infrastructure Archives

Musk runs into a snag on data center gas turbines — they are now illegal

xAI fits gas turbines on flatbed trucks and calls them «temporary» to avoid regulation. No more, says the EPA. (Picture: generated)

Apparently, a new EPA rule on stationary and «temporary» gas turbines for energy generation has made them illegal, according to CNBC.

They produce much too high levels of nitrogen oxides, and must be regulated as combustion engines, the EPA says.

That means Musk’s and xAI’s Colossus plant will have to rethink their energy use, as they make widespread use of natural gas turbines to generate electricity for their facilities.

They will basically have to get Clean Air Act permits, and prove they aren’t harmful.

The local population have long been complaining of a rotten-egg-like stench in the atmosphere, CNBC writes, and smog is supposedly prevalent.

xAI uses 15 turbines for Colossus 1 and at one point had 59 turbines for Colossus 2, The Guardian writes.

Read more: The EPA Rule. Writeups on CNBC, Gizmodo and The Guardian.

OpenAI to spend $10 billion on compute from AI chip startup Cerebras

Cerebras makes powerful inference chips, for when an AI needs to think a little deeper. (Picture: generated)

The deal will land OpenAI with an added power capacity of 750 megawatts — but it’s not just any kind of compute.

Cerebras makes wafer-scale inference chips (for generating the response an AI gives after a query), looping in networking and high-bandwidth memory on the same die.

This makes for much faster thinking on complex tasks, should enable real-time reasoning, and OpenAI could possibly route those kinds of queries to this kind of compute.

The deal is worth over $10 billion, Reuters writes, and the capacity should come online in multiple tranches to be fully delivered in 2028, OpenAI says.

Sam Altman is an early investor in the company, and OpenAI once considered buying it, TechCrunch reports.

Read more: OpenAI’s announcement, writeups on Reuters, CNBC and TechCrunch.

Zuckerberg unveils «Meta Compute» initiative to secure infrastructure

Meta is serious about investing in data centers in the coming years. (Picture: generated)

The Meta CEO announced his new «top level» compute team on threads, saying «Meta is planning to build tens of gigawatts this decade.»

He also said they will expand to «hundreds of gigawatts,» «over time.»

— How we engineer, invest, and partner to build this infrastructure will become a strategic advantage, Zuckerberg continues.

There are lots of obstacles to build global infrastructure like this, and challenges in «operating our global datacenter fleet and network,» Zuckerberg says.

The team will be headed up by Meta’s head of global infrastructure Santosh Janardhan and Daniel Gross, in close collaboration with newly hired Dina Powell McCormick.

Meta hasn’t released a frontier model since the Llama series in April, 2025, which saw lots of controversies. They have since formed a Superintelligence Lab from across the industry to develop the next generation of models. There is no timeline for when they’ll be ready, but the ambition is real, Zuckerberg says:

— [we want to] deliver personal superintelligence to billions of people around the world, he closes his message.

Meta buys into 6 GW of nuclear to power its Prometheus data center

Meta is betting on nuclear for its AI power needs in Ohio, US. (Picture: Adobe)

Three nuclear power companies have been selected by Meta to supply its «supercluster» being built in New Albany, Ohio.

Together, they are offering up to six gigawatts of power from established nuclear reactors to more experimental small modular reactors.

The deals are a result of a request for proposals from Meta in December 2024, TechCrunch writes, and they have now selected Vistra (2.1 GW) for their existing nuclear plants, and Oklo (1.2 GW) and TerraPower (up to 2.1 GW) for their modular reactors.

TerraPower was co-founded by Bill Gates, while Sam Altman is the biggest investor in Oklo.

TerraPower and Oklo are still in the startup phase, with capacity for TerraPower expected to come online in 2032, while Oklo’s «nuclear technology campus» won’t be coming online until 2030, CNBC writes.

No financial details of the deals have been made public, but Meta has committed to spending $600 billion on infrastructure over the next three years.

Read more: Writeups on TechCrunch, CNBC.

Nvidia responds to report that Meta might use Google’s TPU chips

A deal between Meta and Google might see custom TPU chips installed on Meta's data centers, as the industry looks for Nvidia alternatives. — Talks between Meta and Google on custom chips has Nvidia slightly worried, but not much. (Picture: generated)

News on the talks was posted by The Information on tuesday, and caused Nvidia to drop 3% and Google to tease a $4 billion valuation in the markets.

Under the deal being discussed, Meta would start renting compute on Google’s Tensor Processing Units as early as next year, Reuters reports.

Continue reading “Nvidia responds to report that Meta might use Google’s TPU chips”

Cloud honcho at Google says they need to double capacity every six months

Google needs to reach 1000x capacity in the next 4-5 years, they say at an all hands meeting. — Google is not immune to the ever increasing demand on AI infrastructure. (Picture: generated)

The demand for AI infrastructure is «the most critical and also the most expensive part of the AI race,» said Vice President at Google Cloud, Amin Vahdat, at a recent all hands meeting, reported by CNBC.

This comes as almost every other Big Tech company is increasing data center spending and Google has set aside $93 billion in «capital expenditures» this year to do the same.

This will be followed by a «significant increase» in 2026, but likely not matching OpenAI’s enormous $1.4 trillion data center spending.

They are aiming to «spend a lot,» and hit «the next 1000x in 4-5 years,» Vahdat is reported to have said.

That’s one thousand times more «capability, compute and storage networking» that he aims to add for «essentially the same cost, power and energy spend.»

Read more: The scoop at CNBC, writeups at Ars Technica and Gizmodo.

