Intel (INTC) under Lip-Bu Tan — Six Earnings Calls: The Pattern, the Man, and the Q3 2026 Call (Thu Oct 29, 2026)

Research completed Oct 22, 2026 (CDT). Method: official Intel earnings releases (SEC 8-K Ex. 99.1 / Intel newsroom) for all six quarters; full call transcripts read via MarketBeat (complete, timestamped) for Q2 2025, Q3 2025, Q4 2025, Q1 2026, Q2 2026, cross-checked against StockAnalysis, Motley Fool, Seeking Alpha and ROIC transcripts and Intel's own published CEO/CFO prepared-remarks PDFs; reactions from contemporaneous Reuters/CNBC/IBD/Fool/Investopedia coverage. Where a number could not be pinned to two sources it is flagged. Verified fact / reported characterization / inference are labelled throughout, and collected again in §12.

Summary

1. The streak is real and it is the story. In all six Tan-era quarters Intel beat the midpoint of its own prior guidance on revenue, and beat its guided non-GAAP gross margin and EPS in five of six — the only exception, Q2 2025, missed only because of $800M impairment + $200M one-time costs that Intel explicitly had not included in that guidance (excluding them, it beat on all three lines). Management counts the streak itself on every call: "fourth consecutive quarter" (Q3'25) → fifth (Q4'25) → sixth (Q1'26) → seventh (Q2'26, counting from Q4'24). Guidance under Tan is best described as honest-to-conservative for the quarter being guided, with the bad news concentrated in the comparison to consensus and in full-year items (capex, supply) rather than in the guided range itself.
2. What moves the stock is, in order: (1) next-quarter guidance vs consensus, (2) capex/FCF commitments made live on the call, (3) the size of the reported beat when it breaks expectations, (4) foundry-customer news — which has never once been delivered on an earnings call. Evidence: −17.03% on Jan 23, 2026 despite a Q4 beat (Q1 guide); +23.6% on Apr 24, 2026 (print and guide smashed a low consensus); −7.9% on Jul 24, 2026 after an initial +5–12% after-hours pop, reversed during the call when 2026 capex was raised above $20B with 2027 "significantly higher."
3. Foundry-customer language has escalated in confidence but not in specificity. Jul 2025: 14A needs "a meaningful external customer" and "there are no more blank checks." Jan 2026: customers "will begin to make firm supplier decisions starting in the second half of this year and extending into the first half of 2027." Apr 2026: Tan refused to name anyone — "my style is under-promise, over-deliver… we have no plan to announce the customer unless the customer wants to announce it" — and the only name, Tesla/Terafab, was supplied by Elon Musk, with no terms, under a framework SpaceX's own filing says binds nobody. Jul 2026: Intel "made the decision in Q2 to fully commit to high-volume ramp in 2028," but the Q2 10-Q still gates capacity scale on "committed demand… design wins with potential significant external customers." External foundry revenue in Q2 2026 was $293M of $5.8B segment revenue (~5%).
4. Tan the operator (see §10–11): a venture capitalist-turned-operator whose one prior CEO job, Cadence 2009–2021, is the template — revenue <$1B → ~$3B, operating margin >35%, stock +3,200% — built on customer intimacy, subscription/ratable revenue, disciplined bolt-on M&A and cost restructuring. At Intel he has run the same playbook at 20× the scale: flatten (layers ~12→6, VPs ~450→~200 per Zinsner at Deutsche Bank), cut (core workforce target 75,000), cancel (Germany/Poland), gate capital on customer commitments, and convert political attack (Trump's Aug 7, 2025 "must resign") into an $8.9B passive government stake within 15 days. His failure modes so far are equally specific: no signed leading-edge external wafer customer with terms, a supply-planning miss that produced the Jan 2026 guidance shock, and a capex "discipline" story that has inverted into >$20B (2026) / higher (2027) funded by a dilutive $20B ($23B with greenshoe) equity offering in Aug 2026.
5. Forecast for Oct 29 (inference, §9): Base case (~55%): an eighth straight beat vs guidance (revenue ~$16.6–17.2B, non-GAAP EPS ~$0.42–0.50, GM ~42–43.5%), a Q4 guide framed around supply improving into Q4, no new named 14A/18A wafer customer from Tan, capex reiterated not cut — stock reaction decided by the Q4 guide, plausibly flat to +5%. Bull (~25%) requires a customer-attributed commitment and/or ≥44% GM with a raise-style Q4 guide. Bear (~20%) is the January template: a print that "beats guidance" but lands below the $16.5B consensus, GM <42%, a soft Q4 guide and/or another capex step-up → −10% to −17%.

Findings

1. Q1 2025 — reported Thu Apr 24, 2025 (Tan's first call; in seat five weeks)

Print vs guidance (given Jan 30, 2025, pre-Tan) and consensus — verified from the releases: - Guidance for Q1'25: revenue $11.7–12.7B; non-GAAP GM 36.0%; non-GAAP EPS $0.00; GAAP EPS −$0.27. - Actual: revenue $12.7B (flat YoY, at the top of the range); non-GAAP GM 39.2% (+~3 pts vs guide, on Raptor Lake demand + Meteor Lake cost); non-GAAP EPS $0.13; GAAP EPS −$0.19. Segments: CCG $7.6B (−8% YoY), DCAI $4.1B (+8%), Foundry $4.7B, All Other $0.9B. Operating cash flow $0.8B; gross capex $6.2B, offsets $1.7B, adj. FCF −$3.7B; cash $21B. - Consensus (Reuters/LSEG): revenue est. $12.30B, adj. EPS est. ~$0.01 → beat on both. - Guidance given for Q2'25: revenue $11.2–12.4B; GM 34.3% GAAP / 36.5% non-GAAP; EPS −$0.32 GAAP / $0.00 non-GAAP. Also: 2025 non-GAAP opex target cut to $17B (from $17.5B), 2026 target $16B; 2025 gross capex cut to $18B (from $20B), net capex $8–11B. Altera: 51% sale to Silver Lake (~$4.46B) announced pre-call.

Reaction — guidance-driven (verified): Shares fell ~5% after hours and closed −6.70% on Apr 25 (AInvest market recap; Reuters: Q2 revenue/profit forecast below Wall Street, cons. ~$12.9B revenue / $0.06 EPS, "casting a shadow over Tan's first earnings"). Zinsner's tariff/recession-risk language ("very fluid trade policies… probability of a recession growing") and the disclosure that Q1 was flattered by tariff pull-ins were the story — not the Q1 beat.

Tan themes / signature language (full transcript, StockAnalysis/MarketBeat): Opens: "I joined Intel five weeks ago. The reason I'm here is simple: I love our company… there are no quick fixes. We must remain laser-focused on execution." Culture/bureaucracy ("organizational complexity and bureaucracies have been suffocating innovation"), flattened his leadership team in week five, 4-day return-to-office from Q3, personal review culture begins. Foundry = trust/service business (see §8). Bad news style: his employee letter "Our Path Forward" was published the same day as earnings — layoffs/restructuring announced by him, in writing, up front, not buried in Q&A.

Zinsner: macro conservatism explicitly "reflected in our outlook"; spending down $400M sequentially / $700M YoY highlighted as proof of discipline.

Hardest Q&A: foundry external customers (Arcuri, UBS) — Tan deflected to process: priority #1 is ramping the internal customer (Panther Lake), "next step is build the trust with our foundry customer" on PDK/schedule/yield. No external name, no timeline. Panther Lake/Clearwater Forest leadership (Moore, Morgan Stanley) — answered jointly with Holthaus, commitment limited to "first SKU by year-end."

