Product +22% YoY, 238 Design Wins, NOR Flash Tailwind Emerging — But Flat Q1 Guide + Litigation Overhang Drive -8.5% T+1 Sell-Off; Maintaining Hold

Results vs. Consensus

Year-over-Year Comparison

Quarter-over-Quarter Comparison

Full-Year 2025 Summary

Revenue Assessment

The Q4 revenue print at $14.8M (+12% YoY; +5% QoQ) closes a clean four-quarter product acceleration arc for FY25. MRAM product sales of $13.5M (+22% YoY; +6% QoQ) is the structural story — Toggle MRAM + STT-MRAM combined posted the strongest product Q4 since 2023. The composition: Data Center continuing as the largest contributor (IBM FCM4 steady + new FCM5 demand + RAID broad customer set at top-5 hyperscalers); Energy Management + Industrial Automation back to normal levels after the multi-quarter inventory correction in Asia completed; ongoing LEO satellite contribution; and the modest Lucid Motors Gravity SUV automotive ramp. The 238 FY25 design wins (+34% YoY from 178) is the most-encouraging leading indicator — design wins typically convert to revenue 18-24 months later, suggesting FY27-FY28 are the harvest years.

The licensing line at $1.3M continued the normalization narrative — down from Q3 $1.4M and materially below Q2's $2.1M peak. This is the third consecutive sequential quarter where licensing declined, with the FY25 total of ~$7M representing -30% YoY from FY24's ~$10M (which included Frontgrade Phase 1 revenue + larger Purdue contract contribution + completed AFRL projects). The structural read: licensing is the diversifier/IP-validator line, not the growth driver.

The Q1 2026 guide of $14-15M (midpoint $14.5M; flat sequential vs Q4 $14.8M) implies the multi-quarter product acceleration pattern is taking a pause. The CFO's explicit framing: "We expect a sequential decline in non-product revenue due to a project completion in Q4 '25, which will result in a gross margin headwind." Translation: product revenue is likely growing $0.3-0.5M sequentially in Q1, offset by a $0.4-0.6M sequential licensing decline = flat consolidated.

Margin Assessment

GAAP gross margin compressed -50bps QoQ to 50.8% and -50bps YoY from Q4'24's 51.3% — the first sequential margin decline in 4+ quarters. The CFO attributed it to "lower licensing and other revenue" — with the high-margin licensing line declining as a percentage of total revenue from 16% in Q2 to ~9% in Q4. The structural product margin held steady (yield improvements continuing per Q3 commentary), but the mix shift compressed the consolidated number. Management's explicit Q1 2026 guidance: "We expect a sequential decline in non-product revenue due to a project completion in Q4 '25, which will result in a gross margin headwind. However, we are still targeting gross margin to be in the 50% range." Translation: Q1 26 GM at 49-50% is expected.

GAAP OpEx of $8.6M (down $0.2M sequentially) is the operating-discipline data point of the quarter — OpEx actually decreased while revenue grew. The non-GAAP OpEx remained at the ~$7.5M Q3 level. Q1 2026 will see OpEx pressure from (a) the newly disclosed $1.6M/quarter patent litigation costs continuing for "next couple of quarters," and (b) increased Microchip Foundry Services activity. The CFO will exclude patent litigation costs from non-GAAP going forward — but they remain real GAAP-bridge items.

EPS & Cash Flow Assessment

Non-GAAP EPS of $0.11 lands in the middle of the $0.08-0.13 guide. The structural marker: this is the first YoY non-GAAP EPS decline in 5+ quarters (Q4'24 was $0.13). The driver is lower DoD other income contribution ($2.0M Q4'25 vs $3.1M Q4'24). The Q1 2026 guide of $0.07-0.12 (midpoint $0.10) implies continued YoY EPS pressure as the DoD contract approaches completion (mid-2027 estimated).

Cash position declined $0.8M sequentially to $44.5M — the first sequential cash decline in multiple quarters. Operating cash flow recovered to $2.8M (from Q3's $0.9M weak AR-timing quarter), but capex of $3.6M (the highest single-quarter capex of FY25, related to manufacturing equipment upgrades at the Chandler facility for the DoD-funded MRAM line) absorbed the operating cash. The CFO's explicit framing on capex: "We had a unique period of capital spend... related to some of the improvements that we saw in the Chandler facility primarily across a couple of different contracts. So that flurry of activity, I think, will start to settle down until we get into the real heart of this Foundry Services Agreement [with Microchip]." Translation: Q1 26 will see another capex-heavy quarter, then more spread out through 2026-2027.

The cash position is still healthy ($44.5M is 30%+ of market cap; debt-free), but the FY26 outlook for cash generation is more pressured: lower DoD contribution + Microchip Foundry Services capex + ongoing patent litigation = FY26 cash position likely flat-to-down vs FY25.

Segment Performance

MRAM Product Sales — $13.5M, +22% YoY, +6% QoQ (4th consecutive double-digit YoY quarter)

The product line at $13.5M is the strongest single-quarter Q4 in Everspin's recent history. The +22% YoY growth rate is consistent with Q3's +22% — sustained acceleration on a now-larger base. The composition is increasingly diversified: Data Center (Toggle in RAID + STT in IBM FCM4/FCM5) remains the largest contributor; Industrial Automation has fully recovered from the Asia inventory correction; Energy Management is back to normal levels; LEO Satellite is contributing meaningfully now that PERSYST EM064LX HR ramped to full production in Q4; and continuing modest contributions from Lucid Motors automotive + casino gaming.

