Everspin’s Print Shows a Real Product Recovery, but Not Yet the Earnings Leverage the Long-Term Target Requires

The clean read from this earnings event is that Everspin has moved past the worst of the 2024 revenue trough, but has not yet proven the model can compound from here without help from non-product items and future platform ramps. On the street-comparison basis, Q4 FY2025 revenue was $14.8 million against an estimate of $14.4 million, a +2.4% surprise, while EPS was $0.05 against an estimate of $0.12, a -54.6% surprise. That split matters more than either headline alone. Investors got confirmation that demand and backlog conversion are good enough to put revenue near the high end of the company’s own range, but the earnings miss says the current mix still does not deliver clean operating leverage. This was not a broken quarter. It was a transitional quarter, and the transition remains expensive, timing-sensitive, and dependent on the company converting product architecture into broader commercial revenue.

That framing is visible in the quarterly sequence, where the revenue line now looks healthier than the earnings line. Revenue fell from $16.7 million in Q4 FY2023 to $10.6 million in Q2 FY2024, then recovered through $12.1 million in Q3 FY2024, $13.2 million in Q4 FY2024, $13.1 million in Q1 FY2025, $13.2 million in Q2 FY2025, $14.1 million in Q3 FY2025, and $14.8 million in Q4 FY2025. The year-over-year comparison turned from -20.7% in Q4 FY2024 to +11.8% in Q4 FY2025, and the sequential growth rate was +5.3% in Q4 FY2025 after +6.5% in Q3 FY2025. That is a legitimate recovery pattern after the 2024 reset. The problem is that gross margin has not followed the earlier peak profile. Gross margin was 58.1% in Q4 FY2023, fell to 49.0% in Q2 FY2024, and has clustered around 51.3%, 51.4%, 51.3%, 51.3%, and 50.8% across the last several reported quarters through Q4 FY2025. In other words, the revenue recovery is real, but the mix and fixed-cost absorption are not yet restoring the prior margin regime.

The margin story explains why the print can be both encouraging and unsatisfying. Management’s own reported basis showed Q4 revenue of $14.8 million, and CEO Sanjeev Aggarwal framed the result as revenue “towards the high end of our guidance range and EPS in line with our expectations.” That wording is important because it clarifies the company was not surprised by its own profit outcome, even though the street-comparison EPS result disappointed. CFO William Cooper gave the bridge: MRAM product sales in the fourth quarter, including both Toggle and STT-MRAM revenue, were $13.5 million, up 22% over the fourth quarter of the prior year, while licensing, royalty, patent and other revenue decreased to $1.3 million from $2.2 million in Q4 '24 due to completed projects. The gross margin decline to 50.8% from 51.3% in the fourth quarter of 2024 was attributed to lower licensing and other revenue. That is the core issue. The healthiest part of the quarter was product revenue, but the highest-margin support from non-product revenue stepped down, muting the P&L benefit.

That same mix dynamic carries into the near-term guide, which looks steady on sales but softer in quality. Cooper guided Q1 total revenue to be “consistent with Q4 '25” and in the range of $14 million to $15 million, with GAAP net loss per fully diluted share between $0.03 and net income of $0.02. On a non-GAAP basis, the company anticipates net income per fully diluted share between $0.07 and $0.12. The revenue guide is not bearish because it preserves the Q4 run-rate range after a recovery from the $10.6 million trough in Q2 FY2024. But the guide also embeds a known gross margin headwind from a sequential decline in non-product revenue due to project completion in Q4 '25, while management is still targeting gross margin in the 50% range. The signal is that product demand is good enough to hold revenue, but not yet strong enough in mix or volume to fully offset the fade of project-linked revenue.

The balance sheet and cash discussion reduce financial stress, but they do not change the operating question. Cooper said Everspin ended the quarter with cash and cash equivalents of $44.5 million, down $0.8 million from $45.3 million at the end of the prior quarter, while cash flow generated from operations increased to $2.8 million for the fourth quarter from $0.9 million in the third quarter. That gives management room to invest through the transition without turning the quarter into a liquidity story. It also means investors should not overread a single EPS miss as a capital problem. The more relevant question is whether cash funding, strategic awards, and operating expense discipline can bridge the company to a larger product cycle before revenue growth stalls in the mid-teens quarterly range. GAAP operating expenses were $8.6 million in Q4 FY2025, down sequentially and up slightly from $8.4 million in the fourth quarter of 2024, so the company is not spending as if demand is collapsing, but neither is it showing the type of revenue scale that would make that expense base easy to absorb.

That bridge increasingly depends on the long-term product roadmap, which management made unusually explicit on the call. Aggarwal said the company’s long-term strategy entails “reaching $100 million in annual revenue over the next 3 to 5 years.” The quote matters because it is not just directional language about growth. It sets a revenue destination and a time frame that can be tested against the current base. Q4 FY2025 revenue was $14.8 million, and the Q1 guide is $14 million to $15 million, so the company is still operating well below the implied annual scale of that ambition. Management identified PERSYST as the major contributor, with licensing and UNISYST also expected to matter. Aggarwal said the first enhanced Serial NOR like UNISYST product family is expected to be in production in 2027 and anticipated to contribute to the $100 million revenue target. He also said the company is on track to tape out a monolithic 256-megabit xSPI STT-MRAM device on a 16-nanometer FinFET node at TSMC in the second half of this year. The investment debate therefore shifts from near-term revenue stabilization to execution risk around new product qualification, production timing, and adoption.

