The collaboration will see large-scale integration of NVIDIA’s computing and networking technologies into Meta’s global data centers, supporting AI training, inference, and core service operations.
Both companies emphasized that the partnership aims to enhance performance efficiency per watt, enabling more sustainable and scalable AI deployments.
“No organization deploys AI at Meta’s scale—combining cutting-edge research with industrial-grade infrastructure to drive the world’s largest personalization and recommendation systems for billions of users,” said Jensen Huang, founder and CEO of NVIDIA.
“Through extensive co-design across CPUs, GPUs, networking, and software, we are delivering the full NVIDIA platform to Meta’s teams as they lay the groundwork for the next wave of AI innovation,” Huang added.
Focus on Confidential Computing and High-Performance Networking
Under the expanded agreement, Meta will implement NVIDIA’s Confidential Computing solutions to power AI features in WhatsApp, prioritizing user data privacy and security.
Additionally, Meta will adopt NVIDIA’s Spectrum-X Ethernet networking platform to ensure low-latency, high-performance connectivity tailored for AI workloads across its infrastructure.
“We’re thrilled to deepen our collaboration with NVIDIA to build state-of-the-art clusters using their Vera Rubin platform, bringing personal superintelligence to people worldwide,” said Mark Zuckerberg, founder and CEO of Meta.