2. Q2 2025 — reported Thu Jul 24, 2025 (the "no blank checks" call)

Print vs guidance/consensus — verified from the release (Ex. 99.1) and Zinsner's prepared remarks: - Guidance for Q2'25: revenue $11.2–12.4B; non-GAAP GM 36.5%; non-GAAP EPS $0.00. - Actual: revenue $12.9B (above the top of the range); GAAP GM 27.5%, non-GAAP GM 29.7%; non-GAAP EPS −$0.10; GAAP EPS −$0.67 (incl. −$0.45 from $1.9B restructuring charges). The miss vs guide is fully explained in the release and by Zinsner on the call: $800M non-cash impairment/accelerated depreciation on excess tools + ~$200M one-time period costs (−800 bp of GM; −$0.20 non-GAAP EPS) that "were not incorporated into the guidance Intel provided for Q2." Excluding them: GM 37.5%, EPS +$0.10 — a beat on both. - Consensus (LSEG): revenue $11.92B, adj. EPS +$0.01 → revenue beat, headline EPS miss. - Segments: CCG $7.9B, DCAI $3.9B, Foundry $4.4B (operating loss $3.2B incl. the impairment), All Other $1.1B. OpCF $2.1B; gross capex $4.5B; adj. FCF −$1.1B; cash & ST investments $21.2B. - Guidance given for Q3'25: revenue $12.6–13.6B; GM 34.1% GAAP / 36.0% non-GAAP; EPS −$0.24 GAAP / $0.00 non-GAAP; guided below-seasonal H2 on tariff pull-in payback. FY items: opex $17B (2025) / $16B (2026) reiterated; 2025 gross capex ~$18B, net $8–11B; NCI guided ~$250–300M in each of Q3/Q4 (GAAP); core workforce to ~75,000 by year-end (majority of actions done in Q2; management layers cut ~50%); Germany and Poland fab projects cancelled, Costa Rica assembly/test consolidated into Vietnam/Malaysia, Ohio construction slowed; Mobileye stake sale raised ~$900M; Altera close expected Q3.

Reaction — reported-quarter + strategy driven (verified): Shares fell ~8.2% pre-market and ~8.5% on Jul 25 (AP/Reuters coverage: "steeper losses than expected," workforce cuts, foundry future hinging on a 14A customer). Unlike Q1, this drop was about the print itself (the un-guided charges) and the conditional framing of 14A — Reuters' headline that day was that Intel could pause/exit leading-edge development without a significant external customer (the 10-Q risk language filed alongside).

Tan themes: Four-part agenda — organization/culture, foundry strategy, x86 franchise, AI strategy — in that order. New policy: every major chip design personally reviewed/approved by him before tape-out. Foundry passage is the most-quoted of his tenure (verbatim §8). On AI: admitted Intel had approached AI "with a traditional silicon and training-centric mindset without a cohesive… stack"; pivot to inference/agentic. Balance-sheet bluntness: "our last full fiscal year of positive free cash flow was 2021. This is completely unacceptable."

Hardest Q&A: Rasgon (Bernstein) — can Intel run a sustainable business on internal 18A volume alone if 14A fails? Tan: 18A carries three generations of internal product and peak volumes "probably the beginning of the next decade"; on 14A, "We are not going to put any CapEx until we see the yield performance and also our internal customer and external customer feedback and the volume commitment, then we will put the CapEx in… We are committed to the foundry business. Meanwhile, we are going to be very disciplined in the CapEx deployment." Zinsner added 18A external "probably won't get a lot in wave one." Direct on the gate; evasive (by design) on what happens if no customer comes.

3. Q3 2025 — reported Thu Oct 23, 2025 (the balance-sheet call)

Print vs guidance/consensus — verified from SEC Ex. 99.1: - Guidance for Q3'25: revenue $12.6–13.6B; non-GAAP GM 36.0%; non-GAAP EPS $0.00. - Actual: revenue $13.7B ($13.65B, +3% YoY, above the top of the range); GAAP GM 38.2%, non-GAAP GM 40.0% (+400 bp vs guide); non-GAAP EPS $0.23; GAAP EPS $0.90 — GAAP flattered by the Altera divestiture gain (reconciliation shows −$1.20/sh divestiture gain) after the 51% sale closed Sep 12, 2025 ($5.2B received incl. Mobileye stake sale). Segments: CCG $8.5B, DCAI $4.1B, Foundry $4.2B (operating loss $2.3B), All Other $1.0B. OpCF $2.5B. - Consensus (LSEG): revenue $13.14B, EPS $0.01 → large beat. - Frank-disclosure flag in the release itself: results labelled preliminary, subject to possible revision pending an SEC staff consultation on accounting for the U.S. government transactions (shutdown had stalled it) — bad/uncertain news disclosed in the release, page 3. - Guidance given for Q4'25 (ex-Altera): revenue $12.8–13.8B; GM 34.5% GAAP / 36.5% non-GAAP; EPS −$0.14 GAAP / $0.08 non-GAAP. Zinsner: midpoint revenue roughly flat QoQ adjusted for Altera; CCG down modestly, DCAI "up strongly" as wafer capacity is prioritized to servers over entry-level client; NCI ~$350–400M in Q4; diluted shares ~5.0B; 2025 gross capex ~$18B reiterated, "deploy more than $27B of capex in 2025 vs $17B in 2024" (deployment incl. partner/government offsets); demand "outpacing supply… persist into 2026."

Reaction — beat faded by guide (verified): +7–8% after hours (Reuters: profit beat, cost cuts, strategic investments); on Oct 24 the stock spiked above $41 intraday (18-month high) but faded to roughly +$1–2% (~$38.60) by the close per contemporaneous intraday reporting — a U.S. closing % was not cleanly reported in the sources checked (flagged in §12). Attribution: better-than-feared Q4 guidance ex-Altera, margin/cost progress and the ~$15B of strategic funding (U.S. government $5.7B received in Q3, SoftBank $2B, NVIDIA $5B pending) vs a Q4 guide whose midpoint sat slightly below consensus with GM guided down to 36.5%.

Tan themes: "AI is accelerating demand for compute…" across x86, purpose-built ASICs/accelerators and foundry; balance-sheet repair as a CEO-level priority "since becoming CEO in March"; Panther Lake architecture unveiled, Fab 52 (Arizona) fully operational on 18A. Foundry: conviction in the foundry TAM "continues to grow," capacity added only with "committed external demand"; emphasis visibly shifting from marketing 18A to external wafer customers toward 14A + advanced packaging (EMIB-T) pipeline; 14A described in Q&A (per Atlas Peak's cross-quarter comparison) as "off to a great start… better in terms of performance and yield [than 18A at the same maturity]," engagements with "multiple customers… milestone basis." Still no name.

Zinsner: quantified the supply constraint (Intel 10/7 capacity limits; substrate shortages industry-wide; pricing/mix actions to shift demand to available supply) — the first appearance of the demand > supply framing that dominates 2026. NCI modeling question (Rakers) answered with a 2026 range of $1.2–1.4B.

Hardest Q&A: the legacy $30B-by-2030 foundry revenue target — Atlas Peak's debrief records Zinsner as evasive when Arcuri pressed it; supply sustainability; and SEC/govt-deal accounting.

4. Q4 2025 — reported Thu Jan 22, 2026 (the guidance shock)

Print vs guidance/consensus — verified from the release: - Guidance for Q4'25: revenue $12.8–13.8B; non-GAAP GM 36.5%; non-GAAP EPS $0.08. - Actual: revenue $13.7B ($13.67B, −4% YoY, top end); GAAP GM 36.1%, non-GAAP GM 37.9% (+140 bp); non-GAAP EPS $0.15; GAAP EPS −$0.12. FY25 revenue $52.9B, flat YoY. Segments: CCG $8.19B, DCAI $4.74B (+15% QoQ, fastest sequential growth in a decade per Atlas), Foundry $4.5B (loss $2.51B), custom ASIC business at a $1B annualized run-rate. Cash & ST investments $37.4B at year-end. NVIDIA's $5.0B share purchase closed Dec 29, 2025. - Consensus: revenue ~$13.41B, EPS $0.08 → beat on both. - Guidance given for Q1'26: revenue $11.7–12.7B; GM 32.3% GAAP / 34.5% non-GAAP; EPS −$0.21 GAAP / $0.00 non-GAAP. FY26 color from Zinsner: supply improves from Q2 and each quarter after; opex target $16B; capex "flat to down slightly" vs 2025 (revised up from an earlier "down" intention; tools up, space down, H1-weighted); positive adjusted FCF expected for 2026; retire $2.5B of maturing debt; NCI ~$325M in Q1, ~$1.2B for 2026, growing "meaningfully" in 2027; shares ~5.1B. Tan promised an Investor Day in H2 2026 in Santa Clara (occurrence not verified — §12).

Reaction — the purest guidance event of the six (verified): −17.03% on Jan 23, 2026 ($54.32 → $45.07), worst session since Aug 2024, >2× the implied options move, despite the beat. Attribution (CNBC/Reuters/FXStreet/Atlas): Q1 guide midpoint $12.2B vs ~$13.3B consensus (−8%), GM 34.5%, breakeven EPS; management's explanation — buffer inventory depleted (finished goods ~40% of peak), "hand to mouth," wafer mix shift to servers not exiting fabs until late Q1, 18A yields "in line with internal plans but still below what I want" (Tan) — landed as a supply-execution failure, not a demand problem. The stock had also run ~+47% YTD into the print.