The most strategically important product disclosure of the call: PERSYST 64-megabit xSPI STT-MRAM (the EM064LX HR variant) ramped to full production in Q4 — meeting the late-2025 production target management has set since Q1 2025. AEC-Q100 Grade 1 qualified; 125°C operating temperature; minimum 10-year data retention. The CEO disclosed: "These capabilities are demanded by our customers to secure critical data in a variety of systems from aerospace and defense to industrial applications, including automotive. We are taking orders to support high-volume production from our customers and began shipping in the current quarter." Higher-density variants (128Mb and 256Mb) are in qualification with H2 2026 high-volume availability targeted.

"We are pleased to announce that during the fourth quarter, we ramped our PERSYST 64-megabit xSPI STT-MRAM high reliability product to full production and saw strong demand driven by new customer interest and design wins, specifically in the low earth orbital or LEO satellite market. These devices are AEC-Q100 Grade 1 qualified and ideally suited for use in harsh conditions such as 125°C operating temperature with a minimum 10 years of data retention." — Sanjeev Aggarwal, CEO

The 238 FY25 design wins (vs 178 FY24, +34% YoY) is the structurally bullish indicator. Design wins typically convert to revenue 18-24 months later, suggesting the FY27-FY28 revenue acceleration ramp is well-positioned. The categories: industrial automation, casino gaming, energy management, military & aerospace.

Assessment: The product line trajectory is structurally healthy. +22% YoY for two consecutive quarters on a growing base + +34% design-win pipeline expansion + xSPI HR variants ramping to full production = the FY26 product growth trajectory should sustain mid-teens YoY. We model FY26 product revenue at $54-58M (+13-21% YoY from FY25 ~$48M).

NOR Flash Supply-Chain Shock — New Tailwind Emerging

The single most strategically important new disclosure of the call: NOR Flash supply-chain pressure is emerging industry-wide. The CEO disclosed that NOR Flash suppliers are reallocating fab capacity to DRAM to capture higher margins from AI-driven DRAM demand. This creates a structural supply gap in NOR Flash, and Everspin is being engaged by customers to evaluate xSPI STT-MRAM as a replacement.

"As has been widely publicized, the industry is expecting — experiencing memory shortages. Memory suppliers who have for decades been pushed into commoditization have been — have seen a shift based on unprecedented memory shortages driven by the demands of AI. As a result, they have gone into allocation mode and are moving their capacity up the food chain. Companies that can make NOR flash, NAND and DRAM are shifting those capacities to where they can get more margin out of their fixed capacity. NOR suppliers, for example, are converting their lines to support DRAM to maximize their margins and generate more revenue. This has created a gap in the supply for NOR flash and driving customers to look for alternatives. We are in conversations with customers to evaluate our xSPI STT-MRAM to replace NOR flash. We have the capacity to support such demand and our parts are compatible with NOR flash." — Sanjeev Aggarwal, CEO

The strategic significance: the standalone NOR Flash market is ~$3B+ TAM. The CEO disclosed the UNISYST family target is 5-10% market share over multiple years. Even partial conversion of NOR Flash demand to MRAM would represent a multi-hundred-million dollar TAM expansion for Everspin specifically. The structural caveat: qualification cycles for replacing NOR Flash with MRAM typically run 12-18 months — meaning the revenue conversion is FY27-FY28, not FY26.

Assessment: NOR Flash supply-chain tailwind is the highest-quality new forward catalyst we have heard from Everspin in the four-quarter coverage period. The market dynamic is genuine, the strategic positioning is right (xSPI STT-MRAM IS the natural NOR Flash replacement for higher-reliability applications), and the timing (NOR Flash allocation already underway) is current. The qualification-cycle lag means it's a 2027+ revenue contributor, not a 2026 catalyst — which is why the stock didn't fully credit it in the Q4 reaction. We treat NOR Flash conversion as the structural multi-year revenue acceleration lever, with the first measurable contribution in late 2027.

DoD Sustainment Contract — $2.0M Q4 (Below Expectations)

The DoD contract recognized $2.0M in Q4 (vs $1.2M Q3) — meaningful sequential acceleration, but below the $3M+ market expectations had built in. Cumulative recognition at $10.5M of $14.6M (72%); $4.1M of contract value remaining. Management expects contract completion in "first half of 2027" — implying ~$0.7M per quarter for the remaining 6 quarters, a clear deceleration from the FY25 H2 peak rate.

"We recognized $2 million in other income in the fourth quarter and $10.5 million to date from the $14.6 million contract we have with a DoD contractor to develop a sustainment plan for our MRAM manufacturing facilities to provide continuous onshore MRAM capabilities to their aerospace and defense customers. We expect this business to progress on schedule with estimated completion in the first half of 2027." — Sanjeev Aggarwal, CEO

Assessment: The DoD contract is moving toward completion at a steady pace, but the FY26 contribution will be materially smaller than FY25. The market was modeling a back-end-loaded acceleration that didn't materialize. The follow-on opportunities (the $40M defense-prime subcontractor agreement disclosed at Q1 2026) are the structural replacement, but they ramp on a different timeline.

