AMD’s Revenue Jumps 38% as Its AI Chips Close the Gap

Advanced Micro Devices reported first-quarter revenue of $10.25 billion, up 37.85% from a year earlier, the fastest growth in years, as its AI accelerators and server processors found buyers across the data center market. The results, reported May 5, beat analyst expectations, and the company’s outlook for the current quarter came in ahead of estimates as well.

The engine is data center computing. That segment brought in $5.8 billion in revenue, up 57% from a year earlier, powered by AMD’s EPYC server processors and the continuing ramp of its Instinct line of AI accelerators, led by the MI300 family. Lisa Su, AMD’s chief executive, put the shift in plain terms: data center has become the primary driver of the company’s revenue and earnings growth, a role it did not hold two years ago.

The financials improved across the board. GAAP gross margin came in at 53%, operating income at $1.48 billion and net income at $1.38 billion, or 84 cents a share. On a non-GAAP basis, gross margin was 55%, operating income $2.54 billion and earnings $1.37 a share. The company generated record quarterly free cash flow, according to chief financial officer Jean Hu.

The AI numbers are where the market is looking. AMD has spent two years positioning its Instinct accelerators as the second choice in a market dominated by Nvidia, and the strategy is producing revenue even if the market share remains lopsided. Nvidia still controls roughly 94% of the AI GPU market, according to industry estimates, a dominance that makes AMD’s progress a matter of taking slivers at the margin. But the pace of the chase is accelerating: AMD’s shares have climbed about 280% over the past twelve months, and Su said customer engagement around the next-generation MI450 series and the Helios platform is strengthening, with leading customers’ forecasts exceeding the company’s initial expectations.

The demand picture helps. AI workloads are shifting from training, which runs in bursts on clusters of GPUs, to inference, which runs constantly and needs a mix of CPUs and accelerators. Agentic AI, software that acts on users’ behalf, is adding workloads that spend heavily on CPU cycles. Su said the company sees strong momentum as inferencing and agentic AI drive demand for high-performance CPUs and accelerators, and she said server growth should accelerate meaningfully as the company scales supply to meet demand.

The rest of the business held up. Client and gaming revenue was $3.6 billion, up 23%, driven by Ryzen processor share gains; embedded revenue was $873 million, up 6%. The growth in the core data center and client lines more than offset the modest contributions from legacy businesses.

The quarter’s numbers also show how far AMD has come from the dark years of the late 2010s, when the company was selling assets to stay alive. The recovery was built on a series of bets that all paid off: the Zen processor architecture that made EPYC a credible server challenger, the acquisition of Xilinx that added programmable chips, and the early commitment to AI accelerators at a moment when rivals treated the market as a niche. Su, who took over in 2014, has presided over a roughly tenfold increase in the share price and a transformation of the company’s revenue mix, with data center now the largest and fastest-growing segment.

The comparison to Intel has flipped. Two years ago, Intel out-earned AMD in every segment and held the server market in a grip that seemed permanent. Intel’s struggles with its manufacturing transition, and its costly foundry turnaround under chief executive Lip-Bu Tan, have opened the door, and AMD has walked through it with EPYC, taking server share quarter after quarter. AMD’s revenue is still smaller than Intel’s, but the gap is narrowing, and in the segments that matter for AI, AMD is growing faster.

The question is whether AMD can convert momentum into share. Nvidia’s 94% grip on AI GPUs is defended by a software ecosystem, years of customer relationships and a supply chain that AMD cannot match at scale. AMD’s answer has been to sell into the customers who want a second source: hyperscale cloud operators, enterprise buyers and national champions that do not want to depend on a single vendor. The Instinct ramp suggests the pitch is landing.

The guidance for the current quarter adds to the case. AMD said it expects revenue above the Street’s estimate at the time, a signal that the data center ramp has not hit a demand ceiling, and executives pointed to a pipeline of large-scale deployments that extends visibility into next year. The company is also benefiting from the broad move by cloud providers to diversify chip suppliers: every hyperscaler wants a second source for accelerators, and AMD is the only merchant vendor besides Nvidia that can supply them at scale. That structural position, more than any single quarter, is what the stock price reflects.

For investors, the quarter answered a narrower question: AMD can grow fast while the AI boom is running. The harder question is what happens to the 94% number. Every quarter of 38% revenue growth and 57% data center growth chips at it, but slowly. AMD’s share price has already priced in a lot of success. The company’s job now is to keep delivering the numbers that make the chase look like a race rather than a long pursuit.

Related Posts

  • September 6, 2026
  • 10 views
Anthropic Moves Its IPO Filing to Late September

The bankers and lawyers running Anthropic’s initial public offering had told investors to expect the company’s registration documents as soon as this week. The calendar has moved. Anthropic now plans…

  • September 6, 2026
  • 12 views
OpenAI Quietly Revises GPT-6 Astra Scores After Launch

When OpenAI released GPT-6 Astra on Sept. 3, the launch post carried the usual furniture of a modern model debut: coding results, speed comparisons and a figure for how often…