Intel: The Party Is Only Getting Started

Intel retained a Buy rating after Q2 2026 revenue reached a record $16.13B, up 25.4% year over year, while adjusted EPS beat expectations by more than 90%. AI-driven data-center CPU and AI PC demand supported rapid margin expansion, and foundry losses narrowed by $1B as external sales increased sharply. Potential upside catalysts include an SK hynix partnership, Altera IPO value realization, and reported customer wins with Tesla and SpaceX.

Analysis

The investable question is whether Intel’s apparent earnings inflection is a sustainable mix shift or a low-quality recovery amplified by cost actions and favorable comparisons. CPU demand tied to agentic workloads is incrementally constructive, but it does not yet displace NVIDIA’s accelerator economics or AMD’s server-share opportunity; the key read-through is whether Intel can convert AI-PC attach rates into higher platform ASPs rather than merely defend unit volumes. A durable gross-margin recovery would support multiple expansion, while revenue growth without margin durability leaves INTC exposed to a post-earnings de-rate.

Foundry progress matters more for valuation than near-term EPS because it determines whether Intel’s capex burden becomes a strategic asset rather than a balance-sheet drag. External foundry traction could pressure TSMC’s long-duration scarcity premium only if customers commit meaningful leading-edge volumes and prepayments; isolated design wins have limited financial relevance. The SK hynix linkage is potentially more material for Intel’s platform competitiveness and memory supply security than for SK hynix itself, whose AI-memory economics remain primarily governed by NVIDIA/ASIC demand and HBM pricing.

Near term, bullish positioning is vulnerable to any evidence that the improvement is client-led rather than data-center-led, or that gross-margin expansion relies on underutilized-capacity absorption rather than structural yields. Over 6-18 months, the upside case requires credible proof of foundry customer economics, capital-intensity discipline, and product execution against AMD. Tesla and SpaceX references should not be capitalized into estimates absent disclosed volumes, qualification status, or contract commitments; neither is a reliable standalone earnings catalyst.

Key Decisions for Investors

• Initiate a 1-3 month tactical long INTC only on confirmation that next-quarter gross-margin guidance rises sequentially and data-center revenue outgrows client revenue. Target a 12-15% upside rerating; exit if margin guidance is flat/down or management attributes gains primarily to cost controls.
• Express the execution thesis as long INTC / short AMD in equal dollar risk for 3-6 months, rather than an outright semiconductor beta trade. The pair works only if Intel demonstrates share stabilization and improving server-product mix; stop out if AMD raises data-center guidance while Intel’s server growth decelerates.
• Do not treat SK hynix as a direct Intel proxy. Retain exposure through 000660.KS only if HBM contract pricing and supply commitments remain tight; a broad DDR/NAND inventory recovery would be a weaker and more cyclical thesis than AI-memory demand.
• Set an event-driven alert around disclosed foundry customer prepayments, wafer-volume commitments, and capex guidance. Upgrade the structural long case only if external-foundry bookings are accompanied by cash commitments and improving utilization; otherwise, avoid underwriting foundry value unlock beyond 12 months.