OpenAI and partners pursue $500 billion Stargate megadatacentre project for next-generation AI compute
Crusoe Energy got its start harnessing oilfield flare gas to mine bitcoin. At the 1,000-acre construction site codenamed Project Ludicrous in Abilene, Texas, cofounders Chase Lochmiller and Cully Cavness are now building the first phase of Project Stargate, the $500 billion AI datacenter being developed for OpenAI, Oracle, and Microsoft. With 2,000 workers on site, the companies claim that Stargate could be the first generation of AI models to fully surpass human cognition. Lochmiller, CEO of lead developer Crusoe Energy, describes it as a "factory of factories," meaning that AIs trained here will be used to develop more such systems, producing an endless stream of intelligence-as-a-service.
By the end of this year, Crusoe expects 100,000 Nvidia graphic processing units to be installed into each of the first two 500,000 square foot "H"-shaped buildings. Each Stargate datacenter building will draw 100 megawatts of electricity, enough for a city of 100,000. Most power will come from hundreds of 300-foot-tall wind turbines in west Texas, which often generate far more electricity than the region needs. Windmill owners sometimes sell excess power into the grid at a loss while collecting federal green energy tax credits, prompting Lochmiller, 38, to note "Not a great use of those electrons." Crusoe has set up five gas turbines from General Electric Vernova for backup power.
Crusoe raised $3.4 billion in joint venture funding with asset management company Blue Owl Capital and data center-focused investment firm Primary Digital Infrastructure for these first two buildings. This is only a portion of the $500 billion that Oracle founder Larry Ellison, SoftBank CEO Masayoshi Son, and OpenAI CEO Sam Altman promised in January to invest over the next five years. The announcement appealed to President Trump, who in his February speech to Congress praised the investment, saying "Which they wouldn't have done if Kamala had won." Yet some Crusoe clients are frustrated with the company's cloud computing offerings, and the first phase of Stargate—which will eventually comprise 10 buildings representing 1.2 gigawatts of computing power at an estimated cost of $15 billion—faces significant execution risks.
Trump's tariffs, imposed on some 90 countries last week and now paused, could make materials like aluminum and structural steel more expensive. Lochmiller countered that Crusoe views this as "a competitive moat," noting "From the beginning, Crusoe made the decision to onshore the manufacturing of critical electrical components including switchgear because we believe in building things in America." Industry skepticism persists: consultant Daniel Golding calls Crusoe "the flavor of the month," and analyst Zach Krause at East Daley Analytics questions whether all announced AI infrastructure will materialize. Meta, Amazon, and Alphabet have separately earmarked more than $200 billion for AI infrastructure, and Krause estimates those that do get completed will add some 80 gigawatts of electricity demand nationwide by 2030—an unprecedented 10% growth, or as much as all of Texas uses on a hot summer day.
Lochmiller and Cavness, 37, went to high school together in Denver, where they spent weekends rock climbing. Lochmiller, 6-foot-1 with black hair, developed his interest in AI at Stanford, where he attended grad school starting in 2013 and was struck by Professor Andrew Ng's assertion that "AI is the new electricity." Lochmiller experienced AI's massive power demands firsthand when running stock forecasting and trading algorithms at hedge fund KCG Holdings in the early 2000s. Meanwhile, Cavness became obsessed with locating the world's cheapest, greenest energy sources, visiting 25 countries including Chile, Argentina, and Iceland.