Intel's confident 2023 promise — five new manufacturing process generations in four years, restoring 'process leadership' by 2025 — collapsed by mid-2024 into a dividend suspension, 32% headcount reduction, and CEO departure, followed by the U.S. government, Nvidia, and SoftBank all becoming new shareholders in 2025, even as guidance beats have widened sharply since the new CEO's arrival.
The story
2023: a confident four-year promise
Two years ago, Intel was confident. CEO Pat Gelsinger's "IDM 2.0" turnaround strategy rested on one specific, dated promise: five new manufacturing process generations in four years, restoring "process leadership" by 2025. On the Q4 FY2023 call, he called it "the culmination of a year of tremendous progress," and set an even bigger long-term target — becoming the world's #2 external foundry business by 2030. Across this period, Intel hit its own quarterly revenue guidance, usually near the top of the range, in every one of 8 tracked quarters.
Mid-2024: the promise breaks
On the Q2 FY2024 call, the same CEO's opening words were "Q2 profitability was disappointing." In that same release, Intel announced plans to cut more than 15% of its workforce by the end of 2025 and suspended its dividend entirely — ending decades of uninterrupted payouts. Counting mentions of "IDM 2.0" across successive 10-Ks tells the story precisely: 26 times (FY2021), 21 (FY2022), 20 (FY2023), then zero in both FY2024 and FY2025. The specific "five nodes in four years, leadership by 2025" language disappeared from the filings at the same time. Sometime between the Q3 FY2024 call (Gelsinger's last as CEO) and the Q4 FY2024 call, Gelsinger left the company — the next call's speakers were introduced simply as "interim co-CEOs Michelle Johnston Holthaus and David Zinsner."
The response: a new CEO promising no quick fixes
Lip-Bu Tan, the former Cadence CEO, took over in March 2025. On his first earnings call he said, "I joined Intel Corporation five weeks ago... I saw the challenges we were facing and I could not sit on the sidelines," adding repeatedly that "there's no quick fixes." He flattened management layers, lowered 2025-2026 operating-expense targets, and introduced a four-day in-office policy. Headcount fell from 124,800 (FY2023) to 108,900 (FY2024) to 85,100 (FY2025) — a 32% reduction in two years. At the same time, Intel traded equity for cash: an 8.4% stake went to the U.S. Commerce Department in exchange for CHIPS Act funding, new shares were issued to Nvidia ($5B) and SoftBank ($2B), and a 51% stake in subsidiary Altera was sold off. The last piece of the "five nodes in four years" promise, 18A, quietly slipped from "second-half-2024 production readiness" to "2025 production" in the FY2024 10-K, while 20A was abandoned for internal products entirely, its resources folded into 18A.
Now: beating guidance again, but on a lower bar
Guidance itself changed shape under the new CEO: starting with Tan's first quarter, Intel explicitly widened its own revenue ranges, describing them in its own words as "wider than normal." Since Q3 FY2025, results haven't just landed inside those ranges — they've blown through the top of them, beating the guided midpoint by $1.4B in Q1 FY2026 and $1.8B in Q2 FY2026, even as the company says a shortage of older Intel 7 wafers is limiting how much of that demand it can actually fill. Intel Foundry's operating margin has also improved, from -77% (2024) to -58% (2025). But the FY2025 10-K still states plainly that Intel has not secured a single meaningful external customer for its next-generation 14A process.
