OpenAI has announced a sweeping expansion of its AI infrastructure through the Stargate project, a $500 billion initiative to build a network of AI data centers across the United States in partnership with Oracle and SoftBank. The company revealed plans for five new data center sites, bringing the project’s total capacity to nearly 7 gigawatts—enough to power over five million homes. Three sites are being developed with Oracle in Shackelford County, Texas; Doña Ana County, New Mexico; and an undisclosed Midwest location. The other two are being built with SoftBank, in Lordstown, Ohio and Milam County, Texas. The flagship Stargate facility in Abilene, Texas, is already operational, featuring Oracle Cloud infrastructure and racks of Nvidia AI chips. OpenAI’s finance chief, Sarah Friar, said the Abilene campus could eventually scale beyond one gigawatt, enough to power about 750,000 homes. The project is part of a broader push to meet the soaring demand for computing power needed to train and run advanced AI models. Friar emphasized that the U.S. is still in the early stages of this infrastructure boom, with construction beginning now to deliver capacity by 2026. The Stargate project is being accelerated by a landmark $100 billion investment from Nvidia, announced on Monday, which will provide OpenAI with access to tens of thousands of AI processors and help build additional data centers. This investment, combined with existing agreements, brings the total projected investment in OpenAI’s infrastructure to over $400 billion in three years. Nvidia CEO Jensen Huang personally engaged in last-minute negotiations with OpenAI CEO Sam Altman to secure the deal, underscoring the high stakes involved. Altman outlined an ambitious vision in a blog post titled "Abundant Intelligence," calling for the creation of a “factory” capable of producing one gigawatt of new AI infrastructure every week. He described this as “the coolest and most important infrastructure project ever,” though he acknowledged the execution will be extremely difficult and take years. While he did not specify exact locations, the Stargate project is central to achieving this goal. OpenAI plans to fund the buildout using its own cash flow, debt financing, and partnerships. Although Nvidia is investing equity, it will be paid for all GPUs deployed. OpenAI expects to generate $13 billion in revenue this year. The company also continues to rely on Microsoft’s Azure cloud credits and contracts with smaller cloud providers for additional compute. The project has significant political and economic implications. Stargate was first unveiled alongside President Donald Trump at the White House in January, with Trump calling it a national priority. Oracle projects the initiative will create over 6,000 construction jobs and nearly 1,700 long-term positions. Critics have raised concerns about the circular nature of the funding—where suppliers like Nvidia invest in the same infrastructure they are building for OpenAI. Friar countered that history shows technology booms require bold infrastructure bets, citing the early internet as a precedent. A major challenge remains: energy. Experts warn that securing the electricity needed for these data centers—especially with Nvidia’s 10-gigawatt commitment—could be a “silent bottleneck.” The U.S. power grid may need to be restructured to support this level of demand. Despite these hurdles, OpenAI, Oracle, and SoftBank are moving ahead with plans to meet a full 10-gigawatt commitment by the end of 2025. The Stargate project, along with partnerships and massive investments, signals a pivotal moment in the race to scale AI infrastructure and position the U.S. as a global leader in artificial intelligence.