Will Intel publicly name at least one external 14A foundry customer commitment by Q4 2026 earnings?

The resolution bar is intentionally low — ANY one of six tier-1 names (Apple, Qualcomm, Broadcom, Amazon, Google, Microsoft) publicly named as a 14A customer suffices. That asymmetry argues against a very low probability. However, the committee established that foundry design-win lead times are typically 2-3 years from first contact to public announcement, and zero such names have surfaced after 12 months of Tan's customer-gated framework. Any commit landing by Q4 2026 earnings would already need to be deep in technical qualification today, which would likely have leaked through supplier channels. External foundry revenue has declined for three consecutive years in a boom market — a strong prior against imminent commercial inflection. Weighted against the breadth of eligible customers and the Tan board-pressure incentive to produce at least one name, I land around 28%.

The reflexive customer-gating trap is the dominant structural factor. Customers demand yield credibility before committing; yield credibility requires running volume at scale; running volume at scale requires customer commits. Tan's Q4 2025 comment that 18A yields are 'still below what I want them to be' signals the credibility prerequisite is not yet met at the predecessor node, let alone 14A. TSMC N2 is commercially derisked and is the default alternative — any 14A commit must beat TSMC on process + price + geopolitics simultaneously. The committee's moat-mapper convergence (~55% CONTESTED / 27% ERODING / 18% DEFENSIBLE / 0% DOMINANT) already incorporates the optimistic case that at least some commit materializes; the probability of public naming specifically by Q4 2026 earnings is a subset of even that optimistic branch. I land at 22%.

Tan faces mounting board and market pressure to produce at least one named commit before the window closes — this is not just a technical question but a reputational forcing function for the CEO. The Foundry Stranding Cascade scenario at 25-35% probability is itself a strong incentive for Tan to extract a commit on asymmetric-negative terms if necessary just to defuse the central bear case. Historical base rate for customers signing under NDA is real but partially offset by the fact that Intel itself has every incentive to announce even a limited commit publicly. Strategic dark matter (NVIDIA, SoftBank, USG) could also reclassify — a single 8-K reveal of embedded 14A take-or-pay would resolve YES. I weight these upside paths and land higher than the base-rate-only estimate, at 32%.

Five of six discourse lenses independently converged on this exact binary as the central monitoring trigger — that level of cross-lens agreement is itself a signal the committee views the downside path as the dominant scenario. The decision window has already slipped once (Tan Q3 2025 'decisions in 2H 2025' became Q4 2025 'second half of 2026 extending into first half of 2027'), which is the single most bearish direct signal in the context. Windows that slip once tend to slip again. Combined with the structural customer-gating reflexivity and TSMC N2 alternative, I see roughly 1 in 4 odds of a public name landing on schedule.

Calling this one direct. The context says ZERO commits after 12 months of Tan trying. The context says external foundry revenue has DECLINED three years running. The context says design wins take 2-3 years to surface publicly. The context says TSMC N2 is already commercially derisked. These are all pointing the same direction. The offsetting factor is that the resolution bar is low (one name from a broad eligible pool) and Tan has a strong incentive to produce at least one announcement. The committee assigned 25-35% to the Foundry Stranding Cascade — that is approximately the probability space for the NO outcome on this specific market. The inverse is not exactly YES probability because neither branch captures the full probability mass, but it anchors my estimate. I land at 30%.

The direction-of-change facts (FY25 segment op loss -$2.2B vs FY24 -$11.7B, Panther Lake shipping, DCAI +15% QoQ) are real positives but the committee explicitly notes profits are protected by cash reserves and divestiture gains — NOT by competitive advantage. That distinction matters: operational improvement does not translate into foundry customer pull. The Tan quote 'I do not subscribe to the belief that if you build it, they will come' is effectively telling the market that Intel will not push; it will wait. A CEO waiting is a CEO assigning low probability to inbound commits on his preferred timeline. Combined with the window already slipping, my estimate is 27%.

Zero commits after 12 months. Window already slipped once. External foundry revenue declined three years. Design-win lead times 2-3 years. TSMC N2 is the commercially derisked alternative. All point to NO. Offset: low bar (one name from broad pool), Tan board pressure. 25% feels right.

The most salient fact: 18A yields 'still below what I want them to be' per Tan Q4 2025. 14A is a node beyond that. Customers cannot commit to a node whose predecessor has not yet demonstrated yield credibility. Reflexive gating plus TSMC N2 availability means the path to a Q4 2026 public name is structurally narrow. 22%.

Committee base case is that COMPETITIVE_POSITION is CONTESTED with bearish tilt (~55% CONTESTED / 27% ERODING / 18% DEFENSIBLE). Public naming of a tier-1 customer by Q4 2026 is roughly equivalent to being in the DEFENSIBLE branch or part of the CONTESTED branch. That gives a rough upper bound around 30%. Combined with the breadth of eligible customers offering some optionality, I land at 28%.