Intel Q2 2026 Results (Actual Print)

Filed: 8-K, 2026-07-23, Exhibit 99.1 earnings release. Quarter ended June 27, 2026. All figures [primary: 8-K Ex-99.1, 2026-07-23] unless noted. Fetched directly from SEC EDGAR (accession 0000050863-26-000155) with a declared User-Agent — the full P&L, segment table, cash-flow detail and Q3 guide are all in the one exhibit, so nearly everything below is primary, not attributed.

The supply-side bookend to the week's demand-side prints. Where Alphabet and Tesla printed roughly-doubled capex financed into negative free cash flow (the "who pays the rent" demand story), Intel is the merchant/foundry supplier on the other side of that spend — and its print says two things a hyperscaler print cannot: the core operating business inflected sharply positive, and its own capital outlay is falling, not rising.

The Print — Load-Bearing Actuals

Consolidated P&L

• Total revenue: $16,128M, +25% YoY (was $12,859M in Q2 2025) — Intel's own release calls this its "strongest revenue growth in more than fifteen years" [primary, CEO Lip-Bu Tan]. It landed $1,328M above the high end of the company's own $13.8–14.8B guide (given 2026-04-23) — the seventh consecutive quarter of revenue above Intel's guidance range. [primary: 8-K Ex-99.1, 2026-07-23]
• Gross margin: GAAP 40.4% / Non-GAAP 41.8% — both ~2.8–2.9pp above the 37.5% / 39.0% guide. [primary: 8-K Ex-99.1, 2026-07-23]
• GAAP operating income: POSITIVE $1,796M (operating margin 11.1%) — a swing from an operating loss of $(3,176)M in Q2 2025, a ~$5.0B YoY improvement at the operating line. [primary: 8-K Ex-99.1, 2026-07-23]
• Consensus context (attributed, not primary): Street sat near ~$14.4B revenue / ~$0.21 non-GAAP EPS into the print — the $16.128B actual beat revenue by ~12% and non-GAAP EPS by roughly double. [web: TradingKey / aggregated coverage, 2026-07-23]

The one fact the headline EPS number requires — the CHIPS Act Escrowed Shares mark

• GAAP diluted EPS: $(2.16) (a huge "miss" vs an $0.08 guide) vs Non-GAAP diluted EPS: $0.42 (+110% above the $0.20 guide). Both are true; the entire gap resolves to a single non-operating reconciling line. [primary: 8-K Ex-99.1, 2026-07-23]
• "(Gains) losses on mark-to-market of Escrowed Shares: $12,529M" (quarter; $13,619M on the H1 cash-flow basis). Escrowed Shares are "shares of Intel common stock held in escrow to be released to the U.S. Department of Commerce as we perform and receive cash proceeds in connection with our CHIPS Act Secure Enclave agreement with the U.S. Government." [primary: 8-K Ex-99.1, 2026-07-23 — GAAP-to-non-GAAP reconciliation + cash-flow footnote]
• Mechanically: the US government holds an equity-linked position in Intel tied to the CHIPS Act. Intel's stock rose sharply this cycle (one secondary source: INTC +163% YTD into the print [web, 2026-07-23]), which raises the mark-to-market value of the shares Intel must eventually deliver — GAAP books that as a non-cash loss to Intel. It is a capital-structure mark on a government instrument, not a demand or execution signal. Reading the GAAP net loss as an operational miss reads the print exactly backwards.

Foundry — the dollar-loss trend just inverted

• The absolute dollar loss narrowed both YoY (−$1,079M, −34.1%) and QoQ (−$348M, −14.3%) — the first time in this KB's tracking that the margin and the absolute number improved in the same quarter. Foundry revenue grew +30.5% YoY. [primary: 8-K Ex-99.1, 2026-07-23, segment table]
• Panther Lake / High-NA EUV — the cleanest technical upgrade of the cycle: "Intel Foundry has entered high-volume manufacturing for a subset of Intel® Core™ Ultra Series 3 processors, code-named Panther Lake, using ASML's EXE High NA EUV technology." A step up from the prior "qualified on select layers" framing (per ASML's own 6-K) to a named product in actual HVM — though still scoped ("a subset of"). [primary: 8-K Ex-99.1, 2026-07-23, "Foundry Momentum"]
• Still NOT disclosed: any 18A yield percentage, defect density, or Foundry break-even figure — CFO Zinsner attributes part of the beat to "higher factory yields and improved cycle times" but gives no number; "break-even"/"breakeven" return zero hits in the release. The KeyBanc 85%-yield claim remains an analyst estimate, unconfirmed by Intel. [primary-negative + web]

