OpenAI has announced the launch of five new AI data centre sites in the US, developed in partnership with Oracle and SoftBank Group Corp., as part of its $500 billion Stargate Project. This expansion brings the initiative close to 7 gigawatts of planned capacity, moving the company nearer to its target of 10 GW by the end of 2025, ahead of schedule.
“These new sites put us on a clear path to securing the full $500 billion, 10-gigawatt commitment we announced in January,” OpenAI said. Three of the new facilities, powered by Oracle, are located in Shackelford County, Texas, Doña Ana County, New Mexico, and an upcoming Midwest site. Collectively, these centres are expected to provide 5.5 GW of capacity, supplemented by a 600 MW expansion near the flagship Stargate site in Abilene, Texas.
The remaining two sites, developed in collaboration with SoftBank, are based in Lordstown, Ohio and Milam County, Texas, and are projected to scale to 1.5 GW over the next 18 months. OpenAI CEO Sam Altman highlighted the scale of the vision, stating, “Our vision is simple: we want to create a factory that can produce a gigawatt of new AI infrastructure every week,” while acknowledging that execution would be extremely challenging. He added, “Maybe with 10 gigawatts of compute, AI can figure out how to cure cancer. Or with 10 gigawatts of compute, AI can figure out how to provide customised tutoring to every student on earth.”
Alongside this expansion, OpenAI has strengthened its partnership with NVIDIA, signing a letter of intent to deploy at least 10 GW of NVIDIA systems for its next-generation AI infrastructure. The agreement includes up to $100 billion in investment from NVIDIA, with the first gigawatt of capacity expected to go live in 2026 on OpenAI’s Vera Rubin platform.
This ambitious rollout positions OpenAI at the forefront of high-scale AI compute infrastructure, aiming not only to accelerate AI development but also to tackle some of the world’s most complex challenges, from healthcare to personalized education.
