AMD & Meta expand AI partnership with Helios systems - IT Brief Australia
AMD and Meta have expanded their strategic partnership across chips, systems, networking and software, with AMD's Helios rack-scale systems now running at Meta.
The work also includes support for Meta's AI models and infrastructure. Under the arrangement, AMD's EPYC processors support the computing environments behind Meta's Muse AI agent, while AMD Instinct GPUs are used in Meta production systems for training, ranking and recommendation workloads. Meta has also installed initial Helios rack-scale systems for testing ahead of a wider deployment.
Strategic Hardware Collaboration
Meta has deployed millions of EPYC processors across multiple generations and is a lead customer for AMD's sixth-generation EPYC CPUs. The two companies are also working on future processor generations, including a platform referred to as Verano. On the graphics side, Meta is described as an early definition customer for AMD's Instinct MI450 architecture. It will also deploy a custom Instinct GPU tailored to its workloads.
Meta is a strategic partner for AMD, and we have built this partnership through years of deep technical collaboration - from EPYC processors powering Meta's infrastructure to Instinct GPUs and open software supporting its AI workloads.
Software Integration
The partnership extends into software, with AMD working with Meta on ROCm through PyTorch so developers can use AMD GPUs from the outset. It also delivered Day-0 support for Meta's Muse Spark Open models on AMD Instinct GPUs, and for Muse Glimmer on AMD Ryzen AI PCs and AMD Radeon GPUs. Beyond model support, the companies have collaborated on large-scale AI training and optimizing recommendation models on AMD Instinct MI350X GPUs.
Financial Investment
A financial element accompanies the engineering tie-up: AMD will invest USD $1 billion in Meta's equity offering. The announcement offers a clearer view of how Meta is spreading its AI infrastructure across multiple layers of hardware and software as it expands the computing base behind its frontier models and consumer services.
Related
