Revenue +24% YoY at Guide High End, Non-GAAP Profitable, $45M Cash — Initiating at Hold on Design-Win-Ramp Timing

Results vs. Consensus

Year-over-Year Comparison

Quarter-over-Quarter Comparison

Revenue Assessment

The $13.2M print at the guide high end is the third consecutive quarter Everspin has hit the top of its self-guided range, signaling the company has converged on a guidance-discipline pattern that we can model with reasonable confidence. The +24.5% YoY headline is flattered by an easy comp ($10.6M Q2'24 included some inventory-correction softness in industrial customers); the underlying organic-growth rate is closer to +12% on the MRAM product line. That product growth is structurally driven by three independent vectors: (1) data center / RAID with broad customer set (Dell, Supermicro, Broadcom), (2) industrial automation PLC recovery as inventory destocking in Asia completes, and (3) the steady-state IBM FlashCore Module 4 PERSYST 1Gb STT-MRAM contribution. The licensing/royalty/patent/other line's surge to $2.1M is concentrated in the Purdue contract reaching steady-state delivery + Frontgrade Phase 1 completion + QuickLogic ramp.

The Q3 guide of $13.5-14.5M (midpoint $14.0M; +6% QoQ at midpoint) tells us the company is modeling continued sequential MRAM product growth with a modest tailwind from the DoD contract picking up. Management's "second-half-weighted" framing is consistent across quarters; we model FY25 at $54-56M total revenue (vs FY24 $51M), implying ~$13.5-14M H1 + ~$14.5-15M H2 average quarterly run-rate.

Margin Assessment

GAAP gross margin of 51.3% (+230bps YoY) is the structural surprise. The driver per the CFO is "a larger mix of high-margin licensing and other revenue" — the $2.1M licensing line carries materially higher gross margin than the $11.1M MRAM product line (estimated product GM ~46-48% based on prior-quarter disclosures). The risk going forward is that licensing is structurally lumpy (Purdue contract is finite; QuickLogic project completes in ~2 years; Frontgrade Phase 1 done with Phase 2 contingent on government renewal). Management's deliberately-cautious 45-50% structural product-margin commentary signals they don't expect to repeat 51%+ consistently.

Operating expenses of $8.7M (flat QoQ; +9% YoY) reflect deliberate investment in sales and business development (the newly added VP of Sales role) and continued R&D investment in the xSPI STT-MRAM product family. The OpEx growth rate is below the revenue growth rate — operating leverage is positive but modest. Q3 EBITDA-margin guidance implied by the EPS range works out to ~12-15% non-GAAP — a low-double-digit margin profile that does not yet justify a software multiple but does support a "premium-specialty semi" multiple if the growth rate sustains.

EPS & Cash Flow Assessment

Non-GAAP EPS of $0.03 (at the guide high end of breakeven-to-$0.05) reflects the operating-leverage benefit of higher revenue against a controlled OpEx base. The GAAP net loss of $(0.7)M / $(0.03) EPS narrows materially from $(2.5)M / $(0.11) Q2'24, with SBC ($1.4M Q2) the primary GAAP-to-non-GAAP reconciliation item. Diluted weighted-average shares of 22.6M is up ~4% YoY on continued SBC dilution — typical for a microcap semiconductor company in the growth phase. Cash flow from operations of $5.0M is the single most encouraging operating metric — it shows the business can self-fund growth without dilution or debt, and supports continued reinvestment in product development.

The $45M cash position is the structural defensive feature of the investment case. At ~30-35% of market cap (depending on the trading price), Everspin has multi-year runway to execute on the product-roadmap thesis even if revenue softens. The debt-free balance sheet eliminates capital-structure tail risk that would otherwise be a meaningful overhang for a microcap of this size. The CFO's explicit framing of cash as "an effective hedge against macroeconomic uncertainty" is strategically correct — the company is not pursuing aggressive growth-at-any-cost; it is preserving optionality.

Segment Performance

MRAM Product Sales — $11.1M, +12% YoY, +1% QoQ

The MRAM product business is the structural growth engine, currently $44M+ annualized and growing low-to-mid teens YoY. The end-market mix is diversifying: data center (RAID + IBM FCM4) is the largest single contributor; industrial automation is the fastest-recovering sub-segment in Q2; aerospace/LEO is the highest-growth-rate emerging vertical. The CEO's framing of "depletion of inventories at our customers, almost all across the world, specifically in the Asian region" is the cleanest forward indicator — industrial inventory normalization in Asia is the leading trigger for Everspin's industrial-customer order patterns.

