firms such as Microsoft Corp., Oracle Corp. and OpenAI Group PBC. According to Bloomberg, Meta Platforms Inc. is also a customer. The social media giant has reportedly signed contracts to use two Crusoe data centers undergoing construction in Texas and Missouri.

• Meta data center water discharges suspended after contaminating the city's reclamation water supply with bacterium — system offline for months for cleaning, closed-loop cooling system purge spread rare metal-resistant bacteria in Cheyenne’s water system

  Fill-and-flush is a commissioning step in which crews fill a cooling loop's piping with water, flush it to clear debris before the system is run, and then send the used water to drain. Goat Systems routed that flush water, which contained Cupriavidus gilardii, into Cheyenne's sanitary sewer, Frank Strong, the Board's engineering and water resource division manager, told the Wyoming Tribune Eagle. Strong said the fill water had been purchased from the Board itself and that the origin of the bacterium remains unknown, but said that lab staff caught it in February during routine fecal-bacteria sampling. "This isn't something we normally test for," Strong told the paper.

• Cheyenne Won’t Take Data Center Wastewater After Meta Contractor Contaminated System | Cowboy State Daily

  Cheyenne’s Board of Public Utilities isn’t accepting industrial wastewater associated with data center systems until further notice after a contractor for Meta Platforms contaminated the city’s wastewater system, prompting months of cleanup.

  The announcement was made by the Board of Public Utilities (BOPU) on Thursday in conjunction with naming the Meta company as the source of the initial contamination.

• EU appears to find datacenter emissions easier to offset than lobbyists

  According to the Financial Times, the European Commission is weakening its original proposals after pressure from datacenter operators and tech giants. The newspaper claims to have seen the revised draft of the regulations, set to be discussed by representatives of member states on Thursday.

• Looming Data Center Boom Tests Wyoming’s Promise Not To ‘Piss It All Away'

  Now that data centers are finally racing toward the state’s wide-open prairies and cheap power, however, the rush is forcing a choice. How much growth is Wyoming will to accept, and at what cost to power bills, water supplies, and its wide-open spaces?

• Wisconsin residents file class-action lawsuit against Microsoft's 'world's most powerful AI data center' due to data center noise — plaintiffs also mention construction noise and extreme light pollution from $7.3 billion facility

  Controversy due to AI data center buildouts generally centers around their massive power usage and draining of local water reserves, but noise is a third torment to nearby residents, and one that is arguably much harder to correct. Residents of Sturtevant, slightly south of Milwaukee, Wisconsin, filed a class-action lawsuit against Microsoft due to the excessive noise produced by the company's Fairwater facility. CEO Satya Nadella described the project as "the world's most powerful AI data center," projected to generate 865,000 tokens per second and have a final bill of $7.3 billion.

• Nashville Mayor to Use Eminent Domain to Stop Data Center Near Zoo

  Nashville Mayor Freddie O’Connell plans to use eminent domain to stop a data center from moving in near the Nashville Zoo, reports WKRN-TV. In Tennessee, eminent domain allows the government the power to take property but under the Fifth Amendment, eminent domain must be for a “public use,” which traditionally means infrastructure projects like roads or bridges. The government must then pay the owner compensation.

• Microsoft sued by Sturtevant residents for noise from data center

  Residents in Sturtevant have filed a class action lawsuit against Microsoft over noise from its data center.

  The lawsuit alleges the facility generates excessive and pervasive noise from its HVAC systems and diesel generators.

• Meta's AI Data Center Caught Infecting Town Water Supply With Deadly Bacteria

  That decision came after one bad actor, the Meta-affiliated data center company Goat Systems LLC, flooded local waste water pipes with fill-and-flush swill containing a rare and deadly bacterium known as Cupriavidus gilardii. Per Cowboy State, Goat Systems was found to be in “significant noncompliance” with Cheyenne’s industrial waste regulations after a months-long investigation traced the bacteria to Meta’s discharge.

• 'Data Centres Will Add To Summer Heat, Public Hearing Must'

  The Union ministry of environment, forests and climate change must make environmental impact assessments for data centers mandatory, Donthi Narasimha Reddy, environment and public policy expert, said. But this cannot happen as at present, data centres are screened for environmental clearance under a category meant for townships and area development projects, he said.

• The People Who Want to Stop AI by Any Means Necessary

  It would be easy to dismiss these episodes as the work of zealots. That would be a mistake. They are the most dramatic expression of a broader skepticism about artificial intelligence. Canadians are hugely wary: a Leger poll found 85 percent of respondents want the government to regulate the technology. But that number doesn’t convey just how frightened many are. One week, we learn an overly “sycophantic” ChatGPT may have aided and abetted a school shooting in Tumbler Ridge. The next, that data centres could consume as much water as the entire population of Sub-Saharan Africa does. Even as we absorb reports that generative AI may exacerbate delusional thinking—psychiatrists are now encouraged to screen patients for “chatbot psychosis”—we learn armies are expanding the use of autonomous AI-powered killer drones.