Guidance scorecard
| Cycle | Revenue guidance | Actual result | Result |
|---|---|---|---|
| 23Q3 → 23Q4 | $14.6-15.6B | $15.4B | Met, upper range |
| 23Q4 → 24Q1 | $12.2-13.2B | $12.7B | Met, mid-range |
| 24Q1 → 24Q2 | $12.5-13.5B | $12.8B | Met, lower range |
| 24Q2 → 24Q3 | $12.5-13.5B | $13.3B | Met, upper range |
| 24Q3 → 24Q4 | $13.3-14.3B | $14.26B | Met, upper range |
| 24Q4 (interim co-CEOs) → 25Q1 | $11.7-12.7B | $12.7B | Met, top end |
| 25Q1 (Tan's first) → 25Q2 | $11.2-12.4B ("unusually wide") | $12.9B | Beat the range |
| 25Q2 → 25Q3 | $12.6-13.6B | $13.7B | Beat the range |
| 25Q3 → 25Q4 | $12.8-13.8B | $13.7B | Met, upper range |
| 25Q4 → 26Q1 | $11.7-12.7B | $13.6B | Beat by $1.4B vs. midpoint |
| 26Q1 → 26Q2 | $13.8-14.8B | $16.1B | Beat by $1.8B vs. midpoint |
All 11 tracked quarters landed inside or above their guided range — but the pattern shifted sharply starting with the new CEO's second quarter (25Q2 onward): from "inside the range, near the top" to "blowing through the range entirely." Whether that shift reflects real demand strength or a bar set deliberately low enough to clear easily is the open question.
Source: CFO guidance and reported-results sections of each cited earnings call transcript, Q3 FY2023 through Q2 FY2026.
Timeline
- Jan 2024Q4 FY2023 call: "tremendous progress" declared; IDM 2.0 framed as on track; #2-foundry-by-2030 goal set.
- Aug 2024Q2 FY2024 call: "disappointing" profitability; 15%+ workforce cut and full dividend suspension announced the same day.
- Jan 2025Q4 FY2024 call: speakers introduced as interim co-CEOs, confirming Gelsinger's exit sometime after the prior quarter's call.
- Mar 2025Lip-Bu Tan becomes CEO; first call: "there's no quick fixes."
- Aug 2025U.S. Commerce Department receives an 8.4% equity stake in exchange for CHIPS Act funding; Nvidia and SoftBank make new equity investments.
- Q3 FY2025 onwardGuidance beats turn into large, range-exceeding surprises; Foundry losses narrow from -77% to -58%.
Our read
Intel's last three years trace a collapse and a rebuild happening almost simultaneously: a confident, specific four-year technology promise broke in mid-2024, triggering a dividend suspension, mass layoffs, and a CEO exit, followed by the U.S. government, Nvidia, and SoftBank all becoming new shareholders within the same year. The recent run of large guidance beats is a genuinely positive signal, but it arrived alongside deliberately widened guidance ranges — meaning it's not yet possible to fully separate real demand recovery from a bar that simply got easier to clear. The company's own admission that it still has zero meaningful external customers for 14A is the detail that will ultimately decide which story this turns out to be.
What we still don't know
- Whether the recent guidance-beating streak reflects genuine demand recovery or a deliberately lowered bar can't be fully separated using this filing history alone — it would require watching whether the pattern continues into Q3 FY2026 and beyond.
- Whether Intel has landed any meaningful external foundry customers for 14A since the FY2025 10-K (which says "none") would require checking recent news or subsequent earnings calls.
- The precise circumstances of Gelsinger's departure (the specific nature of the board conflict) isn't disclosed in earnings calls or 10-K filings — only the shift to interim co-CEOs in the speaker introductions confirms the transition happened.
Frequently asked questions
Did Intel deliver on its 'five nodes in four years' promise?
Numerically, close to it — 18A reached production in 2025, roughly a year later than the 'second half of 2024' originally promised. But the specific language and the accompanying goal of restoring 'process leadership by 2025' vanished from Intel's 10-K filings entirely starting in FY2024, and the FY2025 10-K states plainly that Intel had not secured a single meaningful external customer for its next-generation 14A process.
Why did Intel suspend its dividend and lay off so many workers?
On the Q2 FY2024 earnings call, after calling that quarter's profitability 'disappointing,' Intel announced plans to cut more than 15% of its workforce by the end of 2025 and suspended its dividend entirely, redirecting cash toward its costly Intel Foundry manufacturing buildout. Headcount fell 32% between FY2023 and FY2025.
What sources does this analysis draw from?
This piece is built from Intel's 10-K filings for FY2021 through FY2025 and 12 quarters of earnings call transcripts from October 2023 to July 2026.