The Lucid Motors Gravity SUV PERSYST MRAM deployment is the highest-visibility auto-design-win, but the revenue contribution is currently small (Lucid Gravity is a low-volume premium SUV; per-vehicle MRAM content is in the $5-20 range). Management's framing "expect volumes to increase as the automaker ramps production" is appropriately measured. The IBM FlashCore Module 4 PERSYST 1Gb STT-MRAM contribution is the largest single design win in the company's STT-MRAM portfolio — but is "steady state" and not expected to grow further in 2025, suggesting H2 2025 growth will need to come from new product introductions (xSPI HR variants) and continued Toggle-MRAM data-center momentum.

"The good news for Everspin from Everspin's point of view is that we are actually seeing depletion of inventories at our customers, almost all across the world, specifically in the Asian region, where we were concerned about inventory overbuild. So I think we are actually seeing some good number of orders come in from that region. So that's good." — Sanjeev Aggarwal, CEO

Assessment: The MRAM product business is steady but unspectacular. The +12% YoY product growth is consistent with what an established niche memory franchise should produce in a mid-cycle environment. The structural acceleration thesis depends on (a) the new xSPI HR products reaching full production in late 2025, (b) the UNISYST family launching in 2027 to address the NOR Flash $3B TAM, and (c) the DoD contract milestones unlocking additional revenue. None of these is reflected in the Q2 product number; all are forward catalysts.

Licensing, Royalty, Patent & Other — $2.1M, +200% YoY, +91% QoQ

The licensing line is the surprise contributor of the quarter, more than tripling YoY and nearly doubling QoQ. The composition disclosed: Purdue University contract (STT-MRAM technology for energy-efficient AI applications) reaching steady-state delivery; Frontgrade project Phase 1 successfully completed and recognized; QuickLogic AgILYST MRAM technology contract continuing on milestone delivery; and a TMR sensor foundry services project for an unnamed leading provider continuing through the Chandler facility.

The structural caveat is that licensing revenue is project-based and lumpy. The Purdue contract has a defined endpoint (continuing development of low-power magnetic tunnel junction devices); the Frontgrade Phase 1 is complete with Phase 2 contingent on AFRL government funding renewal; the QuickLogic project has roughly 2 years of additional deliverables. We model the licensing line at $1.5-2.0M per quarter average through FY25, with structural decay risk into FY26 unless new contracts replace expiring ones.

"We completed the first phase of the front grade project successfully meeting all our deliverables in Q2. During this phase, we recognized revenue related to delivering the process design kit or PDK. This contract allows for the award of future optional phases based upon successful performance of all parties contributing to this phase and at the discretion of the U.S. government." — Sanjeev Aggarwal, CEO

Assessment: Licensing is structurally strategic but financially lumpy. The +200% YoY is not a sustainable growth rate; we treat the steady-state contribution as $1.5-2M per quarter ($6-8M annualized) with upside if new licensing deals (e.g., the Microchip relationship building, the Lattice xSPI/FPGA companion solution evaluation) convert to revenue. The line provides gross-margin uplift to the consolidated business and is the cleanest demonstration that Everspin's IP has third-party value.

Defense / DoD Sustainment Contract — $0.8M other income (cumulative $7.4M of $14.6M)

The $14.6M DoD contract awarded in August 2024 to develop a long-term sustainment plan for Everspin's MRAM manufacturing facilities is the most important multi-quarter "below-the-line" revenue lever. The contract is structured as "other income" (not above-the-line revenue), reflecting government-services accounting treatment. The Q2 contribution was $0.8M; cumulative through Q2 is $7.4M (51% of the contract value). Management's explicit commentary on the Q2 call: "We expect this business to pick up meaningfully in the fourth quarter." Translation: the remaining $7.2M of contract value is back-half-loaded.

