Meta Platforms has signed a major deal with AMD to deploy 6 gigawatts of AI data center infrastructure, marking a significant expansion of Meta’s AI capabilities.
Meta Platforms has announced a major deal to deploy 6 gigawatts of AMD-based AI data center infrastructure, marking a significant win for the chipmaker as it competes with Nvidia.
The multi-year agreement will power Meta’s next-generation AI systems across multiple generations of AMD Instinct GPUs and 6th Gen AMD EPYC CPUs, codenamed “Venice” and “Verano,” running ROCm software on AMD Helios rack-scale architecture. The first gigawatt deployment is scheduled for the second half of 2026.
AMD and Meta will collaborate across silicon, systems, and software to optimize AI platforms for Meta’s workloads. As part of the deal, AMD has issued Meta a performance-based warrant for up to 160 million shares of AMD stock, vesting as specific shipment milestones are met and tied to both technical and commercial achievements.
“This partnership accelerates one of the industry’s largest AI deployments and places AMD at the center of the global AI buildout,” said AMD CEO Lisa Su. Meta CEO Mark Zuckerberg called it a “long-term partnership” to deploy efficient inference compute and diversify Meta’s AI infrastructure.
The agreement allows Meta to take a financial stake of up to 10 percent in AMD, reflecting a growing trend in the AI industry where chipmakers and large AI users invest mutually to secure supply and scale operations. Meta will purchase AMD chips and AI-ready computing systems over five years, with each gigawatt deployment valued in the double-digit billions of dollars.
This deal builds on an existing strategic partnership between the two companies and represents a major step in Meta’s expansion of AI infrastructure, ensuring efficient, scalable, and purpose-built compute for AI workloads globally.
End of Article
