AMD Just Out-Earned Intel in the Data Center. Here's What That Means for Both Stocks.

AMD’s Q1 2026 data-center revenue surged 57% YoY to $5.8B, overtaking Intel’s $5.1B (up 22%), while AMD grew total revenue 38% to $10.3B and reported non-GAAP EPS of $1.37 with gross margin above 50%. Intel remains unprofitable and is still constrained by its foundry ramp (18A yields not expected until 2027), with its stock down ~21% over the past week. Despite Intel’s cheaper valuation (100+× expected earnings) and AMD’s premium multiple (≈59× forward earnings), the article favors AMD as the better risk given stronger profit conversion and data-center share gains.

Analysis

The real economic story is not “AMD is better,” it is that the server market is re-pricing around scarcity of credible alternatives. AMD’s mix shift into higher-ASP CPUs plus accelerator attach implies its incremental margins should stay well above the semiconductor average as long as hyperscalers keep dual-sourcing to de-risk supply. That creates a second-order loser in Intel’s ecosystem: not just Intel itself, but any vendor whose valuation depended on a faster foundry re-rating or on a dominant x86 incumbent preserving pricing power.

The immediate trade is less about the reported quarter and more about what comes next: the next 1-3 months are likely driven by guidance quality, EPYC/accelerator backlog commentary, and any fresh evidence that Intel’s manufacturing slip extends the investment cycle. Intel’s equity remains highly duration-sensitive; if the foundry path keeps pushing out, the market will keep treating earnings as optically cheap but economically unavailable. AMD’s risk is the opposite: at this multiple, the stock needs continued beat-and-raise behavior, not just good revenue growth.

Contrarian take: consensus may be underestimating how much of AMD’s run is already in the price while still underestimating how long Intel can stay “cheap.” A real Intel rally would require two things the market can verify: improving external foundry revenue and visible yield progress, not narrative. Until then, the cleaner expression is relative value, not outright conviction, because the first sign of deceleration in AMD or validation in Intel could compress the spread sharply over a 3-6 month horizon.

Key Decisions for Investors

• Long AMD / short INTC as a 1-3 month relative-value pair; thesis is margin durability and share gains versus a delayed Intel turnaround. Risk: spread narrows if Intel gets credible 18A/yield validation or AMD guidance merely meets, not beats.
• Use call spreads on AMD rather than outright stock for the next earnings cycle: 3-6 month bull call spread to express continued beat-and-raise potential while capping valuation risk if growth cools.
• For investors already short INTC, keep size modest and use a catalyst-based stop: cover on any externally verifiable foundry wins, yield improvement disclosures, or a durable re-rating in gross margin guidance.
• Alert rather than trade: watch hyperscaler capex commentary and AI accelerator order flow over the next 1-2 quarters. If AMD’s AI mix slows while server CPU share holds, the multiple can compress faster than fundamentals deteriorate.