Tan themes / candor: The bad news was delivered in his prepared remarks, not extracted in Q&A: "In the short term, I'm disappointed that we are not able to fully meet the demand in our markets… While yields are in line with our internal plans, they are still below what I want them to be." Foundry passage — the timeline the whole 2026 year now hangs on (verbatim §8): PDKs "now viewed by customers as industry standard"; "Engagements with potential external customers on Intel 14A are active. We believe customers will begin to make firm supplier decisions starting in the second half of this year and extending into the first half of 2027."

Hardest Q&A: Rasgon again — how does a company with $11.6B of inventory end up hand-to-mouth? Zinsner answered directly and in detail (demand inflected in H2'25 after hyperscalers had signalled flat units; the inventory is the wrong mix), and conceded "I do believe the first quarter is the trough." Arcuri on why not order tools more aggressively — answer: yields/cycle-time first; aggressive tool orders for Intel 7/3/18A, but 14A capacity held until customer commitments. The $30B foundry target and external wafer revenue timing (risk production late 2027, volume 2028) were deferred, not resolved.

5. Q1 2026 — reported Thu Apr 23, 2026 (the blowout; Terafab arrives — from the customer)

Print vs guidance/consensus — verified from the release (8-K Ex. 99.1): - Guidance for Q1'26: revenue $11.7–12.7B; non-GAAP GM 34.5%; non-GAAP EPS $0.00. - Actual: revenue $13.6B ($13.58B, +7% YoY — above the top of the range by $0.88B, +$1.38B vs midpoint); GAAP GM 39.4%, non-GAAP GM 41.0% (+650 bp vs guide; Zinsner bridge: higher volume incl. previously-reserved inventory, mix, pricing, better 18A yields; includes a ~$0.06 one-time gain in interest/other); non-GAAP EPS $0.29; GAAP EPS −$0.73, burdened by a $3.9B goodwill impairment (mostly Mobileye). Segments: CCG $7.7B, DCAI $5.05B (+22% YoY), Foundry $5.4B (+20% QoQ; loss $2.44B; external revenue only $174M). OpCF $1.1B; gross capex ~$5.0B; adj. FCF −$2.0B. AI-driven businesses ~60% of revenue, +40% YoY. - Consensus (FactSet): revenue $12.42B, EPS $0.01–0.02 → the largest beat of the six. - Balance-sheet action in the release (announced Apr 8): repurchase of Apollo's 49% of the Fab 34 (Ireland) entity for $14.2B, plus $6.5B senior notes (Apr 2026). - Guidance given for Q2'26: revenue $13.8–14.8B; GM 37.5% GAAP / 39.0% non-GAAP (guided down on 18A mix — Panther Lake volume guided up ~6–7× QoQ); EPS $0.08 GAAP / $0.20 non-GAAP. FY color: FY26 gross capex ~flat YoY (~$18B) with tool spend +25%; server CPU units up double digits in 2026 with momentum into 2027; PC TAM down low-double-digits; NCI ~$250M/qtr (post-Fab 34 buyback).

Reaction — print + guide, both directions favourable (verified): +23.6% on Apr 24, 2026 ($66.78 → $82.54), biggest one-day gain in nearly four decades, record close. Attribution (IBD/WSJ/TIKR): EPS $0.29 vs ~$0.01–0.02 consensus, DCAI +22%, and a Q2 guide ~$1.3B above consensus at midpoint; the Apr 7–8 Terafab announcement had already broken the stock out before the call (IBD), i.e., the foundry-customer catalyst arrived mid-quarter, from outside Intel.

Tan themes: Triumphal but disciplined framing: "A year ago the conversation was about survival. Today it's about how quickly we can add capacity" (as reported from the call); culture now "data-driven, paranoid and engineering-centric" (the Grove vocabulary, deliberately); CPU:GPU ratio improving 1:8 → 1:4, "could move towards parity"; semiconductor TAM "approaching $1 trillion."

The foundry-customer exchange of the year (Q&A, verified via Fortune's call coverage): Asked on Terafab/14A customers, Tan described Terafab as a broad relationship to "refactor" silicon fab technology and said of 14A customers: "We're making great progress in terms of yield and cycle time. And clearly we're engaging with multiple customers; heavy engaging. My style is under-promise, over-deliver, so we have no plan to announce the customer unless the customer wants to announce it, and we support that." Days later, on Tesla's call, Musk supplied the name himself: "We plan to use Intel's 14A process, which is state-of-the-art and, in fact, not yet totally complete." Note what Intel has never disclosed: price, volume, timing commitment, or a signed volume agreement — and SpaceX's May 2026 IPO filing describes Terafab as a general framework, specific projects subject to separate agreements, with neither Tesla nor Intel obliged to remain.

Other Q&A: Google multi-year LTAs (Reitzes) — Tan/Zinsner described volume/pricing product agreements (Xeon/ASIC/IPU), i.e., product LTAs, not foundry wafer commitments; advanced-packaging demand "in the billions of dollars per year" (Zinsner to Arcuri).

6. Q2 2026 — reported Thu Jul 23, 2026 (the capex reversal)

Print vs guidance/consensus — verified from the release: - Guidance for Q2'26: revenue $13.8–14.8B; non-GAAP GM 39.0%; non-GAAP EPS $0.20. - Actual: revenue $16.1B ($16.13B, +25% YoY — fastest growth in >15 years; +$1.83B vs midpoint, +$1.33B above the top of the range); GAAP GM 40.4%, non-GAAP GM 41.8% (+280 bp vs guide; bridge: volume, yields, ASP/mix/pricing); non-GAAP EPS $0.42; GAAP EPS −$2.16 — GAAP net loss $11.0B driven by a $12.5B mark-to-market loss on the Escrowed Shares (stock held for release to the U.S. Department of Commerce under the CHIPS Secure Enclave agreement; the liability was $15.6B at Jun 27, 2026). Segments: CCPG (renamed Client Computing & Physical AI) $8.88B (+13% YoY), DCAI $6.26B (+59% YoY, operating margin ~40%), Foundry $5.8B (loss $2.09B; external revenue $293M, ~5% of segment, up from $174M); 18A output ~25% above target; AI-driven businesses >70% of revenue, +>70% YoY. OpCF $7.0B; gross capex $2.7B; adj. FCF −$8.4B, but that figure includes −$12.2B net partner contributions (the Apollo/Fab 34 unwind) — ex-that item, cash generation was strongly positive. Cash & ST investments ~$29.7–30B. - Consensus: revenue $14.42–14.50B, EPS $0.21–0.22 → beat by ~$1.7B / ~2×. - Guidance given for Q3'26: revenue $15.8–16.8B; GM 41.0% GAAP / 42.0% non-GAAP; EPS $0.31 GAAP / $0.38 non-GAAP (the GAAP guide carries zero escrowed-share MTM in the reconciliation — any Q3 MTM is un-guided GAAP noise). Call color (Zinsner, via ROIC transcript): supply growth "non-linear," skewed to end-Q3/into Q4, especially servers; PC market down low-double-digits in 2026; NCI ~$250M in each of Q3 and Q4, ~$1.1B in each of 2027 and 2028 (post-Fab 34); 2026 capex raised to "more than $20 billion" (from ~$18B entering the year); 2027 capex "significantly above" 2026, vast majority in the U.S.; cumulative U.S. tools+space 2021–2026 approaching $100B.

Reaction — the call, not the release, did the damage (verified): Shares rose +5% to +12% after hours on the print and the above-consensus Q3 guide, then closed −7.9% on Jul 24 ($100.23 → $92.32). Attribution (Fool/Investopedia/ValueSlate): the reversal tracked the call's capex escalation and FCF math, the still-~$2.1B foundry loss, and external foundry revenue of only $293M — against a stock already +163% YTD and near a $140.94 record close (Jun 22).

Tan foundry language (verbatim, MarketBeat transcript): "I'm pleased to see the increasing momentum on customer engagements for Intel 14A, and I'm increasingly confident that the 14A will be highly competitive… With encouraging external customer progress and increased demand for our internal products, we remain on track for 14A risk production for our internal products in the second half of 2027, and we made the decision in Q2 to fully commit to high-volume ramp in 2028." The Q2 10-Q is more precise than the call: Intel "committed to completing development of Intel 14A," while "the scale and pace of our manufacturing expansion projects will ultimately be dictated by the amount of committed demand for Intel 14A… from our Intel products roadmap and design wins with potential significant external customers," under "a disciplined approach of investing capital… only where we have a clear line of sight to an acceptable return." Development: committed. Capacity: still gated.