Capex and financing — DOWN, not up

• Quarterly gross capex fell 41% YoY — the second straight quarter of YoY decline, contrary to CFO Zinsner's on-record forward framing ("we are meaningfully increasing our investments in equipment, clean room space, and substrates" to support "expected growth this year and next"). The increase is described as coming, not yet visible. Watch whether Q3/Q4 gross capex actually inflects up. [primary: 8-K Ex-99.1, 2026-07-23, Adjusted FCF reconciliation + cash-flow statement]
• No FY2026 capex-dollar guide anywhere in the release. The only FY2026 dollar figure given is operating expenses (GAAP ~$23.0B / Non-GAAP ~$16.5B) — a different line, not capex. The ~$17B FY2026 capital-plan figure this KB carried is web-sourced (TrendForce/wccftech) and remains unconfirmed by any primary Intel document. [primary + web]
• The huge negative Adjusted FCF is a partner buyout, not capex. "Partner contributions, net" swung to $(12,216)M (from +$1,250M in Q2'25), driving Adjusted FCF to $(8,419)M; the H1 cash-flow statement shows "Partner distributions $(14,339)M" (vs $(91)M). This is the Apollo/Fab 34 (Ireland) 49% minority-stake repurchase — a capital-structure reallocation, not incremental fab spending. Reading the negative FCF as "capex is exploding" would be the wrong read — gross capex is down; a one-off JV buyout moved the cash line. [primary: 8-K Ex-99.1, 2026-07-23]

Q3 2026 guidance (new, primary)

The Q3 revenue midpoint ($16.3B) guides above a pre-print Street figure around ~$15.06B [web: aggregated coverage, 2026-07-23]. [all primary: 8-K Ex-99.1, 2026-07-23, "Business Outlook"]

Reaction Delta — What Printed vs What the Pre-Print Watched

Value-Capture Trace (Money/Quinn)

Intel is the supply-side mirror of the week's demand-side prints, and the mechanism is the opposite. Alphabet and Tesla ran capex up into negative free cash flow and plugged the gap from capital markets — the buyer taking on funding risk to pay the compute/power supply chain. Intel's headline free-cash-flow figure is also deeply negative, but for a structurally different reason: its gross capex fell 41% YoY, and the cash-line damage is a one-off buyout of Apollo's 49% Fab 34 stake, i.e. Intel retiring an external partner's claim on its Irish fab rather than pouring new money into silicon. The GAAP net loss, meanwhile, is a non-cash mark on shares owed to the US government under the CHIPS Act — value the state captures if Intel's equity keeps rising, at zero cash cost to operations. So the reader's decision-relevant read is: the operating business inflected positive (op income +$5.0B YoY; Foundry loss finally shrinking in absolute dollars) while two large, scary-looking negative numbers on the statements are both non-operational — one a government equity instrument, one a JV buyout. The supply side is healing its unit economics precisely as the demand side levers up to pay for capacity.

Source: Intel Corporation Q2 2026 8-K Ex-99.1 — SEC accession 0000050863-26-000155, filed 2026-07-23. First-party earnings release; fetched via EDGAR with declared User-Agent. Compiled from the research-layer reaction delta menfem-research/companies/intel/deep-dive-2026-07-23-reaction.md (itself primary-grounded on the same exhibit). Still pending the Q2 2026 10-Q: 18A yield/defect-density, Altera-reconciled external foundry revenue, and the 14A pause/discontinue risk language.