Timing clarification: Intel reports today, Thursday, July 23, 2026, after the U.S. market closes—not tomorrow. The earnings call is scheduled for 2:00 p.m. PDT. (newsroom.intel.com)
Intel enters this report with substantially more optimism—and a much higher valuation—than it carried into Q1. The company’s April results materially exceeded expectations, data-center CPU demand has accelerated, manufacturing output is improving, and recent foundry developments have added credibility to the turnaround.
That also raises the bar. A small Q2 beat probably will not be enough by itself. The more important issues are:
| Metric | Intel Q2 guidance | Current consensus/expectation |
|---|---|---|
| Revenue | $13.8B–$14.8B | Approximately $14.4B |
| Adjusted EPS | $0.20 at midpoint | Approximately $0.21–$0.22 |
| Adjusted gross margin | 39.0% | Near management’s 39% guide |
| DCAI revenue | Sequential growth, up double digits | Approximately $5.6B, +42% YoY |
| CCG revenue | Sequential growth | Approximately $8.0B, +1.3% YoY |
| Intel Foundry revenue | Not specifically guided | Approximately $5.5B, +25% YoY |
Intel’s official outlook calls for $13.8 billion to $14.8 billion of revenue, 39% adjusted gross margin and $0.20 of adjusted EPS. Visible Alpha expects approximately $14.4 billion of revenue, including $5.6 billion from Data Center and AI, $8.0 billion from Client Computing and $5.5 billion from Intel Foundry. Other published consensus figures place adjusted EPS around $0.21–$0.22. (intc.com)
Because consensus revenue is only modestly above Intel’s $14.3 billion guidance midpoint, the market appears to expect a beat—but not necessarily a repeat of Q1’s very large upside surprise.
Intel’s Q1 Data Center and AI revenue was $5.1 billion, up 22% year over year. Management subsequently guided DCAI to grow by a double-digit percentage sequentially in Q2, implying revenue of at least roughly $5.6 billion. That is essentially where current segment consensus sits. (intc.com)
The bull case rests on the idea that AI infrastructure is becoming more CPU-intensive as workloads move from model training toward inference and agentic applications. Intel has argued that CPUs act as the orchestration and control layer around accelerators, while Xeon 6 has secured positions including the host CPU for Nvidia’s DGX Rubin NVL8 systems. The company also signed multiyear agreements intended to improve supply and pricing visibility.
A DCAI result around $5.6 billion would satisfy the prior outlook. For the stock to respond strongly, investors may want material upside to that level or a clear indication that momentum is carrying into Q3 and 2027.
Intel generated a 41% adjusted gross margin in Q1 but guided Q2 to 39%. Management attributed the expected decline to two factors:
Management said Panther Lake volume could rise six- to sevenfold sequentially in Q2. Although 18A yields were reportedly improving faster than Intel’s internal plan, the increased volume mix still creates a near-term margin headwind. Rising memory, substrate and other input costs add another risk for the second half. (download.intel.com)
The most important manufacturing disclosure may be the expected date at which Panther Lake and Intel 18A cease being a material drag on consolidated gross margin.
Consensus calls for CCG revenue of approximately $8.0 billion, up only 1.3% year over year. That contrasts sharply with the projected 42% growth in DCAI. (spglobal.com)
Intel previously said it was prudently planning for the full-year PC unit market to decline by a low-double-digit percentage. Management nevertheless expected Intel’s revenue to perform better than end-market units because of customer inventory replenishment and pricing. It also suggested that quarterly client revenue could remain roughly flat from Q2 through year-end. (download.intel.com)
The key question is whether server strength can continue to outweigh:
A solid CCG quarter would help, but investors are unlikely to reward Intel heavily for client upside if it comes from short-term inventory replenishment rather than sustainable sell-through.
Intel Foundry reported $5.4 billion of Q1 revenue, but most of that was generated internally, while the segment posted a roughly $2.4 billion operating loss. Therefore, headline foundry revenue is less informative than:
Intel recently announced that Fortinet’s next-generation SP6 security processor will use Intel design, packaging and manufacturing capabilities. SP6 is expected to be built on Intel 4, making Fortinet the first named external customer for that node. The agreement is a useful commercial proof point, but no production timing was disclosed and it does not, by itself, settle the larger question of leading-edge 18A and 14A demand. (tomshardware.com)
Investors should distinguish among three levels of foundry news:
The highest-value disclosure would be a significant external customer commitment for 18A or 14A, accompanied by enough information to evaluate volume, capital requirements and eventual returns.
Intel’s Q1 performance showed that demand was exceeding available supply. The company expected output to continue increasing through the year through additional wafer starts, higher yields, better cycle times and selective use of external foundries. It also raised its full-year capital-spending expectation from “flat to down” to approximately flat year over year. (download.intel.com)
Since then, Intel has announced a €5 billion investment in its Leixlip, Ireland campus to expand Intel 3 capacity for Xeon 6 and future Xeon products. The investment primarily uses existing cleanroom space, which should be more capital-efficient than constructing an entirely new fab, but it reinforces the importance of disciplined execution and demand visibility. (newsroom.intel.com)
Intel previously maintained that it could produce positive adjusted free cash flow for 2026, excluding the Fab 34 joint-venture repurchase. That commitment deserves close attention after:
A further increase in capital spending would be easier to justify if accompanied by contracted customer demand. Capex rising on the basis of anticipated rather than committed demand would increase execution risk.
INTC was recently around $102.62. The shares have risen dramatically since the Q1 report and are up more than 150% in 2026, although they have also fallen materially from their June peak amid a broader semiconductor pullback. (kiplinger.com)
That creates an unusual setup:
This would represent solid execution, though the stock reaction could be muted given the prior rally.
The principal question is no longer whether Intel’s business is stabilizing. It is whether the company can convert strong AI-related CPU demand and improving process execution into durable margins and cash flow.
Q2 revenue and EPS expectations are not especially aggressive relative to management’s own guidance. The harder hurdle is qualitative: Intel needs to show that server strength will persist, that 18A’s economics are improving, and that capital-intensive foundry expansion is increasingly supported by real customer commitments.
The cleanest positive report would combine DCAI upside, gross margin at or above 40%, constructive Q3 guidance and tangible foundry validation. A revenue beat accompanied by weak margins, higher capex and vague foundry commentary would be considerably less compelling.