Advanced Micro Devices (AMD) has secured a major partnership with AI firm Anthropic, committing up to $5 billion in strategic equity investment and agreeing to supply substantial AI computing power. The agreement, announced on Wednesday, positions AMD as a key hardware provider for Anthropic's growing demand for computational resources to train and operate its large language models, including its Claude AI system.
Under the terms of the multi-year agreement, Anthropic will deploy up to two gigawatts of AMD's Instinct MI450 Series GPUs, utilizing AMD's Helios rack-scale solutions. The initial deployment of one gigawatt is slated to begin in the first half of 2027. These Helios systems will incorporate AMD Instinct MI455X GPUs, alongside AMD EPYC "Venice" CPUs, AMD Pensando networking, and ROCm software. This move marks an expansion of Anthropic's existing use of AMD hardware, as the company also utilizes AMD Instinct MI355X GPUs.
The partnership extends beyond hardware supply, including a multi-year engineering collaboration. Both companies will work together to use Anthropic's Claude models to optimize workloads for AMD Instinct GPUs and accelerate the development of AMD's ROCm software. AMD also plans to adopt Claude across its own engineering and product development teams.
AMD's commitment of up to $5 billion in Anthropic represents its first direct equity investment in an AI company. This investment is contingent upon Anthropic reaching certain computing power deployment milestones. The deal aims to challenge the dominance of Nvidia in the artificial intelligence chip market, where its GPUs have historically been the primary choice for training large AI models. Leading AI companies, including Anthropic, have relied heavily on Nvidia for their computing needs.
AMD Chief Executive Dr. Lisa Su stated that the collaboration will help "establish Helios as a major platform for the next generation of AI infrastructure." She added that AMD has "very much wanted to be a major part of their infrastructure." Tom Brown, co-founder and chief compute officer at Anthropic, emphasized the importance of compute access for maintaining Claude's capabilities and meeting customer demand, noting that "running across a diversified range of hardware lets us map the right workloads to the right hardware."
The agreement is part of Anthropic's strategy to diversify its computing infrastructure and secure capacity as demand for its AI services grows. The company has previously announced significant infrastructure deals with other major players, including Google and Amazon, as well as SpaceX. The scale of the chip purchase is anticipated to amount to tens of billions of dollars over time.
AMD's stock performance has shown volatility in response to the news, with some reports indicating a slight rise in premarket trading on Wednesday following the announcement. The company's data center business has become a significant driver of growth, with customer forecasts for MI450 systems reportedly exceeding initial expectations.