Ambiq Doubled Revenue and Halved Its Stock — Then Quietly Retired the Breakeven Date
A 110% growth quarter, a raised full-year guide and $363 million of net cash... against a 40% ninety-day drawdown, a diversification story that shrank when restated in revenue, and a profitability timeline nobody mentioned.
AMBQ · Earnings Call · 2026-11-10
The quarter the numbers and the tape disagreed
Ambiq's third quarter was, on the page, the best quarter in the company's short public life. Net sales of $38.1 million rose 110% year-over-year — the sixth consecutive quarter of sequential growth — with non-GAAP gross margin at 48.7%, up 390 basis points, and non-GAAP gross profit up 128%, faster than revenue. Management now expects roughly “$140 million in net sales for the full year 2026, nearly doubling from 2025” — Fumihide Esaka, Chief Executive Officer · 2026-11-10, versus an initial expectation of "more than $100 million." The fourth-quarter guide of $43.5–44.5 million implies a seventh straight sequential record, and the CFO's characterisation was unusually crisp: the company's performance has “significantly exceeded the expectations we set at the beginning of 2026” — Jeffrey Winzeler, Chief Financial Officer · 2026-11-10. And yet the equity has been cut nearly in half. AMBQ is down 40.1% over the last ninety trading days and sits 50% below its June 18 peak of $90.48 — on a $1.1 billion market cap, against $363.5 million of cash and no debt. The company is not being singled out. Across the market's own price tape, the AI-hardware complex has been de-rating for a full quarter: high bandwidth memory names are down 32% over ninety days, co-packaged optics down 32%, high performance computing down 31%. Ambiq sells ultra-low-power microcontrollers, not accelerators, but it trades in the same sentiment bucket. The read is that this report landed into a sector-wide multiple compression, not a company-specific loss of faith — which is precisely why the things management chose to stop saying matter more than usual.The diversification story shrank when it was measured in dollars
For four consecutive calls, the answer to "how is the non-wearables push going?" was a funnel statistic. In August, the CFO said “roughly 25% plus of that funnel is nonwearable devices” — Jeffrey Winzeler, CFO · 2026-08-11. In March and May it was the same "about 1/4 of our pipeline" framing, and back at the IPO it was 17% of funnel. The number never moved much, but it always sounded like progress. This quarter, that framing vanished and was replaced — for the first time — with a revenue-based quantification, and it is a much more sobering number. Non-wearables are on track to double this year, but will represent only a mid-single-digit percentage of total revenue in 2026, will remain mid-single digits in 2027, and only step up to low double digits in 2028. Doubling a base that is mid-single digits gets you... high single digits. That is not a criticism of the engineering; it is a criticism of a narrative that leaned on a funnel percentage precisely because the revenue percentage was so small. The company deserves credit for finally putting a dollar-denominated shape on it — but investors who underwrote "diversification" in 2025 are now told the payoff year is 2028, and even then it is a low-teens mix. What is genuinely encouraging is the underlying breadth. Seven customers generated more than $1 million each this quarter versus three a year ago; the total active customer base grew roughly 50%; and sales outside the major customers rose more than 270%. Sales to end customers in China jumped to 17.3% of net sales from 6.7% — good growth, and a concentration fact worth watching given how tariff- and export-sensitive the current keyword tape is. The new proof points are also specific rather than generic: a design win for an AI-enabled glass-break detection sensor with a large cloud-based physical-security provider, using both Apollo hardware and the HELIA software stack, plus an industrial AI partner whose predictive-maintenance software already sits with major U.S. energy operators. That is real, and it is the kind of infrastructure monitoring wedge that could compound — in 2027, they say.A software business model appears; a profitability date disappears
The freshest thing on this call, and the one that barely registered in revenue terms, is that Ambiq is now selling its own models. A customer began deploying the first Ambiq-developed AI model for a voice use case in the quarter. The CTO was unusually direct about why it matters:This is a genuine extension of the model, not a re-labelling. Historically the company sold silicon plus development tools; now it sells the AI model itself, aimed at exactly the customers who lack in-house AI resources — which, tellingly, is the non-wearables customer set, not the giant wearable accounts. The keyword's elevation to the company's top-ranked theme this quarter is deserved. The corollary, though, is that Ambiq is now competing partly on software, where its differentiation is far less protected than low power compute at 300 millivolts. Meanwhile, the thing that has conspicuously fallen out of the conversation is profitability. In May, the CFO laid out arithmetic that was easy to hold management to: “you need revenues roughly $47 million a quarter to get to that point... we're hopeful that, that investment will allow us to pull our cash flow breakeven and our P&L profitability point from, call it, mid-2028 into early 2028 or potentially into the second half of 2027” — Jeffrey Winzeler, Chief Financial Officer (CFO) · 2026-05-12. On this call, with revenue guided to a $44 million midpoint and non-GAAP opex guided to $25.5–26.5 million, the arithmetic has moved the wrong way: at a ~49% gross margin, the operating breakeven revenue is now north of $50 million a quarter. The hurdle rose because operating expense compounded faster than gross margin did. The loss is narrowing — operating margin improved 23.5pp year-over-year to −25.7% — but free cash flow less stock compensation was −$13 million last quarter, and nobody on either side of the line chose to revisit the 2027 breakeven aspiration. Its absence is the loudest sentence on the transcript.that is a big moment for the company. It's kind of, let's call it, immaterial right now in terms of contribution to revenue, but it has a potential to be significant. So we see that as a proof point.