Intel Foundry Turnaround Progress and Challenges
Intel Foundry Turnaround Progress and Challenges
I'll pull the latest Intel quote, then the most recent earnings and foundry disclosures.
Conclusion: Intel's foundry turnaround is making progress on execution (yields, cost, packaging, process readiness) but has not yet shown the two things that would make it a real turnaround: outside customers committing volume, and a foundry P&L that moves toward breakeven. Intel Foundry still lost $1.85B in Q3 2026, and management has given no firm timeline for external customer announcements and does not expect positive free cash flow in 2027. The next two to three quarters, with the January update and the first signed external wafer or packaging deals, will decide which view is right.
As of: 2026-11-11 (Shanghai). Quote data is the 2026-11-10 4:00 PM ET close. The latest earnings release is the Q3 2026 8-K dated 2026-10-29.
Evidence
1. Foundry losses are narrowing, but the margin is still deeply negative.
- Q3 2026 Intel Foundry revenue was , up from a year earlier.
- Operating loss was , against in Q3 2025. That is an improvement of .
- Operating margin was in Q3 2026, versus in Q3 2025.
- Most foundry revenue is internal. Rackline reports that most of it comes from Intel's own products, and the segment table shows of intersegment eliminations against total revenue of .
2. External demand is the weak link.
- Outside-customer foundry revenue was , about of foundry revenue.
- Rackline reports that the sequential increase of was mainly a one-off customer program, according to the CFO.
- On the call, management gave no clear timeline for external foundry customer announcements (GuruFocus transcript summary). The CEO said he expects announcements within the next two quarters. Packaging (EMIB and EMIB-T) is the nearest external opportunity.
3. Execution progress is real.
- The CFO said 18A yields improved at a double-digit rate and 18A costs fell about since Q1. Intel 3 costs improved about 15% (call commentary, GuruFocus summary).
- The 14A 0.9 process design kit was released in October, and the 1.0 kit is targeted for Q1 2027 (release and call).
- Intel Products are strong: Data Center and AI revenue rose 78% year over year. That is the main reason the company's overall results beat guidance.
4. Cash and capex are the main constraint.
- Q3 operating cash flow was , against gross capex of , up year over year. Adjusted free cash flow was .
- Full-year 2026 gross capex is guided at . With spent through September, that implies about in Q4.
- The CFO said 2027 capex will rise significantly and free cash flow probably will not be positive in 2027, as 18A, Intel 3 and 14A ramp together. Intel reported about of cash and short-term investments after a $23B equity raise.
5. Near-term guidance is weak.
- Q4 2026 revenue guidance is to . The midpoint is below Q3.
- Non-GAAP gross margin guidance is , and non-GAAP EPS guidance is $0.25 (GAAP: 37.1% and $0.10).
- A Fab 28 shutdown in Israel is expected to cut about of Q4 revenue, with a similar hit in Q1.
6. Market context. INTC closed at on 2026-11-10, below its 52-week high of . The market is pricing in the cash-flow problem more than the process progress.
Reasoning
The foundry loss is shrinking because of cost and yield gains and because Intel Products are now absorbing more of the cost base. Those gains are real, but they come mostly from captive volume. Until outside customers commit meaningful wafer or packaging volume, the segment is a cost centre funded by the Products business and by equity. Management's 2027 cash-flow guidance suggests the company expects the investment to pay back later, not in the next year. The bull case depends on Intel 3 turning profitable in 2027 and on 14A customers, which management says are in active discussions. The bear case is that another round of losses and capex increases arrives before those customers sign.
Milestones that would show the turnaround is working
Signals that would weaken the thesis: no named external wafer customer by mid-2027; foundry losses widening again; 2027 capex raised without a matching customer commitment; or further equity issuance to fund the build-out.
Risks and gaps
- I did not retrieve the Q3 10-Q, which was due by 5 November. Balance sheet and cash-flow details come from the 8-K.
- Call commentary comes from a GuruFocus transcript summary and a Rackline recap, not a full verified transcript. Items attributed to the call should be checked against the replay.
- Secondary reports say a 14A customer conversation is growing ahead of the 2028 ramp (TradingView, TrendForce). Neither is a signed commitment, so I have not counted them as evidence of demand.