Intel’s latest results finally gave investors something the company has lacked for years: evidence that the artificial intelligence boom may be helping its core business instead of passing it by. The chipmaker reported first-quarter revenue of $13.6 billion, up 7% from a year earlier, while non-GAAP earnings per share rose to $0.29 from $0.13. Just as important for the market, Intel forecast second-quarter revenue of $13.8 billion to $14.8 billion, comfortably ahead of the range analysts had been expecting.
The reaction was immediate. Reuters reported that Intel shares surged more than 24% in early trading on Friday to about $83, pushing the stock above its dot-com era peak. That move says as much about the market’s hunger for a credible AI infrastructure second act as it does about one strong quarter. For much of the past two years, investors treated Nvidia as the clearest beneficiary of AI spending and viewed Intel as a costly turnaround story with too little exposure to the most profitable part of the boom. This quarter suggested that view may have become too narrow.
The most important number in Intel’s report was not the headline revenue figure but the strength in its Data Center and AI segment. That business generated $5.1 billion in first-quarter revenue, up 22% from a year ago, while Client Computing Group revenue rose 1% to $7.7 billion. In prepared remarks, CEO Lip-Bu Tan argued that the industry is moving from training giant models toward inference, agentic systems, robotics and edge deployments, a shift that raises the importance of CPUs as the orchestration layer around accelerators. CFO Dave Zinsner said demand for Intel’s products still exceeded supply and that AI-driven businesses now account for 60% of revenue and grew 40% year over year.
That matters because Intel’s comeback case has increasingly depended on a simple proposition: AI infrastructure will require far more than GPUs. Someone still has to supply the server CPUs, packaging, memory orchestration and factory capacity that allow those systems to run at scale. Intel appears to be capturing more of that spending than skeptics expected, especially in Xeon server processors. The company also said Google expanded its use of Intel Xeon chips and that Nvidia selected Xeon 6 as the host CPU for DGX Rubin NVL8 systems. Those are not marginal endorsements. They suggest Intel is regaining relevance inside the plumbing of the AI buildout, even if it is not leading the accelerator market itself.
Still, the quarter did not resolve Intel’s hardest problem. Intel Foundry reported $5.4 billion in revenue, but only $174 million of that came from external foundry customers, according to the company’s prepared remarks. The unit still posted a $2.4 billion operating loss in the quarter. That is the central tension in the Intel story. The manufacturing platform is improving, and Tan said 18A yields are now ahead of internal projections, while 14A is progressing well enough that multiple customers are evaluating it. But better execution is not the same thing as a fully proven foundry model. Intel said it expects early design commitments to emerge in the second half of 2026 and into the first half of 2027, which means investors are still being asked to underwrite a business whose most important outside customer wins remain mostly ahead.
That is why the broader strategic moves Intel has made this month matter. The company agreed on April 1 to repurchase Apollo’s 49% stake in the Fab 34 joint venture in Ireland for $14.2 billion, restoring full control over a key manufacturing asset tied to Intel 4 and Intel 3 production. It also joined the Terafab project alongside SpaceX, xAI and Tesla, giving it another chance to frame its manufacturing and packaging capabilities as part of the next wave of AI infrastructure investment. Neither move guarantees future foundry success, but together they show a company acting with more confidence in its balance sheet and industrial position than it had even a year ago.
The bigger shift is that Intel no longer looks like a company trying merely to survive long enough to finish a turnaround plan. It looks like a company racing to add enough supply to serve demand that has already arrived. That does not eliminate the execution risks, especially in foundry economics, external customer adoption and the capital intensity of staying competitive. But for the first time in a long while, Intel’s revival is being driven by real demand rather than only by hope. If that continues, the company’s next challenge may not be convincing investors that the comeback is possible. It may be proving that it can scale fast enough to meet it.