Today, we’re announcing a multi-year agreement with AMD to energy our AI infrastructure with as much as 6GW of AMD Instinct GPUs, the silicon computing expertise used to assist trendy AI fashions.
At Meta, we’re working to construct the following technology of AI and allow private superintelligence for all. To do that, we’d like huge, scalable compute energy that may deal with the rising calls for of our AI workloads. Our partnership with AMD, which builds on our current collaboration, will assist us meet these wants.
Working With an Industry Leader
Under our new settlement, we will even be working with AMD on alignment with our roadmaps throughout silicon, methods and software program enabling vertical integration throughout our infrastructure stack. This collaboration throughout each software program and {hardware} will allow us to innovate rapidly and at scale.
“We are proud to expand our strategic partnership with Meta as they push the boundaries of AI at unprecedented scale,” stated Dr. Lisa Su, chair and CEO, AMD. “This multi-year, multi-generation collaboration across Instinct GPUs, EPYC CPUs and rack-scale AI systems aligns our roadmaps to deliver high-performance, energy-efficient infrastructure optimized for Meta’s workloads, accelerating one of the industry’s largest AI deployments and placing AMD at the center of the global AI buildout.”
Shipments to assist the primary GPU deployments will start within the second half of 2026, and can be constructed on the Helios rack-scale structure, a rack that we developed and announced eventually yr’s Open Compute Project Global Summit in collaboration with AMD.
“We’re excited to form a long-term partnership with AMD to deploy efficient inference compute and deliver personal superintelligence,” stated Mark Zuckerberg, Founder and CEO of Meta. “This is an important step for Meta as we diversify our compute. I expect AMD to be an important partner for many years to come.”
Our Portfolio-Based Approach
Our settlement with AMD is a part of our Meta Compute initiative, an effort to massively scale our infrastructure for the period of private superintelligence, future-proofing our management in AI. By diversifying our partnerships and expertise stack, we’re constructing a extra resilient and versatile infrastructure. We’re combining {hardware} sourced from a spread of companions with our personal quickly advancing Meta Training and Inference Accelerator (MTIA) silicon program.
We consider this portfolio method will allow us to advance and innovate at an unmatched tempo, rolling out highly effective, environment friendly new {hardware} co-designed with our software program stack to deal with huge development. We stay up for working with AMD to energy our AI improvements and safe our means to ship world-class AI experiences to billions of individuals globally.
This submit accommodates forward-looking statements, together with about Meta’s enterprise.You mustn’t depend on these statements as predictions of future occasions. Additional info concerning potential dangers and uncertainties might be present in our most up-to-date Form 10-Ok filed with the Securities and Exchange Commission. Meta undertakes no obligation to replace these statements on account of new info or future occasions.
