Meta, Nvidia Forge Landmark Multi-Year AI Infrastructure Deal

Meta and Nvidia have announced a sweeping multi-year, multi-generational strategic partnership that will see Meta deploy millions of Nvidia Blackwell and Vera Rubin GPUs across hyperscale AI data centers one of the largest infrastructure commitments in semiconductor history

Meta and Nvidia have announced a multi-year, multigenerational strategic partnership spanning on-premises, cloud, and AI infrastructure. The deal, confirmed via official press releases from both companies on February 17, 2026, marks one of the largest hardware commitments in the history of the technology industry.

Under the agreement, Meta will deploy millions of Nvidia Blackwell and next-generation Vera Rubin GPUs across its hyperscale data centers, which are optimized for both AI training and inference. Meta will also expand its use of Nvidia Grace CPUs, in what Nvidia described as the first large-scale Grace-only deployment, delivering significant improvements in performance per watt.

The partnership also brings Nvidia Spectrum-X Ethernet networking to Meta’s infrastructure, designed to support AI-scale connectivity at low latency. Additionally, Nvidia Confidential Computing will be adopted for WhatsApp, enabling AI-powered features while protecting user data privacy.

While neither company disclosed exact financial terms, analysts estimate the deal stretches into the tens of billions of dollars. The move forms part of Meta CEO Mark Zuckerberg’s commitment to spend up to $135 billion on AI in 2026 alone, as part of a broader $600 billion US infrastructure pledge through 2028.

Nvidia CEO Jensen Huang said the companies are bringing the full Nvidia platform to Meta’s engineers to help build the foundation of the next AI frontier. Zuckerberg described the deal as central to his stated mission of delivering personal superintelligence to every user worldwide across WhatsApp, Instagram, and Facebook.