AMD & Anthropic Ink 2GW GPU Deal

AMD and AI developer Anthropic have announced a sweeping strategic partnership that will see Anthropic deploy up to 2 gigawatts of AMD Instinct MI450 Series GPUs. This massive deployment, utilizing AMD Helios rack-scale solutions, is set to begin in the first half of 2027.

The deal signifies a significant expansion for AMD in the burgeoning AI infrastructure market, driven by the insatiable demand for AI training and inference capacity. Anthropic is scaling its compute power to meet the growing demand for its Claude AI models.

The partnership goes beyond hardware supply. AMD and Anthropic will collaborate on software, with Anthropic's Claude AI being used to optimize workloads for AMD Instinct GPUs and accelerate the development of AMD's ROCm software stack. AMD will also integrate Claude across its own engineering and product development teams.

As part of the agreement, AMD has committed to a strategic equity investment of up to $5 billion in Anthropic. This move underscores AMD's confidence in Anthropic's future and its role in the AI landscape.

AMD's Helios at Scale

Anthropic will be deploying AMD Helios rack-scale solutions, featuring the latest AMD Instinct MI455X GPUs, alongside AMD EPYC CPUs, Pensando networking, and ROCm software. This comprehensive infrastructure is designed for the most demanding AI training and inference tasks.

This marks a significant scaling of Anthropic's compute infrastructure, building upon their existing use of AMD Instinct MI355X GPUs. The sheer scale of this deployment, reaching gigawatt scale GPU deployment, highlights the intense competition and investment in AI hardware.

Dr. Lisa Su, AMD's CEO, emphasized the synergy between Anthropic's AI leadership and AMD's high-performance computing capabilities. She stated that the collaboration will accelerate AI adoption and establish Helios as a pivotal platform for future AI infrastructure.

Tom Brown, Anthropic's co-founder and chief compute officer, highlighted the critical need for compute access to maintain Claude's leading edge and satisfy customer demand. He noted that partnering with AMD across the stack ensures necessary capacity and optimizes it for Claude's specific training and serving needs.

This extensive collaboration, covering both massive hardware deployment and deep engineering integration, aims to create a robust, long-term foundation for next-generation AI infrastructure development and deployment.