Chiplet Ecosystem Engagement — UNISYST Family Strategic Positioning

The Q4 call surfaced multiple new strategic partnerships positioning Everspin's UNISYST family (forthcoming 2027 NOR Flash replacement product line) within the chiplet ecosystem:

• Fraunhofer Chiplet Center of Excellence: Engaged on next-generation automotive compute platforms; system-level simulations using MRAM simulation models
• IMEC Automotive Chiplet Forum: Open chiplet ecosystem for automotive applications
• Physical AI Chiplet Ecosystem (PACE): Newly joined; co-developing interoperable chiplets with MRAM as nonvolatile memory layer for boot, weights, and code storage in physical AI systems
• Quintauris (continued): RISC-V automotive reference designs combining MRAM as persistent working memory

The strategic logic: chiplets are the architectural future of semiconductor design — heterogeneous mix-and-match across process nodes, with MRAM positioned as a high-density nonvolatile memory chiplet for AI accelerators, automotive compute, and edge AI applications. Everspin's UNISYST family (in design phase; production 2027) is being positioned for these ecosystems before launch.

Assessment: The chiplet positioning is multi-year strategic value, not 2026 revenue. The UNISYST family launch in 2027 + the design-in cycles for chiplet integration imply meaningful revenue contribution starting 2028-2029. We treat this as structural moat-building, not near-term EPS.

Key Topics & Management Commentary

Overall Management Tone: Management's tone shifted somewhat from prior quarters. The CEO was more expansive on multi-year strategy (chiplet ecosystem, NOR Flash supply-chain tailwind, $100M revenue target, UNISYST family) — signaling confidence in the long-term thesis. The CFO was more measured on near-term execution — Q1 26 flat sequential guide, $1.6M/Q litigation cost overhang, and the cash drawdown were the new realities the market had to absorb. The structural framing: "things are getting better long-term; near-term is choppier than the market wants." The tension between the long-term optimism and the near-term caution is the central narrative of the print.

1. NOR Flash Supply-Chain Tailwind — Most Important New Disclosure

The NOR Flash supply allocation dynamic surfaced for the first time on this call (referenced briefly in Q3 industry color but not as a Everspin-specific opportunity until now). The CEO's framing positions Everspin as a beneficiary of structural industry capacity reallocation — NOR Flash suppliers (Macronix, Winbond, etc.) prioritizing higher-margin DRAM over NOR Flash, leaving NOR customers seeking alternatives.

The strategic implication: Everspin's xSPI STT-MRAM portfolio (Toggle MRAM has different positioning; xSPI STT-MRAM is the direct NOR Flash replacement category) is the natural alternative for customers seeking a same-form-factor, higher-reliability persistent memory. The Distributor listings as "alternate for NOR flash" is the leading indicator that conversion conversations are real.

Assessment: This is potentially the most consequential new dynamic for Everspin in our coverage period. The qualification cycles (12-18 months) gate the revenue conversion timing, but the structural addressable market expansion is meaningful. We model $5-10M of incremental NOR-replacement revenue by FY27 and $20-40M by FY29 if the supply-chain dynamic persists.

2. PERSYST 64Mb xSPI STT-MRAM HR Ramp to Full Production

The PERSYST 64-megabit xSPI STT-MRAM high reliability product reached full production in Q4 — meeting the late-2025 target management has reiterated across Q1-Q3 calls. AEC-Q100 Grade 1 qualification + 125°C operating temperature + 10-year data retention = automotive- and aerospace-grade specifications. The 128Mb and 256Mb variants are in qualification with H2 2026 high-volume availability targeted.

"We are pleased to announce that during the fourth quarter, we ramped our PERSYST 64-megabit xSPI STT-MRAM high reliability product to full production and saw strong demand driven by new customer interest and design wins, specifically in the low earth orbital or LEO satellite market." — Sanjeev Aggarwal, CEO

Assessment: The xSPI HR full production is the operational milestone of FY25. The strategic value is that Everspin now has production-grade high-reliability STT-MRAM available for the most demanding applications (aerospace, defense, automotive, extreme industrial). The 128Mb/256Mb variants in H2 2026 are the next product-roadmap milestones. Revenue contribution from xSPI HR in FY26 is likely $2-4M; FY27 $6-10M.

3. UNISYST Family — 2027 Launch and $3B TAM Opportunity

The UNISYST family was reiterated as the structural multi-year growth lever. First product is a monolithic 256-megabit xSPI STT-MRAM device on a 16-nanometer FinFET node at TSMC, taped out in H2 2025. The product positioning: unified code-and-data memory architecture combining persistent storage + high-bandwidth read/write speeds + fast boot capability — a direct alternative to NOR Flash for high-density applications.

