AMD is turning AI demand into long-term customer commitments
Advanced Micro Devices is deepening its push into artificial intelligence infrastructure with a major deal to supply Anthropic with tens of billions of dollars’ worth of AI servers.
The agreement gives Anthropic access to AMD’s latest-generation Instinct MI450 chips, starting in the first half of 2027. It also includes a potential AMD investment of up to $5 billion in the Claude maker, tied to certain deployment milestones.
The deal is significant because it shows AMD is no longer only trying to win share through better chips. It is using large customer partnerships, financing structures and long-term deployment commitments to build a stronger position in a market still dominated by Nvidia.
For AMD, Anthropic is exactly the type of customer it needs: fast-growing, compute-hungry and under pressure to secure enough infrastructure to keep its AI models competitive.
Anthropic needs compute to keep Claude at the frontier
Anthropic has become one of the leading AI companies in Silicon Valley, helped by strong adoption of enterprise products such as Claude Code.
That growth has created a familiar problem across the AI sector: demand for the product is rising faster than available computing capacity. Anthropic co-founder and compute head Tom Brown said access to compute is central to keeping Claude at the frontier and meeting customer demand.
The AMD deal gives Anthropic another major supply channel at a time when frontier AI labs are racing to lock in chips, data-centre capacity and cloud availability years in advance.
Under the agreement, Anthropic will buy up to 2 gigawatts of AMD chips. AMD executives have said 1 gigawatt of computing capacity can cost double-digit billions of dollars, which helps explain why the deal could become one of the largest AI infrastructure commitments in the market.
The deal strengthens AMD’s challenge to Nvidia
Nvidia still controls the AI accelerator market, but AMD is becoming more aggressive in trying to win large strategic customers.
The Anthropic agreement follows AMD’s October deal with OpenAI, which was also expected to generate tens of billions of dollars in annual revenue and gave the ChatGPT creator the option to buy up to roughly 10% of AMD.
The pattern is becoming clearer. AMD is trying to position itself as the most credible second major platform for hyperscalers and AI labs that need alternatives to Nvidia.
That matters because the AI market is too large for customers to rely on one supplier. Model developers need more capacity, cloud providers need more pricing leverage and large enterprises want more choice in the infrastructure supporting AI workloads.
AMD does not need to overtake Nvidia to create a major business. If it can win even a meaningful minority share of frontier AI deployments, the revenue opportunity could be enormous.
Circular AI deals are becoming part of the market
The Anthropic deal is also part of a broader trend in AI financing, where chipmakers invest in companies that are also among their biggest customers.
Nvidia has reportedly been in talks to invest $30 billion in OpenAI. AMD is now investing in Anthropic while also selling it a massive amount of AI infrastructure.
These deals can look circular, but they reflect the current economics of AI. Model companies need more compute than they can easily finance on traditional terms, while chipmakers want to lock in demand for future hardware generations.
The structure gives both sides something they need:
- AI labs get access to scarce compute capacity
- Chipmakers secure long-term demand
- Cloud providers gain more infrastructure partners
- Investors get clearer visibility into future AI spending
The risk is that these arrangements make it harder to separate organic demand from demand supported by vendor financing or strategic investment. That will become a larger market question as the dollar amounts grow.
Anthropic is moving quickly to secure capacity
Anthropic has already been working aggressively to solve its compute bottleneck.
In May, the company agreed to rent the full computing power of SpaceX’s Colossus 1 facility in Memphis, which houses more than 220,000 Nvidia processors and added 300 megawatts of capacity. Meta is also reportedly in talks to lease computing power to Anthropic in a deal worth up to $10 billion over two years.
The AMD agreement adds another major layer to that strategy. According to the Wall Street Journal, Anthropic would use some of the AMD chips in its own data centres and lease additional capacity through cloud providers and new AI cloud companies.
AMD was also reportedly in talks to provide a financial backstop for Anthropic’s future data-centre leases.
That shows how tight the AI infrastructure market has become. Anthropic is not relying on one data-centre partner, one chip supplier or one financing path. It is building capacity through every available channel.
AMD’s stock has already priced in a stronger AI story
AMD shares ticked up 0.5% in early trading after the announcement. The muted reaction likely reflects how much optimism has already moved into the stock.
Shares have more than doubled this year as investors price in AMD’s growing role in the AI server market. The company’s opportunity is real, but the market is now looking for proof that major commitments can translate into deployments, revenue and margin expansion.
The Anthropic deal helps that case because it gives AMD another marquee AI customer. The next test is execution.
The key questions now are straightforward:
- Can MI450 chips perform at scale against Nvidia systems
- Can AMD deliver enough capacity on schedule
- Can Anthropic deploy the systems efficiently
- Can AMD turn these large agreements into durable data-centre revenue
If the answer is yes, AMD’s AI business could move from challenger narrative to major revenue engine.
WSA Take
AMD’s Anthropic deal is another sign that the AI infrastructure race is shifting from chip announcements to long-term capacity control. Anthropic needs compute to keep Claude competitive, and AMD needs flagship customers to prove it can challenge Nvidia at scale.
The opportunity for AMD is not replacing Nvidia overnight. It is becoming the second major AI infrastructure platform for customers that need more supply, more choice and more leverage. The risk is execution, but the size of the commitment shows how serious AMD’s AI opportunity has become.
Disclaimer
WallStAccess is a financial media platform providing market commentary and analysis for informational and educational purposes only. This content does not constitute investment advice, a recommendation, or an offer to buy or sell any securities. Readers should conduct their own research or consult a licensed financial professional before making investment decisions.