"We recognized $0.8 million in other income in the second quarter and $7.4 million to date from the $14.6 million contract we have with the 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 pick up meaningfully in the fourth quarter." — Sanjeev Aggarwal, CEO

Assessment: The DoD contract is the highest-visibility forward EPS lever for FY25. If Q4 recognizes $3M+ of additional contract value (consistent with Q4'24's $3M+ recognition rate), the consolidated FY25 EPS picture brightens materially. The contract also serves as strategic validation — the U.S. government has identified Everspin as a domestic MRAM manufacturing supply-chain priority, which carries long-term implications for follow-on contracts (the $40M FY26 subcontractor agreement disclosed in April 2026, per management's later disclosures, is one such follow-on).

Key Topics & Management Commentary

Overall Management Tone: Management's tone was measured and consistent — the CEO and CFO both used the framing "consistent financial performance reflects the strength of our product portfolio and ramping design wins" without hyperbole. The forward-looking commentary was deliberately conservative on structural margins (45-50% range despite 51%+ actuals), on the back-half FY25 acceleration (committed but not quantified), and on tariff exposure (no impact observed; no material future impact expected). The areas of greatest forward-leaning conviction were the design-win pipeline (multiple end-markets cited with named customer references), the new VP of Sales hire (Sean Dougherty from Intel signaling intent to accelerate revenue monetization), and the DoD sustainment contract's Q4 acceleration commitment. There were no defensive callouts or downbeat commentary — the call was professionally calm and execution-focused.

1. Data Center Toggle MRAM Strength in RAID Memory

Data center RAID memory was called out as the largest single growth contributor in Q2, with high-single-digit sequential growth driven by demand from a broad set of customers including Dell, Supermicro, and Broadcom. The CEO's framing emphasized the breadth — this is not a single-customer concentration story but a structural product-category strength.

"We saw high single-digit sequential growth in the data center business. This growth was driven by strong demand on the redundant array of independent disks or RAID from a broad selection of data center customers, including Dell, Supermicro and others." — Sanjeev Aggarwal, CEO

The strategic significance of RAID-card MRAM deployment is twofold: (a) data-center storage build-outs are a secular growth area independent of macro cyclicality (AI infrastructure buildout in particular drives RAID-card demand), and (b) the Toggle MRAM franchise is the mature, profitable, repeatable revenue base that funds Everspin's STT-MRAM development. Toggle is the cash-cow that subsidizes the bet on STT-MRAM as the future.

Assessment: The RAID-card data-center business is the most structurally stable revenue line at Everspin. We model it as a low-single-digit-growth franchise tied to data-center storage hardware build cycles. Not the bull-case driver, but the load-bearing floor.

2. Industrial Automation Recovery on Asia Inventory Normalization

Industrial automation (PLCs and similar automation equipment) grew >20% sequentially in Q2, driven by inventory destocking completing at customers — particularly in Asia, where Everspin had been flagging concerns about inventory overbuild through 2024 and early 2025. The CEO's "depletion of inventories at our customers, almost all across the world" framing is the cleanest macro datapoint on this business in many quarters.

"We saw good momentum from our customers who build industrial automation equipment like programmable logic controllers or PLCs, with sequential growth in excess of 20% from the first quarter. Everspin has a significant historical business here, and we are seeing momentum from current customers and recent design wins with our industrial xSPI product." — Sanjeev Aggarwal, CEO

The industrial-automation customer base is structurally fragmented (no single dominant customer; rather, many mid-sized OEMs across automotive parts, factory robotics, energy management, and HVAC). This means Everspin's industrial business is a leveraged play on global industrial-capex cycles. The Q2 +20% sequential is the strongest single recovery quarter we have observed.

Assessment: Industrial recovery is genuinely positive. If sustained through Q3-Q4, it represents the cleanest above-trend growth lever in the FY25 picture. We monitor distributor inventory data (typically reported by tier-2 industrial distributors) for confirmation through Q3-Q4 cadence.

3. LEO Satellite Aerospace Applications — Engineering Samples Shipped, Late 2025 Production

The Low Earth Orbit (LEO) satellite market is the highest-growth-rate emerging vertical for Everspin. Q2 saw engineering samples shipped for the two new high-reliability xSPI products announced in Q1: PERSYST EM064LX HR (64Mb) and PERSYST EM128LX HR (128Mb), featuring expanded temperature range for aerospace/defense and "extreme industrial" environments. Full-production ramp target is late 2025.

