AMD to invest up to $5 billion in Anthropic under AI infrastructure deal
| Source: AI News (ainews.com)
Tags: AMD, Anthropic, AI infrastructure, Instinct MI450, AI hardware, capital investment
AMD has agreed to invest up to $5 billion directly in Anthropic under a deal committing Anthropic to deploy 2 gigawatts of AMD Instinct MI450 infrastructure — the third major AMD financial-plus-hardware agreement after similar deals with OpenAI and Meta, structured this time as direct equity rather than warrants.
Details
AMD and Anthropic have announced a combined equity-and-infrastructure deal: AMD commits up to $5 billion in direct strategic equity, tied to Anthropic meeting unspecified deployment milestones, while Anthropic agrees to deploy up to 2 gigawatts of AMD-based AI compute. The first gigawatt — running on AMD Helios systems with MI455X GPUs (Instinct MI450 series), EPYC 'Venice' processors, Pensando networking, and ROCm software — is scheduled to begin deployment in H1 2027. AMD's investment has not yet transferred; it will be released as milestones are met. This follows AMD's pattern of combining hardware commitments with financial stakes at frontier AI labs. In October, AMD struck a deal with OpenAI for up to 6 GW of capacity alongside warrants allowing OpenAI to acquire roughly 10% of AMD shares. A February deal with Meta covers 6 GW with performance-based warrants. The Anthropic deal differs structurally — AMD invests equity directly rather than issuing warrants — suggesting a deeper financial alignment with the Claude maker. Anthropic already runs AMD MI355X accelerators in production, so this extends an existing relationship. Critical terms remain undisclosed: Anthropic's purchase obligations, cancellation rights, the AMD investment valuation, and treatment of delayed deployments. Nvidia has separately held talks about investing up to $30 billion in OpenAI — signaling that chip suppliers broadly are using financial stakes to anchor AI infrastructure relationships.