OpenAI and Oracle have launched the first data center of their $500 billion Stargate project in Abilene, Texas, with plans to expand to New Mexico, Ohio, and other locations across the U.S. The facility, powered by Oracle Cloud Infrastructure and Nvidia chips, marks a significant step in the companies' ambitious push to secure the compute power needed for the future of artificial intelligence.
The flagship site in Abilene, about 180 miles west of Dallas, is now operational. OpenAI CFO Sarah Friar tells CNBC, 'We're just getting going here in Abilene, Texas, but you'll see this all around the United States and beyond.' The data center, leased by Oracle, is part of a broader initiative to address the growing demand for AI infrastructure, which has seen over $2 trillion in planned investments globally, according to HSBC estimates.
In addition to the Stargate project, OpenAI announced an equity investment deal with Nvidia, adding an estimated $500 billion worth of data centers in the coming years. Nvidia CEO Jensen Huang personally engaged in negotiations with OpenAI CEO Sam Altman to ensure Nvidia's involvement. 'People are starting to recognize just the sheer scale that will be required,' Friar says.
OpenAI, Oracle, and SoftBank, a key funder, have announced five additional Stargate sites across Texas, New Mexico, Ohio, and an unnamed location in the Midwest. This brings the total initiative to nearly 7 gigawatts and more than $400 billion in investment over the next three years. OpenAI will fund the computing capacity as an operating expense, with Nvidia being compensated for the GPUs it provides.
Friar projects that OpenAI will generate $13 billion in revenue this year, using its cash flow and debt financing to support the construction. The Stargate name will encompass all of OpenAI's infrastructure projects moving forward. With partners like CoreWeave, the companies are on track to meet their full 10-gigawatt commitment by the end of 2025. Construction is already underway for compute capacity that will come online in 2026, starting with Nvidia's next-generation Vera Rubin chips.
Subscribe to our newsletter for the latest AI news, tutorials, and expert insights delivered directly to your inbox.
We respect your privacy. Unsubscribe at any time.
Comments (0)
Add a Comment