"We shipped engineering samples for the 2 new products we announced last quarter as part of our xSPI family, the PERSYST EM064LX HR and EM128LX HR. These parts feature an expanded temperature range to address the growing demand for persistent high-speed memory in aerospace, defense and extreme industrial environments and provide designers with a robust, fast and scalable alternative to static RAM or NOR Flash. We remain on track to ramp to full production in late 2025." — Sanjeev Aggarwal, CEO

The strategic positioning: LEO satellites have short lifecycles (3-5 years due to atmospheric drag), with multiple MRAM parts per satellite. As the LEO market scales (SpaceX Starlink, Amazon Kuiper, Eutelsat OneWeb, plus smaller commercial constellations), the unit-volume implications for Everspin's high-reliability xSPI portfolio are meaningful. Management has previously disclosed design wins with Astro Digital and Blue Origin in this segment.

Assessment: LEO aerospace is the longest-runway revenue lever in Everspin's portfolio. The 2025 contribution is small (engineering samples → late-2025 production); the 2026-2028 contribution could be material if LEO satellite deployments compound at projected rates. Watch item: production-ramp pacing in Q4 2025 / Q1 2026.

4. IBM FlashCore Module 4 PERSYST 1Gb STT-MRAM — Steady State

The IBM FlashCore Module 4 deployment (PERSYST 1Gb STT-MRAM in IBM's high-end storage products) reached steady-state contribution in Q2 and is expected to remain at that level through year-end. This is the largest single STT-MRAM design win the company has disclosed publicly.

"During the second quarter, we reached a steady state of revenue from the sale of our PERSYST 1 gigabit STT-MRAM into IBM's FlashCore Module 4 or FCM4 for data center applications and continue to anticipate product revenue from this ongoing project to remain consistent for the remainder of the year." — Sanjeev Aggarwal, CEO

The strategic importance is twofold: (a) IBM's continued specification of PERSYST 1Gb in mission-critical enterprise storage products validates Everspin's STT-MRAM as production-ready for tier-1 customers, and (b) the "steady state" disclosure tells us this is a multi-year revenue contributor at a roughly predictable run-rate (we estimate $1.5-2.5M per quarter; not separately disclosed).

Assessment: The IBM deployment is one of Everspin's marquee customer references — tier-1 hyperscaler-adjacent enterprise storage. The steady-state disclosure suggests no near-term upside from the existing FCM4 deployment, but follow-on FCM5 / FCM6 design wins (referenced in later quarters) represent the natural upgrade path. The reference value alone is worth more than the current revenue contribution.

5. Lucid Motors Gravity SUV — Automotive Design Win Ramping

The Lucid Motors Gravity SUV deployment using Everspin's PERSYST MRAM is the highest-visibility automotive design win. Q2 saw continued shipments with management expecting volume to "increase as the automaker ramps production." Lucid's production trajectory has been below initial expectations (Gravity is a premium SUV at $80-100K+), but the per-unit MRAM content + the validation value of an automotive OEM-spec design win is strategically meaningful.

"We continue to ship and recognize revenue for our PERSYST MRAM solution from Lucid Motors for the Gravity SUV and expect volumes to increase as the automaker ramps production." — Sanjeev Aggarwal, CEO

Assessment: Lucid Gravity is a strategic-validation win more than a revenue-driver. Per-unit MRAM content is small; Lucid's production pace is modest. The value is in the design-win reference for other automotive OEMs evaluating MRAM for their next-generation vehicle architectures. Watch item: any new automotive OEM design-win disclosures in Q3-Q4.

6. DoD Sustainment Contract — Q4 Acceleration Committed

The $14.6M DoD contract (awarded August 2024 to develop a long-term plan for providing onshore MRAM manufacturing services to aerospace and defense customers) has $7.4M cumulative recognized through Q2, with management explicitly committing to "meaningful pick-up in Q4." The Q2 contribution was $0.8M; Q4 2024 was the highest-recognition quarter at $3M+. The pattern suggests Q3 will be modest ($1-1.5M) with Q4 ramp to $3-4M.

"We had a strong back half last year. We'll expect to see a strong second half and particularly more toward Q4 as well." — Bill Cooper, CFO

The strategic significance of this contract goes beyond the dollar value. It positions Everspin as a U.S. government-designated domestic MRAM supplier for aerospace and defense applications — a position that has compounded into multiple follow-on opportunities (the Microchip Foundry Services Agreement in 2026, the $40M defense-prime subcontractor agreement disclosed in April 2026). The DoD relationship is the structural moat-builder that's compounding behind the scenes.

