Meta is doubling down on its relationship with Nvidia in what the AI chip giant calls a “multi-generational” deal.
The deal, announced Tuesday, calls for Meta to build data centers equipped with millions of Nvidia’s current and next-generation chips for AI training and inference.
The move highlights how Meta has become increasingly reliant on Nvidia, even as the social networking giant develops its own chips and works with competing suppliers such as AMD. Reports also suggest that Meta is considering using TPUs, chips designed by rival Google.
Patrick Moorhead, principal analyst at Moore Insights and Strategy, said the partnership with Nvidia could ease speculation surrounding the allegedly meta TPU talks. However, large tech companies often test multiple suppliers at the same time.
The deal comes amid increasing competition in AI infrastructure. While Nvidia leads the market, rivals like Google, AMD, and Broadcom are working to chip away at its lead.
Importantly, this partnership will see Meta deploy not only Nvidia’s GPUs but also its CPUs.
The CPU, long dominated by Intel and AMD, is the central processor that works in conjunction with the GPU in data centers. GPUs are used for general computing tasks and are essentially the core of all modern computing systems, whereas GPUs are used in specialized cases that require more computing power, such as AI training or game graphics. By offering both, Nvidia can capture even more spending and deepen Meta’s role within the AI stack.
While this increases competitive pressure, the demand for infrastructure is so high that NVIDIA’s rivals are unlikely to completely decline in the short term, Moorhead said.
Moorehead said NVIDIA is making its CPU ambitions more clear, including selling its upcoming Vera CPU as a standalone product. This emphasis reflects how CPUs will play a larger role as AI workloads move beyond model training to inference.
“CPUs tend to be cheaper and a little more power efficient for inference,” said Rob Enderle, principal analyst at Enderle Group.
Moorhead and Enderle said Meta’s decision to source both GPUs and CPUs from a single vendor also has the potential to reduce complexity, as chief information officers often prefer a “blow-and-blow” approach to problem-solving.
In addition to GPUs and CPUs, Meta will use Nvidia networking equipment in its data centers as part of the deal and sensitive computing technology to run AI functions within WhatsApp.
According to Nvidia, the two companies will also work together to deploy Nvidia’s next-generation Vera CPUs beyond the current Grace CPU model.
Any tips? To contact this reporter via email, please specify the following address: gweiss@businessinsider.com Or send a signal at @geoffweiss.25. Use a personal email address, non-work WiFi network, and non-work device. here is our guide Share information securely.
