AMD shared a revenue forecast of roughly $17 billion for the fourth quarter of 2026. Intel, for the same period, gave a figure in the $15-16 billion range. The gap — which could mark the first time in company history that AMD has surpassed Intel in quarterly revenue — is being driven by rising sales of EPYC server processors and Instinct AI accelerators.

What happened?

AMD management's fourth-quarter revenue guidance came in above Intel's own forecast. The figures disclosed are each company's own projections, not finalized results yet; but the gap suggests that the decades-long balance between the two rivals could be reversing for the first time. AMD's growth is concentrated in two products: EPYC processors used in data center servers, and the Instinct series of accelerators built for AI training and inference.

Why does it matter?

AMD's revenue claim against Intel is the natural extension of the turnaround the company launched in 2017 with the Zen architecture. Before that, AMD was nearly written off in the server market due to the failure of the Bulldozer architecture. With successive EPYC generations after Zen (Milan, Genoa, Turin), AMD steadily grew its server CPU market share; but overtaking Intel on revenue is a different threshold, because Intel — as both a CPU maker and a foundry operating its own fabs — has a much larger revenue base.

This time, the real factor reshaping the picture is AI demand. Major cloud providers — companies like Microsoft, Meta, and Oracle — want to reduce their dependence on Nvidia GPUs for training and inference workloads, both because of supply constraints and to gain price negotiating leverage. AMD's Instinct MI series stands out as the most mature alternative in that search. At the same time, Intel is grappling with delays in its 18A manufacturing process and losses on the foundry side, which has left the company late to market with next-generation data center products.

Who does it affect, and how?

CompanyServer CPUAI ChipManufacturing Model
AMDEPYC, growing market shareInstinct MI series, an alternative to NvidiaFabless, dependent on TSMC
IntelXeon, declining market shareGaudi series, limited tractionIDM + Foundry, 18A delayed
Nvidia—Still dominant in data center GPUsFabless, TSMC

The table shows that AMD's advantage isn't confined to a single product — it's growing on two fronts at once, in gaps left open by both Intel and Nvidia. The risk is heavier for Intel, since the data center segment is the company's most profitable business line; losing share there hits margins directly. For Nvidia, the risk is smaller — its position as market leader remains strong — but AMD's presence gives cloud companies negotiating leverage, which could push prices down over the long run.

What's next?

The key question to watch is whether this revenue edge will also translate into a profit-margin edge. The revenue crossover carries symbolic weight, but as long as Intel's foundry business keeps losing money, the company's overall profitability could remain below AMD's. On the other hand, whether Intel's next-generation server products tied to the 18A process arrive on time could shift the balance again. On AMD's side, there's a sustainability question: the rapid growth in Instinct demand depends heavily on the investment cycles of a limited number of cloud customers; if that cycle slows, so does the growth.

What deserves discussion here isn't a shift in power, but a moment of momentum.

The opportunity now in AMD's hands is real, and it proves that both EPYC and Instinct have matured into fully realized products. But there is no transformation yet that erases Intel's decades of scale advantage in a single quarter; the real test is whether this growth also shows up in margins over the next several quarters.