Autoresearch: As of 19 Sep 2026, has Intel named an outside 18A/14A volume customer, or is TSMC saturation still routing only evaluations?

Generated by /autoresearch on 2026-09-19. Synthesized across 2 rounds from Bits&Chips / Piper / Futurum. See Provenance. Treat as raw material — review before promoting into a project or thread. Context: vault/projects/stock-market Research-only. No buy/sell/size.

Summary

The hinge is unchanged. CFO Zinsner (Deutsche Bank, reported 2 Sep): external engagements “have significantly increased”; customers are “moving away from just looking at data to thinking about, well, how much capacity can I get?” He “stopped short of announcing a formal engagement.” Piper Sandler (18 Sep tweeted note): Amazon, Apple, AMD, Google, Tesla, Microsoft, Nvidia, Qualcomm are evaluating 14A; “TSMC is sold out through 2028” is the broker line. Futurum: external foundry revenue still $293M vs $2.1B Foundry op-loss; “no external 14A anchor customer has been named.” Named 14A path remains Terafab, first product mid-2028. Step 4 stays partial. October / PDK 0.9 checkpoints stand. Do not re-date INTC/TSM.

Findings

Capacity talk is not a purchase order

Bits&Chips (2 Sep): “Intel has given its clearest indication yet that external chipmakers are preparing to manufacture products on its 14A process.” Zinsner: “The engagements with customers externally from a foundry perspective have significantly increased.” Tan’s team meeting weekly. “They’re moving away from just looking at data to thinking about, well, how much capacity can I get? What does that supply look like?” “Although he stopped short of announcing a formal engagement, Intel has now reached a point where it has ‘conviction’ for securing 14A external customers.” 14A risk 2027, HVM 2028 (Bits&Chips).

Wccftech / Piper (18 Sep): “While Intel’s next-gen 14A process currently lacks a definitive anchor customer, it is nonetheless being evaluated by” the named US majors. John Intel tweet: “TSMC is sold out through 2028, so customers want a…” — Piper/X, not a TSMC filing. Intel reaffirmed 14A volume Q1 2028, risk Q2 2027 (Wccftech).

Futurum: “External foundry revenue was $293 million in Q2 against a $2.1 billion Foundry operating loss and no external 14A anchor customer has been named.” Internal DCAI +59% to $6.3B still the wafer sink (Futurum).

Clarity: Tan can name Terafab (Tesla/SpaceX-led), first product mid-2028; “What would show buyers actually paying for diversification is a second named 14A customer with a product date before 2028. Or a prepayment reaching Intel” (Clarity).

Q2 $293M / no named 18A-14A customer is already on the page (07-23/07-24). This pass does not graduate it.

Contradictions and open questions

• Wei’s EMIB-T “welcome the competitor” firewall (already on-page) still bounds packaging ≠ front-end share.
• [[intel-18a-yield-vs-tsmc-yield]] stays.
• Checkpoints: October PDK 0.9 / Tan update; H2’26–H1’27 14A decision window.

Provenance

Rounds run: 2 of 3 (early-exit — Zinsner + Piper + $293M answered the named-order question in the negative). X sources: attempted via X search_news; spend-cap 403. Piper tweet is cited only via the Wccftech reprint, not as an X permalink. Grokipedia: not used for any 2026 claim.

Web sources:

• Bits&Chips, 2 Sep — Zinsner; no name.
• Wccftech / Piper, 18 Sep — evaluators; sold-out-through-2028 is broker.
• Futurum — $293M / no 14A anchor.
• Clarity — Terafab mid-2028.

Generated: 2026-09-19