"We are 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. This part will be our first product in the UNISYST family, unifying code storage and data memory in a high-density, nonvolatile architecture for edge AI, industrial and mission-critical designs." — Sanjeev Aggarwal, CEO

The strategic positioning: UNISYST family targets the standalone NOR Flash market ($3B TAM), with Everspin targeting 5-10% share over multiple years post-launch. Target use cases: AI at the edge, military and aerospace, automotive, industrial, casino gaming. Engineering samples available Q4 2026; production 2027.

Assessment: UNISYST is the 2027+ revenue acceleration story. We model first meaningful UNISYST revenue contribution in late 2027 ($2-5M), scaling to $15-25M by FY29 as customer qualification cycles complete. The $3B TAM target at 5% share = $150M peak revenue opportunity — but realistically 3-5 years from launch.

4. $100M Revenue Target — 3-5 Year Horizon Reaffirmed

The CEO reaffirmed the long-standing $100M annual revenue target with a 3-5 year horizon. The composition the CEO outlined: (1) PERSYST family is the largest contributor (Toggle MRAM + xSPI STT-MRAM + IBM ST-DDR products); (2) Licensing portfolio (Microchip Foundry Services, Quintauris, future deals); (3) UNISYST family kicking in volume by 2027 with first contribution to the $100M.

"Before I close, I would like to discuss our long-term strategy, which entails reaching $100 million in annual revenue over the next 3 to 5 years. We believe this growth will be driven by the ramp of new products, most notably our new XY parts in our PERSYST product portfolio, such as the 64-megabit part I described earlier and continued solid growth in our Toggle MRAM and licensing business." — Sanjeev Aggarwal, CEO

The math: FY25 revenue ~$55M; FY28 target ~$100M = +82% over 3 years or ~22% CAGR. Achievable but requires both (a) product growth sustaining +20%+ YoY across the period, AND (b) licensing/non-product revenue contribution maintaining $5-7M annual. The first is plausible given the +34% design-win expansion in FY25; the second requires new contract wins replacing the legacy Frontgrade/QuickLogic/Purdue base.

Assessment: The $100M target is a meaningful multi-year north star but the path is gated by NOR Flash conversion conversion timing + UNISYST production ramp + new licensing contract wins. We model FY28 revenue at $80-95M (vs the $100M target) — below the high end of management's range but recognizing the structural trajectory.

5. Memory Shortage Industry Dynamic

The broader memory shortage industry context (DRAM allocation, NAND tightness, NOR Flash supply gap) was a featured topic in the prepared remarks. The CEO's framing positions Everspin as a beneficiary of multiple supply-chain dynamics: NOR Flash gap (most direct), broader memory cost pressure (drives customers to evaluate alternatives), and AI-driven demand growth (boosts the entire memory category).

"The industry is expecting — experiencing memory shortages. Memory suppliers who have for decades been pushed into commoditization have been — have seen a shift based on unprecedented memory shortages driven by the demands of AI. As a result, they have gone into allocation mode and are moving their capacity up the food chain." — Sanjeev Aggarwal, CEO

Assessment: The macro framing is favorable. Memory category tightness benefits Everspin both directly (NOR conversion) and indirectly (broader specialty-memory pricing power). We don't model specific revenue contribution from the macro shift, but treat it as structural tailwind through 2026-2027.

6. Cash Decline and Capex Pressure

The cash position declined $0.8M sequentially to $44.5M — the first sequential cash decline of FY25. Q4 operating CF of $2.8M was offset by capex of $3.6M (Chandler facility manufacturing improvements). The CFO's explicit Q1 26 commentary suggested continued capex pressure into the early heart of the Microchip Foundry Services Agreement.

"We had a unique period of capital spend. And that, again, was related to some of the improvements that we saw in the Chandler facility primarily across a couple of different contracts. So that flurry of activity, I think, will start to settle down until we get into the real heart of this Foundry Services Agreement... there will be some significant capital spend over the next 2 years. Again, it's going to be spread out over time a little bit, probably some later this year as well as early next year." — Bill Cooper, CFO

Assessment: The FY26 cash trajectory is more pressured than FY25. With DoD contract winding down (lower other income), patent litigation costs continuing at $1.6M/Q, capex elevated through 2026-2027 for Microchip Foundry Services, and operating CF still recovering — the cash position likely declines $2-4M during FY26 before stabilizing in FY27. Still healthy ($40-42M at FY26 exit), but the cash story is no longer "growing each quarter."

7. Patent Litigation — New $1.6M/Quarter Overhang

The CFO disclosed a new GAAP-to-non-GAAP reconciliation item: patent litigation costs, which are expected to continue at approximately $1.6M per quarter for "the next couple of quarters." Going forward, the company will exclude these costs from non-GAAP results — but they remain real GAAP-margin and cash-flow headwinds.

"We do show the $1.6 million that we had to expend in Q1 on litigation costs. And what I would say is, unfortunately, litigation is expensive, and I think we're kind of expecting it to continue in that range for at least the next couple of quarters. But again, we'll see how that ultimately pans out." — Bill Cooper, CFO

Assessment: $1.6M/quarter = $6.4M annualized — meaningful for a microcap with $55M revenue. The non-GAAP exclusion is structurally correct (litigation costs are non-recurring) but the GAAP impact is real. We model FY26 GAAP EPS to remain pressured by ~$0.20-0.25 per share from litigation alone; the GAAP-to-non-GAAP bridge widens materially through FY26.