Hardest Q&A (per contemporaneous call analysis): Arya (BofA) — when does customer enthusiasm become announcements? Answer returned to process milestones (PDK 0.9 due October, risk production H2 2027). Reitzes (Melius) — does the higher capex reflect hard external orders? Answer broadened to confidence across Intel's businesses. This is the established deflection pattern: precise where evidence is strong (yields, output, cost), general where it is weak (external orders).

Post-call balance-sheet action (pattern-consistent): On Aug 10–12, 2026 Intel priced an upsized common-stock offering: 210,526,315 shares at $95 = $20.0B base (upsized from $15B); underwriters fully exercised the greenshoe (+31,578,947 shares), taking the total to 242,105,262 shares / $23.0B gross / ~$22.6B net (424B5). Stock fell −4.1% on the initial announcement (Aug 10). Stated use: general corporate purposes incl. capex/working capital.

7. Guidance behaviour — is there systematic sandbagging?

Scoreboard vs Intel's own prior guidance (non-GAAP):

Verdict (inference from the verified table): - Revenue: 6/6 above midpoint; 4/6 at or above the top of the range. Average beat vs midpoint ≈ +$0.95B, and the beats grow through 2026 (+$1.38B, +$1.83B) as demand outruns a supply-constrained forecasting process. - GM/EPS: 5/6 beats, 1 miss — and the miss (Q2'25) was an explicitly un-guided charge, not an operating miss. Ex-charges it is 6/6 on all three lines. - This is consistent with deliberate conservatism ("under-promise, over-deliver" is Tan's stated motto — §10) compounded by genuine forecast difficulty: management repeatedly guided for tariff payback (Q2'25), below-seasonal halves, and a Q1'26 supply trough that demand then overwhelmed. It is not costless credibility: the same conservatism produced the Jan 2026 shock, because the Q1'26 guide was set ~$1.1B below consensus and the call had to explain a supply failure. The variance lives in (a) guide-vs-consensus gaps and (b) un-guided items — charges (Q2'25), goodwill (Q1'26), escrow MTM (Q2'26) — not in the guided operating lines. - Corollary for Oct 29: consensus ($16.5B / $0.39–0.40) already sits above the guide midpoint ($16.3B / $0.38) precisely because the Street now prices the streak. A print that merely "beats guidance" at, say, $16.4B / $0.38 would be a consensus miss.

8. Call structure, and the foundry-customer language ladder (verbatim)

Structure — identical all six calls (verified across transcripts): 1. Pitzer (IR) — safe harbor, non-GAAP pointer. 2. Tan prepared remarks (scripted; Intel publishes the CEO/CFO remarks as a PDF — he reads them closely): (i) headline "revenue, gross margin and EPS above guidance" + streak count; (ii) demand/supply framing; (iii) organization/culture (2025) → foundry (18A yields/output, 14A status, packaging) → products (client, then DCAI, then ASICs/AI); (iv) partnerships/customers; (v) closes with a candour item ("no quick fixes" / "disappointed we cannot meet demand" / "multiyear journey"). 3. Zinsner prepared remarks: market conditions → print-vs-guide bridge (always quantified, incl. one-offs) → segment P&L → guidance (range, GM, tax, EPS, NCI, share count) → full-year color (opex, capex, FCF, debt) → "manage what's in our control / high say-do ratio." 4. Q&A (~30 min): Tan takes strategy/foundry/customers in short, qualitative answers; Zinsner then quantifies or re-answers — the complement is systematic (e.g., Q4'25 inventory interrogation, Q3'25 NCI range). Repeat interrogators: Rasgon (foundry viability, inventory), Arcuri (foundry timing/tools), Arya (customer conversion), Reitzes (LTAs, capex-vs-orders), Seymore (guidance bridge). 5. What he never does (six calls, no exception found): name an unannounced external leading-edge wafer customer; attach dollars/volume to a foundry engagement; give full-year revenue/EPS guidance; give a foundry breakeven date; commit 14A capacity dollars ahead of customer commitments; discuss a dividend (suspended since Q4'24).

The ladder — exact customer language, quarter by quarter: - Q1'25 (prepared + Q&A): Foundry is "first and foremost a customer service business built on the foundational principle of trust… we need to instill a customer service mindset." On external customers: priority one is "ramping our internal customer" (Panther Lake); "next step is basically build the trust with our foundry customer." No name, no timeline. (The Microsoft/AWS 18A deals are Gelsinger-era and were not re-announced.) - Q2'25 (prepared): *"The increase in capital costs at Intel 14A makes it clear that we need both Intel products and a meaningful external customer to drive acceptable returns on our deployed capital. *I will only invest when I'm confident those returns exist… I do not subscribe to the belief that if you build it, they will come. Under my leadership, we will build what customers need, when they need it, and earn their trust through consistent execution." Employee letter, same day: "Going forward, our investment in Intel 14A will be based on confirmed customer commitments. There are no more blank checks." The accompanying 10-Q warned Intel might pause/discontinue 14A and successors absent a significant external customer (Reuters). - Q3'25: Capacity only with "committed external demand"; 14A "off to a great start," engagements on a "milestone basis" with multiple customers; centre of gravity shifts to 14A + EMIB-T packaging pipeline. No name. - Q4'25 (prepared, verbatim): "Engagements with potential external customers on Intel 14A are active. We believe customers will begin to make firm supplier decisions starting in the second half of this year and extending into the first half of 2027." Q&A added: two potential customers working with the PDK 0.5 and test chips (per call summaries); PDKs "now viewed by customers as industry standard." No name — and the decision window is now on the clock. - Q1'26 (Q&A, verbatim): "…we're engaging with multiple customers; heavy engaging. My style is under-promise, over-deliver, so we have no plan to announce the customer unless the customer wants to announce it." The customer then announced itself: Musk/Tesla on 14A for Terafab — a plan, no terms, under a non-binding framework (SpaceX filing, May 2026). Also in Q1'26: Google product LTAs; (Jul 21, 2026) Fortinet SP6 — design/packaging/manufacturing on Intel 4, reported as the first named foundry customer of the Tan era, on a mature node. - Q2'26 (prepared, verbatim): "…increasing momentum on customer engagements for Intel 14A… increasingly confident… we made the decision in Q2 to fully commit to high-volume ramp in 2028" — for development/internal risk production H2 2027; the 10-Q keeps expansion scale gated on committed demand. External foundry revenue: $293M. - Aug 2026, Deutsche Bank conference (Zinsner, reported by Fool, Oct 3, 2026): customer conversations have shifted from evaluating the technology to discussing how much capacity they can secure; Zinsner has reached "conviction" about landing external 14A customers; 14A defect-density improvement ahead of target. Zinsner's standing rule (Deutsche Bank remarks reported by DataCenterDynamics): "He's not declaring a customer win until it's a customer win with a real commitment and signature behind it… we're not declaring anything until we've got somebody signed on the dotted line." - ⚠ Oct 2026 headline trap (verified as a discrepancy, not a fact): A wave of Oct 17–20, 2026 headlines ("Intel secures Tesla deal," "Tesla selected 14A") presents Tesla as a new foundry win. Traced back, the underlying public statement remains Musk's April 2026 comment; at least one earnings preview explicitly notes "some Oct 19 headlines say 'agreement,' while the public statement was Musk's April comment that Tesla plans to use 14A; no terms were disclosed." Treat Tesla/Terafab as a named intention, not a signed volume commitment, unless Intel files terms.

9. What has moved the stock — ranked, with evidence

Positioning context for Oct 29 (verified): INTC closed Oct 22, 2026 at $85.49, −8.04% that day, ~30% below its late-Sept level and far below the Jun 22 record close of $140.94 — i.e., the stock comes in de-risked, unlike Jan (+47% YTD) or Jul (+163% YTD). One options preview derives an implied move of ~±9% (single source, flagged). The FOMC decision lands Oct 28, the day before (per mission; a Fed statement that week is also noted in earnings previews).

10. Lip-Bu Tan — the person and the operator

Labels: [FACT] = verified from primary/reputable sources · [REPORTED] = characterization by named reporting/analysts · [INFERENCE] = my synthesis.

