AMD Buys Into Anthropic With $5B Equity Deal Tied to 2GW of AI Chips

AMD and Anthropic unveiled the deal on July 22, 2026. Under the terms, AMD commits up to $5 billion in Anthropic Class N common stock, released in stages as delivery and deployment milestones get hit. The first milestone is expected in 2027, with the full investment spread through fiscal year 2028. The shares carry no voting rights, but match Anthropic's Class A stock on economic rights.

The hardware side carries the weight. Anthropic plans to deploy up to 2 gigawatts of AMD's Instinct MI450 Series GPUs, housed in AMD's Helios rack-scale systems, with the first phase targeted for the first half of 2027. That makes AMD a compute supplier to Anthropic for the first time.

The structure flips AMD's usual AI deal. With OpenAI and Meta, AMD handed out equity warrants to lock in hardware commitments. Here it's the other way around: AMD is buying a slice of its customer. The agreement also includes a multi-year engineering partnership, with AMD planning to run Anthropic's Claude models for its own internal software development and both sides tuning workloads for ROCm, AMD's GPU software platform.

Anthropic's numbers give the check some scale. The company was valued at $965 billion after its Series H round, and its total committed spending on compute and infrastructure sits around $895 billion. AMD's up-to-$5 billion lands as a real check but a small slice of that appetite, and it diversifies Anthropic's supply beyond Amazon and Google.

Timing ties the deal to Anthropic's march toward public markets. The company filed its S-1 on October 26, 2026, ahead of a planned Nasdaq listing, and the Class N stake gives AMD exposure to Anthropic's fortunes before any debut. For AMD, the 2-gigawatt commitment is a high-profile reference point in AI compute. If deployments slip, the investment timeline likely slips with them — making the first-half 2027 rollout and the 2027 first milestone the markers to watch.