Nvidia and Meta expand GPU team up with millions of additional AI chips

Nvidia (NVDA) and Meta (META) announced an expanded multiyear, multi-generational partnership that will see Nvidia provide the social media giant with millions of its Blackwell and Rubin GPUs [graphics processing units], as well as its CPUs [central processing units] and networking offerings.

The plan calls for Meta to use the products within its data centers for both training and running AI models.

Nvidia didn’t reveal a dollar amount for the expanded deal, though it said that Meta will roll out the chips in its own data centers and lean on those available via Nvidia Cloud Partners. Nvidia’s Cloud Partner program includes companies like CoreWeave (CRWV) and Crusoe, which host Nvidia chips for other companies to rent and use.

“No one deploys AI at Meta’s scale — integrating frontier research with industrial-scale infrastructure to power the world’s largest personalization and recommendation systems for billions of users,” Nvidia CEO Jensen Huang said in a statement.

“Through deep codesign across CPUs, GPUs, networking, and software, we are bringing the full Nvidia platform to Meta’s researchers and engineers as they build the foundation for the next AI frontier.”

In addition to the GPUs, Nvidia says Meta is rolling out its first large-scale Grace CPU-only servers with plans to launch Vera CPU-only systems in 2027. These would be similar to more traditional severs Meta operates using Intel chips and wouldn’t include any GPUs.