Intel (NASDAQ: INTC) has returned approximately 380% over the trailing twelve months, yet the stock still sits 26% below its 52-week high, leaving a significant gap between performance and potential.
The core story driving Intel’s next move is not a product launch but a straightforward manufacturing bottleneck, with demand for its server chips running well ahead of what its factories can currently produce.
Intel’s data center and AI segment grew 59% year over year in Q2 2026, following 22% growth in Q1 2026, and management has been clear that this acceleration reflects supply improvements rather than the full picture of underlying demand.
In April, management quantified the demand it could not fulfill in a single quarter at billions of dollars, and by July it acknowledged Intel would still be running behind on server processor shipments even heading into Q4 2026.
The server processor built on Intel 3 is described as running extremely tight, with wafers, externally procured substrates, and memory all flagged among the company’s most challenged supply areas.
The PC market softness that many analysts view as a headwind may actually work in Intel’s favor, as the company expects client volumes to fall low double digits in calendar 2026 and has said it will redirect as much production capacity as possible toward higher-value data center CPUs.
The financial logic behind that pivot is clear, with the data center group posting an operating profit equal to 40% of its revenue in Q2 2026, compared with 26% at the client segment.
Capital spending plans have been revised sharply upward in a short window, moving from a flat year-over-year projection in April to more than $20 billion for calendar 2026 by July, with tooling expenditure up 40% year over year and 2027 spending guided even higher.
Intel reinforced that commitment in August by pricing a $20 billion stock offering at $95 per share, upsizing it from an initial $15 billion proposal and issuing 210,526,315 new shares to fund the expansion.
Management does not typically commit capital of this magnitude without corresponding customer commitments behind it, making the scale and speed of the revision a meaningful signal about the strength of Intel’s order book.
Intel Foundry, where the bulk of this capital is being deployed, remains the primary source of losses, though its operating loss narrowed to $2.1 billion in Q2 2026 from $2.4 billion in Q1 2026 as yields improved across the business.
The cost of producing Panther Lake, Intel’s lead product on its new 18A process, has already fallen roughly 50% in 2026, with management targeting an additional 20% reduction as the ramp matures.
Net margin remains negative at minus 19.8% over the trailing twelve months, and management guided gross margin flat for Q3 2026 even as revenue climbs, largely because a Q2 inventory write-down is not expected to recur.
Intel has now beaten its own guidance for seven consecutive quarters, and the broader test going forward is whether the company continues to raise its own outlook as supply constraints ease and yields improve.
The investment case rests on a multi-year manufacturing bet carried by a single company’s execution, with the payback tied directly to Intel’s ability to close the gap between what customers want and what its factories can deliver.
