Meta(META.US) joins forces with NVIDIA(NVDA.US) to expand strategic cooperation and accelerate AI data center deployment. AMD(AMD.US) stock price drops accordingly.

Meta (META.US) announced a multi-year major partnership with NVIDIA (NVDA.US) to deploy millions of NVIDIA chips in its AI data centers, including NVIDIA’s latest standalone CPUs and the next-generation Vera Rubin system. The deal was announced on Tuesday, but specific financial terms have not been disclosed.

Meta CEO Mark Zuckerberg stated in a release that this expanded collaboration will continue the company’s long-term vision of “providing personal-level superintelligence for everyone around the world,” a vision officially introduced in July of this year. Following the announcement, Meta’s stock rose over 1% in after-hours trading, NVIDIA’s stock increased nearly 1%, while AMD (AMD.US) stock fell about 4%.

In January, Meta disclosed plans to increase AI-related capital expenditures to a maximum of $135 billion by 2026. Creative Strategies chip analyst Ben Bajarin noted that this partnership “likely involves hundreds of billions of dollars,” and expects a significant portion of Meta’s capital spending to be directed toward NVIDIA’s computing infrastructure.

Although Meta has used NVIDIA GPUs for at least a decade, the scope of this collaboration is clearly expanding. The most notable development is the standalone CPUs. Meta will become the first company to deploy NVIDIA’s Grace processors at scale in data centers, rather than integrating them with GPUs in servers as before. NVIDIA stated that this is the first time the Grace CPU has achieved true large-scale standalone deployment.

Bajarin pointed out that these CPUs are primarily designed for inference and “agent” workloads, serving as key components for Grace Blackwell or Vera Rubin racks. “Meta’s adoption at scale validates NVIDIA’s full-stack infrastructure strategy of deploying both CPUs and GPUs.” The next-generation Vera CPU is scheduled to be deployed in Meta’s data centers in 2027.

This multi-year agreement is also part of Meta’s overall infrastructure expansion plan. The company has committed to investing approximately $600 billion in data centers and related infrastructure in the United States by 2028. Meta plans to build 30 data centers, with 26 located in the U.S. Currently, two major AI data center projects are under construction: the 1 GW Prometheus project in New Albany, Ohio, and the 5 GW Hyperion project in Richland Parish, Louisiana.

In addition to computing chips, the agreement covers NVIDIA’s networking technology, including Spectrum-X Ethernet switches used for large-scale AI data center GPU interconnects. Meta will also leverage NVIDIA’s security technology to support AI features in products like WhatsApp.

However, Meta has not solely relied on NVIDIA. In November last year, reports indicated that Meta was considering introducing Google’s (GOOG.US, GOOGL.US) tensor processing units (TPUs) into its data centers by 2027, which temporarily caused NVIDIA’s stock to drop about 4%. Additionally, Meta is developing its own chips and continues to use AMD products. As AI computing demands surge and NVIDIA’s supply remains tight, industry giants are actively seeking “second sources.” Previously, OpenAI reached a significant partnership with AMD in October.

Currently, NVIDIA’s Blackwell architecture GPUs have been backlogged for months, and the next-generation Rubin GPUs have only recently entered mass production. Through this agreement, Meta is believed to have secured sufficient current and next-generation computing capacity. The engineering teams from both sides will also collaborate closely to optimize and accelerate Meta’s cutting-edge AI models.

On the model front, Meta is developing a new foundational model called “Avocado,” intended as the successor to the Llama series. However, the latest version released last spring did not generate strong reactions within the developer community. In recent months, Meta’s stock has experienced significant volatility, and its AI strategy has continued to divide Wall Street. The company suffered its worst single-day decline in three years last October due to aggressive AI investments, but after announcing better-than-expected revenue guidance in January, the stock surged by about 10%.