Nvidia will invest up to $100 billion in OpenAI as the artificial intelligence firm prepares to build massive data centers powered by the chipmaker’s processors, both companies said Monday. The plan involves systems requiring 10 gigawatts of power—equal to 4 to 5 million GPUs, according to Nvidia CEO Jensen Huang.
Calling the effort “a giant project,” Huang said the partnership underscores the deep ties between Nvidia and OpenAI, whose release of ChatGPT in 2022 ignited demand for Nvidia’s chips. OpenAI now reports 700 million weekly users and is pushing to expand capacity.
Just in: Nvidia plans to buy equity in OpenAI progressively over time as its systems are deployed. The US chipmaker's shares rose nearly 3% on the announcement. https://t.co/hfySGmCKlE pic.twitter.com/1hP4iB3UiQ
— Financial Times (@FT) September 22, 2025
The first phase of the investment, using Nvidia’s Vera Rubin systems, is expected online in 2026. The deal comes after investors valued OpenAI at $500 billion and reflects its collaborations with Microsoft, Oracle, and SoftBank.
Nvidia, meanwhile, has been on a spending spree, including a $5 billion stake in Intel and nearly $1 billion in AI startup acquisitions.