8. Microchip PIC64-HPSC Aerospace-Defense Partnership

Everspin qualified its PERSYST 64-megabit xSPI STT-MRAM for Microchip's PIC64-HPSC series of 64-bit microprocessors. The partnership is positioned for aerospace and defense applications where MRAM's persistent + harsh-environment characteristics align with the MCU's target market.

This sits alongside the broader Microchip Foundry Services Agreement (10-year strategic partnership to establish an MRAM line at Microchip's Oregon fab, with first product shipments expected H2 2027). Two separate Microchip relationships: (1) the chip-level MCU + MRAM qualification (PIC64-HPSC); (2) the 10-year foundry services agreement to expand Everspin's onshore manufacturing capacity.

Assessment: The dual-track Microchip relationship is structurally important. PIC64-HPSC creates direct chip-level revenue opportunity in aerospace/defense; the foundry services agreement creates supply-chain redundancy that the U.S. government (a key customer) values. We don't model specific FY26 contribution from either, but treat both as structural multi-year revenue drivers.

9. Lattice and Microchip Co-Package Partner Programs Progressing

The two named partner programs the CEO is "really excited about" are Lattice Semiconductor and Microchip — both targeting MRAM as a companion solution to their core products. The Lattice partnership (xSPI/FPGA companion solution) was disclosed Q2 2025; progress is described as "steady" but multi-quarter to material revenue. The Microchip PIC64-HPSC qualification is more advanced — Everspin's parts have been formally qualified for the MCU ecosystem.

"There are 2 partner programs that we're really excited about, in particular. One is the one with Lattice and the other one is this Microchip. And I think we are seeing steady progress with both those partners, trying to get our product qualified and integrated into their standard offering. And I think that's where this PIC64 at Microchip comes into play. There — the markets that they are targeting align very well with Everspin in the aerospace and defense market." — Sanjeev Aggarwal, CEO

Assessment: Both partner programs are multi-quarter / multi-year revenue contributors. Material contribution from Lattice + Microchip co-package solutions is FY27+ — not a 2026 catalyst.

Guidance & Outlook

The Q1 2026 guide is the central reason the stock dropped -8.5% on T+1. Revenue at the high end of guide is fine; flat sequential is not. Non-GAAP EPS midpoint $0.10 is fine; the $1.6M litigation drag introducing new uncertainty is not. Cash declining for the first time in FY25 is not. The collective effect: the multi-quarter acceleration narrative has paused.

Implied FY26 trajectory: Q1 26 ~$14.5M + Q2 ~$15M + Q3 ~$15.5M + Q4 ~$16M = $61M FY26 revenue (+11% YoY from FY25 ~$55M). Non-GAAP EPS ~$0.40-0.45 FY26 (+90-110% from FY25 ~$0.21). Street at: Pre-print consensus FY26 ~$63-65M; post-print rebase to ~$58-62M expected. Guidance style: Pattern holds — high end of guide on revenue; non-GAAP EPS landing toward high end. But the sequential growth trajectory has flattened.

Analyst Q&A Highlights

NOR Flash Replacement Opportunity Timing

The opening question of Q&A pressed on the timing and sizing of the NOR Flash replacement opportunity — the most strategically important new disclosure of the call. The CEO's response was deliberately measured — opportunity is real, sizing is difficult, qualification cycles gate timing.

Q: "Great to hear about the NOR flash opportunity. I'm curious sort of — you're talking about that you're in conversations. I guess, sort of how fast or how quickly do you think you could see upside from that? And sort of if you could any way size the upside, that would be really helpful of possible upside in revenue."
— Neil Young, Needham & Company

A: "Like I said in the prepared remarks, I think it really depends on the qualification cycle for our potential customers. I can say that we are now getting listed as an alternate for NOR flash at various distributors worldwide because of the tight supply chain issues that we are seeing with NOR flash. So we do expect some upside, but it's very difficult to quantify today as to what that upside can be. But we are obviously available to meet the demand if the requests come in, and it just depends on the call cycle at the customers."
— Sanjeev Aggarwal, CEO

Assessment: The "distributor listings as NOR Flash alternate" is the cleanest leading indicator we have — it signals that the conversation has moved from theoretical to operational. Qualification cycles of 12-18 months gate the revenue, putting first conversions in FY27. We model $3-7M of NOR-replacement revenue contribution in FY27 and $10-20M+ in FY28.

Industrial Inventory and 2026 Demand Visibility

The follow-up question asked whether the industrial inventory normalization (called out in prior quarters) is now structurally resolved or whether risk remains for renewed inventory build. The CEO's response confirmed the structural recovery and signaled confidence in FY26 demand visibility.

Q: "Just one more question for me. You talked about the inventory levels in energy management and industrial automation. They're starting to look pretty healthy or you think they do look healthy at this point. I guess what gives you confidence that, that shouldn't be an issue next quarter and going forward?"
— Neil Young, Needham & Company

A: "Based on the backlog that we are seeing at our distributors and the forecast that we're seeing at our customers, we do feel that they have burned through the inventory that they had overbuilt over the last year or so. So we're pretty confident that going forward, we would not run into that same issue at least in 2026."
— Sanjeev Aggarwal, CEO

Assessment: The CEO's "at least in 2026" framing on industrial demand visibility is the cleanest forward signal we have on the industrial vertical. Combined with distributor backlog data, this supports a sustained industrial recovery through FY26. We model industrial automation as a steady +10-15% YoY contributor through 2026.