• The Popup That Says the Quiet Part Out Loud

  Here's why that's a bigger deal than a UI notification. When Apple introduced Private Cloud Compute back in 2024, the whole pitch was that the heavy AI lifting would still happen on Apple silicon, in Apple's own data centers, verifiable by outside security researchers. That was the differentiator. Google, Amazon, Microsoft — everyone else was shipping your prompts off to whatever GPU farm was available. Apple said: not us, we built our own.

  That promise just changed. Apple confirmed in a security blog post published the day of the WWDC26 keynote that its newest Foundation Models include a fifth, top-tier model called AFM Cloud Pro, built for agentic tool use and complex reasoning — and it doesn't run on Apple silicon at all. It runs on Nvidia Blackwell GPUs, hosted inside Google Cloud. Apple's own words: this is the first time PCC has extended beyond Apple's own data centers into third-party ones. Google isn't just supplying a model anymore. They're supplying the racks.

• Trump Tells Big Tech He’ll Fast-Track Fossil Fuel Plants to Power AI

  While the administration has sought to waive environmental protections, expedite permits and loosen construction rules for gas plants and data centers, there are a slew of state and local requirements both power plants and data centers must satisfy that even in the fastest permitting environments take months.

• Calls Grow for Pause on Scotland's Largest Data Centre Projects - The Highland Times

  The intervention comes amid increasing debate over the impact major data centres could have on energy supplies, water resources and climate ambitions.

  According to the Scottish Greens, there are currently 24 proposed hyperscale data centres across Scotland.

• UK gives data centers option to apply for 'national importance' status that overrides local regulations, cuts timeline by a year — eligible projects to bypass local councils, save more than a billion dollars in NIMBY fights

  The UK just gave Nationally Significant Infrastructure Projects (NSIPs) the right to bypass pre-application consultation with local council planning processes, meaning the green light will come directly from the national government. According to The Register, NSIPs are major projects that the state considers to be nationally important, like power plants and railways. Data centers were included in this list earlier this year, meaning these developments could now get approval directly from the national government instead of going through hoops in local councils. Still, this is an opt-in process requiring developers to apply for the status.

• Google’s exponential path to climate-wrecking digital bloat

  Every time I look at this chart I have to go and double check every single Google number, because it just looks so ridiculous

  The power grids that Google’s data centres are plugged into have to increase generation to match this new demand – and that includes rising use of coal and gas, and as a consequence, worse climate disasters like deadly heatwaves. Google’s consumption is rising way faster than the grids are being cleaned up with renewables, and that means their emissions number is going up fast, too. It’s the steepest rise on record: [...]

• ODC-Noord: bouwstenen voor een overheidscloud die al draait

  ODC-Noord ontstond in 2012 uit de Compacte Rijksdienst, het programma dat het aantal rijksdatacenters moest terugbrengen van 65 naar 4. In Groningen zag een klein team de kans om dat niet alleen efficiënter, maar ook toekomstbestendig te doen. Het resultaat is inmiddels veel meer dan een technisch platform: ODC-Noord vormt een belangrijk fundament onder de digitale overheid en daarmee onder de publieke dienstverlening. 

• Satellite 6.19 delivers Red Hat Lightspeed on premise security monitoring

  The big news: Red Hat Lightspeed’s vulnerability service is now generally available (GA) for disconnected RHEL environments in Satellite 6.19, thanks to Red Hat Lightspeed on premise. Red Hat Lightspeed on premise is more than just another dashboard; it brings the full power of a SaaS-based vulnerability monitoring service directly into your air-gapped data center, giving you the ability to identify, triage, and remediate Common Vulnerabilities and Exposures (CVEs) without ever sending data outside your environment.

• To Reduce Electrical Grid Strain Amid Heat Wave, Data Centers Are Ordered to Use Backup Power

  Orders to use such generators are most likely to come in 13 Mid-Atlantic States, which share a strained grid and are home to the largest concentration of data centers in the world. That region has also been subject to this week’s heat wave. The area’s primary grid manager, PJM, runs a system that stretches from Chicago to Virginia Beach and covers most of Pennsylvania, Maryland and New Jersey. PJM expects that demand for electricity on its system will hit its highest level on Thursday, exceeding a previous record set two decades ago.

• Microsoft Reports a Massive 25 Percent Jump in Emissions

  Microsoft’s greenhouse gas pollution increased by roughly 25 percent last year, the company says in its new sustainability report released Thursday.

