Intel posts fastest revenue growth in 15 years as AI demand lifts its CPU business

Intel's Fastest Growth in 15 Years: AI Demand Is Lifting Its CPU Business

For three years, the story around Intel and AI has been simple: NVIDIA won, Intel missed it. Intel watched from the sidelines as GPU accelerators became the foundational hardware of the generative AI boom.

The second quarter of 2026 just complicated that narrative. Intel reported revenue of $16.1 billion, up 25% year-over-year, its fastest growth rate for any quarter since the third quarter of 2011 1. The Data Center and AI segment, which houses Intel's server processors, grew 59% to $6.3 billion 1.

That 59% figure is the number to watch. It suggests AI infrastructure spending is no longer confined to GPUs. Every time a data center deploys GPU accelerators to train or run models, those chips need companion CPUs to handle data preparation, workload orchestration, and inference serving. Intel's entrenched position in data-center CPUs means that if AI demand is structurally increasing the volume of general-purpose compute alongside accelerators, its installed base becomes a second derivative of AI spending that most analyst models still price at zero.

CEO Lip-Bu Tan framed it in similar terms. "AI is driving unprecedented demand for compute," he said in the earnings statement, adding that "Intel is well-positioned to capture sustainable growth across our CPU franchise" 1.

The rest of the financials add depth. Adjusted earnings came in at 42 cents per share, double the 21 cents analysts expected 1. Gross margin recovered to 42%, which Intel attributed to greater operating scale and selling chips at higher prices 1. The client computing group, Intel's PC chip division and still its largest unit by revenue, grew 13% to $8.9 billion 1.

Forward guidance reinforced the momentum. Intel projected third-quarter revenue between $15.8 billion and $16.8 billion, above the $15.1 billion analysts were expecting 1. The stock rose about 4% in extended trading 1.

Supply constraints tell you something about demand intensity. CFO David Zinsner said Intel is supply constrained, with data-center customers requesting more chips than the company can currently manufacture. "Customers continue to signal a strong and sustainable spending environment," Zinsner said on the earnings call 1. Intel has reached 10 long-term agreements for server CPUs, some with pricing locked in, a move becoming common among memory and semiconductor vendors looking to preserve pricing power if the AI market cools 1.

Context keeps the celebration in check. Intel shares are up over 170% in 2026 as of Thursday's close, but the stock dropped 28% in July before this report 1. The foundry business, which Intel is betting its long-term relevance on, posted $5.8 billion in sales, up 31%, but has not yet landed a major customer for its most advanced manufacturing process 1. Intel named Fortinet as its first foundry customer under Tan earlier this week, but that deal uses an older manufacturing technology to make security chips 1.

The signal for builders and investors is narrower than "Intel is back." It is that the AI infrastructure build-out is broadening beyond accelerators. When the thesis was GPUs only, Intel was a spectator. Now that AI workloads demonstrably require more CPU compute alongside GPUs, Intel's data-center franchise has a tailwind it did not have eighteen months ago. Whether that tailwind is large enough to offset the foundry's open questions and the capital required to compete with TSMC is a separate debate. But the re-rating is grounded in real demand: AI compute is not monolithic, and Intel's CPU business is the proof.