OpenAI is ramping up its AI infrastructure with five new data center sites across the U.S., developed in partnership with Oracle and SoftBank through the Stargate project.
Quick Summary – TLDR:
- Five new data centers announced in Texas, New Mexico, Ohio, and an undisclosed Midwest site
- Expansion pushes Stargate capacity to nearly 7 gigawatts, representing $400 billion in investment
- Project expected to create 25,000 onsite jobs plus tens of thousands more across the U.S.
- Part of the $500 billion commitment to reach 10 gigawatts of AI infrastructure by 2025
What Happened?
OpenAI, Oracle, and SoftBank have revealed plans to build five new Stargate AI data centers across the U.S. The sites will contribute to a total planned capacity of nearly 7 gigawatts and investments surpassing $400 billion over the next three years, putting the project ahead of schedule to meet its $500 billion, 10-gigawatt target by the end of 2025.
Together with @sama and the @OpenAI team, we’re developing three additional new AI data center sites in Shackelford County, Texas; Doña Ana County, New Mexico, and a site in the Midwest.
— Oracle (@Oracle) September 24, 2025
Here’s the latest news on Stargate and the rapid progress at the flagship campus in… pic.twitter.com/xm9K2n4DQ3
Expanding Stargate’s Reach
Three of the new facilities are being developed with Oracle, located in Shackelford County, Texas, Doña Ana County, New Mexico, and an undisclosed site in the Midwest. Oracle is also expanding near Stargate’s flagship campus in Abilene, Texas, adding 600 megawatts of potential new capacity.
The remaining two sites come through OpenAI’s partnership with SoftBank. These include a data center in Lordstown, Ohio, and another in Milam County, Texas, which will be developed alongside SB Energy, a SoftBank Group company. SoftBank has already broken ground in Lordstown, aiming to bring its advanced design online as early as next year.
A Massive Infrastructure Push
The Stargate initiative was first unveiled in January at the White House, with U.S. President Donald Trump hosting tech CEOs to launch the private-sector project. The plan calls for $500 billion in AI infrastructure spending to deliver 10 gigawatts of compute capacity nationwide.
So far, Stargate has reviewed more than 300 proposals from over 30 states, selecting the first five projects as part of its expansion strategy. The initiative is also working with CoreWeave on additional projects to accelerate progress.
Beyond scale, the project emphasizes job creation. OpenAI said the five new sites alone will generate more than 25,000 onsite roles, with many more indirectly supported across the country.
Fueling AI’s Next Phase
The announcement comes just a day after Nvidia pledged up to $100 billion in investment in OpenAI, supplying high-performance AI processors to power these massive data centers. OpenAI and its partners are also pursuing debt financing to lease chips for Stargate, ensuring that the rollout can keep pace with surging AI demand.
Microsoft, OpenAI’s largest backer, is already investing heavily in data centers to support generative AI products like ChatGPT and Copilot, signaling an industry-wide race to build capacity.
Voices from the Leaders
Sam Altman, CEO of OpenAI, said:
Clay Magouyrk, CEO of Oracle, added:
SoftBank’s Chairman and CEO Masayoshi Son said:
SQ Magazine’s Takeaway
I see Stargate as one of the most ambitious infrastructure projects in modern tech history. A half-trillion-dollar plan to power AI is staggering, and it shows how serious the industry has become about scaling compute. What stands out most is the sheer speed of expansion. OpenAI and its partners are not just talking about future capacity, they are already breaking ground and rolling out hardware. To me, this feels like a clear signal: the AI race is no longer just about models, it is about building the world’s most powerful digital factories to run them.