  The report follows similar ones released by Google and Amazon last week. Together, they show a troubling trend of rising tech company emissions, driven by the global race to build out power-hungry data centers.

• Data Centers Are Quietly Taking Over Texas. The Pollution Could Be Catastrophic

  With some 300 data centers already in operation and 200 more in development, Texas could surpass Virginia as the nation’s leading data center market by 2030. Amidst the frenzy to capitalize on the AI boom, a regulatory loophole has allowed dozens of data centers like Stargate to quietly construct massive power sources that emit harmful pollutants with little to no public notice, a Floodlight investigation has found.

  Typically, before you can build a major source of new emissions, you have to get a major air permit, which includes extensive environmental reviews and engagement with the local community. But in Texas, regulators have allowed some data centers like Stargate to avoid that process by first obtaining so-called minor air permits—the kind more commonly associated with dry cleaners and autobody shops and rubber-stamped with minimal review.

• AI-driven datacenter builds drive Microsoft's emissions up a quarter in one year

  Microsoft says it matched its entire electricity consumption with renewable energy last year. The bad news is it also increased greenhouse gas (GHG) emissions by 25 percent due to datacenter construction.

• Irish datacenters now guzzle 23% of the country's electricity

  The latest figures from Ireland's Central Statistics Office (CSO) show that giant server farms now account for nearly a quarter of the country's metered electricity consumption.

  Their share rose to 23 percent in 2025 after passing 20 percent in 2023 and 14 percent in 2021 – up from just 5 percent way back in 2015.

• Stories Told About Data Centres

  If you read a website like this one, you were probably aware that data centres were commonplace twenty or more years ago. Like the one near King’s Cross, some were hidden in plain sight, while others were purpose-built facilities that look like hangars stuffed with servers. But the A.I. boom has meant rapid increases in the speed, scale, and quantity of data centres. People quickly learned not only of their existence, but how much pressure they put on local resources. Tech companies, it seemed, were caught by surprise; and as someone who spends a lot of time immersed in this world, so was I.

• Ireland’s data centers consumed nearly as much electricity as every home in the country combined in 2025 — server farms gulped 23% of national power despite years of grid restrictions

  Data centers accounted for 23% of Ireland's total electricity consumption in 2025, according to data released by the country's Central Statistics Office last week. The report revealed that data center consumption rose to 7,663 GWh in 2025 from 6,973 GWh in 2024, a 10% rise in a single year. Meanwhile, consumption by the rest of the country increased by just 2% within the same period.

• Memory crisis sends smartphone market into steep decline

  Memory prices extended their climb as suppliers prioritised AI data centre customers over consumer electronics, forcing manufacturers to pass higher component costs on to consumers through price hikes, particularly for entry-level and mid-range devices.

• New York becomes first US state to impose a data center moratorium

  Executive Order No. 62 took effect on signing and covers hyperscale data centers that draw 50 megawatts of power or more. While it’s in force, the state Department of Environmental Conservation will not issue discretionary permits that have not already been deemed complete.

• Lies Told About Data Centres

  That is one reason why we are seeing a bunch of these proposals here. And these companies benefit from a lack of transparency.

• AI Data Centers and the Concentration of Wealth

  Opposition to AI data centers has emerged as a primary theme in US politics, one that—surprisingly—doesn’t fall along party lines. We applaud people coming together for constructive debate on any issue, and agree that communities need to evaluate whether any economic benefits these data centers bring is worth their costs. Still, we worry that a focus on data centers obscures the larger impacts of AI on people’s lives: the concentration of power of AI companies, and their widespread political and financial influence.

• The Economics of Data Centers Creating Jobs Are So Bad That They Sound Like a Joke

  In theory, more tax revenue means money for schools, roads, and first responders. In practice, states and counties are waving those taxes in order to entice tech companies to set up shop — while existing campuses put major strain on local communities. One analysis found that Georgia, Virginia, and Texas each lose over $1 billion a year to state data center incentives, while at least fourteen states don’t even disclose data center tax breaks in the first place.

  The down-stream impacts don’t really materialize either. Researchers at Georgia Tech found that in rural areas, data centers typically employ fewer than 100 permanent workers and are likely to import specialized services from outside the community.

• The Lies They're Telling Towns And Tribes About The Benefits Of AI Data Centers

  These organizations are useful to corporate power because they parrot the goals and aims of corporate power, but they do so under a pseudo-intellectual, pseudo-academic veneer designed to trick news audiences into believing they're reading the scientific insights of an objective subject matter expert.

