Intel Corporation

Intel today is three separate bets bundled into one ticker: a slowly shrinking legacy CPU franchise generating real but declining cash flows, a government-subsidized foundry megaproject with binary commercial outcomes, and a data center CPU comeback story facing an enormous performance gap against AMD. The current price implies all three work. The numbers suggest all three are struggling simultaneously. What makes Intel genuinely interesting is the intersection of strategic necessity and operational degradation — the US government needs Intel to succeed, which meaningfully reduces catastrophic downside, but strategic importance and return on invested capital are entirely different quantities, and five years of near-zero ROIC is the proof. The most credible path forward is a decade-long slog rather than a dramatic inflection. Tan's appointment is legitimately positive, and the EUV ramp from near-zero to double digits is real manufacturing progress worth acknowledging. The advanced packaging opportunity — where EMIB engagements can exceed a billion dollars per customer — may be the underappreciated catalyst, because it offers a path to foundry revenue before leading-edge wafer wins materialize. But the core CPU business will keep contracting through that entire period, meaning growth in the new bucket perpetually offsets attrition in the old one. The single most specific and dangerous risk is hyperscaler ARM adoption crossing an enterprise software ecosystem threshold. The x86 switching cost survives only as long as Microsoft, SAP, and Oracle maintain x86 as their primary optimization target. The moment those vendors start shipping ARM-native builds as the performance leader rather than a compatibility port, the gravitational lock that keeps Intel in enterprise server rooms begins a five-year fade — quiet, structural, and nearly impossible to reverse once it reaches escape velocity.