Before Intel [FACT unless noted]: - Born in Malaysia, raised in Singapore. B.S. physics, Nanyang Technological University (1978); M.S. nuclear engineering, MIT (1980); MBA, University of San Francisco (1983). Early career at EDS Nuclear and ECHO Energy. (Intel 8-K director bio, Aug 11, 2022.) - Founded Walden International in 1987 (chairman since), named for Thoreau's Walden — "contrarian, rather than just following the trend." Walden grew from ~$3M at founding to ~$5B by 2001, focused on Asian/U.S. tech, especially semiconductors; Forbes dubbed him "the pioneer of Asian VC" in 2001. Later founding managing partner of Celesta Capital and Walden Catalyst Ventures; >$5B under management across his firms (CMU bio). Seed investor in SMIC (2001), on its board until 2018 (Reuters factbox). - Cadence Design Systems: board member Feb 2004; interim co-CEO Oct 2008; CEO Jan 2009 – Dec 15, 2021; executive chairman until 2023, having recruited and trained his successor (Anirudh Devgan). The company he inherited had revenue <$1B, a restatement hangover and was losing to Synopsys [REPORTED — BBN/Needham]. Results, per Cadence's own transition release: revenue grew from <$1B to ~$3B (~$2.99B FY21), operating margins expanded to >35%, and the stock appreciated >3,200%. Intel's appointment release states the same trajectory ("more than doubled revenue, expanded operating margin, stock +3,200%"); one secondary source cites ~4,000% total shareholder return — I use the company's 3,200% figure. - What he actually did at Cadence [FACT + REPORTED]: (i) Rebuilt customer intimacy personally — "leapfrog innovation and customer delight" as the stated culture; deep partnerships with market-shaping customers (Cadence release). (ii) Shifted the model to recurring revenue: multi-year subscription contracts with ratable recognition — by 2013 Cadence calls report ~90% of revenue/orders ratable, ~2.3–2.5-year weighted contract life; analysts credit the 3-year subscription shift with ending quarter-end discounting games [REPORTED — Expanse/Fortune-cited analysis]. (iii) Expanded the market from EDA tools into IP and system-level analysis ("Intelligent System Design"): bolt-on M&A, not mega-deals — Denali and Tensilica ($380M cash, 2013) as the foundation of the IP business, plus AWR, Sigrity, Celsius-type additions; characterized by analysts as disciplined $50–500M add-ons [REPORTED]. (iv) Early cost restructuring (a ~225-position reduction / ~$30M opex saving in year one is reported by a secondary profile — lower confidence, flagged). (v) Received the SIA's Robert N. Noyce Award (2022), the industry's highest honour (Intel 8-K). - Board seats [FACT — Intel 8-K, Aug 2022]: at appointment to Intel's board he sat on Cadence, Schneider Electric and Credo Technology; previously SoftBank Group (2020–2022), Hewlett Packard Enterprise (2015–2021), AMEC (2005–2020), Flextronics (2003–2012); Carnegie Mellon trustee. He sat on Intel's own board Sep 2022 – Aug 2024, then resigned — [REPORTED, Reuters] over disagreements with CEO Pat Gelsinger about workforce size, bureaucracy/risk-aversion and AI strategy, including frustration at being unable to push a third-party manufacturing-contracts initiative. He returned as CEO effective Mar 18, 2025, succeeding interim co-CEOs Zinsner/Holthaus. - Reputation at appointment [REPORTED]: Intel chair Frank Yeary: "an exceptional leader… proven track record of creating shareholder value… puts customers at the heart of everything he does." TD Cowen: the appointment was "as good as stakeholders could have hoped for" (Reuters); the stock jumped ~10%+ on the announcement (Mar 12–13, 2025). Zinsner at a May 2025 financial conference (CNBC): "Lip-Bu's Rolodex is like nobody else's in the semiconductor industry" — 22 customer/partner meetings in a single day. CNBC's framing of the open question is the fairest one-line summary of the bull/bear split: he "has a long track record… now he needs a big customer."

Decision/working style at Intel [FACT from his letters/calls + REPORTED]: - Engineering-first, personally. Elevated core engineering to his executive team; flattened his own leadership team in his first five weeks; personally reviews/approves every major chip design before tape-out (Q2'25 call); twice-weekly 18A yield reviews in 2025 (Q2'25 call). His Apr 24, 2025 employee letter (published by Intel, same day as earnings): teams "eight or more layers deep" must be flattened; "the most important KPI for many managers has been the size of their teams. Going forward, this will not be the case… the best leaders get the most done with the fewest people." By 2026 Zinsner could cite layers cut ~12→6 and VP ranks ~450→~200 (Deutsche Bank remarks, reported). - Speed and finality on cuts. Opex target cut ($17.5B→$17B, then $16B for 2026) and capex target cut ($20B→$18B) on his first call; Germany/Poland cancelled, Costa Rica consolidated, Ohio slowed on his second; layoffs communicated by his own signed letter on earnings day (Apr and Jul 2025), with the numbers in the release — "hard but necessary decisions." Restructuring charges ($1.9B in Q2'25) were taken in a bunch, early, rather than dribbled out. - Capital discipline as identity — with a 2026 twist. The 14A gate (§8) is the purest expression: commitment follows signed customer demand, not strategy decks. But note the evolution [INFERENCE from verified facts]: in 2026 the same man raised capex twice (Jan: "flat to down slightly" replacing "down"; Apr: tools +25%; Jul: >$20B and 2027 "significantly higher") and funded it with a $20–23B equity offering — because, in his framework, demand signals he trusts (server CPU orders, substrate lock-ins) had arrived. Discipline, for Tan, means "spend follows committed demand," not "spend less." The Aug 2026 offering — upsized $15B→$20B into strength at $95, greenshoe fully exercised — is Cadence-era opportunistic capital allocation executed at Intel scale. - Political operator when forced. [FACT] Aug 7, 2025: President Trump posted that Tan is "highly CONFLICTED and must resign, immediately," following Sen. Cotton's letter citing the China investments and Cadence's China export case (below). Tan's public response was a single employee message — there was "misinformation" about his past roles and he had "always operated within the highest legal and ethical standards" — no public fight. [REPORTED — Reuters' Dec 2025 reconstruction] He sought the meeting, was vouched for, and in a ~40-minute White House meeting on Aug 11 converted the attack into the agreement announced Aug 22: the U.S. government bought 433.3M shares at $20.47 = $8.9B for ~9.9% (funded by $5.7B of unpaid CHIPS commercial grants + $3.2B Secure Enclave), passive — no board seat, no governance/information rights, votes with the board with limited exceptions — plus a 5-year warrant for +5% at $20 only if Intel's foundry ownership falls below 51% (Intel 8-K). SoftBank ($2B, closed Sep 29, 2025) and NVIDIA ($5B + product collaboration, closed Dec 29, 2025) followed. In Oct 2026 he was still publicly insisting the stake carries "no policy… to interfere" with strategy (Nikkei interview, via Reuters Oct 16, 2026). - Bad news handling [pattern, §1–6]: His own decisions — announced by him, in writing, in prepared remarks/letters, quantified, on day one (layoffs, fab cancellations, impairments bridged by Zinsner in prepared remarks, the Q1'26 supply failure confessed in his opening: "I'm disappointed that we are not able to fully meet the demand"). Market/accounting bad news — disclosed fully but placed where technicians look: the SEC-consultation caveat inside the Q3'25 release; the $12.5B escrow MTM inside the Q2'26 GAAP reconciliation, essentially unmentioned in the beat-focused prepared remarks. He does not hide bad news; he sequences it.

