AMD will invest up to $5 billion in Anthropic under a deal to deploy two gigawatts of AI computing capacity using AMD’s next-generation chips.
AMD has agreed to invest up to $5 billion in Anthropic as part of an infrastructure agreement covering tens of billions of dollars’ worth of AI computing systems. The deal marks one of AMD’s largest commitments yet to a single AI company.
What the AMD-Anthropic Deal Covers
Under the agreement, Anthropic will deploy up to two gigawatts of computing capacity using AMD’s Instinct MI450-series accelerators. Deployment of the first gigawatt is expected to begin in the first half of 2027.
AMD’s investment is tied to Anthropic meeting certain deployment milestones, though the specific conditions have not been disclosed. The companies have also not said whether Anthropic must use the investment funds to pay for the systems itself.
- Up to $5 billion equity investment from AMD
- Two gigawatts of planned AI computing capacity
- First gigawatt deployment starting in early 2027
- Investment tied to undisclosed deployment milestones
How This Compares to AMD’s Other AI Deals
The Anthropic agreement follows similar large-scale arrangements AMD has struck with other AI developers. In October, AMD agreed to supply OpenAI with systems supporting up to six gigawatts of capacity and issued warrants that could let OpenAI acquire about 10% of AMD’s shares.
Those OpenAI warrants vest in stages tied to deployment, commercial, and share-price conditions. AMD has said that agreement could generate tens of billions of dollars in revenue.
AMD reached a separate agreement with Meta in February covering up to six gigawatts of GPU capacity, which also included performance-based warrants linked to shipment and purchase targets.
The structures differ across the three deals. With Anthropic, AMD is making a direct equity investment. With OpenAI and Meta, those companies instead receive rights to acquire AMD shares once specific conditions are met.
| Partner | Capacity | Financial Structure |
|---|---|---|
| Anthropic | Up to 2 gigawatts | Direct equity investment up to $5B |
| OpenAI | Up to 6 gigawatts | Warrants for ~10% of AMD shares |
| Meta | Up to 6 gigawatts | Performance-based share warrants |
Separately, Nvidia has held talks about investing up to $30 billion in OpenAI, according to Reuters, underscoring a broader trend of chip suppliers pairing equity stakes with commercial supply agreements involving major AI developers.
The Technology Behind the Deployment
Anthropic will deploy AMD Helios systems built around MI455X GPUs from the Instinct MI450 series, alongside AMD’s EPYC “Venice” processors, Pensando networking hardware, and ROCm software. This builds on Anthropic’s existing use of AMD’s earlier MI355X accelerators.
The arrangement covers an integrated rack-scale platform spanning accelerators, CPUs, networking, and software, rather than a standalone GPU purchase.
AMD chair and chief executive Lisa Su said the companies are “thrilled to deepen our partnership with Anthropic and deploy AMD Helios at gigawatt scale.” She added that the collaboration brings together Anthropic’s frontier AI work with AMD’s high-performance computing capabilities, and said the companies intend to establish Helios as a platform for future AI infrastructure deployments.
Scale of the Commitment
The two-gigawatt Anthropic commitment represents 2,000 megawatts of power capacity. AMD executives have said that building one gigawatt of AI computing infrastructure can cost tens of billions of dollars, depending on equipment and facilities.
Anthropic has separately secured more than 300 megawatts through SpaceX’s Colossus 1 facility in Memphis, which houses more than 220,000 Nvidia GPUs. At full scale, the planned AMD deployment would represent more than six times the power capacity of Colossus 1, though the two figures are not directly comparable in computing performance since they involve different hardware and deployment models.
According to The Wall Street Journal, Anthropic plans to install some AMD systems in its own data centers, while others will be hosted by cloud providers and specialist AI infrastructure companies. AMD and Anthropic are jointly seeking operators capable of hosting the planned systems.
Su noted that gigawatt-scale capacity typically requires 12 to 24 months of planning before deployment. The Wall Street Journal also reported that AMD had discussed providing financial guarantees for Anthropic’s future data-center leases, though neither company has confirmed the proposed terms.
Anthropic’s Broader Compute Strategy
The AMD systems will operate alongside infrastructure based on Nvidia GPUs, Amazon Trainium processors, and Google tensor processing units. Amazon remains Anthropic’s primary cloud and training partner, having secured Anthropic up to five gigawatts of capacity, with more than one million Trainium2 processors currently in use and Trainium3 deployments planned during 2026.
Anthropic has also secured multiple gigawatts of next-generation TPU capacity from Google and Broadcom, with deployments expected to begin in 2027. Claude is available through Amazon Web Services, Google Cloud, and Microsoft’s Azure-based Foundry platform.
- Up to 5 gigawatts from Amazon, including Trainium2 and upcoming Trainium3 chips
- Multiple gigawatts of TPU capacity from Google and Broadcom starting 2027
- More than 300 megawatts via SpaceX’s Colossus 1 facility
- Ongoing discussions to lease infrastructure from Meta
Anthropic co-founder and chief compute officer Tom Brown said running across a range of hardware lets the company “map the right workloads to the right hardware.” He added that access to compute is central to keeping Claude competitive and meeting customer demand, and that the AMD partnership helps secure additional capacity optimized for training and serving Claude.
Reuters has reported that a potential agreement to lease infrastructure from Meta could be worth up to $10 billion over two years, though terms have not been finalized.
An Engineering Partnership Beyond Hardware
The AMD agreement also includes a multi-year engineering programme in which the companies will use Claude to help optimize workloads for AMD’s Instinct accelerators and support development of AMD’s ROCm software platform. AMD plans to deploy Claude across its own engineering and product-development teams.
This gives Anthropic a role in adapting Claude workloads to AMD’s hardware and software environment, making Anthropic both a customer of AMD’s infrastructure and a contributor to the software ecosystem that runs on it.
Market Reaction
AMD shares rose 2.4% following the announcement. The company’s stock has more than doubled since the start of the year, outperforming the broader Philadelphia Semiconductor Index.
Frequently Asked Questions
AMD has agreed to invest up to $5 billion in Anthropic, tied to deployment milestones that have not been publicly detailed.
Anthropic will deploy AMD Helios systems featuring MI455X GPUs from the Instinct MI450 series, along with EPYC "Venice" processors, Pensando networking, and ROCm software.
AMD is making a direct equity investment in Anthropic, while its OpenAI and Meta agreements instead grant those companies warrants to acquire AMD shares based on performance conditions.
