U.S. tech giant Meta announced a multi-year strategic partnership with Nvidia to advance its long-term AI infrastructure roadmap. Meta stated that the extensive use of Nvidia technology will further strengthen the existing relationship, enabling the company to build data centers optimized for AI training and inference, supporting core operations.
The announcement highlighted that these developments will significantly improve performance per watt, supporting more efficient and large-scale AI operations.
Meta also revealed it will use "Nvidia Confidential Computing" for its WhatsApp messaging app, enabling AI-powered features across the platform while safeguarding user data privacy.
Meta will deploy the "Nvidia Spectrum-X Ethernet network platform" across its infrastructure to provide large-scale AI network communication, predictable low-latency performance, maximize utilization, and enhance both operational and energy efficiency.
As part of the collaboration, engineers from both companies will optimize and accelerate state-of-the-art AI models in core workloads to improve performance and efficiency of new AI capabilities used by billions globally.
Meta CEO Mark Zuckerberg said they are excited to expand the partnership with Nvidia to build cutting-edge clusters using their "Vera Rubin" platforms and deliver personal superintelligence to everyone worldwide.
Nvidia CEO Jensen Huang noted that no company deploys AI at Meta's scale, emphasizing that Meta integrates pioneering research with industrial-scale infrastructure to support the world's largest personalization and recommendation systems for billions of users.
Huang added that through deep co-design across CPUs, GPUs, network infrastructure, and software, Nvidia will provide its full platform to Meta's researchers and engineers to lay the foundation for the next frontier in AI.
v
