AMD And Anthropic Strike Huge 2-Gigawatt MI450 Chip Deal And Up To $5B Stake
AMD and Anthropic have just announced a strategic partnership that will see the deployment of up to 2 gigawatts of AMD Instinct MI450 series GPUs in Helios rack-scale solutions. Also as part of the partnership, AMD said it committed to a potentially massive equity stake investment of up to $5 billion in Anthropic as the AI race heats up.
"We are thrilled to deepen our partnership with Anthropic and deploy AMD Helios at gigawatt scale," said Dr. Lisa Su, chair and CEO, AMD. "This collaboration brings together Anthropic’s leadership in frontier AI with the full strength of AMD high-performance computing. Together, we will accelerate AI adoption at scale and establish Helios as a major platform for the next generation of AI infrastructure."
This is another vote of confidence in AMD's AI hardware. Back in February, AMD landed a multi-year deal with Meta to provide key hardware for its AI infrastructure build-out, starting with a custom GPU based on its Instinct MI450 accelerators.
Under terms of this latest agreement, Anthropic will scale its own infrastructure by integrating AMD Helios rack-scale solutions to power its next-generation AI training and inference models for Claude. Deployment of the first gigawatt is scheduled for the first half of 2027.
The agreement builds on Anthropic's earlier utilization of AMD's Instinct MI355X GPUs. Looking ahead, the plan is to leverage next-generation Instinct MI455X GPUs, though it doesn't stop there. Anthropic will also deploy AMD's 6th Gen EPYC 'Venice' server processors, along with AMD's Pensando networking and ROCm software.
Notably, the breadth of this deal is similar to the one AMD just struck with Microsoft for its Azure cloud platform, in terms of a full-stack infrastructure. Like that deal, this one with Anthropic underscores how hyperscalers are evaluating complete systems rather than taking a piecemeal approach. The GPU is obviously an important piece of the puzzle, but so are things like networking, software, and ability to deploy at scale, and so forth.
"Access to compute is central to keeping Claude at the frontier and meeting demand from our customers," said Tom Brown, co-founder and chief compute officer, Anthropic. "By partnering with AMD across the stack, we are securing the capacity we need and optimizing it for training and serving Claude. Running across a diversified range of hardware lets us map the right workloads to the right hardware."
What this massive deal with Anthropic and others like it indicate is that AMD is confident it can execute on its Helios deployment strategy. Now it just has to prove it.