The defense and aerospace angle gives that roadmap a second leg, but it should be treated as support rather than proof of scale. Aggarwal noted $2 million in other income in the fourth quarter and $10.5 million to date from the $14.6 million contract with a DoD contractor to develop a sustainment plan for MRAM manufacturing facilities to provide continuous onshore MRAM capabilities to aerospace and defense customers. Cooper separately tied other income of $2 million to the strategic award won in mid-2024 to upgrade manufacturing equipment in the existing Chandler, Arizona facility. The language is strategically useful because it reinforces onshore capability and high-reliability positioning, both of which fit Everspin’s differentiated MRAM identity. But the accounting also reminds investors that such awards can help the income statement without being the same as recurring product demand. The best version of the story is that defense-linked funding improves capability and customer confidence while product families create repeatable revenue. The risk is that investors capitalize award income as if it were already a durable product stream.

That distinction matters most in the end-market commentary around satellites and high reliability. Cooper described the LEO satellite market as a “burgeoning market,” adding that management sees it in both orders and backlog and feels confident about the company’s products and position there. The useful part of the comment is not the adjective, but the linkage to orders and backlog. Everspin’s products have a credible fit where persistence, endurance, and reliability matter, and the company is introducing high reliability products that management says fit that market space. Still, the call did not provide named customer breakouts in the data set, and the supply-chain read-through is therefore deliberately narrow. With no named customers of MRAM and no named suppliers to MRAM in the supplied read-through table, this print should not be used to infer specific revenue momentum for any external company. The only defensible supply-chain conclusion is that Everspin’s product recovery and onshore manufacturing work point to better internal demand visibility in specialized MRAM markets, not a confirmed order signal for a named customer or supplier.

The tone of the call was also telling because it was more controlled than promotional. The tone history shows Q4 FY2025 sentiment at 0.21, down from 0.35 in Q3 FY2025 and 0.38 in Q2 FY2025, while guidance_tone was 0.26 versus 0.27 in Q3 FY2025 and 0.43 in Q2 FY2025. Prepared_sentiment was 0.22, qa_sentiment was 0.16, and ai_optimism was 0.08 in Q4 FY2025, a sharp contrast with 0.93 in Q3 FY2025. At the same time, qa_evasiveness dropped to 3.1 in Q4 FY2025 from 51.2 in Q3 FY2025, and uncertainty was 56.3 versus 58.4. That combination is consistent with a management team that had a concrete quarter to discuss, but fewer near-term catalysts it was willing to overstate. It was less upbeat in measured language, yet more direct in Q&A delivery. For a small-cap technology supplier with long-cycle product ramps, that is preferable to a call that leans too hard on aspiration.

The subsequent tone data reinforce the same interpretation rather than overturn it. In Q1 FY2026, sentiment was 0.22, guidance_tone was 0.13, tone_confidence was 0.41, prepared_sentiment was 0.27, qa_sentiment was 0.22, ai_optimism was 0.56, uncertainty was 54.7, and qa_evasiveness was -13.8. The call-over-call delta from Q4 FY2025 to Q1 FY2026 shows sentiment +0.00, guidance_tone -0.12, tone_confidence +0.10, prepared_sentiment +0.06, qa_sentiment +0.06, ai_optimism +0.47, uncertainty -1.6, and qa_evasiveness -16.9. That means management’s delivery became more confident and less evasive even as explicit guidance tone softened. The likely read is that management has better command of the execution path, but the near-term guide remains constrained by non-product revenue timing and gross margin headwinds. This is exactly the tone one would expect from a company exiting a revenue trough but not yet entering a broad product inflection.

The peer context makes the same point from a different angle. Everspin is in memory, but it is not participating in the scale economics visible in larger or more cyclical memory peers. In the latest peer table, MU reported $41,456.0 million of revenue, 84.6% gross margin, and +345.7% revenue YoY, while WDC reported $3,337.0 million of revenue, 50.2% gross margin, and +45.5% revenue YoY, and STX reported $3,112.0 million of revenue, 46.5% gross margin, and +44.1% revenue YoY. Everspin’s Q4 FY2025 revenue was $14.8 million with gross margin of 50.8% and revenue YoY of +11.8%. The comparison is not meant to benchmark end markets one for one, because Everspin’s MRAM exposure is specialized and far smaller. It does, however, show that Everspin is not simply a levered play on the broader memory upcycle. Its valuation case has to rest on differentiated non-volatile memory products, high-reliability niches, licensing, and platform execution, not on the expectation that industry beta alone will carry the model.

The investment conclusion is therefore narrower and more demanding than a simple revenue beat narrative. Q4 FY2025 showed that Everspin can grow product sales again, stabilize quarterly revenue around the $14 million to $15 million guidepost, generate operating cash flow, and sustain gross margin in the 50% range even as licensing and other revenue decline. Those are meaningful improvements from the 2024 trough. But the EPS miss on the street-comparison basis, the 50.8% gross margin, the sequential non-product revenue headwind, and the dependence on future PERSYST, UNISYST, licensing, and defense-linked milestones keep the burden of proof on management. The quarter supports the strategic thesis that MRAM has defensible niches and that Everspin has credible assets in those niches. It does not yet prove that the company can reach $100 million in annual revenue over the next 3 to 5 years with attractive earnings conversion. Until product ramps broaden and margins move beyond merely holding the 50% range, the right stance is constructive on recovery, selective on the long-term option value, and disciplined on what this quarter actually proved.