Oracle and OpenAI Expand Partnership, Plan 4.5 GW of US Data Center Capacity

Oracle and OpenAI are expanding their partnership to develop 4.5 gigawatts of additional US data center capacity, with sites in Texas, Michigan, Wisconsin, and Wyoming under consideration. This adds to OpenAI's existing 5-gigawatt capacity, running on over 2 million AI chips. The expansion is part of the Stargate project, which aims to power AI workloads. The partnership is expected to create over 100,000 jobs in construction and operations in the US.

Oracle and OpenAI have announced an expansion of their partnership to develop an additional 4.5 gigawatts of US data center capacity. This new development will bring the total capacity of OpenAI's Stargate project to over 5 gigawatts, running on more than 2 million AI chips [1]. The expansion is part of the Stargate project, which aims to power AI workloads and accelerate America's reindustrialization.

The new data center capacity is expected to create over 100,000 jobs in construction and operations across the US. The sites under consideration for the new development include Texas, Michigan, Wisconsin, and Wyoming [2]. The expansion is part of a broader commitment to invest $500 billion in US-based AI infrastructure over the next four years, a commitment that OpenAI now expects to exceed [1].

The Stargate project is an ambitious undertaking designed to meet the historic opportunity in front of us. The project encompasses ongoing data center partnerships with Oracle, SoftBank, and CoreWeave, as well as international investments in US infrastructure through OpenAI for Countries. Microsoft will continue to provide cloud services for OpenAI, including through Stargate [1].

The expansion with Oracle is a significant milestone for Stargate, OpenAI's AI infrastructure platform and long-term vision to deliver the benefits of AI to everyone. The project has already created thousands of jobs, with more expected as operations expand. The Stargate I site in Abilene, Texas, has been up and running for a few weeks, with Oracle delivering the first Nvidia GB200 racks in June [2].

OpenAI expects the additional data center development to create more than 100,000 jobs in construction and operations in the US. Historically, data centers don't employ many full-time workers, but the Stargate project is expected to create a significant number of jobs across various roles [2].

The partnership with Oracle is moving forward with strong momentum, complementing the ongoing partnership with SoftBank. Both partnerships are critical to meeting OpenAI's continually expanding compute needs. The new data center capacity will be used to run and train AI algorithms, pushing the limits of OpenAI's next-generation frontier research [1].

The expansion of the Stargate project is a testament to the growing interest and support for AI infrastructure. The White House has recognized the critical role AI infrastructure will play in driving innovation, economic growth, and national competitiveness. The project builds on the longstanding and ongoing compute partnership with Microsoft, which continues to serve as a technology partner to Stargate [1].

References:
[1] https://openai.com/index/stargate-advances-with-partnership-with-oracle/
[2] https://www.bloomberg.com/news/articles/2025-07-22/oracle-to-supply-openai-with-2-million-ai-chips-for-data-centers