Intel surprised some on Thursday by beating Wall Street on headline Q4 revenue and adjusted EPS — $13.7 billion in sales and an adjusted 15 cents a share — yet the stock plunged roughly 13% after the company gave a softer-than-expected outlook for the new quarter. The disconnect is simple and oddly human: Intel says demand is there, but the chips aren’t ready fast enough.
The numbers were muddy. The company posted a roughly $600 million quarterly net loss and flagged first-quarter revenue guidance between $11.7 billion and $12.7 billion with breakeven adjusted EPS — below the consensus analysts were using. Investors who bid the shares up dramatically over the past year had been pricing in a smooth transition to volume production, new foundry customers and a big slice of the AI server boom. For now, Intel says it can’t supply all of that demand.
Why the market sold off
On the conference call CEO Lip‑Bu Tan and CFO David Zinsner put the problem plainly: Intel underestimated how quickly customers would want chips for data-center AI builds, and internal yields and capacity improvements haven’t yet ramped to capture the wave. Zinsner told reporters the supply crunch should ease in the second quarter, but the company will miss some near-term seasonal demand.
The result is a self-inflicted revenue gap. Data Center and AI sales rose about 9% year-over-year to roughly $4.7 billion in the quarter — healthy growth, but not the tidal wave some had hoped for. Client (laptop) sales fell, and Intel’s foundry business, which brought in about $4.5 billion in the quarter, still includes a meaningful portion of chips made for Intel itself rather than outside customers.
Investors have been feverish about Intel’s foundry potential — rumors of big customers like Apple have circulated among analysts — but building a competitive contract-manufacturing business is capital- and time-intensive. Tan said the company hit an important technical milestone with its 18A process during 2025 and that 14A customers could appear later in the year or in 2027, but he also acknowledged that turning technology wins into steady supply takes a while.
Competing for the AI pie
Intel is racing in a landscape shaped by enormous new demand for specialized compute. Companies building large AI models and infrastructure are rewriting capacity plans — an evolution that reaches beyond corporate campuses into ambitious infrastructure plays, from hyperscalers to proposals that imagine data centers off-planet. That kind of scale helps explain why projects such as Google’s Project Suncatcher are being discussed alongside factory roadmaps.
Meanwhile, software and model advances — and their appetite for chips — are part of the backdrop. As data and AI workloads grow more sophisticated, features like tighter integration between models and productivity tools (seen in efforts like Gemini’s Deep Research) increase demand for dense, high‑performance server silicon. Intel says its server CPUs are increasingly relevant in AI systems, but it’s still working to scale the production that would let it fully participate.
Beyond technical and supply issues, Intel’s corporate story has been dramatic: leadership changes, large workforce reductions, and major outside investments — including capital from the U.S. government, SoftBank and a $5 billion stake purchased by Nvidia during the quarter. Those moves have helped the stock rally over the past year, but they also set high expectations.
Foundry hopes, and a longer runway
Analysts still see the potential: if Intel nails yields and attracts big fab customers, the economics shift. But executives were candid that foundry revenue figures in the quarter reflect both external work and chips Intel makes for itself, and they declined to name imminent anchor customers. Zinsner said that when customers commit to next‑gen nodes like 14A, Intel will need to accelerate capital spending to support them — a milestone investors will watch.
For now the market punished the gap between optimism and near-term operational reality. The company’s long-term story — rebuilding manufacturing prowess, closing the performance-per-watt gap versus rivals, and scaling foundry customers — remains intact on paper. The immediate story is more tactile: missed capacity means missed sales, and in the world of AI infrastructure that can translate to lost share to rivals who can ship on schedule.
The next few quarters will be telling. If yields and fab throughput improve as Intel expects in the spring, the company could convert technical progress into tangible growth; if not, shareholders may demand faster proof that the turnaround is more than a technological roadmap.