  I've run into the PPI consistently during my research on broadband. They can often be found breathlessly insisting that modest oversight of shitty telecom monopolies will stifle innovation, or that popular, community-owned and operated broadband networks are some kind of reckless socialism run amok.

  Anyway, the Institute's spokesperson, Mr. Mandel, is lying. Plunking down an AI data center in the middle of rural West Virginia will not magically draw a bunch of innovative new startups to the region. And the 5ms or so lower latency from having a server farm in your backyard offers no tangible difference to AI service quality.

  That's simply not an actual thing.

• No ban against data centre development after Hamilton council votes down bylaw

  Hamilton won't be imposing a temporary ban on the development of data centres, council has decided.

  The vote was split 10 against and six in favour of what they were calling an interim control bylaw that would have paused development on data centres in the city for one year. Such a ban would have given the city time to assess and develop further governance rules around facilities that house equipment such as network servers, storage drives and information technology (IT).

• Massive data center wins local tax break — but 9 times smaller than requested

  Saline Township had been bound by a court order to approve a 12-year, 50% tax break for “The Barn,” a 250-acre project developed by The Related Cos. for Oracle and its customer, OpenAI.

  But the township said the deal, which followed a lawsuit by Related, had been struck when the project was valued at $4.8 billion — and that’s the amount the township voted to approve as the basis for the tax break.

  Residents spoke against the project for nearly two hours, just as they had a week earlier when the township decided to delay the vote until Tuesday.

• Jackson Puts Six-Month Pause on Data Center Construction

  After weeks of residents appearing before the Jackson City Council to voice concerns and warnings about the effects of data center construction, the council passed a six-month moratorium on data center-related decisions at its Tuesday meeting, following a final public comment session. The moratorium passed with a 5-2 vote, with Ward 1 Councilman Ashby Foote and Ward 5 Councilman Vernon Hartley, the Council’s new president, opposed.

  Although the moratorium did not come to a vote until Tuesday, the city already promised that zoning changes and permit approvals for any potential data centers will be delayed until Jackson passes regulations on data centers at its first related public hearing on June 22.

• NY state pauses new large data center projects in US first

  AI already accounts for between 15 and 20 percent of electricity consumption at data centers, and this share could climb to 40 percent by 2030, the report said.

• Google Announces Support for Nation’s Largest Solar Farm, Located in Arkansas

  Google is building an AI data center in West Memphis, Arkansas, across the Mississippi River from Memphis. (I know, it’s a very original city name.) Google joins SpaceXAI and others in the Mid-South for cheap land and abundant water, but it faces the same core challenges — mainly, how to feed its huge collection of power-hungry hardware.

• Generative AI Is an Engineering Disaster

  The memory is being put into data centers, which tech firms are expanding at incredible speed. They are planning to multiply total U.S.-data-center capacity by a factor of eight over the next few years. The demand for electricity at these sites is already so great that some companies are repurposing jet engines to power them.

  The problem is not simply that AI is being deployed so widely or quickly. Other computer technologies have seen similarly massive growth without triggering such a large spike in electricity or a shortage of computer components: Video and music are now streamed around the globe, accounting for many terabytes of internet traffic daily; the smartphone boom required the manufacturing of billions of devices that are now transferring huge amounts of data; billions of household devices are also now part of the Internet of Things; and whole industries have moved their operations to cloud software, which is hosted not in the sky but in, yes, data centers.

• AI data centers must produce as much power as they use, Australia PM says — new national AI framework will also ensure water efficiency and protect intellectual [sic] property [sic] rights

  Aside from concern for data centers’ use of natural resources, the Australian government also wants to ensure that the intellectual [sic] property [sic] rights of its people are protected. Albanese said that Australian creators, including writers, musicians, artists, and news reporters, should “retain control of the price and value of their work” when used for AI training. “Anything less is theft. No country has got this right yet,” says the prime minister.

• Australia to Put Environmental Brakes on A.I. Data Centers

  The announcements on Wednesday in a speech by Prime Minister Anthony Albanese came as Australia draws significant interest from A.I. companies because of its size and the availability of renewable energy, and as resistance to data centers builds in many parts of the United States and Europe.

• NY Is ‘First’ State to Pause Large Data Centers

  New York has become the ‘first’ state to impose a one-year moratorium on new large-scale data centers (50 MW or larger), according to Datacenter Digest. Signed by Governor Kathy Hochul, the one-year pause halts environmental permits while the state develops regulations to evaluate data centers’ impacts on energy demand, water use, air quality, and local communities.

  The new framework includes a statewide environmental impact study and guidance for local governments to negotiate community benefits with developers. The state is also considering requiring data centers to help fund grid upgrades and may eliminate sales tax exemptions for these facilities.

