SANTA CLARA, Calif., April 23, 2026, 04:13 PDT
Intel goes into first-quarter earnings after the bell on Thursday with investors betting it can turn renewed demand for AI-linked server processors into a cleaner turnaround story. Analysts still expect sales to slip and adjusted profit to fall sharply, even after the stock climbed more than 70% this year.
Why this matters now is simple enough. Central processing units, or CPUs, are becoming more important again as AI work shifts from training models to inference — the stage when systems answer user requests — and those chips sit alongside graphics processors from Nvidia in the data center. That gives Intel an opening it did not have at the start of the AI boom, but it also means Thursday’s report will be judged heavily on supply, pricing and how durable that demand really is.
Analysts tracked by LSEG expect first-quarter revenue of about $12.42 billion, down 1.9% from a year earlier, while adjusted earnings are seen falling by nearly 90%. Intel said in January it expected revenue between $11.7 billion and $12.7 billion and adjusted earnings to break even, and it is due to discuss the figures on a conference call at 2 p.m. PT.
The hardest read may come from data-center chips. Intel’s data center and AI unit is expected to grow 6.8% to $4.41 billion, but investors also want to know whether first-quarter supply snarls are starting to ease. “Rising demand for CPUs in AI data centers gives the company a steadier revenue lifeline,” eMarketer analyst Jacob Bourne said. Reuters
The stock’s run has made that bar tougher to clear. Shares are up more than 70% this year heading into the report, after fresh analyst upgrades from BNP Paribas and HSBC. Wedbush analyst Matt Bryson said the stock “has gotten ahead of reality,” and RBC Capital Markets analyst Srini Pajjuri has warned that enthusiasm around Intel’s foundry business may be running ahead of the actual revenue opportunity. MarketWatch
Intel has spent the month adding to the bullish case. Google expanded its AI infrastructure work with Intel on Xeon CPUs and custom infrastructure processing units, or IPUs, which move some work off the CPU, and Tesla said on Wednesday it plans to use Intel’s next-generation 14A process in its Terafab project, giving Intel its first major disclosed 14A customer. “Having a customer is more important than the timing,” Seaport Research Partners analyst Jay Goldberg said. Reuters
Even so, the contract chipmaking business remains the bigger test. Investors are watching Intel’s 18A manufacturing process for better yields — the share of usable chips from each silicon wafer — and for signs that 14A can attract more external customers as Intel tries to challenge Taiwan Semiconductor Manufacturing. Intel also continues to face AMD in server and PC chips, while Nvidia still dominates the AI accelerator market.
The company has also moved to strengthen its finances. On April 1, Intel agreed to pay $14.2 billion to buy back Apollo Global Management’s 49% stake in its Ireland fab, and Chief Financial Officer David Zinsner said the move reflected a “stronger balance sheet” and tighter financial discipline. Reuters
But the risks are not hard to find. Intel has warned that memory shortages are making PCs more expensive, which could weigh on its client-computing business, and any fresh delay in easing server-chip constraints would land badly with a market already pricing in more than a routine recovery. Material foundry revenue from outside customers may still take years to show up.
Intel shares were last at $65.27, giving the company a market value of about $296 billion. The quarter on Thursday will matter, but the tone around second-quarter supply and whether the AI CPU demand story is turning into repeatable business may matter more.