Assessment: The Q4 DoD acceleration is the central back-half FY25 catalyst. If the company recognizes $3M+ in Q4 (vs $0.8M Q2), the consolidated EPS picture for FY25 improves materially. Watch item: Q3 trajectory will tell us whether Q4 is the steep ramp or a more measured curve.

7. Sales Team Expansion — New VP of Sales (Sean Dougherty from Intel)

Everspin announced the hire of Sean Dougherty as dedicated VP of Sales, with David Schrenk transitioning to focus exclusively on Business Development. This is the most meaningful executive-team change Everspin has disclosed in multiple quarters and signals intent to accelerate revenue monetization on the existing design-win pipeline.

"In order to ensure that we meet the future demand for our products, we are expanding our executive team with the addition of a dedicated VP of Sales, Sean Dougherty, who joined us recently from Intel. With Sean's addition, David Schrenk, who has been our VP of Sales and Business Development for the last 3 years, will be able to focus his efforts exclusively on business development." — Sanjeev Aggarwal, CEO

The strategic split — sales execution vs business development — is the right architecture for a company that has 178+ design wins (per FY24 disclosure) but limited revenue conversion. The role separation should drive better focus on (a) closing existing pipeline opportunities and (b) developing new strategic partnerships and licensing deals.

Assessment: Executive-team additions in microcap semis often correlate with revenue acceleration 2-4 quarters out. The Dougherty hire is too new to evaluate operationally in Q2; we monitor design-win conversion data in Q3-Q4 for early efficacy signals.

8. Frontgrade / QuickLogic / Purdue Licensing Portfolio

The licensing/services portfolio comprises three named active contracts and one unnamed foundry services project: (1) Frontgrade Phase 1 — completed Q2, awaiting AFRL renewal for Phase 2; (2) QuickLogic AgILYST MRAM technology — ongoing, ~2 years remaining; (3) Purdue University STT-MRAM for energy-efficient AI — steady state; (4) TMR sensor foundry services for unnamed leading sensor provider — ongoing in Chandler facility.

The Frontgrade project is particularly strategic — it focuses on enabling embedded radiation-hard STT-MRAM macros for aerospace applications, which is the technology foundation for future LEO satellite + defense system memory deployments. Phase 1 completion in Q2 was a significant milestone; Phase 2 renewal depends on AFRL funding decisions.

Assessment: The licensing portfolio collectively contributes $1.5-2M per quarter at the steady-state run rate, with structural decay risk as projects complete. The strategic value (IP validation, customer relationships, defense-supply-chain positioning) materially exceeds the revenue value. Watch item: AFRL renewal of Frontgrade Phase 2.

Guidance & Outlook

The Q3 guide implies modest sequential growth driven by (a) industrial automation continuing the Q2 recovery, (b) LEO satellite engineering-sample-to-production transition beginning, (c) DoD sustainment contract picking up modestly, and (d) licensing line normalizing from the Q2 Purdue-spike high. The non-GAAP EPS range of $0.02-0.07 spans a wide range, reflecting uncertainty on gross-margin mix and OpEx timing — typical for a microcap with project-based revenue.

Implied FY25 trajectory: Q1 $13.1M + Q2 $13.2M + Q3 ~$14M + Q4 ~$14.5M = $54.8M FY25 revenue (+7% YoY from FY24 $51M). Non-GAAP EPS ~$0.10-0.15 FY25 (vs ~$0.00 FY24 estimate). Street at: Consensus is thin given limited coverage; we model Q3 at $14.0M and Q4 at $14.7M with Q4 DoD acceleration. Guidance style: Everspin has guided conservatively under this management team — three consecutive quarters at the high end of revenue guide; non-GAAP EPS landing at or near the high end of the range. Pattern suggests Q3 will likely print at $14.0-14.3M with non-GAAP EPS of $0.05-0.07.

Analyst Q&A Highlights

Product Gross Margin Trajectory and the Path to 50%+

The opening question of Q&A asked about product gross margin dynamics in the quarter — specifically whether the modest sequential decline reflected yield or mix issues. The CFO's response was deliberately measured, anchoring expectations at 45-50% as the structural product gross margin range despite Q1/Q2 actuals at 51%+.

