OpenAI and NVIDIA have signed a landmark agreement to build one of the world’s largest AI infrastructures, deploying at least 10 gigawatts of NVIDIA systems over the coming years. This effort represents millions of GPUs dedicated to training and running OpenAI’s next-generation models, including work toward artificial general intelligence (AGI).
As part of the deal, NVIDIA plans to invest up to $100 billion in OpenAI, tied to each gigawatt of systems deployed. The first wave- one gigawatt of NVIDIA’s new Vera Rubin platform – is scheduled to go live in the second half of 2026.
“This partnership marks the next leap forward — deploying 10 gigawatts to power the next era of intelligence,” said Jensen Huang, NVIDIA founder and CEO.
“Compute is the foundation of the future economy,” said Sam Altman, OpenAI cofounder and CEO. “With NVIDIA, we’ll push new AI breakthroughs and deliver them at scale.”
OpenAI, which now has over 700 million weekly active users, will designate NVIDIA as its preferred compute and networking partner. Both companies will also co-develop their roadmaps—aligning OpenAI’s AI models and infrastructure software with NVIDIA’s hardware and systems.
The partnership builds on years of collaboration and complements OpenAI’s wider network of infrastructure partners, including Microsoft, Oracle, SoftBank, and Stargate.
More details of the agreement are expected in the coming weeks.
Press Release: NVIDIA
