Bottom line: AMD and Anthropic are deepening their partnership around agentic AI workflows in software development, with AMD targeting a tenfold productivity increase while Claude is used in parallel to optimize AMD Instinct hardware and ROCm.
In a joint blog post, AMD and Anthropic describe how AI agents are taking over complete development workflows – from planning through coding to testing. In the medium term, AMD is aiming for a tenfold increase in its own developer productivity.
In a joint blog post, AMD and Anthropic explained how agentic AI systems are affecting software development. Rather than mere code completion, AI agents are increasingly taking over entire workflows: planning, writing, testing, and debugging code now run largely autonomously, while humans provide goals, context, and final evaluation. AMD executive Mark Papermaster emphasizes that the real challenge lies not in completing existing tasks faster, but in fundamentally rethinking how work is organized. Anthropic says it now deploys Claude broadly across the company – from onboarding to product development, code review, security testing, and incident response.
AMD itself has set a goal of increasing its own developer productivity tenfold in the medium term with the help of agentic AI. As evidence, the company cites complex design problems that previously took several weeks to solve and can now be handled within a single day. For CTOs, this signals that the shift from Copilot-style code completion to agentic workflows is not merely an efficiency issue but an organizational one – for example, regarding role distribution, review processes, and accountability for automatically generated code.
The blog post follows the strategic partnership between the two companies announced in July. Under the agreement, Anthropic will deploy up to 2 gigawatts of AMD Instinct MI450 accelerators in AMD Helios rack systems, complemented by AMD EPYC processors from the Venice series and AMD Pensando networking technology. Deployment of the first gigawatt is scheduled to begin in the first half of 2027. In return, AMD has committed to a strategic equity investment in Anthropic of up to $5 billion.
A multi-year technical collaboration was also agreed upon: Claude is to be used to optimize workloads on AMD Instinct hardware and to accelerate development of AMD’s ROCm software platform, while AMD rolls out Claude extensively across its own development teams. For decision-makers in infrastructure and platform teams, this links two levels – hardware commitment and productive AI use in everyday development – into a shared strategic narrative between chipmaker and model provider.