OpenAI's Stargate Project Surpasses 10GW AI Infrastructure Goal Ahead of Schedule

OpenAI has revealed that its Stargate initiative has exceeded the 10GW goal for AI computing capacity in the U.S., initially aimed for 2029. In just the past 90 days, over 3GW has been added. This long-term project seeks to establish the computational groundwork for AGI, collaborating with local communities, energy providers, chip manufacturers, cloud services, and construction companies. The inaugural site in Abilene, Texas, employs closed-loop cooling to reduce water usage and has trained OpenAI's latest model, GPT-5.5, on Oracle Cloud Infrastructure using NVIDIA GB200 systems. Additionally, OpenAI initiated a community engagement program, contributing to the Port Washington-Saukville Education Foundation in Wisconsin, in partnership with Vantage Data Centers and Oracle. The organization stresses that robust computing is essential for developing superior models and reducing expenses, highlighting the need for coordinated infrastructure development in the Intelligence Age.