RAD-Hard Project + LEO Satellite Market Sizing

The Craig-Hallum analyst probed the LEO satellite sizing — how meaningful is the contribution today and how does it scale. The CEO acknowledged the LEO market is still "burgeoning" with confidence in product positioning but appropriate caution on quantification.

Q: "On the strategic RAD-Hard project you've been working with your partner, and this has talked about a much better year. I'm wondering if that's something similar that you're expecting as well. And then also this quarter and a couple of past ones, you've been talking about some increasing contributions from the LEO satellite market. Great to get a sense of how kind of — what's kind of the sense of scale of that today? And do you see that increasing over this year and over the next couple of years?"
— Richard Shannon, Craig-Hallum

A: "As far as the LEO satellite market is concerned, I'll let Bill address it. On the QuickLogic project, I think the award that QuickLogic talked about does not relate to the project that we've been working on jointly. And in fact, I think that's the project that Bill was talking about in his prepared remarks that is not going to renew in the near future, and we're going to see some decline in that non-product revenue in Q1 of this year. That is still waiting for some milestones to be met by the other partners in the program, and we expect that to be kicking in again towards the second half of this year, but not in the first half."

[Bill on LEO]: "Yes, I would say that the LEO satellite market is still that burgeoning market, and that's what we see both in terms of our orders and our backlog, and we feel confident about our products and our position there. And we expect to again kind of move up with that increased demand, especially as we've introduced our high reliability products as well that sort of fits perfectly in that particular market space as well."
— Sanjeev Aggarwal, CEO & Bill Cooper, CFO

Assessment: Two important data points emerge: (1) QuickLogic-related project revenue is not renewing in the near future — translating to the Q1 26 non-product revenue decline the CFO disclosed; (2) LEO satellite contribution is growing but still "burgeoning" — material but not yet a primary revenue driver. The combination explains why Q1 26 product growth needs to fully offset the licensing decline to hold consolidated revenue flat sequentially.

NOR Flash Partner Programs and Microchip PIC64

The Q&A continued on the NOR Flash replacement theme, specifically asking about Microchip's PIC64-HPSC qualification and the Lattice partnership update. The CEO's response confirmed both partner programs are progressing steadily with Microchip's aerospace/defense market alignment particularly close to Everspin's existing strategic positioning.

Q: "On your NOR flash replacement products and maybe taking a different angle than one of the last questions here and also your very interesting comments. But obviously, you've been targeting NOR flash replacement in certain markets. And if I caught your comments right, Sanjeev, you're talking about, I think, a win with — I think it was Microchip on an MCU. I'd love to get a sense of when you see that becoming a material contributor."
— Richard Shannon, Craig-Hallum

A: "There are 2 partner programs that we're really excited about, in particular. One is the one with Lattice and the other one is this Microchip. And I think we are seeing steady progress with both those partners, trying to get our product qualified and integrated into their standard offering. And I think that's where this PIC64 at Microchip comes into play. There — the markets that they are targeting align very well with Everspin in the aerospace and defense market. And then I think we can then expect to take into the commercial market as well. But right now, the PIC64 is targeted towards the aerospace and defense as well."
— Sanjeev Aggarwal, CEO

Assessment: Both partner programs are aerospace/defense first, commercial second. The strategic alignment to Everspin's existing high-reliability MRAM positioning is clean. Revenue contribution is multi-quarter / multi-year — likely first material contribution in FY27.

$100M Revenue Target Composition

The closing analyst question explicitly sized out the $100M revenue target — what contributes most? The CEO's response provided the cleanest decomposition we have heard, naming PERSYST (Toggle + STT-MRAM + IBM ST-DDR) as the largest contributor, with UNISYST and licensing as roughly equal secondary contributors.

Q: "Sanjeev, you talked about a goal of driving towards or driving to $100 million of revenues within, I think, 3 to 5 years. I'd love to get a sense any way you quantify or at least rank order of the contributors of that revenue. I think the big picture here that I think of this as the toggle, the the STT products and licensing, I guess, if there's any other way you'd categorize the contributors, that would be very helpful."
— Richard Shannon, Craig-Hallum

A: "The way I look at it is the major contributor is going to be the PERSYST products, followed by perhaps equal contributions from our licensing and UNISYST in 3 to 5 years down the road towards the $100 million. In the PERSYST, I'm including the Toggle MRAM market products as well as the xSPI STT products that we are shipping today as well as the ST-DDR products that we're shipping to IBM. So I think those 3 would form the portion of the HR — from the portion of the PERSYST products that are going to contribute. And this high reliability product family that we have just introduced is going to actually give us a nice boost over there. And then in 2027, we expect UNISYST to kick in some volume. And so I think between that, UNISYST and our licensing is what's going to get us to the $100 million mark down the road."
— Sanjeev Aggarwal, CEO

Assessment: The composition math is illuminating. Implied: PERSYST family ~$65-70M (~65-70% of total); UNISYST ~$15M; Licensing ~$15M. This implies FY28/29 PERSYST contribution growing from FY25's ~$48M to $65-70M = +35-46% over 3-4 years = ~10-13% CAGR. Plausible given the design-win pipeline + xSPI HR contribution + IBM FCM5 contribution + NOR Flash conversion revenue. UNISYST at $15M assumes ramped to ~$15M annualized by FY28, which requires production in 2027 + 18-24 month qualification cycle = realistic.

