Intel Q3 Preview: Server CPU Shortage Sets Up A Pricing Tailwind, Strong Buy

The analyst reiterates a Strong Buy rating on Intel with a $164-per-share price target, citing AI inference-driven server CPU demand and a forecast FY26 revenue of $63.66B. The bullish turnaround thesis also rests on improving margins, foundry expansion, and strategic partnership talks; these are outlook and analyst claims rather than reported results.

Analysis

The bullish case hinges on CPU inference demand translating into incremental Intel server-CPU share—not merely a broader workload increase captured by AMD, Arm-based designs, or cloud providers’ custom silicon. Even if demand is real, customers can dual-source and shift workloads; monitor Intel’s data-center product mix and share, not just aggregate AI-related demand. For Intel, foundry expansion adds a separate execution and capital-intensity risk: external customer commitments, yield, and utilization matter more than partnership discussions. Treat the stated revenue outlook and target as analyst assumptions, not independently established outcomes.

Near term, this is a weak standalone catalyst absent disclosed contract terms or a measurable guidance revision. Over 1–3 months, evidence of server-CPU share gains, improving gross margin, or binding foundry commitments could support a re-rating. Over 6–18 months, foundry economics and customer adoption are the larger swing factors. The contrarian risk is that enthusiasm for inference overlooks workload substitution and the time required for foundry investment to earn returns. Falsify the thesis if Intel’s data-center growth fails to outpace the relevant market, margins deteriorate, or foundry customer commitments and utilization disappoint.

Key Decisions for Investors

• Do not trade the bullish rating alone; keep INTC on a catalyst watchlist pending verifiable customer commitments, guidance changes, or data-center share evidence.
• For a relative-value expression, consider INTC versus AMD only after comparing server-CPU share and product mix; avoid assuming CPU inference demand accrues disproportionately to Intel.
• Track quarterly data-center revenue and gross-margin direction, plus foundry external-customer commitments and utilization. Reassess if these indicators fail to improve over the next 1–3 quarters.
• Treat strategic partnership talks as non-binding until terms, expected revenue contribution, and execution milestones are disclosed; absent that evidence, there may be no actionable trade.