Nvidia Commits $100B to Power OpenAI’s Next-Gen AI

Nvidia has announced plans to invest up to $100 billion in OpenAI to support the construction of large-scale data centers designed for training and running advanced AI models. The companies said Monday they signed a letter of intent to deploy 10 gigawatts of Nvidia-powered systems, an amount of energy comparable to powering millions of homes, to fuel OpenAI’s next generation of artificial intelligence infrastructure.

The investment signals a shift in OpenAI’s cloud strategy as it reduces its dependence on Microsoft, which has been its largest backer and provider of computing resources. Earlier this year, Microsoft adjusted its partnership terms to allow OpenAI to collaborate with additional infrastructure partners. Since then, OpenAI has pursued projects such as Stargate and forged agreements with Oracle and SoftBank to expand its AI compute capacity.

By naming Nvidia a “preferred strategic compute and networking partner,” OpenAI is aligning itself with the world’s dominant AI chipmaker at a time when demand for high-performance hardware continues to soar. Nvidia’s contribution will complement OpenAI’s existing partnerships, giving the AI company more options as it scales globally.

Details of how Nvidia’s investment will be delivered remain unclear. The companies did not specify whether the $100 billion commitment will take the form of chips, cloud credits, cash, or a mix of resources. What is clear is that the deal underscores both firms’ ambitions to accelerate AI infrastructure development on a scale few others can match.

For OpenAI, the expanded collaboration provides critical computing power to advance new models while balancing its reliance on multiple partners. For Nvidia, the move cements its role not just as a chip supplier but as a strategic backbone for the world’s most influential AI companies. As both organizations work to build the infrastructure of the AI era, the partnership highlights how the race to dominate artificial intelligence is increasingly tied to massive investments in energy and compute capacity.