Q: "I'd love to ask about product gross margins in the quarter. If I did my numbers right, it seems like they're in line or a little bit lower than last quarter and kind of a bit lower than maybe what the trend we've seen in the last several quarters here. So I want to get a sense of whether there's any yield or mix dynamics going on in here... Is there a path or a view to getting those product gross margins up above 50% anytime soon?"
— Richard Shannon, Craig-Hallum Capital Group

A: "They're earlier in the life cycle. So we're always looking for ways to improve those gross margins and working with the foundries and folks in manufacturing to improve — continuous improvement on the yields... that is our target. We expect to be north of 50% in total for product gross margins. But as always, those things take a little bit of time. So we — but we do expect to be solidly in that 45% to 50-ish percent margin range for products."
— Bill Cooper, CFO

Assessment: The CFO's deliberately-cautious 45-50% framing — vs Q1/Q2 actuals at 51%+ — is consistent with management's overall conservative posture. The implication is that the consolidated gross margin uplift is partly licensing-mix-driven and partly yield-driven; as licensing normalizes (next 1-2 quarters), GAAP gross margin could compress toward the 49-50% range. We model FY25 GAAP gross margin at 50%.

New Product Ramp Trajectory and FY26 Visibility

The follow-up question pressed on the new xSPI HR product family ramp — specifically whether the late-2025 production ramp would translate to material revenue contribution in 2026. Management's response noted that Q2 industrial-automation +20% sequential growth was partly attributable to the xSPI family already, and that incremental Q3-Q4 contribution is implicit in the guide.

Q: "I'd love to ask about kind of progress in the new product area here. I think you mentioned a couple of brief comments in this area, but it doesn't sound like we should expect a notable increase here as we exit the year. I think the phrase I heard used was going to full volume production with these products late this year. So if I caught that right, maybe you can help us understand what kind of contributions we should look for as we get into next year."
— Richard Shannon, Craig-Hallum

A: "On those newer products, right, they are ramping, and they are kind of in the burgeoning aerospace segment. That's one of the key segments, key industries for that area. And we definitely are starting to see some uptick and again, some good traction on some of the newer parts and products. So I think we're moving along healthily with those products... we did see a significant sequential growth, right, almost 20% compared to Q1. And a good portion of that is actually attributed to this new products of the xSPI family that we brought to market over the last couple of years. So we are seeing good traction, and we are seeing some volume pickup."
— Bill Cooper, CFO & Sanjeev Aggarwal, CEO

Assessment: The CEO's acknowledgment that xSPI contributed materially to the Q2 +20% sequential industrial growth is the cleanest forward signal we have on the product family. The implication is that xSPI is already contributing to revenue in Q2 — full-volume late-2025 production will accelerate the trajectory, not initiate it. We model xSPI HR variants contributing $2-4M of incremental annual revenue by FY26 vs FY25.

Industrial and Geographic End-Market Dynamics

The questioning continued on broader end-market dynamics, with the CEO providing the most comprehensive geographic update of the call — specifically the inventory normalization in Asia and the breadth of customer traction across data-center and automation.

Q: "Let's ask on the product side here from an end market perspective. You mentioned some nice dynamics here in data center and you called out some stuff in automotive here. Maybe just talk about the dynamics you're seeing in the broader industrial markets and even by geographies, as I think you've talked about from time to time in the past here, what the trends are looking like."
— Richard Shannon, Craig-Hallum

A: "The good news for Everspin from Everspin's point of view is that we are actually seeing depletion of inventories at our customers, almost all across the world, specifically in the Asian region, where we were concerned about inventory overbuild. So I think we are actually seeing some good number of orders come in from that region. So that's good. And again, most of these are either in the automation or in the data center... most of these are either in the automation or in the data center. And by data center, I mean the RAID memory or general memory applications. So the Dells, the Supermicros, the Broadcoms. And then as far as automation, those are our traditional customers from across the world that use them for programmable logic controllers."
— Sanjeev Aggarwal, CEO

Assessment: The Asia inventory normalization is the macroeconomic data point that matters most for Everspin's industrial business through H2 2025 and into 2026. The CEO's "good news... depletion of inventories" framing is the cleanest leading indicator we've heard from this management team on industrial demand pacing. If sustained, this could drive industrial-automation revenue growth materially above the consolidated growth rate through FY26.

DoD Sustainment Contract Q4 Pickup

The most specific forward-EPS-related Q&A exchange of the call addressed the timing of the DoD contract acceleration. The CFO's response was deliberately measured — confirming the Q4 weighting without committing to specific dollar amounts.

