Better Buy: AMD vs. Intel After Intel's Monster Run

The article argues AMD is the better buy versus Intel, citing AMD's first-quarter data center revenue of $5.8 billion, up 57% year over year, and its stronger profitability and cash flow. Intel's turnaround remains credible, with revenue up 7% to $13.6 billion and data center and AI revenue up 22%, but its foundry business still posted a $2.4 billion operating loss and the stock now trades at over 120x forward earnings versus AMD at about 73x. Overall tone is constructive on both companies but more favorable to AMD on growth and valuation.

Analysis

The market is no longer pricing this as a simple x86 share-wars story; it is increasingly a capital-allocation contest around who can monetize AI inference more efficiently. That matters because inference is a higher-volume, lower-margin workload than training, so the CPU attach rate can expand unit demand without necessarily expanding industry-wide profit pools at the same pace. In that setup, AMD’s cleaner P&L and cash conversion create more operating leverage per incremental data-center dollar, while Intel’s foundry re-rating is still being powered more by narrative than by demonstrated external customer traction.

The second-order read-through is that Intel’s foundry needs proof within the next 2-4 quarters, not years. With outside revenue still tiny relative to foundry sales and operating losses still large, the equity is effectively underwriting a future mix shift that has not yet arrived; any disappointment on customer wins, yields, or capex intensity could compress the multiple quickly. AMD, by contrast, is already getting paid for execution, but that also means the stock is more vulnerable to a deceleration in hyperscaler spending or a slip in the 2026 accelerator ramp, since the valuation leaves little room for schedule risk.

A useful contrarian angle is that the bullish AI-CPU thesis may be more crowded than it looks. If agentic/inference workloads boost server CPU demand across the board, the near-term beneficiaries may include the broader server ecosystem and component suppliers before either incumbent captures enough incremental economics to justify current multiples. Meanwhile, the Apple/foundry chatter around Intel could be a classic rerating catalyst that stays headline-positive but earnings-negative unless it converts into multi-year, high-volume wafers; until then, the stock is trading more like an option on future share gains than a cash-flow compounder.

Key Decisions for Investors

• Long AMD / short INTC as a 3-6 month relative-value pair: AMD offers execution-backed growth and cash flow, while INTC’s valuation embeds a foundry turnaround that still lacks external customer proof. Risk: a major Intel foundry win or sudden AMD AI ramp disappointment.
• Buy AMD on pullbacks only, with a 6-12 month horizon: prefer entries after 5-8% drawdowns caused by sentiment rather than fundamentals. Upside is driven by continued data-center reacceleration; downside is a spending pause at hyperscalers that could compress the multiple materially.
• Avoid fresh outright long INTC until there is evidence of external foundry revenue scaling above a token level for at least 2 consecutive quarters. At current valuation, the risk/reward is skewed toward multiple compression if the turnaround becomes a story without operating follow-through.
• Consider a call spread on AMD into the next 2 earnings prints to express upside with defined risk. This fits a thesis of continued execution but acknowledges that the stock already discounts a strong 2026 accelerator ramp.
• For more conservative exposure, pair AMD long with a hedge in NVDA or a broad semis basket if you want to isolate the CPU-share thesis from AI capex beta. This reduces the risk that a sector-wide rerating masks relative outperformance.