Communication/personality [FACT from transcripts + REPORTED]: - Prepared remarks are scripted and read closely — Intel publishes them verbatim as PDFs; his delivery is low-verbosity, short declarative sentences, few ad-libs, no hype vocabulary, no jokes at rivals' expense. Recurring phrases across all six calls: "laser-focused," "no quick fixes" / "multiyear journey — it will take time," "build what customers need, when they need it," "earn (back) the trust of our customers," "delight our customers," and his stated motto (Intel Vision, Mar 31, 2025): "My motto is very simple: under-promise and over-deliver. That has been my trademark. I will not be satisfied until we delight all of you." - In Q&A he is direct on internal failure, non-committal on customers — answers are shorter and less syntactically polished than his script; he frequently answers the principle rather than the number, after which Zinsner supplies the number (the transcripts' most reliable rhythm; Zinsner also re-states Tan's points in model-ready language for NCI, share count, capex phasing). He does not spar with hostile questioners; Rasgon's sharp inventory challenge (Jan 2026) was absorbed and answered by Zinsner with data, and Tan's own prepared remarks had already conceded the point. - On the accent/language question the mission raises: I found no reputable reporting that analysts cannot understand Tan or that comprehension is a noted issue. What is verifiable is narrower: (i) his Q&A syntax is compressed and occasionally garbled in third-party transcripts — e.g., MarketBeat's Q2'25 transcript renders "Secure Enclave" as "Secure and Grave programs," a transcription artifact, not a Tan quote; (ii) StockAnalysis's Q3'25 transcript intro erroneously names "Pat Gelsinger" as CEO — vendor error, a warning against over-reading any single transcript; (iii) he is a native Mandarin speaker who has delivered speeches in Chinese for Intel's China anniversary (TipRanks/DEXWire, 2025). Treat any stronger claim as unverified. - Off-call candour is a trait, not an accident: at the DBS panel on Oct 12, 2026 (Reuters/Bloomberg reporting) he and Broadcom's Hock Tan publicly questioned AI-lab ROI and token economics — the style of a man who says the quiet part in prepared settings too (cf. "our last year of positive FCF was 2021 — completely unacceptable").

Criticism and failure modes — verified facts, separated from opinion: - [FACT — Reuters investigation, Apr 10, 2025] Tan invested at least $200M between Mar 2012 and Dec 2024 in hundreds of Chinese advanced-manufacturing and chip firms, including PLA contractors/suppliers, via Walden and two Hong Kong holding companies (Sakarya, Seine). Reuters identified 20 funds/companies where Walden was a joint owner alongside Chinese government funds/SOEs, and six firms co-invested alongside China Electronics Corp (sanctioned 2020); one co-investment (QST) has had sensors found in Russian military drones per a Ukrainian government database; another (Intellifusion, ~2%) was U.S.-blacklisted in 2020. Walden was a seed investor in SMIC; Tan exited in 2021 per the House Select Committee. Reuters alleged no illegality by Tan; its expert frame was political sensitivity and board diligence (a law professor: something "responsible business leadership… would at least have a serious conversation about"). Most Intel directors did not respond to Reuters. - [FACT] The political sequel: Cotton's letter (also citing Cadence's 2025 guilty-plea settlement over EDA exports to a Chinese military-linked university — conduct from the period overlapping Tan's Cadence tenure; the plea was Cadence's, and I found no reporting charging Tan personally) → Trump's Aug 7, 2025 resignation demand → Aug 11 meeting → reversal ("an amazing story," AP) → the Aug 22 stake deal. Whether one reads this as capitulation or statecraft is opinion; the sequence and terms above are the verified facts. - [REPORTED skepticism at appointment] Analysts questioned whether a foundry can win orders from designers wary of entrusting production to a rival (Reuters); Morgan Stanley held a Hold citing strategic ambiguity; Loop Capital's "catch-22" (Intel Products needs TSMC to compete, but Foundry needs Products' volume to cover fixed costs). The EDA/VC-background question — he had never run a fab — was the standing doubt behind CNBC's "now he needs a big customer" framing. - Where the style has NOT worked yet at Intel [FACT]: (1) No signed, termed, leading-edge external wafer customer as of Oct 22, 2026 — external foundry revenue $293M in Q2'26 (~5% of the segment; mostly mature-node/Altera-related per filing analysis), Foundry still losing ~$2.1B/quarter ($10.32B in FY25). (2) Demand forecasting/supply planning failed publicly once (Q4'25 buffer depletion → Jan 2026 −17%) — under his "say-do" banner. (3) The client business keeps shrinking as a share (CCG $8.5B→$7.7B across Q3'25–Q1'26 before the Q2'26 rebound; PC TAM guided down low-double-digits for 2026; AMD server unit share 34.5% in Q2'26 per Mercury via The Register). (4) Capex discipline inverted into dilution — the Aug 2026 offering increased share count ~4.8%; every EPS guide now carries ~5.3B shares. (5) The stock itself: −30% in the month into this print (close $85.49 on Oct 22, −8.04% on the day, on an analyst downgrade per MarketBeat) — the market is re-testing the narrative he built.

11. Forecast — Q3 2026 call, Thu Oct 29, 2026, after close (call 5:00 p.m. ET)

Everything in this section is INFERENCE/forecast except the setup facts, which are cited. Mission-provided setup: Intel guided (Jul 23, 2026) Q3 revenue $15.8–16.8B, non-GAAP EPS $0.38, non-GAAP GM 42.0% (GAAP: $0.31 / 41.0%); consensus ~$16.5B / $0.40 as of Oct 20, 2026 — corroborated by MarketBeat consensus via Primexbt ($16.5B/$0.40) and Benzinga ($16.45B / $0.39). 2026 capex guided above $20B. Aug 2026 offering: $20B base at $95 (242.1M shares incl. greenshoe, $23.0B gross). Demand said to exceed supply (Q2 call). FOMC decision Oct 28 (mission). Management's own Q2-call framing for Q3: supply growth skewed to late Q3/Q4; NCI ~$250M in Q3; client soft on memory constraints.

Probability-weighted scenarios (each tied to a named precedent):

What Tan will most likely say / not say (persona-mapped): - Named new 14A/18A customer: NO (from him). Nineteen months, six calls, zero exceptions; his stated rule is that the customer announces, and Zinsner's rule is signature-or-silence. If a name comes, expect it to arrive the Musk way — from the customer, or as Tan confirming a customer's announcement. A Tan-initiated naming with volume/dollar terms would be a deliberate style break — treat it as the single most bullish tell available, precisely because it violates his trademark. - Q4 guidance style: a $1.0B-wide range, GM/EPS at midpoint, with the explanation front-loaded (supply skew, memory) — Jan taught him the cost of an unexplained guide; Jul taught the market to listen to the bridge, so the bridge will be detailed. Expect "better than seasonal / supply improving each quarter" language carried from Q4'25/Q2'26. - Capex: reiterated >$20B for 2026; 2027 "significantly higher" — not cut (he has never cut a number he raised on demand grounds; Cadence pattern is the same: spend follows committed demand). A further raise in prepared remarks (vs Q&A) would signal demand conviction but, per Jul'26, may still sink the stock intraday. - Gross margin: the 42.0% guide is only +20 bp vs Q2's 41.8% — expect him to let Zinsner carry the bridge (18A cost-down: Panther Lake SKU cost −50% YTD, −20% more guided in Q2 remarks) and to frame >42% as yield execution, not pricing. - Supply constraints: "demand continues to outpace our growing supply" is now his signature sentence (Q2'26); expect it verbatim, with the late-Q3/Q4 skew repeated. A shift to "supply is catching up" would be a bearish demand tell in his vocabulary, not a relief. - GAAP escrowed-share charge: expect near-silence in prepared remarks (Q2'26 pattern) and the number in the release reconciliation. Direction is mechanical: the charge/gain marks the $15.6B (Jun 27) escrow liability to the share price — with the stock down sharply during Q3 from its June levels, a material GAAP gain is arithmetically plausible, which would make GAAP EPS look far better than the $0.31 guide (which assumed zero MTM). Either way: trade the non-GAAP line; a shocked GAAP headline is the known trap (Primexbt's preview makes the same warning). - Deviation signals to price immediately: (i) Tan names/commits a customer himself → regime change, bullish; (ii) Tan delivers a new restructuring/charge in his opening minutes → his Cadence/Q2'25 pattern: painful day, credibility up, historically a buy-the-credibility setup only after the guide is digested; (iii) Tan defers the 14A decision window ("into 2027" without "second half of this year") → the Jan timeline has slipped, bearish for the foundry thesis regardless of the print; (iv) Zinsner, not Tan, softens the customer language → the two-man pattern breaking down is itself information.

