AMD just delivered a harsh reminder that beating earnings isn't enough when you're chasing Nvidia in the AI chip wars. The chipmaker's shares tumbled more than 6% in after-hours trading Tuesday despite crushing Q4 expectations, as its first-quarter revenue forecast came in softer than some analysts hoped. With AI spending hitting unprecedented levels across the industry, the cautious outlook raises questions about whether AMD can truly capitalize on the boom while Nvidia continues to dominate the market for graphics processors powering AI models.
AMD thought it had all the right numbers. The chipmaker posted fourth-quarter revenue of $10.27 billion against expectations of $9.67 billion, while earnings per share hit $1.53 versus the $1.32 analysts were modeling. Net income more than tripled year-over-year to $1.51 billion. But none of that mattered once Wall Street saw the Q1 forecast.
The company projects first-quarter revenue of $9.8 billion, give or take $300 million. On paper, that beats the consensus estimate of $9.38 billion. In reality, it fell short of what some analysts were banking on as companies like Microsoft, Google, and Meta continue pouring tens of billions into AI infrastructure. Shares dropped more than 6% in extended trading Tuesday, erasing what should have been a victory lap.
The disconnect reveals the pressure AMD faces in its uphill battle against Nvidia, which still commands the lion's share of the AI chip market. While AMD has landed some high-profile customers recently - including OpenAI and Oracle - it's still playing catch-up in a race where expectations have become unmoored from traditional growth metrics. When you're growing revenue 34% year-over-year and investors still sell off, you know the bar has moved.
AMD's data center segment tells the story of both opportunity and limitation. Sales hit $5.4 billion in the quarter, up 39% annually, driven by demand for both its EPYC central processors and Instinct AI GPUs. CEO Lisa Su has been aggressively positioning the company's MI300 series chips as viable alternatives to Nvidia's H100 and H200 accelerators, and the numbers show some traction. But "some traction" isn't the same as "market share gains" when is projected to capture more than 80% of AI chip spending this year.