• Elon Musk’s Colossus 2 data center installed 59 natural gas turbines without permission, report claims — thousands of tons of pollutants reportedly impact black communities in Mississippi already suffering from elevated lung disease rates

  Elon Musk's Colossus 2 xAI data center, which runs independently of the power grid through on-site natural gas turbines, is said to be releasing thousands of tons of nitrogen oxide and carbon monoxide every year. According to Reuters, the company has installed 59 temporary, mobile natural gas turbines without permission, and these unpermitted portable units mostly heavily affect the communities surrounding the site, which happen to be predominantly black and are already suffering disproportionately high lung disease rates. xAI has claimed that it’s running 27 turbines without any permits, saying that it’s exempted because of their temporary nature.

• AI in Australia's interests | Prime Minister of Australia

  The Government will introduce a set of Australian Standards for AI, building on the Data Centre Expectations, with our values as the benchmark to ensure AI works in Australia’s interests.

  The new standards will set out clear rules for large data centres— including a legal obligation to underwrite their own new power supply, pay their full share of connection costs so energy bills are not impacted, reduce power when needed to strengthen the grid, and be as water efficient as possible.

• Eric Trump Has Lost $600 Million of Daddy's Money Betting on [Cryptocurrency]

  After merging with an existing mining company, the company made the questionable decision last year to pivot from building data centers — an extremely lucrative industry that has seen massive interest as of late — to [cryptocurrency] mining, a gambling-like practice that’s far from guaranteed to make a profit.

• Memory card prices might not fall until after 2030

  The chip shortage is due to the unprecedented demand for HBM (high bandwidth memory) chips, which are being voraciously consumed by AI datacenters. SK Hynix produces around 58% of the global HBM chip supply and is investing heavily in in new semiconductor production facilities, with $4 billion allocated to build an advanced chip packaging plant in Indiana.

• ‘Slough is like an experiment’: Europe’s largest datacentre hub leaves town sweltering

  Ten miles (16km) west of Heathrow, Slough has become one of the largest datacentre hubs in the world, hosting an estimated 30 to 40 huge facilities, many of them on a campus in the centre of town. These – owned and maintained by companies like Equinix and Digital Realty – serve dozens of clients, including the world’s biggest tech companies: Amazon, Google, Oracle and Microsoft.

  More are still to be built in a planned development on the edge of the same campus.

  Emerging research – including a preprint of a paper led by researchers at Cambridge earlier this year – suggests datacentres create a heat island effect, pushing up temperatures in their immediate vicinity by an average of 2C, and as much as 9C.

• Wyoming Residents Join National Day Of Protest Against Data Centers

  A few dozen Wyoming residents gathered at the State Capitol on Saturday as part of a national day of protest against data centers. Organizer Taylor Haynes called data centers "digital prisons" and said the need for artificial intelligence has taken away Americans’ freedom.

• Data center protests go national as backlash grows

  • Protests against data centers occurred in 125 locations nationwide on Saturday.
  • HumansFirst, a grassroots group, criticized data centers' unaccountable expansion.
  • Polls show only 14% support data centers in communities for AI projects.

• How data centers became a symbol of Americans’ rage

  Feeling ignored by political and economic powers, people are rallying against computer hubs over concerns about their communities’ future.

• Data Centers Are Making Electricity Brutally Expensive for the Public

  According to a report by Monitoring Analytics, an independent monitor for the largest transmission company in the US, PJM, data center demand is expected to drive over $23 billion in customer price increases by 2028. That eye-watering figure, first spotted by Fortune, is a direct result of the country’s old and confusing electrical infrastructure, which ultimately leaves regular people, not the tech industry, holding the bag.

• Family Horrified When They're Forced to Sell Their Beloved Home to Make Way for AI Data Centers

  Like a tornado descending from the heavens, data centers are tearing through homes across the US, roaring like a jet engine, leaving schools without power, and decimating entire communities unlucky enough to stand in their path.

  Case in point, one Georgia family has been forced to sell their home, after the state utility concern decided it needed to build a new transmission line to connect new AI data centers.

• My land is my land

  If Mississippi utilities ever use eminent domain against a farmer or homeowner to build a transmission line for a hyperscale data center, it will be no different in principle from what happened to Susette Kelo in New London or what is happening to Ansley Brown in Georgia. The state will be taking someone’s land for someone else’s profit.

• Georgia family says they're forced to sell home to help power AI data centers: "It's theft"

  Utility giant Georgia Power is planning to build a new transmission line to in part help power new data centers. It estimates 70-80% of the power on the new line will help serve data centers and the remaining 20-30% of power will serve the state's growing residential and commercial demand.

