Meta said it has signed a multi-year strategic partnership with Nvidia to expand the company’s AI infrastructure, as Big Tech ramps up spending on the chips and data centers needed to train and run new models.
Meta said Nvidia technology will underpin AI-optimized data centers designed for both training and inference, with the companies targeting “substantial improvements” in performance per watt as workloads scale.
Under the agreement, Meta will deploy Nvidia CPUs and millions of Nvidia’s current Blackwell and next-generation Rubin graphics processing units (GPUs), Nvidia said.
The partnership also includes Nvidia’s Spectrum-X Ethernet platform to support AI-scale networking across Meta’s infrastructure.
Meta said it has also adopted Nvidia Confidential Computing for WhatsApp, aimed at enabling AI-powered features while protecting user data confidentiality and integrity. Nvidia said the two companies will work to expand confidential computing beyond WhatsApp to other use cases across Meta’s platforms.
Nvidia CEO Jensen Huang described Meta as operating AI at unmatched scale, saying the companies are “deep” co-designing across CPUs, GPUs, networking and software to raise performance and efficiency.
Meta CEO Mark Zuckerberg said the expanded partnership will help Meta build “leading-edge clusters” using Nvidia’s “Vera Rubin” platform to deliver what he called “personal superintelligence” to users worldwide.
Financial terms were not disclosed.
