OpenAI plans five additional Stargate data centers in the US with Oracle and SoftBank partners.
OpenAI announced plans to build five new data center sites in the U.S. as part of its Stargate initiative, which was launched in January with the goal of providing 10 gigawatts' worth of computing infrastructure at an expected cost of $500 billion over four years. The company is partnering with Oracle Corp. on three of the upcoming facilities, which will be located in Shackelford County, Texas; Doña Ana County, New Mexico; and a yet-unspecified Midwest location, with more details about the latter to come. According to the New York Times, Oracle will finance the three data center sites and oversee their construction, with the database maker hoping to cover some costs through "new kinds of financial deals with various partners." This collaboration is part of a $300 billion cloud infrastructure deal that OpenAI signed with Oracle earlier this month, and according to the Wall Street Journal, the database maker expects to start generating revenue from the contract in 2027.
OpenAI will build two additional data center campuses through a partnership with SoftBank Group Corp., one of its largest investors, which led a $40 billion round for the ChatGPT developer in March. The first site is located in Lordstown, Ohio, where SoftBank broke ground earlier this year and expects to bring the data center online in 2026. The second campus will be developed by the company's SB Energy infrastructure business in Milam County, Texas. OpenAI says the two sites can be equipped with 1.5 gigawatts' worth of compute infrastructure within 18 months.
These facilities will join a data center campus in Abilene, Texas that Oracle started building for the ChatGPT developer earlier this year. OpenAI detailed that the facility's first server racks came online in June and are already powering AI training and inference workloads. The server racks are powered by Nvidia Corp.'s GB200 chip, which combines two Blackwell B200 graphics processing units with a central processing unit, and Oracle reportedly expects to install more than 64,000 GB200 chips in the data center by next March. OpenAI will also use newer Nvidia silicon, announcing plans to adopt the chipmaker's upcoming Vera Rubin chip, which includes an 88-core CPU and a GPU based on the next-generation Rubin architecture, a plan disclosed alongside news that OpenAI will raise up to $100 billion from Nvidia to finance data center construction projects.
According to Nvidia Chief Executive Officer Jensen Huang's statement last month, building 1 gigawatt of AI infrastructure costs $50 billion to $60 billion, with the chipmaker's hardware accounting for well over half that sum. In addition to the five data centers announced, OpenAI may build a 600-megawatt site near its Abilene campus, bringing the facilities' combined power draw to over 5.5 gigawatts. OpenAI expects the projects to create more than 25,000 jobs onsite and tens of thousands more nationwide.