  According to the company, increasing demand has outpaced the capacity of its existing grid and building a new transmission line requires acquiring more than 300 parcels of land, including residential properties.

• Climate Activists Pelt Microsoft Data Center With Balloons Full of Acid

  “This acidic mixture attacks concrete, and the hydrogen peroxide causes steel to rust faster,” Extinction Rebellion said in a statement.

  What physical impact the balloons had on the data center itself is unknown, but it may be beside the point. Extinction Rebellion often uses nonviolent civil disobedience and direct action to draw attention to climate issues that would otherwise be ignored. Previous actions involved spattering the offices of fossil fuel insurers with foul smelling butyric acid, and using superglue and bicycle locks to barricade the entrance of Lloyd’s, a major insurance market in London.

• Who’s Building That Data Center?

  Whether you’re pro, against, or ambivalent about the proliferation of “AI” data centers, the map offers an engaging way to look at what projects are happening around the nation, especially when you start looking at clusters and how that interacts with the power generation and political makeup in a region. It’s particularly interesting how only three states account for roughly 70% of all the projects. Let us know if there’s a similar tracker in your area if you’re from one of the other parts of the globe!

• Project Map — PoweredByWho

  Public-record reporting on the data centers reshaping American grids, communities, and elections. Independent, non-partisan, no paywall.

• Distilling The Moat

  AI companies' intellectual [sic] property [sic] is their models. They spend vast sums funding the technical and human resources to "train" these models, the racks of GPUs in the data centers, and the hordes of workers labeling images, and having "genuine human conversations" with the nascent model. These expenditures are thought to create a "moat" around the value thus generated, because it would be equally expensive for a competitor to create an equivalent model. It is this moat that supports their extraordinary valuations, despite their lack of earnings.

• AI data center builder Fluidstack raises $830M at $7.5B valuation

  The company announced the Series A round on Monday. The lead investor was Situational Awareness, an AI-focused fund backed by Stripe Inc.’s founders. The investment values Fluidstack at $7.5 billion.

• Africa’s data center growth needs power sector overhaul

  The potential upside is significant: The International Monetary Fund, in a paper published on Tuesday, said AI could make sub-Saharan Africa’s economy around 4% bigger over the next decade than its baseline projection — but that requires better ‌electricity supply, internet access, and digital skills. Policymakers and business leaders have long warned that Africa’s power shortfall in particular, if left unaddressed, would starve energy-hungry data centers of the electricity needed to digitize economies.

• Blaming Data Centres, Alberta Electricity Industry Expects Tripling of Power Prices

  White reports this would increase the electricity costs to a typical household by “hundreds of dollars more per year”. However, as White caveats, these prices have precedent as recently as three years ago — and that undersells it. In 2022, electricity costs were as much as five times above current rates. Even without adjusting for inflation, we are currently paying less than we did from 2018–2020, and a tripling of the current rate would have a nominal cost similar to that of 2008. That is not to say this is good, but I think it is worthwhile seeing the “tripling” figure within the context of recent fluctuations.

• Pluralistic: Trump’s America can’t even win a rigged game (22 Jul 2026)

  After that speech, I wrote a book (The Post-American Internet) that Farrar, Straus and Giroux will publish in August 2027. In the book, I describe the role that "trusted third parties" (T3P) play in complex transactions. Think of an escrow agent who holds onto the deed for the house you're buying from the seller until you hand over the money, and then forwards the deed to you and the money to the seller.

  The more complex a transaction is, the more it needs a T3P. For most of the past century, the US has been the world's T3P. Most of the world's transoceanic fiber optic lines make landfall in the US and interconnect to one another in US data centres. Most of the world's international transactions are conducted in dollars and are cleared through US-controlled platforms like SWIFT. Through its aid programs, the US sets the health and public services agenda for billions of non-Americans, and the US has military installations in more than 100 countries. The US trains the world's militaries, it supplies (and withholds) information from the world's intelligence agencies, and it runs the IT infrastructure powering the world's government agencies and critical infrastructure, from tractors to medical equipment.

• Send the data centres north

  The Northern Territory offers significant renewable-energy potential, abundant land, proximity to Indo-Pacific markets and increasing strategic importance. It is also central to Australia’s defence posture and is emerging as a hub for critical minerals processing, space activities, undersea communications and industrial development. While each of these investments serves a distinct purpose, together they form a strategic industrial and national-security ecosystem – one that data centres would strengthen and draw strength from.

• The AI Boom Is a Corporate Project. It Can Be Disrupted

  A lot of AI technology may not work very well, but providing consumers with reliable information was never its major commercial promise. Investors have poured billions into generative AI because of the promise that it will eliminate jobs on a massive scale. Tech journalist Brian Merchant joins Movement Memos host Kelly Hayes to discuss how the AI boom is being used to displace workers, degrade information, expand surveillance and strengthen state violence—and why this brittle corporate project can be disrupted.

