Meta Expands Nvidia Deal to Deploy Millions of AI Chips in Largest-Ever Data Center Build-Out

Meta Expands Nvidia Deal to Deploy Millions of AI Chips in Largest-Ever Data Center Build-Out

Meta has announced a sweeping expansion of its partnership with Nvidia, becoming the first company to deploy Nvidia's standalone Grace central processing units at scale while committing to a massive data center build-out that solidifies its position as one of the world's largest AI infrastructure investors.

The multiyear deal, announced Tuesday, will see Meta deploy millions of Nvidia chips across its data centers, including the tech giant's new standalone CPUs and next-generation Vera Rubin systems. The partnership is valued in the tens of billions of dollars, according to analysts.

"We're delivering on our commitment to bring personal superintelligence to everyone in the world," said Meta CEO Mark Zuckerberg in a statement. The deal marks the most significant expansion of Meta's AI infrastructure strategy to date.

Standalone CPUs: A First for Nvidia

The most notable aspect of the deal is Meta's deployment of Nvidia's Grace CPUs as standalone chips—a first for the GPU maker. Traditionally, CPUs have been paired with GPUs in AI servers, but Nvidia's Grace CPUs are designed to run inference and agentic workloads independently.

"They're really designed to run those inference workloads, run those agentic workloads, as a companion to a Grace Blackwell/Vera Rubin rack," explained chip analyst Ben Bajarin of Creative Strategies. "Meta doing this at scale is affirmation of the soup-to-nuts strategy that Nvidia's putting across both sets of infrastructure: CPU and GPU."

The next-generation Vera CPUs are planned for deployment in 2027.

$600 Billion US Commitment

The deal is part of Meta's broader commitment to invest $600 billion in the United States by 2028 on data centers and infrastructure. The company plans to build 30 data centers, with 26 located in the United States.

Two massive projects are already under construction: the 1-gigawatt Prometheus site in New Albany, Ohio, and the 5-gigawatt Hyperion facility in Richfield Parish, Louisiana.

Meta also announced it would use Nvidia's networking technology, including Spectrum-X Ethernet switches to link GPUs together, along with security capabilities for WhatsApp AI features.

Market Reaction

Shares of both Meta and Nvidia climbed in extended trading following the announcement. Meanwhile, Advanced Micro Devices stock sank approximately 4% on concerns about competition in the AI chip market.

In January, Meta announced plans to spend up to $135 billion on AI in 2026 alone—a figure that underscores the massive infrastructure race underway among tech giants.

Sources

• •CNBC: Meta expands Nvidia deal to use millions of AI chips