Listen-for list — in the order it will occur: 1. Release, ~4:01 p.m. ET: Revenue vs $16.8B (guide top) and $16.5B (consensus) — both matter, in that order. Non-GAAP EPS vs $0.38 guide / $0.39–0.40 cons. GM vs 42.0%. GAAP EPS and the escrowed-shares MTM line (expect a swing; ignore for trading, note for headlines). DCAI revenue vs Q2's $6.26B. Foundry external revenue vs $293M — the quiet number that grades the foundry thesis. Adj. FCF and the partner-contributions line (Q2's −$12.2B distortion should not repeat). 2. Tan, opening minutes: the streak count ("eighth consecutive…") — and whether any bad news is front-loaded here (his tell for his decisions). Bullish phrase: "demand continues to outpace our growing supply." Bearish phrase: any variant of "we are not able to fully meet demand" without the demand-strength framing. 3. Tan, foundry section: PDK 0.9 — was the October milestone met? (A miss/silence = timeline slip.) 14A verbs: "engagements/momentum" (= Q2, neutral) vs "commitments/decisions" (bullish) vs "extending into 2027" (bearish). Tesla/Terafab: "framework/explore" (= status quo) vs "signed/commitment/volume" (bullish). EMIB-T backlog — any dollar figure upgrades packaging from story to business. 18A output vs target (Q2: +25%). 4. Tan, products: CCPG — memory-constraint quantification and whether client revenue holds ~$8.9B; DCAI — Xeon demand, CPU:GPU ratio update (1:4 → ?), ASIC run-rate vs the $1B (Q4'25) / ~3× YoY (Q2'26) markers. 5. Zinsner, print bridge: how much of the beat is volume/yield/price vs one-offs — Q1'26 contained a $0.06 one-time gain and reserved-inventory benefit; a repeat-quality question. Foundry loss vs $2.09B. 6. Zinsner, guidance (the moment that has decided 4 of 6 reactions): Q4 revenue range midpoint vs consensus; GM direction — ≥43% bullish, ≤41% bearish; EPS; NCI (~$250M guided for Q4); share count — now including the 242M offering shares (~5.3B diluted): check the EPS denominator drag explicitly. 7. Zinsner, cash/capex: 2026 >$20B reiterated or raised; any quantification of 2027 (new information either way); deployment of the ~$22.6B net offering proceeds; FCF ex-partner-items; debt maturities. 8. Q&A, in likely order: guidance bridge/supply (Seymore/Rasgon) → foundry-customer conversion (Arcuri/Arya — watch for the milestone-deflection) → capex-vs-hard-orders (Reitzes — a broadened answer was Jul'26's bearish tell) → memory/client (Rasgon) → NCI/escrow modeling (Rakers' slot). A question Tan answers himself, with a number, on customers would be unprecedented — and therefore the loudest signal on the call. 9. Close: Investor Day — promised for H2 2026 on the Jan call; a date/audience would confirm the engagement phase; silence pushes it into 2027 and is a minor negative tell.

12. Verified fact vs inference — and what could not be verified

Verified fact (primary sources — Intel/SEC releases, Intel-published remarks, filed 10-Q/424B5, Cadence release, Intel 8-K bio): all six quarters' revenue/EPS/GM actuals and the guidance given for each following quarter (§1–7 table); the Q2'25 charge bridge; the Q3'25 SEC-consultation caveat; the Altera/SoftBank/NVIDIA/U.S.-government transaction terms and dates; the Fab 34 $14.2B buyback; the Q2'26 $12.5B escrow MTM and $15.6B liability; the Aug 2026 offering share counts and $95 price; the 14A development-commitment vs capacity-gate distinction (Q2'26 10-Q); Tan's biography, Walden/Cadence record and board history; the Reuters China-investigation findings and the Aug 2025 Trump sequence. Reported characterization (named secondary sources, labelled in text): consensus figures (LSEG/FactSet as republished — they vary by vendor by $0.01–0.08 on EPS and ~$0.1–0.5B on revenue in some quarters); day-after % moves for Q1'25 (−6.7%, AInvest recap) and Q2'25 (~−8.5%, AP/Reuters recaps) — the Q4'25 (−17.03%), Q1'26 (+23.6%) and Q2'26 (−7.9%) moves are multiply corroborated with exact closes; Cadence-era operational colour (subscription shift mechanics, M&A size discipline, year-one cost cuts) from analyst/secondary profiles. Inference/forecast (mine): the sandbagging verdict (§7), the mover ranking (§9), all of §11 including probabilities, and the persona→behaviour mapping. Could not verify / gaps: - Q3'25 U.S. closing % on Oct 24, 2025 — sources report +7–8% after hours, an intraday print above $41 and ~$38.60 (+~1%) intraday; no clean closing figure found (one earnings-history source explicitly notes the same gap). The research window also failed to load StockAnalysis's price-history page (DNS), and the structured finance-data tool returned no parseable INTC result, so closes rely on contemporaneous reporting. - Full-transcript caveat: complete transcripts were accessible for all six quarters via MarketBeat, cross-checked against StockAnalysis/Fool/Seeking Alpha/ROIC and Intel's official prepared-remarks PDFs. None of the third-party transcripts is error-free (two vendor errors documented in §10) — verbatim quotes were used only where two sources or the official remarks agree. Seeking Alpha's full text sits behind its standard transcript presentation; the Q1'26 call was verified via MarketBeat + Intel's remarks PDF instead. - Whether the H2 2026 Investor Day promised in Jan 2026 has occurred/was scheduled — not found in sources checked. - The exact date of Zinsner's "signed on the dotted line" Deutsche Bank remark (reported by DataCenterDynamics; the "conviction/capacity" remarks are dated late Aug 2026 by the Fool's Oct 3, 2026 article). - Q1'25 consensus EPS is reported as $0.01 (LSEG via AInvest) and $0.06 (Reuters, in the Q2-guide comparison) by different vendors — both are cited where used; the discrepancy does not change any conclusion. - The mission's FOMC date (Oct 28, 2026) and 5:00 p.m. ET call time are taken from the mission and are consistent with Intel's standard 2 p.m. PT webcast time in its releases; the Oct 29 date itself is confirmed by Intel's earnings calendar coverage and MarketBeat.

Sources

Primary (Intel / SEC / company): - Q1'25 release (Ex. 99.1), Apr 24, 2025 — https://d1io3yog0oux5.cloudfront.net/_3ccf9cf30376bf77bbb27f582e51d00d/intel/db/887/9128/earnings_release/Q1%2725_EarningsRelease.pdf - Q2'25 release (Ex. 99.1), Jul 24, 2025 — https://d1io3yog0oux5.cloudfront.net/_55d7c77db61d0836bb9bcea650fc712a/intel/db/887/9139/earnings_release/Q2%2725_EarningsRelease_FINAL.pdf - Q3'25 release (Ex. 99.1, SEC), Oct 23, 2025 — https://www.sec.gov/Archives/edgar/data/50863/000005086325000169/q325earningsrelease.htm - Q4'25 release, Jan 22, 2026 — https://www.intc.com/news-events/press-releases/detail/1759/intel-reports-fourth-quarter-and-full-year-2025-financial (also Business Wire, Jan 22, 2026) - Q1'26 release (8-K Ex. 99.1), Apr 23, 2026 — https://www.intc.com/filings-reports/all-sec-filings/content/0000050863-26-000077/0000050863-26-000077.pdf ; newsroom: https://www.intc.com/news-events/press-releases/detail/1767/intel-reports-first-quarter-2026-financial-results - Q2'26 release (Ex. 99.1), Jul 23, 2026 — https://www.intc.com/filings-reports/all-sec-filings/content/0000050863-26-000155/q226earningsrelease.htm ; PDF: https://d1io3yog0oux5.cloudfront.net/_91f56640ed59478f64737e64880beea2/intel/db/887/9266/earnings_release/Q2+2026+Earnings+Release.pdf - Intel prepared remarks PDFs (official): Q1'25 — https://d1io3yog0oux5.cloudfront.net/_c138057a45e97cd10ae03f84bac0c3c1/intel/db/887/9128/prepared_remarks/Q1-2025-Earnings-Call-Script-FINAL-FINAL.pdf ; Q3'25 — https://d1io3yog0oux5.cloudfront.net/_12b9c1a6e9c73ed7302f8a8a131a4cfc/intel/db/887/9148/prepared_remarks/3Q2025+Earnings+Call+Prepared+Remarks.pdf ; Q4'25 — https://d1io3yog0oux5.cloudfront.net/_a0dbd358c0dfe27d139bef6d4198de40/intel/db/887/9159/prepared_remarks/Intel-4Q2025-Earnings-Call+1+%281%29.pdf ; Q1'26 — https://d1io3yog0oux5.cloudfront.net/_3ccf9cf30376bf77bbb27f582e51d00d/intel/db/887/9254/prepared_remarks/1Q2026-Earnings-Call.pdf - Tan employee letter "Our Path Forward," Apr 24, 2025 — https://d1io3yog0oux5.cloudfront.net/_bd8a0e7a53ac248169bfb79bb936571e/intel/news/2025-04-24_Lip_Bu_Tan_Our_Path_1738.pdf - U.S. government investment 8-K, Aug 25, 2025 — https://www.intc.com/filings-reports/all-sec-filings/content/0000050863-25-000129/0000050863-25-000129.pdf - Offering prospectus (424B5), Aug 12, 2026 — https://www.intc.com/filings-reports/all-sec-filings/content/0001193125-26-345221/0001193125-26-345221.pdf ; pricing release summary — https://semiwiki.com/forum/threads/intel-announces-upsize-and-pricing-of-20-billion-common-stock-offering.25691/ - Q2'26 10-Q (14A commitment/gating language; escrow liability) via SEC PDF — https://files.sec2pdf.com/Intel%20Corp%20(INTC)/SEC_Filings_PDF/10-Q/20260724-10-Q-INTEL%20CORP-0000050863-26-000157.pdf - Intel 8-K electing Tan to the board (full bio/boards), Aug 11, 2022 — https://www.intc.com/filings-reports/all-sec-filings/content/0001193125-22-218541/0001193125-22-218541.pdf - Cadence CEO transition release (Tan record: rev <$1B→~$3B, OM >35%, stock +3,200%), Jul 26, 2021 — https://www.stocktitan.net/news/CDNS/cadence-announces-anirudh-devgan-to-become-ceo-in-december-2021-lip-sxohw2xtqkgt.html

