OpenAI Announces Five New Stargate AI Data Centres Across the U.S.

Stargate is now set to deliver nearly 7 gigawatts of planned capacity and over $400 billion in investment within three years

OpenAI Announces Five New Stargate AI Data Centres Across the U.S.

September 24, 2025 — Washington, D.C. — OpenAI, Oracle, and SoftBank have announced the selection of five new U.S. data centre sites under Stargate, OpenAI’s massive AI infrastructure programme.

This builds on OpenAI and NVIDIA’s $100 billion partnership to deploy 10GW of systems by 2026, leveraging the Vera Rubin platform to power next-gen AI models and push toward superintelligence.

Recently, Oracle also agreed to provide OpenAI with $300 billion worth of compute power over five years, starting in 2027.

Together with its flagship campus in Abilene, Texas, and ongoing projects with CoreWeave, Stargate is now set to deliver nearly 7 gigawatts of planned capacity and over $400 billion in investment within three years—putting it ahead of its $500 billion, 10-gigawatt target announced in January.

In July, OpenAI and Oracle signed an agreement to build up to 4.5 gigawatts of Stargate capacity, representing more than $300 billion in joint investment over five years. New Oracle-led sites in Shackelford County (Texas), Doña Ana County (New Mexico), and the Midwest—along with a 600-megawatt expansion in Abilene—are expected to deliver 5.5 gigawatts and create 25,000 on-site jobs.

SoftBank is leading two additional sites, with 1.5 gigawatts of capacity planned over 18 months. These include a Lordstown, Ohio campus, already under construction, and a Milam County, Texas site in partnership with SB Energy.

“AI can only fulfill its promise if we build the compute to power it,” said Sam Altman, CEO of OpenAI. Oracle’s Clay Magouyrk added, “We continue to expand OCI’s footprint at an unrivaled pace to deliver the most performant and cost-effective AI training and inferencing.” SoftBank’s Masayoshi Son said, “Stargate is harnessing SoftBank’s innovative data center design and energy expertise to deliver the scalable compute that powers AI’s future.”