Oracle, SoftBank, OpenAI power Stargate expansion

By
Nat Rubio-Licht

Sep 24, 2025

12:00pm UTC

Copy link
Share on X
Share on LinkedIn
Share on Instagram
Share via Facebook

Stargate is going through a growth spurt.

Oracle, OpenAI and SoftBank are building five new U.S. data center sites, bringing the Stargate project to nearly 7 gigawatts and $400 billion in investment deployed over the next three years, the companies announced Tuesday. The announcement puts the project ahead of schedule and a step closer to the initial commitment of 10 gigawatts of capacity and $500 billion investment.

The data center sites will be located in Lordstown, Ohio; Shackelford County, Texas; Milam County, Texas; Doña Ana County, New Mexico and an unnamed site in the Midwest. More sites will be added eventually to complete the commitments, OpenAI noted in a press release.

It’s been a big week for AI infrastructure hot shots. The news follows Nvidia’s $100 billion investment in OpenAI, announced Monday, to develop at least 10 gigawatts in AI data centers, with the first gigawatt expected to be deployed in the second half of 2026.

These astronomical dollar signs and hastened timelines could signal that AI companies are feeling the pressure to make their lofty visions a reality. OpenAI CEO Sam Altman told CNBC on Monday that infrastructure presented an "unprecedented" challenge in AI development. In his personal blog, he emphasized that his goal to create “a factory that can produce a gigawatt of new AI infrastructure every week” could be the key to an AI model that can cure cancer or tutor every student on earth.

“AI can only fulfill its promise if we build the compute to power it,” Altman said in Tuesday’s announcement. “That compute is the key to ensuring everyone can benefit from AI and to unlocking future breakthroughs.”

Tech giants are clearly adopting a “build it and they will come” mentality with AI, but the question of how to generate returns on these massive investments remains unanswered. An August report from MIT found that only 5% of AI pilot programs achieve rapid revenue acceleration, with a large majority having little impact on revenue. At this point, it’s unclear if pouring more money into AI is the key to getting people to actually use it.