This Trillion-Dollar A.I. Data Center Tsunami Could Be Coming To A Field Near You

Forbes

The rapid adoption of generative AI will require more advanced data centers, which in turn, require massive amounts of electricity to power them.  Read the full story on Forbes: https://www.forbes.com/sites/christopherhelman/2025/02/08/the-trillion-dollar-ai-data-center-tsunami--coming-to-a-field-near-you/  Subscribe to FORBES: https://www.youtube.com/user/Forbes?sub_confirmation=1  Fuel your success with Forbes. Gain unlimited access to premium journalism, including breaking news, groundbreaking in-depth reported stories, daily digests and more. Plus, members get a front-row seat at members-only events with leading thinkers and doers, access to premium video that can help you get ahead, an ad-light experience, early access to select products including NFT drops and more:  https://account.forbes.com/membership/?utm_source=youtube&utm_medium=display&utm_campaign=growth_non-sub_paid_subscribe_ytdescript  Stay Connected Forbes newsletters: https://newsletters.editorial.forbes.com Forbes on Facebook: http://fb.com/forbes Forbes Video on Twitter: http://www.twitter.com/forbes Forbes Video on Instagram: http://instagram.com/forbes More From Forbes:  http://forbes.com  Forbes covers the intersection of entrepreneurship, wealth, technology, business and lifestyle with a focus on people and success.

Transcript

00:00Today on Forbes, the trillion-dollar AI data center tsunami coming to a field near you.

00:09Last month on his Facebook blog, Mark Zuckerberg wrote, quote,

00:12This will be a defining year for AI.

00:16He was hyping his plan to invest $65 billion in 2025, most of it in support of growing

00:22Meta's AI capabilities and building upon its Llama 4 model.

00:27The locus of these efforts appears to be on 2,250 acres of flat farmland in northeast

00:33Louisiana, where Meta has begun work on a $10 billion data center project.

00:39This grandest of Meta's 20 worldwide data centers will be called Sucre, or Sweep, and

00:45cover 4 million square feet of floorspace on 2,250 acres.

00:50To power racks lined with thousands of NVIDIA H100 GPUs, Sucre will require 2.23 gigawatts

00:58of 24-7 electricity, enough to power more than 2 million homes, which utility company

01:04Entergy will provide via twin high-efficiency natural gas turbines at a capital cost of

01:11$3.2 billion.

01:14Meta's plans also call for $250 million in related infrastructure spending, including

01:19roads and water systems.

01:21Construction jobs will peak at 5,000.

01:24So is this really going to get built?

01:26And is it even a good idea, given recent revelations from DeepSeek that building advanced machine

01:32intelligence might be cheaper and less energy-intensive than initially believed?

01:37Zach Kraus, an analyst at energy consultancy East Daily Analytics, tracks data center announcements

01:44and works on handicapping both their likelihood of getting built and how developers expect

01:48to provide for their electricity demand.

01:51Kraus says he has been encouraging his institutional investor clients — many natural gas and

01:56power generation investors among them — not to panic over the implications.

02:01Kraus, who includes on his list 290 data center projects with a decent likelihood of getting

02:07built, says, quote,

02:08We've seen a short-term overreaction by the market to new tech.

02:14Even if AI energy efficiency improves exponentially, the silicon brains of the future will need

02:20all of those server farms, and then some.

02:23That's because of a dynamic called the Jevons Paradox, named after English economist William

02:28Jevons, who in 1865 noticed that efficiency breakthroughs in steam engine design did not

02:34reduce coal consumption at all, but instead caused demand for the fuel to surge.

02:40Jevons Paradox has also been observed, for example, in explosive demand for modern conveniences

02:46like air conditioning, passenger airplanes, and MRI machines.

02:51Microsoft CEO Satya Nadella in late January tweeted about Jevons Paradox.

02:56He said, quote,

02:58As AI gets more efficient and accessible, we will see its use skyrocket, turning it

03:03into a commodity we just can't get enough of.

03:07Economist Hugh Wynne at Power Research Group concurs, saying, quote,

03:11The demand for power to train and operate AI models could be materially reduced.

03:16However, if these efficiency gains also drive down the cost of AI, the adoption of AI-based

03:22services could be significantly faster and more widespread.

03:27And that will wipe out any expected energy savings.

03:31Krauss sees America's tsunami of advanced data centers demanding an additional 81 gigawatts

03:36of electricity by 2030.

03:39That's enough to power all of Texas today.

03:42If developers used natural gas to run them all, it would take about 12.9 billion cubic

03:47feet per day, he figures, equal to 10 percent of national natural gas supplies.

03:52He says, quote,

03:54The meta-project is circled in red ink for its eventual impact to the natural gas market.

03:59Krauss figures that Sucre will require 360 million cubic feet of natural gas per day.

04:05The energy equivalent of about 60,000 barrels of oil.

04:10And Sucre is not even the biggest.

04:12OpenAI is partnering with Oracle, SoftBank, Microsoft, and others on the Stargate project,

04:17which they say will be a $500 billion multi-site data center expansion, beginning with Abilene,

04:23Texas.

04:24In time, Stargate will need more than 5 gigawatts of electricity, though Krauss has only factored

04:29in a small fraction of that demand until they build something.

04:34Microsoft plans $80 billion of capital investment on AI and cloud computing this year.

04:39Meanwhile, Amazon has earmarked $100 billion for AI over the next decade, and may spend

04:45$35 billion just in northern Virginia's data center alley.

04:50For full coverage, check out Christopher Hellman's piece on Forbes.com.

04:56This is Kieran Meadows from Forbes.

04:58Thanks for tuning in.

Category

Transcript

Recommended