Anthropic partners with Nvidia, Microsoft in $30 billion Azure deal

Anthropic nets a Microsoft investment, and promises to use it all on Microsoft's Azure. — Nvidia and Anthropic will optimize for each other, and use Microsoft’s cloud. (Picture: generated)

The companies will invest some $15 billion in Anthropic, while the AI lab commits to $30B in spending on Azure, Microsoft’s cloud offering.

— We’re increasingly going to be customers of each other. We will use Anthropic models, they will use our infrastructure and we’ll go to market together, Microsoft CEO Satya Nadella said, according to Reuters.

The agreement means Anthropic will optimize their software stack to better run on Nvidia’s hardware, while Nvidia will «optimize for Anthropic workloads.»

Continue reading “Anthropic partners with Nvidia, Microsoft in $30 billion Azure deal”

Google invests $40 billion in Texas data centers

Google will invest $40 billion on three data centers in Texas, along with an energy buildout of some 6 gigawatts and a new solar plant. — Texas Gov. Greg Abbott and Sundar Pichai announcing the news. (Picture: Google)

Google joins the big players in building out staggering AI capabilities, with a new data center in Armstrong and two in Haskell County, near Abilene, hooked up to a solar and battery plant.

— This investment will create thousands of jobs, provide skills training to college students and electrical apprentices, and accelerate energy affordability initiatives throughout Texas, Alphabet CEO Sundar Pichai said, according to Reuters.

The investments will be made through 2027, but Google says nothing of when they will come online.

They will also bring new funding for the power grid to support 6 gigawatts of «new energy generation and capacity» and will support some 1,700 new electrical apprenticeships with Google support.

This makes Texas the second largest data center state in the USA, after Virginia, notes the Texas Tribune.

— They say that everything is bigger in Texas — and that certainly applies to the golden opportunity with AI, Pichai said.

Read more: Google’s blog, writeups on Reuters, Texas Tribune.

Meta commits to «at least» $600 billion spend on AI infrastructure

Meta commits to spending big its AI infrastructure investments in the next three years. — Meta will be a big AI infrastructure player if their investments pan out. (Picture: Adobe)

The outlay is only for next three years, according to Reuters, who says Meta has been doubling down on AI recently, with massive outlays for their Superintelligence Lab and data center buildouts.

They are building out data centers because «it’s the right strategy to aggressively front-load capacity so we’re prepared for the most optimistic cases,» according to CEO Mark Zuckerberg.

Meta saw an 83% drop in its operating income the last quarter, and plans for capital expenditures of $70 billion just this year, but will have increase it threefold to reach their target by 2028.

Read more: Reuters, The Register and Business Insider.

Weekend roundup: Gemini is going to space, Apple chooses Google, and Amazon’s had it with Perplexity

Gemini is launching to orbit
Google’s latest moonshot might almost be literal. They are preparing for sending their TPU processors into low-earth orbit, and maybe then build a proper AI data center in space — where there is ample sunlight to provide it with energy. They have already tested a TPU in orbit conditions in a particle accelerator and it survived, and the next step is the launch of two prototype satellites in early 2027. They call it Project Suncatcher, and say that «in the future, space may be the best place to scale AI compute.»
More at: Sundar Pichai’s tweet, Google’s announcement blog

Google close to Apple deal for AI Siri
Apparently, Apple has chosen Gemini for its upcoming AI version of the Siri assistant. They will use what is likely a custom version of the model with 1.2 trillion parameters, running on Apple’s Private Cloud Compute servers. Apple supposedly also tested options from OpenAI and Anthropic, but Anthropic’s fees were too high and Apple already partners with Google for search results. The deal will cost Apple $1 billion a year, far less than the $20 billion Google pays Apple to be their search provider.
More at: Bloomberg, MacRumors, TechCrunch.

Read on for more!

Continue reading “Weekend roundup: Gemini is going to space, Apple chooses Google, and Amazon’s had it with Perplexity”

Big Tech doubles down on even bigger AI spend

Big tech will spend about a quarter of a trillion dollars on "capital expenditures" this year, according to quarterly reports. — Data center deals are flourishing and none of the big tech spenders feel they can afford not be in the race. (Picture: Adobe)

Quarterly results are in for Microsoft, Alphabet and Meta — and while the numbers are mixed, they all agree on big capital expenditures — needed for building data centers — for fiscal year 2025.

The coming AI wars will be fought with data centers and gigawatts, and nobody wants to lose out.

Continue reading “Big Tech doubles down on even bigger AI spend”

Broadcom to supply OpenAI with 10 GW’s worth of custom chip capacity

With the deals annouced in just a couple of weeks, OpenAI will, along with Stargate, add a huge level of compute. It will all be online in late 2026. — OpenAI is adding an astonishing level of compute infrastructure over the coming years, and show no sign of stopping. (Picture: generated)

After first announcing their collaboration on an OpenAI designed chip in early September, Broadcom now says it’s ready to deliver.

Starting up in late 2026, just like the AMD and Nvidia deals — OpenAI will have added 26 gigawatts of capacity from these agreements alone, and one can wonder how capable the future of GPT will be.

Continue reading “Broadcom to supply OpenAI with 10 GW’s worth of custom chip capacity”

Altman says more, «aggressive» deals to come off heels of AMD agreement

OpenAI is going to aggressively pursue data center deals in the "coming months." — Thinking two years in the future, Altman sees a massive demand for more number crunching. (Picture: screenshot).

OpenAI recently signed deals with both AMD and Nvidia that were almost polar opposites, but their appetite for data center capacity shows no sign of abating.

Their CEO has previously said that they aim for adding 1 gigawatt per week in the future, adding about a Hoover dam every fourteen days — and says many more deals are yet to come.

In the A16z podcast by Andreessen Horowitz today, Altman says «You should expect much more from us in the coming months», as reported by TechCrunch.

Continue reading “Altman says more, «aggressive» deals to come off heels of AMD agreement”