Full transcripts (complete calls, read Oct 22, 2026): - Q1'25 — https://stockanalysis.com/stocks/intc/transcripts/311075-q1-2025/ ; https://www.MarketBeat.com/earnings/reports/2025-4-24-intel-co-stock/ - Q2'25 — https://www.marketbeat.com/earnings/reports/2025-7-24-intel-co-stock/ (read in full via index) - Q3'25 — https://www.MarketBeat.com/earnings/reports/2025-10-23-intel-co-stock/ ; https://stockanalysis.com/stocks/intc/transcripts/366977-q3-2025/ - Q4'25 — https://www.MarketBeat.com/earnings/reports/2026-1-22-intel-co-stock/ ; https://www.fool.com/earnings/call-transcripts/2026/01/23/intel-intc-q4-2025-earnings-call-transcript/ ; https://stockanalysis.com/stocks/intc/transcripts/396566-q4-2025/ - Q1'26 — https://www.MarketBeat.com/earnings/reports/2026-4-23-intel-co-stock/ ; https://seekingalpha.com/article/4893739-intel-corporation-intc-q1-2026-earnings-call-transcript - Q2'26 — https://www.marketbeat.com/earnings/reports/2026-7-23-intel-co-stock/ ; https://www.roic.ai/quote/INTC.NE/transcripts

Reactions & contemporaneous reporting: - Reuters, Q1'25 (via SRN) — https://srnnews.com/intel-forecasts-second-quarter-revenue-below-estimates-fanning-tariff-worries/ ; AInvest close recap (−6.70%, Apr 25, 2025) — https://www.ainvest.com/news/intel-stock-plummets-6-70-disappointing-outlook-record-16th-highest-trading-volume-2504/ - Reuters, Q2'25 (via TradingView) — https://de.tradingview.com/news/reuters.com%2C2025%3Anewsml_L4N3TL1ZS%3A0-intel-to-slash-workforce-by-year-end-as-it-forecasts-steeper-losses-than-expected/ ; CRN on the 14A condition — https://www.crn.com/news/components-peripherals/2025/intel-future-of-foundry-business-hinges-on-a-significant-external-customer - Reuters, Q3'25 (via The Ledger Asia) — https://theledger.asia/intel-beats-profit-estimates-as-cost-cuts-and-new-investments-fuel-recovery/ ; IndexBox Oct 24 intraday — https://www.indexbox.io/blog/intel-shares-hit-18-month-high-as-turnaround-gains-traction/ - Q4'25 reaction — https://www.fxleaders.com/news/2026/01/23/intel-shares-sinked-17-after-disappointing-revenue-outlook/ ; Atlas Peak debrief — https://www.atlaspeakresearch.com/report/23b4fd - Q1'26 reaction — https://www.investors.com/news/technology/intel-stock-intc-q1-2026-earnings/ ; https://www.tikr.com/blog/intel-stock-rose-24-after-q1-2026-earnings-heres-what-the-market-is-still-missing ; Fortune call coverage (Terafab Q&A) via https://citywingnews.com/intel-ceo-lip-bu-tan-crushed-earnings-targets-on-1-year-anniversary-were-embracing-paranoid-roots/ - Q2'26 reaction — https://www.fool.com/investing/2026/07/24/intel-just-posted-its-fastest-revenue-growth-in-mo/ ; https://valueslate.com/intel-stock-drop-q2-2026-earnings/ ; https://www.investopedia.com/intel-stock-jumps-as-earnings-blow-past-expectations-amid-booming-ai-demand-intc-12025717 - Q3'26 setup — https://primexbt.com/for-traders/intel-earnings-date/ ; https://www.tradingview.com/news/benzinga:f5fb60668094b:0-intel-earnings-are-around-the-corner-here-s-what-to-expect/ ; INTC Oct 22, 2026 close $85.49 (−8.04%) via Finnhub market data in index results.

Tan biography / style / controversy: - Intel appointment release coverage — https://convergedigest.com/intel-names-lip-bu-tan-as-ceo-to-drive-turnaround-strategy/ ; CNBC profile ("now he needs a big customer"; Zinsner "Rolodex") via https://techinvestornews.io/2025/05/29/intel-ceo-lip-bu-tan-has-a-long-track-record-in-the-chip-industry-now-he-needs-a-big-customer/ - Appointment reaction/skepticism (Reuters via i3investor) — https://klse.i3investor.com/web/blog/detail/ceomorningbrief/2025-03-14-story-h498680571-Intel_Jumps_10_as_Investors_Cheer_Appointment_of_New_Malaysian_born_CEO_ - Cadence analysis (subscription shift, M&A discipline) — https://expansestocks.substack.com/p/deep-dive-brief-cadence ; Tensilica deal — https://venturebeat.com/business/cadence-buys-chip-design-firm-tensilica-for-380m - Reuters China investigation + factbox, Apr 10, 2025 (via SRN) — https://srnnews.com/intel-ceo-invested-in-hundreds-of-chinese-companies-some-with-military-ties/ ; https://srnnews.com/factbox-intel-ceo-lip-bu-tans-wide-ranging-investments-in-china/ - Trump resignation demand/reversal (AP via Barchart) — https://www.barchart.com/story/news/34071022/trump-says-intel-ceo-has-an-amazing-story-days-after-calling-for-his-resignation ; Reuters White House reconstruction (via Hindu BusinessLine) — https://www.thehindubusinessline.com/info-tech/how-intel-ceo-lip-bu-tan-turned-trumps-attack-into-a-white-house-deal/article70434118.ece/amp/ - Deutsche Bank Aug 2026 (Zinsner "conviction"/capacity; Fool, Oct 3, 2026) — https://www.fool.com/investing/2026/10/03/where-will-intel-stock-be-5-years-comeback-story/ ; Zinsner "dotted line" rule (DCD) — https://www.datacenterdynamics.com/en/news/us-govt-to-take-extra-5-stake-in-intel-should-chipmaker-sell-majority-stake-in-its-foundry-business/ - Terafab/Tesla 14A — https://www.fool.com/investing/2026/10/05/elon-musk-confirmed-terafab-talks-with-tsmc-intel-has-been-its-only-named-chip-partner/ ; Oct-19 headline caution — https://www.youtube.com/watch?v=OSBVXmZ1bb0 (preview transcript) and https://summamoney.com/investing/the-daily/tesla-rises-on-a-foundry-agreement-with-intel-as-jpmorgan-upgrades-the-stock/ - Tan motto (Intel Vision, Mar 2025) — https://www.hackster.io/news/intel-ceo-lip-bu-tan-admits-intel-has-fallen-short-of-your-expectations-promises-a-new-era-02535829528d.amp ; DBS panel candour (Oct 12, 2026) — https://www.reuters.com/world/china/intel-ceo-says-chinese-frontier-ai-model-adoption-has-been-very-fast-2026-10-12/ - Filing-based synthesis used for cross-checks (segment/external revenue, escrow, capex) — https://sec-api.io/insights/financial-analysis-of-intel-q2-2026-filings-and-scenarios-to-2029