OpenAI and Nvidia announced a partnership to deploy 10 gigawatts of GPU-powered data centers, with Nvidia investing up to $100 billion as infrastructure comes online.

The first phase is set for late 2026 using Nvidia’s Vera Rubin platform. Nvidia CEO Jensen Huang called it “the biggest AI infrastructure project in history,” describing the effort as a leap to bring AI from labs to the real world.

The facilities will support large-scale training and inference for AI models, including OpenAI’s ChatGPT, which serves more than 700 million weekly users. OpenAI CEO Sam Altman said compute power is “the fuel we need” to drive new models and future economic growth.

Nvidia will serve as OpenAI’s preferred compute and networking partner, with investment tied to each gigawatt of capacity deployed. The deal builds on existing collaborations with Microsoft, Oracle, SoftBank and other partners.