  [...]

  What happens when a technology doesn’t have to work particularly well to remake the world? Today, we’re talking about AI in the workplace, Google’s decision to ruin its search engine, robotaxis getting people arrested, Flock cameras getting destroyed, the movement against data centers, and what it means to build power against the tech industry.

• scenes from the bike trail

  The AI bubble is giving us a lot of data centers stocked with GPUs that can only do two things: generative AI and crypto mining. The buildings themselves are massive, but often poorly served (if at all) by mass transit lines. Not like plants that actually make things, which are often served directly by railroad lines or are within a short distance of a port (sea or air).

• Whose AI? Whose Opposition?

  This framing raises serious questions about whether legitimate (“organic”) domestic opposition to AI data centers may risk being discredited and dismissed as the product of foreign influence campaigns.

• Critics of AI Aren’t Anti-Tech Extremists

  Recent US intelligence documents hype up the threat of “anti-tech violent extremist activity” against data centers. Casting protests as ideologically extreme is a shortcut to silencing critics of AI and removing democratic constraints on Big Tech.

• Pluralistic: AI solipsists and AI cynics (24 Jul 2026)

  In service to this bubble, AI companies have suborned regional governments into running roughshod over environmental and planning review in order to build endless acres of data centers, many of which will likely end up casualties of the imminent bubble-pop, never to be switched on or even completed. What an indignity to have your farm or house seized through eminent domain, only to see it razed and replaced by a weed-choked empty field, a lonely foundation slab, or an abandoned empty building that could only ever be repurposed for laser-tag or an ICE concentration-camp: [...]

• Trump pledge on data center power supplies draws skepticism

  Consumer advocacy groups and other ​administration critics, however, dismissed the pledge as an empty promise to consumers as Trump tries to balance his support for power-guzzling data centers against public anger over rising utility bills.

• Clip & Save: Trump Says People Are “Begging” to Live Near Data Centers

  Trump made the remarks at an event promoting the White House’s “Ratepayer Protection Pledge,” in which states and companies pledge to keep consumers from paying higher electricity bills due to data centers. The plan has supposedly gotten the support of companies such as Google, Microsoft, Meta, Oracle, xAI, OpenAI, and Amazon, who vowed to pay for increased power costs in areas where data centers are built. The pledge is completely voluntary—and unenforceable.

  People who live near data centers complain about the overhaul to their lives, from the drain on water supplies to increased noise and higher utility costs. In most areas where data centers are proposed, local opposition quickly becomes protests that unite the left and the right. The economic benefit that comes with data centers is actually minimal, with most providing only a few dozen on-site jobs. Trump is very much out of touch with how Americans, including many of his supporters, view data centers.

• A Cautionary Tale

  This CNBC video provides a look at what happens when an AI data center is built in your neighborhood.

• Right-Wing Bill Mill Meeting Centers Voter Suppression, Protecting Data Centers

  One of the most concerning is the Critical Infrastructure Security Act, which would expand ALEC’s 2018 Critical Infrastructure Protection Act to include enhanced penalties for trespassing at data centers and “government facilities” with the intent of damaging property, inflicting bodily harm, “surveillance of nonpublic security measures,” or “another felony.”

• AI gets pinkwashed as female users are targeted

  While large AI companies are being besieged by a more conscious and critical perception of the risks and impact of their technologies, more and more corporations such as Meta, Google, Anthropic and Open AI are investing in what is being termed “Girlbossification of AI.” If the recent booing of tech CEOs during their speeches at graduations wasn’t evidence enough that not everyone wants or needs AI in their lives – a recent New York Times poll found that 47% of 18-to-29-year-olds consider AI to be a very bad thing – coupled with the rise of anti-data center legislative proposals and the growth of anti-AI movements around the world, the female user is now being schmoozed.

  In a clear example of pinkwashing that has been going on for months, companies have invested in communication campaigns involving businesswomen, fashion and beauty influencers, well-known actresses, and even viral challenges aimed at capturing the attention of their most precious target: educated, professional and socially mobile women.

• OpenAI Close to Landing $500 Billion Data Center With Backing From Nvidia

  OpenAI is close to leasing a $500 billion data center in southern Ohio that could be among the largest of the artificial intelligence boom [sic], according to three people familiar with the project.

• [Old] The Dangers of Data Centers

  Data Centers are large facilities containing computer servers used for data storage, data analytics, generative AI, and streaming services. Data centers represent health risks for their neighbors. These risks are especially high from hyperscale data centers powered by fossil fuels, such as those proposed for some parts of Pennsylvania. Listed below are some of the top problems data centers impose on nearby communities.

