AMD Plans to Invest $1 Billion in Meta and Deepen Strategic Cooperation

On October 27 local time, AMD announced a deep strategic partnership with Meta, announcing that AMD Helios rack-scale solutions have begun running at Meta, that AMD EPYC processors power Meta's Muse AI agents, and that AMD technology provides Day-0 support for Meta models (including Muse Spark Open and Muse Glimmer).

This progress builds on years of joint design and collaboration between the two companies across AMD EPYC processors, AMD Instinct GPUs, AMD Pensando networking, and AMD ROCm AI infrastructure.

"Meta is a strategic partner of AMD, and we have built this partnership through years of deep technical collaboration—from the EPYC processors that power Meta's infrastructure to the Instinct GPUs and open software that support its AI workloads," said Lisa Su, AMD Chair and CEO. "With Helios now running at Meta, we are taking this partnership to a new scale in the age of superintelligence. Together we are building and deploying infrastructure at gigawatt scale and shaping future generations of chips, software, and systems to empower Meta's most ambitious superintelligence models and services."

"AMD has been a key partner in building the compute that runs our frontier models," said Mark Zuckerberg, Meta Founder and CEO. "We are investing heavily in compute to bring personal superintelligence to everyone, and AMD plays an important role in that."

AMD Powers Meta Infrastructure Today

AMD technology powers Meta's infrastructure across compute, networking, systems, and software.

The first AMD Helios rack-scale solutions have been installed at Meta and are being tested, moving toward broader deployment in the near future.

AMD EPYC processors support the dedicated compute environment behind Meta's Muse personal AI agents. Meta has deployed millions of EPYC processors across multiple product generations and is the lead customer for sixth-generation EPYC CPUs. The two companies are also collaborating on future generations, including "Verano."

AMD Instinct GPUs are currently used in production environments in Meta's infrastructure for training, ranking, and recommendation engines. Meta is also an early-definition customer for the AMD Instinct MI450 architecture and will deploy a custom AMD Instinct GPU optimized for its workloads. This collaboration is also informing future AMD platforms, including AMD Instinct™ MI500 series GPUs and next-generation rack-scale systems.

AMD and Meta are jointly advancing ROCm through PyTorch, ensuring developers can leverage AMD GPUs out of the box. In addition, AMD provided Day-0 support for Meta's Muse Spark Open model on AMD Instinct GPUs and for Muse Glimmer on AMD Ryzen™ AI PCs and AMD Radeon GPUs.

AMD and Meta have extended their software collaboration to large-scale AI training, optimizing recommendation models on AMD Instinct MI350X GPUs and contributing improvements to PyTorch, Triton, and ROCm for other developers to use.

As Meta continues to expand its AI capabilities, the two companies are working together across chips, software, and systems to meet the demands of operating AI infrastructure for billions of people. The two sides have aligned product roadmaps and engineering decisions around Meta's current and future workloads, evolving years of collaboration into building and deploying integrated AI systems at gigawatt scale.

In addition, AMD will invest $1 billion in Meta's equity offering.