SoylentNews

upstart writes:

Endgadget reports that Meta is Building "Several" Multi-Gigawatt Compute Clusters

Meta is building several gigawatt-sized data centers to power AI, as reported by Bloomberg. CEO Mark Zuckerberg says the company will spend "hundreds of billions of dollars" to accomplish this feat, with an aim of creating "superintelligence."

The first center is called Prometheus and it comes online next year. It's being built in Ohio. Next up, there's a data center called Hyperion that's almost the size of Manhattan. This one should "be able to scale up to 5GW over several years." Some of these campuses will be among the largest in the world, as most data centers can only generate hundreds of megawatts of capacity.

Meta has also been staffing up its Superintelligence Labs team, recruiting folks from OpenAI, Google's DeepMind and others. Scale AI's co-founder Alexandr Wang is heading up this effort.

However, these giant data centers do not exist in a vacuum. The complexes typically brush up against local communities. The centers are not only power hogs, but also water hogs. The New York Times just published a report on how Meta data centers impact local water supplies.

upstart writes:

"This is a giant project," Nvidia CEO said of new 10-gigawatt AI infrastructure deal:

On Monday, OpenAI and Nvidia jointly announced a letter of intent for a strategic partnership to deploy at least 10 gigawatts of Nvidia systems for OpenAI's AI infrastructure, with Nvidia planning to invest up to $100 billion as the systems roll out. The companies said the first gigawatt of Nvidia systems will come online in the second half of 2026 using Nvidia's Vera Rubin platform.

"Everything starts with compute," said Sam Altman, CEO of OpenAI, in the announcement. "Compute infrastructure will be the basis for the economy of the future, and we will utilize what we're building with NVIDIA to both create new AI breakthroughs and empower people and businesses with them at scale."

The 10-gigawatt project represents an astoundingly ambitious and as-yet-unproven scale for AI infrastructure. Nvidia CEO Jensen Huang told CNBC that the planned 10 gigawatts equals the power consumption of between 4 million and 5 million graphics processing units, which matches the company's total GPU shipments for this year and doubles last year's volume. "This is a giant project," Huang said in an interview alongside Altman and OpenAI President Greg Brockman.

To put that power demand in perspective, 10 gigawatts equals the output of roughly 10 nuclear reactors, which typically output about 1 gigawatt per facility. Current data center energy consumption ranges from 10 megawatts to 1 gigawatt, with most large facilities consuming between 50 and 100 megawatts. OpenAI's planned infrastructure would dwarf existing installations, requiring as much electricity as multiple major cities.

[...] Bryn Talkington, managing partner at Requisite Capital Management, noted the circular nature of the investment structure to CNBC. "Nvidia invests $100 billion in OpenAI, which then OpenAI turns back and gives it back to Nvidia," Talkington told CNBC. "I feel like this is going to be very virtuous for Jensen."

hubie writes:

Cast-off turbines generate up to 48 MW of electricity apiece:

Faced with multi-year delays to secure grid power, US data center operators are deploying aeroderivative gas turbines — effectively retired commercial aircraft engines bolted into trailers — to keep AI infrastructure online.

According to IEEE Spectrum, facilities in Texas are already spinning up units based on General Electric's CF6-80C2 and LM6000, the same turbine cores once found on 767s and Airbus A310s. Vendors like ProEnergy and Mitsubishi Power have turned these into modular, fast-start generators capable of delivering 48 megawatts apiece, enough to support a large AI cluster while utility-scale infrastructure lags.

Fast, loud, and anything but elegant, these "bridging power" units come from vendors like ProEnergy, which offers trailerized turbines built around ex-aviation cores that can spin up in minutes to meet energy demand. Meanwhile, Mitsubishi Power's FT8 MOBILEPAC, which derives from Pratt & Whitney jet engines, delivers a similar output in a self-contained footprint designed for fast deployment.

While this might not be the cheapest, and certainly not the cleanest, way to power racks, it's a viable stopgap for companies racing to hit AI milestones while local substations and modular nuclear power deployments remain years away.

[...] In one of the more visible examples, OpenAI's parent group is deploying nearly 30 LM2500XPRESS units at a facility near Abilene, Texas, as part of its multi-billion-dollar Stargate project. Each unit spins up to 34 megawatts, fast enough to cold-start servers in under ten minutes.

Also see: Data Centers Look to Old Airplane Engines for Power

Original Submission

Microsoft: the Company Doesn't Have Enough Electricity to Install All the AI GPUs in its Inventory

upstart writes:

Microsoft CEO says the company doesn't have enough electricity to install all the AI GPUs in its inventory - 'you may actually have a bunch of chips sitting in inventory that I can't plug in':

Microsoft CEO Satya Nadella said during an interview alongside OpenAI CEO Sam Altman that the problem in the AI industry is not an excess supply of compute, but rather a lack of power to accommodate all those GPUs. In fact, Nadella said that the company currently has a problem of not having enough power to plug in some of the AI GPUs the firm has in inventory. He said this on YouTube in response to Brad Gerstner, the host of Bg2 Pod, when asked whether Nadella and Altman agreed with Nvidia CEO Jensen Huang, who said there is no chance of a compute glut in the next two to three years.

"I think the cycles of demand and supply in this particular case, you can't really predict, right? The point is: what's the secular trend? The secular trend is what Sam (OpenAI CEO) said, which is, at the end of the day, because quite frankly, the biggest issue we are now having is not a compute glut, but it's power — it's sort of the ability to get the builds done fast enough close to power," Satya said in the podcast. "So, if you can't do that, you may actually have a bunch of chips sitting in inventory that I can't plug in. In fact, that is my problem today. It's not a supply issue of chips; it's actually the fact that I don't have warm shells to plug into." [Emphasis added]

Nadella's mention of 'shells' refers to a data center shell, which is effectively an empty building with all of the necessary ingredients, such as power and water, needed to immediately begin production.

AI's power consumption has been a topic many experts have discussed since last year. This came to the forefront as soon as Nvidia fixed the GPU shortage, and many tech companies are now investing in research in small modular nuclear reactors to help scale their power sources as they build increasingly large data centers.

This has already caused consumer energy bills to skyrocket, showing how the AI infrastructure being built out is negatively affecting the average American. OpenAI has even called on the federal government to build 100 gigawatts of power generation annually, saying that it's a strategic asset in the U.S.'s push for supremacy in its AI race with China. This comes after some experts said Beijing is miles ahead in electricity supply due to its massive investments in hydropower and nuclear power.

Aside from the lack of power, they also discussed the possibility of more advanced consumer hardware hitting the market. "Someday, we will make a[n] incredible consumer device that can run a GPT-5 or GPT-6-capable model completely locally at a low power draw — and this is like so hard to wrap my head around," Altman said. Gerstner then commented, "That will be incredible, and that's the type of thing that scares some of the people who are building, obviously, these large, centralized compute stacks."

This highlights another risk that companies must bear as they bet billions of dollars on massive AI data centers. While you would still need the infrastructure to train new models, the data center demand that many estimate will come from the widespread use of AI might not materialize if semiconductor advancements enable us to run them locally.

This could hasten the popping of the AI bubble, which some experts like Pat Gelsinger say is still several years away. But if and when that happens, we will be in for a shock as even non-tech companies would be hit by this collapse, exposing nearly $20 trillion in market cap.