Advanced Micro Devices crossed a $1 trillion market capitalization for the first time on Monday, September 21. Its shares rose 9.9% in a single session and roughly a quarter over five trading days. The milestone comes as AMD begins shipping Helios, its first full rack-scale AI system, to the AI labs and hyperscalers that have spent the past three years buying almost entirely from Nvidia.
Shares hit an intraday high of $615.52, according to CNBC, which pushed AMD's value just past the 13-digit mark. Yahoo Finance reported the stock is now up about 185% year to date, compared with 22% for Nvidia. The Motley Fool counts AMD as the fourth chipmaker to reach a $1 trillion valuation, alongside Nvidia, Broadcom and Micron, after a weekly gain of about 24%.
The Numbers Behind the Run
The rally rests on AMD's fundamentals as much as on market momentum. Earlier this month the company reported second-quarter revenue of a record $11.5 billion, up 50% from a year earlier, against analyst estimates of about $11.34 billion. Non-GAAP earnings came in at $1.66 per share, beating the $1.61 consensus. The data center segment drove most of the gain: revenue more than doubled to $6.7 billion, up 107% year over year, on demand for EPYC server CPUs and Instinct AI accelerators.
The guidance was even more aggressive. AMD said it expects server CPU revenue to grow 80% year over year in the second half of 2026 and another 70% in 2027. It also projects data center revenue to more than double next year, with AI GPU sales growing "well over 100%," according to Yahoo Finance.
CEO Lisa Su has said repeatedly that she sees no slowdown. "We're seeing the returns on investment," Su told Yahoo Finance in a late-July interview. "Demand for compute is at a premium today. We are very confident in the demand picture being there."
Helios Moves From Keynote to Loading Dock
Helios is the product investors are watching most closely. AMD introduced the system formally at a San Francisco event in July. It combines Instinct MI450-series GPUs, sixth-generation EPYC Venice CPUs, Pensando networking and ROCm software in a single integrated rack. Each rack pairs 72 Instinct accelerators with 18 EPYC processors. The design competes directly with Nvidia's rack-scale NVL systems, which have dominated frontier training clusters. AMD has said it will begin shipping Helios to customers, including Microsoft, in the second half of 2026, and Su has pointed to demand from AI labs and hyperscalers deploying the racks as a main reason for the company's confidence.
AMD's customer list is the reason investors see the rack as more than a product launch. Microsoft will deploy Helios on Azure for frontier model inference. Anthropic has agreed to deploy up to 2 gigawatts of MI450-series GPUs in Helios systems, with the first gigawatt coming online in the first half of 2027, and AMD committed a strategic equity investment of up to $5 billion in the lab. AMD also has commitments from OpenAI, which is deploying 6 gigawatts of AMD GPUs, and from Meta, for up to 6 gigawatts of Instinct GPUs.
"Access to compute is central to keeping Claude at the frontier and meeting demand from our customers," Tom Brown, Anthropic's co-founder and chief compute officer, said when the deal was announced. "By partnering with AMD across the stack, we are securing the capacity we need and optimizing it for training and serving Claude." Su said the agreement "brings together Anthropic's leadership in frontier AI with the full strength of AMD high-performance computing."
Why It Matters
For most of the generative AI boom, the accelerator market has effectively had one supplier. Futurum Group estimates cited by CNBC earlier this year put Nvidia's share of data center GPUs above 95%, with AMD at around 4.5%. A $1 trillion valuation does not change that ratio overnight, but it shows that investors now believe a second supplier at scale is possible, and that the largest buyers of compute want one.
The shape of AMD's deals matters as much as their size. Anthropic, OpenAI, Meta and Microsoft are not buying individual chips to test. They are committing gigawatts of capacity across multiple years, in some cases with equity or co-engineering arrangements attached. That gives AMD predictable revenue and gives the labs pricing leverage and supply insurance at a time when compute access decides which models ship. It also puts the whole AMD platform under test, including ROCm, the software layer that has long been the weakest part of its offering. Anthropic's plan to use Claude to speed up ROCm development is a sign of how much depends on closing that gap.
There is a market-structure point as well. With AMD in the club, four chipmakers are now valued at $1 trillion or more, which concentrates even more of the equity market's value in AI hardware. That raises the stakes if hyperscaler capital spending slows.
What to Watch
The next test is execution. Helios shipments are beginning now and ramping through 2027, and any delay in MI450 volume, rack integration or networking would hit a stock priced for flawless delivery. Investors will look for AMD's third-quarter report to show early Helios revenue and to confirm that data center sales are on track to more than double next year.
The Anthropic deployment in the first half of 2027 and Microsoft's Azure rollout will be the first large public tests of whether Helios can run frontier training and inference at the reliability Nvidia customers expect. Nvidia's response on pricing and on its next rack generation will shape how much share AMD can actually take. At a $1 trillion valuation, the market has already priced in a strong second-place finish. AMD now has to deliver it.
“Access to compute is central to keeping Claude at the frontier and meeting demand from our customers.”— Tom Brown, Co-founder and Chief Compute Officer, Anthropic