What They're NOT Saying

1. FY26 revenue or EPS guidance: Management continues to defer formal FY26 guidance. Only Q1 26 was guided. The implicit FY26 trajectory ($58-62M) is derivable but not committed.
2. Specific NOR Flash conversion conversation count or pipeline value: "In conversations with customers" disclosed; no count, no dollar pipeline value, no expected conversion timing. This is intentionally vague.
3. Specific xSPI HR Q1 26 revenue contribution: "Full production" began in Q4; "began shipping in the current quarter" disclosed. No Q1 26 revenue contribution dollar value.
4. Patent litigation specifics: Just the $1.6M/Q cost disclosed. No information on (a) plaintiff/defendant identity, (b) nature of the litigation, (c) expected resolution timeline beyond "next couple of quarters."
5. 238 FY25 design-win revenue conversion timeline: 238 wins disclosed; 18-24 month conversion is standard but per-cohort revenue conversion not disclosed.
6. Microchip Foundry Services Agreement capex magnitude: The CFO confirmed "significant capital spend over the next 2 years" — no dollar total disclosed.
7. UNISYST 256Mb 16nm TSMC tape-out cost: The H2 2025 tape-out at TSMC is a meaningful R&D + NRE expenditure — no dollar disclosed.
8. Chiplet ecosystem partnership economics: Multiple partnerships (Fraunhofer, IMEC, PACE) disclosed; no revenue contribution sized for any.
9. Q2 26 outlook beyond Q1: No commentary on Q2 sequencing or seasonal patterns for FY26.
10. Insider activity or capital allocation: No buyback framework discussed despite $44.5M cash position. Standard microcap reinvestment posture but unaddressed.

Market Reaction

• Pre-print setup (Mar 3 close): $10.61, down -20% from Feb 2 peak of $13.24 (30-day pre-print). YTD from Jan 1 2026 (~$10ish baseline) approximately flat to slightly up. Positioning was net-long but materially de-risked from the early February peak. Average daily volume 350-500K (elevated from Q3's 200-300K).
• Day-of action (Mar 4 — print released after-market): Regular hours close $10.79 (+0.37%) on volume of 755K (~2x trailing 30-day average). Modest pre-print buying anticipation. Print released after-market at ~4:05 PM ET; initial after-hours reaction was mixed.
• T+1 (Mar 5): Open $9.97 (gap DOWN -7.6% from Mar 4 close); intraday high $10.01; intraday low $8.91; close $9.12 (-8.53% reported; -15.5% from Mar 4 intraday high). Volume 1,556,614 shares (~4x trailing 30-day average) — heaviest single-day volume in 6+ months.
• T+2 (Mar 6): Open $8.70; close $8.76 (+0.69% — modest bounce, no recovery). Volume 740K. Stock stabilized at new lower level.
• Sell-side reaction: Craig-Hallum (Shannon) maintained Hold with PT lowered (~$11); Needham (Young) maintained Hold. Two PT cuts in aggregate; no upgrades.
• Peer reaction: SOXX down modestly on macro concerns; no clear sector read-through to MRAM.

The Q4 reaction pattern reveals the central tension of the print: the operational quarter was strong (product +22% YoY; 238 FY25 design wins; xSPI HR ramped; NOR Flash tailwind newly visible) but the forward setup is muted (Q1 flat sequential; DoD Q4 light; cash declined; litigation overhang). The market priced the forward more heavily than the operational quarter — appropriate when the FY26 H1 trajectory looks compressed vs the FY25 H2 acceleration arc.

The 4x volume on T+1 signals institutional repositioning, not retail flow. The stabilization at $8.76 on T+2 suggests the market has settled on a new equilibrium pricing the FY26 H1 caution — with FY26 H2 reacceleration (Q4 26 product growth recovery; potential NOR conversion contribution) as the optionality lever.

Street Perspective

Debate: Is the Q1 26 flat sequential guide a structural deceleration or a one-quarter pause?

Bull view: Q1 is mechanical — non-product revenue declining $0.5-0.7M as QuickLogic-related project completes; product revenue continuing to grow $0.5-0.8M sequentially. The decline in non-product is well-flagged (CFO explicit) and partially recoverable in H2 26 as the partner-program milestones unlock. Q2 26 should resume +5%+ sequential growth on continued product strength. The pattern is "lumpy quarter, not structural deceleration."

Bear view: The Q1 guide is the structural reset to a lower growth trajectory. Through FY25, sequential revenue growth accelerated (+1% Q2 → +7% Q3 → +5% Q4); the Q1 26 flat sequential breaks the pattern. Combined with the DoD contract winding down through 2026-2027 and the licensing line structurally below FY24 levels, the consolidated revenue growth rate is moderating to high-single-digits — not the mid-teens that the bull case requires for the $100M revenue target to be achievable in 3-5 years.