• The fight against AI datacenters is important – but it’s just a starting point

  Local datacenter opposition is grounded in legitimate concerns about misallocation of land resources when housing is at a premium, pressures on already higher energy prices, and localized environmental impact. Unlike other resource-consuming and polluting industrial facilities, datacenters produce very few jobs. The fact that US opposition to datacenters seems to be most fierce among lower-income communities reflects righteous indignation with an inequitable bargain, where tech companies and developers profit from exploiting local resources but offer little in return. On a global scale, their carbon footprint could grow unsustainably if usage accelerates. And all this is in aid of a technology that many fear will propagate misinformation, take their jobs, or even cause existential risks for humanity.

  For some, datacenter opposition may feel like the only tangible mechanism for registering their concern, disapproval, or even anger about AI. The problem is that this may be exactly what the AI companies are banking on. They can overcome the protest when it matters to them, and live with a significant fraction of proposals being defeated. More importantly, focusing political opponents on the datacenter issue obscures the bigger prize they’re after.

• Data center boom sparks lobbying surge

  More than 200 entities lobbied between April and June on proposals to prevent data center electric demand from spiking utility rates, a nationwide moratorium on new facilities and the Federal Energy Regulatory Commission’s actions around how to connect the massive electricity users to the grid, a rev[1;5Hiew of new federal lobbying disclosures by POLITICO showed.

  It’s a sharp increase from the same quarter in 2025, when fewer than 100 companies and groups reported lobbying on data center matters.

• Desert data center developer challenges county moratorium in court

  The project has become a flashpoint in Imperial County, where water-intensive industrial projects like data centers are seen by some as incompatible with the region’s limited water resources and agricultural economy. Rucci, however, claims the data center will provide new jobs and millions in tax revenue in the economically stagnant region.

  The Imperial County Board of Supervisors initially approved Imperial Valley Computer Manufacturing’s lot merger for the project, located near a neighborhood in an industrial area just south of the city of Imperial. But the plan was walked back in June when the county approved the 45-day moratorium on data centers, which was then extended to 10 months and 15 days — or until June 2027 — so that the county could evaluate its land use rules.

• Five things to know as data centers spread across rural America

  More than 1,500 data centers are planned nationwide, many in rural communities. Here’s what to know about their water use, environmental impact and the growing opposition.

• Dozens of utilities join pledge pledge to protect consumers from high bills from AI data centers

  But AI’s increasing ability to perform basic tasks — such as driving, analyzing spreadsheets and writing software — also potentially threatens millions of jobs. That has created mounting public resistance as tech companies concentrate historic levels of wealth in the hands of a select group of tycoons.

  The increased electricity demand could cause monthly utility bills to rise by 15% to 40% by 2030, according to a recent analysis by ICF, a consulting and technology services company.

• Data Centers Are Easy to Build. Powering Them Is Complicated, Slow, and Expensive

  To bypass the issue of grid demand, many data center builders have promised to power the buildings themselves. The problem with that plan is that it’s easy to build a warehouse full of GPUs. Building power plants and transmission lines to power those warehouses will take years and cost billions more than the data centers. A skilled and efficient builder can complete a data center construction in under a year. Building new energy generation to meet that data center's power needs could take a decade.

• Google goes cash flow negative for the first time as AI data center buildout increases capex to a staggering $44.9 billion in a single quarter — CFO warns that capex will increase in 2027 as company banks big on TPUs

  Google's parent company Alphabet recently reported negative free cash flow of $5.9 billion for the second quarter of 2026, the company's first cash-negative quarter since its 2004 IPO, after capital expenditures doubled year-over-year to a record $44.9 billion and exceeded the $39.1 billion its operations generated as it continues its rapid buildout of AI data centers, according to its earnings release.

• Wrangling Datacenter GPUs Into A Desktop

  Even though these don’t perform at the same level as the latest top-tier online models, they’re surprisingly capable and also have the benefit of running completely locally. This might be concerning for those looking at the global economy being propped up by companies that essentially have no moat for motivated users, especially as more and more datacenter hardware becomes available on the secondhand market. While this build by [Andrew] goes into detail on getting the software stack up and running, we recently featured another build using the same GPUs that focuses a bit more on hardware for those looking to get started with local hosting.

• Verizon Signs $1 Billion Google Fiber Deal to Support AI Growth

  Verizon said its more than 500,000 route-mile fiber network positions the company to meet rising demand for high-capacity, low-latency connections between AI data centers. The carrier is also expanding long-haul fiber routes and repurposing former central offices into AI-ready data center sites as it continues replacing legacy copper networks with fiber.