Meta has entered into a multi-year agreement to expand data centers its with millions of Nvidia CPUs and Vera Grace and Blackwell and Rubin GPUs .
Although Meta has long used Nvidia hardware for AI products , this deal “represents the first large-scale deployment solely with Nvidia Grace,” which Nvidia says will deliver “significant performance-per-watt improvements in Meta’s data centers.”
See also: OpenAI and Broadcom will produce their own AI chips
The agreement also includes plans to add Nvidia's next-generation Vera CPUs to Meta's data centers (in 2027)
Meta: Collaboration with Nvidia – Efforts for internal chips as well
It is worth noting that Meta is also working on its own chips to run AI models, but according to the Financial Times, it has faced "technical challenges and launch delays."
See also: AMD and OpenAI challenge Nvidia's dominance in AI chips
On the other hand, Nvidia is also facing some problems related to depreciation and loans for chips, which are used to finance AI buildouts.
Additionally, there is intense competition in the industry. CNBC notes that Nvidia shares fell 4% after a November report about Meta Tensor chips Google’s for AI. Meanwhile, late last year, AMD announced chip deals with OpenAI and Oracle.
See also: Nvidia CEO denies he's "unhappy" with OpenAI – Big investment coming
Nvidia and Meta did not disclose the cost of the deal, but AI spending by Meta, Microsoft, Google and Amazon is estimated to cost more this year than the entire Apollo space.