Our take: Probability is ~60/40 bull vs bear. The product growth +22% YoY across two consecutive quarters supports the bull operational case; the licensing/non-product decline is real and not fully offset by product. We model FY26 revenue at $59-62M (+8-13% YoY) — below the high end of bull-case expectations but above the bear-case stagnation thesis.

Debate: Is the NOR Flash conversion thesis material enough to drive a re-rating?

Bull view: The NOR Flash supply allocation dynamic is structurally durable — major NOR suppliers (Macronix, Winbond, others) are unlikely to reverse capacity reallocation as long as DRAM/HBM demand remains strong (multi-year). Even partial conversion to MRAM (5-10% of NOR demand shifting) represents $150-300M of TAM expansion for the specialty memory category. Everspin's xSPI STT-MRAM portfolio is the most-direct NOR Flash replacement available today. Distributor listings as "NOR alternate" is the cleanest leading indicator that conversion is starting. By FY28-29, $20-40M of incremental Everspin revenue from NOR conversion is plausible.

Bear view: NOR Flash to MRAM conversion is gated by qualification cycles (12-18 months) AND meaningful unit-cost premiums (MRAM is structurally 2-5x the cost of equivalent NOR Flash density). Most NOR Flash customers will tolerate supply allocation longer than they will tolerate a 2-5x cost increase. The conversion conversations Everspin is having are likely high-reliability subset of NOR customers — a smaller TAM than the headline $3B suggests. Realistic incremental Everspin contribution: $5-15M annual by FY28.

Our take: Both sides have merit. The structural opportunity is real; the magnitude is uncertain. We model $5-10M of incremental NOR-replacement revenue by FY27 and $15-25M by FY29 — sufficient to be a thesis driver but not sufficient to fully drive a multi-year re-rating without complementary catalysts (UNISYST production, design-win conversion, DoD follow-on contracts).

Debate: Is microcap balance-sheet integrity becoming a thesis risk?

Bull view: $44.5M cash is still 30%+ of market cap; the $0.8M sequential decline is a single-quarter capex-timing artifact. Q1 26 will see another capex-heavy quarter (Microchip Foundry Services + Chandler facility), but operating cash generation should sustain at $2-4M quarterly and the FY26 exit cash position should remain $40M+. The balance-sheet integrity remains the structural defensive feature; we are not anywhere close to dilution or distress.

Bear view: The cash trajectory inflected this quarter. Through FY25 cash was building (~$2.4M for the year); the Q4 $0.8M decline + Q1 26 expected capex-heavy quarter + $1.6M/Q litigation cost + lower DoD other income = FY26 cash could decline $4-6M cumulatively. At $39-40M FY26 exit, the balance-sheet defensive margin compresses. If Microchip Foundry Services capex extends beyond 2026 into 2027, the cash drawdown could continue. Microcap balance-sheet pressure is the structural risk that emerges in Q4 for the first time.

Our take: The bull view is more probable but the bear view warrants monitoring. We model FY26 cash exit at $40-42M — down from FY25's $44.5M but still healthy. The capex curve through 2026-2027 is the key variable; if Microchip Foundry Services capex is meaningfully larger than current expectations, the cash trajectory worsens. We flag this as the new structural risk that emerged on this print.

Model Update Needed

Valuation impact: At Mar 5 close of $9.12 and ~23.8M diluted shares, market cap is ~$217M; cash $44.5M, debt $0, EV ~$172M. On FY26E revenue of $60M and non-GAAP EPS of $0.28 ($6.7M net income), MRAM trades at 2.87x EV/Sales and ~26x EV/NI. On FY27E revenue of $75M and non-GAAP EPS of $0.40 ($9.5M net income), MRAM trades at 2.29x EV/Sales and ~18x EV/NI. The valuation has compressed materially from Q3's print-day peak ($10.53; 2.9x EV/Sales) reflecting both the price decline and the modest model revisions. Our 12-month fair-value estimate: $9-11 (held vs Q3) — the operational case is intact; the multiple is appropriately rebased.

Thesis Scorecard Post-Earnings

Overall: Q4 is a "mixed signals" quarter. The structural product franchise is healthy and accelerating; the design-win pipeline is expanding meaningfully; the NOR Flash supply shock is the highest-quality new forward catalyst we've seen. But the Q1 26 setup is materially more cautious than the FY25 H2 acceleration suggested: flat sequential revenue, DoD wind-down, litigation cost overhang, and cash decline all introduced new pressure points.

Action: Maintaining Hold. Third consecutive Hold call. The thesis hasn't broken — it has paused. The FY26 H1 cadence (Q1 26 print March 4; Q2 26 print early August) is the critical test of whether the multi-quarter acceleration resumes or the deceleration solidifies. Upgrade trigger: Q1 26 prints at $15M+ revenue with operating cash positive, AND Q2 26 product growth re-accelerates to +10%+ QoQ, AND NOR Flash conversion conversations translate to first orders. Downgrade trigger: Q1 26 revenue below $13.5M OR cash position declines below $42M OR FY26 design-win pace falls below 50/quarter.