Chipmaker NVIDIA announced a multiyear, multigenerational strategic partnership with Meta covering on-premises systems, cloud services, and artificial intelligence (AI) infrastructure.
Meta will build hyperscale data centers designed for both AI training and inference as part of its long-term infrastructure roadmap. The partnership includes large-scale deployment of NVIDIA central processing units (CPUs) and millions of Blackwell and Rubin graphics processing units (GPUs). It also covers the integration of NVIDIA Spectrum-X Ethernet switches into Meta’s Facebook Open Switching System platform.
“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,” said Jensen Huang, founder and CEO of NVIDIA. “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.”
“We’re excited to expand our partnership with NVIDIA to build leading-edge clusters using their Vera Rubin platform to deliver personal superintelligence to everyone in the world,” said Mark Zuckerberg, founder and CEO of Meta.
The companies are continuing their work on Arm-based NVIDIA Grace CPUs for Meta’s production data center applications. Meta said these processors provide improved performance per watt, supporting its long-term strategy to make operations more energy efficient.
The collaboration marks the first large-scale deployment using only NVIDIA Grace CPUs. The companies are investing in codesign and software optimization in CPU ecosystem libraries to improve energy efficiency with each new generation.
Meta and NVIDIA are also working on deploying NVIDIA Vera CPUs, with potential large-scale rollout in 2027. The effort is expected to expand Meta’s energy-efficient AI computing capacity and support the broader Arm software ecosystem.
Meta will deploy NVIDIA GB300-based systems and develop a unified architecture that connects on-premises data centers with NVIDIA Cloud Partner deployments. The goal is to simplify operations while improving performance and scalability.
Meta has also adopted NVIDIA Spectrum-X Ethernet across its infrastructure to deliver low-latency networking and better resource use. In addition, Meta is using NVIDIA Confidential Computing for private processing on WhatsApp, enabling AI features while protecting user data. The companies plan to extend confidential computing capabilities to other services across Meta’s platforms.