Q: "I believe, Bill, you had mentioned something about a pickup in some specific contract — something to pick up in the fourth quarter. Maybe you could repeat that and then describe a little bit more about what's going on there, please... So when you talk about a pickup there, any way you can quantify what we're talking about relative to the third quarter or 2? I think you had a fairly sizable quarter in this contract sometime last year. Maybe you could just kind of size that for us, please."
— Richard Shannon, Craig-Hallum

A: "That is the other income that we have from the [Amentum] contract that we were awarded late last year. And we do expect to see some pickup in activity in that contract in the back half of the year... We had a strong back half last year. We'll expect to see a strong second half and particularly more toward Q4 as well."
— Bill Cooper, CFO

Assessment: The CFO's "strong back half... particularly more toward Q4" is the closest he comes to quantifying the DoD acceleration. The implicit reference to "strong back half last year" — Q3'24 was $1.9M of other income and Q4'24 was $3.1M — sets a pattern that Q4 2025 should land in the $3-4M range, materially above the $0.8-1.2M Q1-Q3 quarterly run-rate. This is the load-bearing FY25 EPS lever.

QuickLogic Project Status and Frontgrade Renewal

The second follow-up of the Q&A clarified the status of two of the three named licensing/services contracts. The CEO disclosed that Frontgrade Phase 1 had been completed but Phase 2 was awaiting AFRL renewal; QuickLogic remained on deliverable execution with roughly 2 years of project life remaining.

Q: "In terms of the QuickLogic relationship there, maybe I caught the language wrong here, but is this something that you're finished the current contract you're waiting for the new one? Or I just want to get a sense of the timing and dynamics where we are with the last announced one, please?"
— Richard Shannon, Craig-Hallum

A: "As far as the completion of the phase, I was referring to the Frontgrade project, Richard. We did complete the first phase of that project in Q2. And now we are dependent on renewal of the project and that is actually funded — I think Frontgrade has actually funded through the AFRL. So that's what we are waiting on next. On the QuickLogic, we continue to deliver on the deliverables that we had signed up for in Q2. There's nothing significant to report over there, except that we have met our deliverables and continue to work on future deliverables... The QuickLogic one, I think the total value of the project and the time is almost another 2 years, if I'm not mistaken."
— Sanjeev Aggarwal, CEO

Assessment: The Frontgrade Phase 2 renewal is contingent on AFRL government funding decisions — a risk factor outside Everspin's control. The QuickLogic ~2-year remaining-project visibility provides moderate revenue stability. The cleanest takeaway: don't model material licensing growth beyond the existing project portfolio.

Lattice Semiconductor Co-Package Solution Update

The closing question of Q&A addressed the Lattice Semiconductor partnership — a co-package solution combining Lattice FPGAs with Everspin's xSPI parts. The CEO confirmed evaluation activity is ongoing through standard distribution channels (DigiKey, GitHub drivers) but noted that traction is a multi-quarter horizon.

Q: "Following up on the press release with Lattice Semiconductor as a companion to the FPGA chips they sell there. I wanted to get kind of an update there on the progress. I know this is something that was going to take a number of quarters to come to fruition, but love to get your update on that one."
— Richard Shannon, Craig-Hallum

A: "We do have some activity going on with Lattice, where they're actually doing a co-package solution using our xSPI part of, I believe, it's either 32 or 64 megabit density. And I think that evaluation is ongoing, and I think they are now — those parts are now available for our customers to evaluate through different distributors. For example, DigiKey, you can get the boards from there for evaluation. And then also on GitHub, we have the drivers available to actually download the drivers to evaluate those parts. So I think the collaboration is going well. It's just like you said, it's going to take a couple of quarters before you see any significant traction."
— Sanjeev Aggarwal, CEO

Assessment: The Lattice partnership is one of multiple strategic-partner programs Everspin has in evaluation phase. None of these is currently a material revenue contributor; collectively they represent the optionality layer that could compound 2-4 quarters out. We don't model any specific contribution for Lattice in FY25-FY26; we treat it as upside optionality.

