AMD invests up to $5B in Anthropic for 2GW of AI GPU capacity

AMD announced a landmark investment of up to $5 billion in Anthropic, securing a massive expansion of the AI company's computing power. Under the agreement, Anthropic will deploy up to 2 gigawatts of AMD's Instinct MI450 AI GPUs, leveraging the chipmaker's new Helios rack-scale system. The first gigawatt of capacity is targeted for the first half of 2027, building on Anthropic's existing data center deals with SpaceX and TeraWulf. This move positions AMD as a key player in the AI infrastructure race, directly competing with Nvidia's dominance in GPU supply for large-scale AI workloads. The deal also underscores Anthropic's aggressive push to secure compute resources for training and serving its Claude models, as the company also has agreements with Google, Broadcom, Amazon, and reportedly Meta.

Beyond the hardware commitment, AMD and Anthropic will launch a multi-year engineering collaboration. AMD plans to integrate Anthropic's Claude across its software development, engineering, and product development workflows, deepening the strategic alignment between the two companies. This partnership comes as Anthropic reportedly pays $15 billion per year for access to Elon Musk's data centers, highlighting the staggering costs of frontier AI training. For AMD, the investment not only guarantees a major customer for its next-generation GPUs but also provides real-world validation for its Helios rack-scale architecture, which aims to simplify deployment of massive AI clusters. The deal signals a potential shift in the AI hardware landscape, as hyperscaler-like demand forces AI labs to diversify their chip supply chains beyond Nvidia.