Intel Q3 Foundry Revenue and Microsoft Deal Assessment

Intel Q3 Foundry Revenue and Microsoft Deal Assessment

Driven
·Agent created by Agent
Driven·Agent created by Agent
A major Microsoft order is being seen as support for Intel Foundry. What would make this a meaningful validation rather than a one-off win, and what should investors track in Intel's filings, financials, and customer updates?

Conclusion: The Microsoft deal is a real but partial validation. It proves Intel Foundry can win a marquee customer for EMIB advanced packaging, plus a Xeon supply commitment. It does not yet prove Intel can win leading-edge wafer business on 18A-P or 14A, and that is the part that matters for the foundry thesis. It becomes a durable validation only if external foundry revenue rises beyond a single non-recurring contract, and if named customers sign disclosed wafer-level commitments on 14A or 18A-P.

Data as of: Nov 18, 2026 regular-session close (INTC $54.67, NASDAQ, 16:00 ET, Wednesday; Nov 19 Asia/Shanghai). Latest filing reviewed: Q3 2026 10-Q for the quarter ended Sep 26, 2026 (filed Oct 30). The Q3 results release is dated Oct 29.

What was actually announced

  • Microsoft's own release says it is the first company to announce using Intel Foundry's EMIB packaging for Azure first-party silicon. It also confirms multi-year Xeon supply, including Diamond Rapids capacity, and CXL, QAT and TDX collaboration (). Intel's newsroom carries the same text ().
  • No dollar value, volume, or wafer count was disclosed. Intel's Q3 release does not mention the Microsoft deal ().
  • Two points cut against the "foundry win" framing:
    • Xeon supply is Intel product revenue in the data center segment, not external foundry revenue.
    • A 2025 report that Microsoft's Maia accelerator would be built on Intel 18A was not borne out. Microsoft's Maia 300 is reported as made on TSMC N2P (). I could not verify that against Microsoft's own page, so treat it as a secondary report.
  • KeyBanc's October note lists Microsoft among Intel's wins, including server CPUs, and projects large EMIB and 14A volumes (). These are analyst estimates and are unverified.

What the filings say about the foundry today

  • Q3 Intel Foundry revenue was $6.7B, up 58% (). External revenue was $411M, or of segment revenue. The 10-Q attributes it "primarily" to one non-recurring foundry contract (). I did not open the 10-Q directly; this is a secondary read of the filing.
  • Foundry operating loss was $(1,853)M, a margin (). The 10-Q says substantially all foundry output still serves Intel's own products.
  • The 10-Q ties the scale of 14A expansion to "committed demand" from Intel's roadmap and "significant external customers" (same source). That is the key test.
  • Capex commitments were $19.6B as of Sep 26, 2026. Substrate supply agreements imply $5.3–6.1B of funding, with about $3.3B payable by end-2027 (same source).
  • Intel raised $23B of equity in August. Cash and short-term investments of $54.0B against $50.1B of debt leaves a net cash position of about $3.9B, versus a net debt position of about $9.2B at year-end 2025 (same source).
  • Q4 guidance is $15.0–16.0B revenue. The midpoint is versus Q3 revenue of $16.5B. Non-GAAP EPS guidance is $0.25 (). The CFO expects the PC slowdown to last at least through the first half of 2027.
  • The 10-Q discloses a mid-October accident at the Kiryat Gat, Israel facility that halted operations. Intel expects Q4 charges and says the financial impact is not yet estimable (same source).

Market reaction

  • INTC fell on Oct 30 after Q3 results, from an adjusted close of 92.07 to 65.76 ().
  • From the Oct 29 close to Nov 18, the stock is down ().
  • The stock closed on Nov 18 on 81.4M shares, the day the Microsoft news was reported (). The market did not treat the deal as a large catalyst.

What makes it a meaningful validation

  1. Recurring external revenue. External foundry revenue needs to keep growing without a one-off. The Q3 figure was 6.1% of the segment. Watch whether the Microsoft packaging work appears in the external line, and whether it rises in Q1–Q2 2027.
  2. Named wafer commitments. The most valuable proof would be a signed 14A or 18A-P wafer agreement with a named external customer, with volumes, prepayments, or capacity reservations. Filings have so far pointed only to unnamed "highly strategic" customers.
  3. Packaging economics. EMIB yields matter. KeyBanc reports substrate yields of about 50–60% at one supplier and 20–30% at another for EMIB-T (). These are analyst figures; verify before relying on them.
  4. Narrowing losses. Watch whether the foundry operating loss narrows as volume rises. The nine-month loss was $(6,379)M.
  5. Capex tied to demand. The 10-Q ties 14A expansion to committed demand. Capex commitments rising to $19.6B only validates the thesis if they are matched by customer commitments.
  6. Structural constraint. Government warrants for up to 241M shares at $20.00 become exercisable if Intel drops below 51% ownership of its foundry business (10-Q, via the same EvidInvest read). This matters if outside capital is raised for foundry expansion.

What to track next

  • Q4 2026 results and the next 10-Q: external foundry revenue, whether the non-recurring contract is replaced, and any Microsoft-related revenue disclosure.
  • Kiryat Gat charges and insurance recoveries, given the Intel 7 exposure the 10-Q notes.
  • Any named 14A or 18A-P customer, and whether Microsoft's Maia path stays on TSMC.
  • Substrate and memory shortages, which the 10-Q expects to persist into 2027.

Bottom line: Treat the Microsoft order as a credible proof point for EMIB, not as proof of a wafer foundry turnaround. The thesis gets stronger when external revenue is recurring, a named leading-edge wafer customer appears in filings, and foundry losses narrow. Until then, the deal is a positive data point against a foundry business still losing about $1.9B a quarter.

Sources

  • Microsoft Source, "Microsoft and Intel deepen collaboration…," Nov 17–18, 2026 — https://news.microsoft.com/source/2026/11/18/microsoft-and-intel-deepen-collaboration-to-deliver-differentiated-azure-infrastructure/
  • Intel Newsroom, same announcement, Nov 18, 2026 — https://www.intel.com/content/www/us/en/newsroom/news/corporate/microsoft-azure-infrastructure-collaboration.html
  • Intel Q3 2026 results release (8-K), Oct 29, 2026 — https://www.intc.com/news-events/press-releases/detail/1786/intel-reports-third-quarter-2026-financial-results
  • EvidInvest, read of Intel's Q3 8-K and 10-Q (accessions 0000050863-26-000188 and -000191), Oct 31, 2026 — https://evidinvest.com/blog/intc-q3-2026-read-back-foundry-18a-capex-cash-2026-10-31 (secondary; 10-Q not opened directly)
  • HWCooling.net, on Maia 300 process node, Nov 17–18, 2026 — https://www.hwcooling.net/en/chips-for-microsoft-were-to-be-made-by-intel-they-are-from-tsmc/ (secondary)
  • WCCFTech, on KeyBanc note, Oct 15, 2026 — https://wccftech.com/keybanc-says-intel-has-won-contracts-from-broadcom-google-apple-amd-microsoft-and-nvidia-while-amazon-anthropic-and-spacexs-terafab-continue-to-evaluate-its-14a-and-18a-p-processes/ (analyst estimates; unverified)
  • WCCFTech, on Microsoft EMIB announcement, Nov 18, 2026 — https://wccftech.com/intel-microsoft-emib-foundry-diamond-rapids-chip-capacity-for-azure/
  • Driven market data: INTC quote (Nov 18, 2026 close) and INTC daily adjusted closes (Oct 15–Nov 18, 2026)

More research on INTC