OpenAI’s Stargate initiative has secured 10 gigawatts (GW) of U.S. AI computing capacity. The target was originally set for 2029.
OpenAI says it reached the milestone just over a year after the program launched in January 2025.
The early achievement marks one of the fastest infrastructure buildouts in the history of the technology industry.
What Is the Stargate Project?
Stargate is a new company that intends to invest $500 billion over four years building AI infrastructure for OpenAI in the United States, with an initial $100 billion deployed immediately.
The venture was formally announced on January 21, 2025, by President Donald Trump, alongside Sam Altman of OpenAI, Larry Ellison of Oracle, and Masayoshi Son of SoftBank.
SoftBank holds financial responsibility for the project, while OpenAI manages operations. Due to its enormous scale, the program has been compared to the Manhattan Project.
How the 10GW Milestone Was Reached
The project expanded aggressively throughout 2025. In July, OpenAI and Oracle entered an agreement to develop 4.5 gigawatts of additional Stargate data center capacity in the U.S. That deal alone was valued at over $300 billion across five years.
Then in September 2025, OpenAI, Oracle, and SoftBank announced five new U.S. data center sites.
Combined with the flagship campus in Abilene, Texas, and ongoing projects with CoreWeave, Stargate reached nearly 7 gigawatts of planned capacity and over $400 billion in investment.
By October 2025, the addition of a new Michigan campus brought Stargate to over 8 gigawatts of planned capacity and more than $450 billion in investment over three years.
OpenAI reported adding more than 3 GW in the prior 90 days alone.
The first active site remains in Abilene, Texas, with additional sites under development in New Mexico, Wisconsin, and Michigan.
The Hardware Behind It
Oracle began delivering the first NVIDIA GB200 racks to the Abilene site in June 2025. Early training and inference workloads are already running, using the new capacity to advance OpenAI’s next-generation research.
Looking ahead, NVIDIA and OpenAI signed a letter of intent to deploy at least 10 GW of NVIDIA systems.
NVIDIA intends to invest up to $100 billion progressively as each gigawatt is deployed, with the first systems targeted for the second half of 2026 on the Vera Rubin platform.
OpenAI is also working to reduce its long-term reliance on third-party chips.
The company is developing its own custom AI chip codenamed “Titan” in collaboration with Broadcom, to be fabricated on TSMC’s 3nm process, with mass production targeted for the second half of 2026.
Scale, Energy, and Jobs
The infrastructure demands of the project are immense. Industry observers expect sites to be built for extreme AI density, in the range of 100–150 kilowatts per rack.
At that level, the facilities would host tens of thousands of racks dedicated to training and inference workloads.
Energy is a central challenge. Potential obstacles include environmental permits, water resource management, grid upgrades, and supply chain delays, which could affect project timelines and operational milestones.
OpenAI has said it will tailor energy commitments site by site, and in some cases fund new generation or transmission capacity to avoid raising local electricity prices.
On the jobs front, the five new campuses announced in September are expected to create over 25,000 construction jobs across the U.S.
The broader project claims it will eventually create more than 100,000 American jobs.
Global Expansion
The project is not limited to the United States. OpenAI has announced Stargate data centers in Norway, the United Arab Emirates, and the United Kingdom. In May 2025, NVIDIA, Cisco, and OpenAI announced plans to help build a “UAE Stargate” AI data center, expected to open in 2026.
A Stargate Argentina site in Patagonia was also announced, representing an estimated investment of up to $25 billion.
What Comes Next
The 10GW announcement represents secured and committed capacity, not fully operational infrastructure.
Only a fraction of the theoretical data center space has been physically delivered to date, and questions remain about how many of the projects will ultimately be funded.
OpenAI currently loses billions of dollars each year, and the company plans to fund construction using a mix of its own cash flow and debt financing.
Still, the pace of progress has surprised even industry insiders.
As OpenAI finance chief Sarah Friar told CNBC in September 2025: “No one in the history of man built data centers this fast.”