What They're NOT Saying

1. FY25 revenue or EPS guidance: Management has not provided FY25 revenue or EPS guidance and did not on this call. The implicit framework (Q3-Q4 weighted higher than Q1-Q2; DoD contract Q4 acceleration; xSPI HR full production late 2025) is enough to model, but no formal full-year commitment.
2. Specific design-win count or revenue contribution from xSPI HR: Management acknowledges xSPI is contributing to Q2 industrial growth but does not quantify. The 178 FY24 design-win number was disclosed but no 2025 YTD design-win count provided.
3. IBM FCM4 revenue contribution dollar amount: Steady state confirmed but no dollar disclosure. We estimate $1.5-2.5M per quarter based on commentary patterns.
4. Lucid Motors Gravity SUV unit-economic content: Management notes shipments are continuing and volume ramping — but no per-vehicle MRAM content disclosed, no Gravity production-volume reference.
5. Specific Q4 DoD contract recognition target: "Meaningful pick-up" confirmed; no dollar target. We infer $3-4M based on the FY24 Q3-Q4 pattern.
6. Frontgrade Phase 2 expected revenue: Phase 1 complete; Phase 2 contingent on AFRL renewal. No dollar value or timing disclosed for the contingent Phase 2.
7. Capex / capital intensity: Q2 six-month operating cash flow disclosed at $6.5M; capex $2.9M. Implied capex intensity ~22% of revenue — high for a fabless / fab-light model. The implied infrastructure investment is structural but not quantified for FY25 in total.
8. Pricing-power dynamics: No commentary on ASP trends across the product portfolio. Industry MRAM pricing is structurally premium to NOR/SRAM but compresses over time as volumes scale. Management has not framed the pricing trajectory.
9. NOR Flash supply-chain pressure as MRAM tailwind: Industry-wide NOR Flash capacity reallocation to DRAM (later disclosed in Q4 2025) was not yet referenced on this call. The dynamic exists but management didn't surface it.
10. Capital return / buyback / dividend: $45M cash position; debt-free; no capital-return framework discussed. Standard microcap posture is reinvestment-only, but the topic was not addressed.

Market Reaction

• Pre-print setup (Aug 5, 2025 close): $5.96, -7.0% YTD (from $6.41 Jan 2, 2025), -11.3% trailing 30 days (from $6.72 Jul 7, 2025). The stock had drifted lower through July ahead of the print, suggesting positioning was light. Trailing 12-month performance was roughly flat to slightly negative — typical for a microcap in a "show-me" thesis stage. Average daily volume ~80-100K shares.
• Day-of action (Aug 6, print-day regular hours): Close $5.94 (-0.17% from Aug 5) on volume of 78.8K. Print released after-market (~4:05 PM ET) with call beginning 4:30 PM ET. After-hours reaction modestly positive on the revenue-beat-at-guide-high-end + non-GAAP profitability.
• T+1 (Aug 7, 2025): Open $6.10 (gap up +2.7%); intraday high $6.65 (+11.9% from Aug 6 close); intraday low $5.97 (essentially flat); close $6.03 (+1.5% from Aug 6; -9.3% from intraday high). Volume 151.4K shares (~1.8x trailing average) — the highest single-day volume in 3+ weeks. The classic "buy the rumor, sell the news" intraday fade pattern.
• T+2 (Aug 8, 2025): Close $5.98 (-0.83% from Aug 7); volume 72.8K (back to baseline). Stock essentially returned to pre-print level by end of the print week.
• Sell-side reaction: No formal PT changes disclosed in the 48-hour post-print window. Coverage is concentrated at Craig-Hallum (Richard Shannon) and Needham (Neil Young, who would re-engage at Q3); both desks treated the print as "in-line execution" rather than a thesis-changing event.
• Peer reaction: MRAM is the only U.S.-listed pure-play MRAM company; peers are diversified memory or specialty-IP companies (e.g., NOR Flash — Adesto/Dialog/Renesas; SRAM — Cypress/Renesas; specialty IP — Synopsys, Cadence). No clear sector read-through from MRAM's print. Semiconductor sector (SOXX ETF) up modestly Aug 7 (+0.5%) on broader chip tape; MRAM did not significantly track sector.

The Q2 reaction pattern is the structural feature of a microcap that hits guidance without a catalyst surprise. The +12% intraday spike to $6.65 reflects how thin the float is — modest buying creates outsized moves; modest selling creates the inverse. The Aug 8 fade back to $5.98 (essentially pre-print) signals the institutional positioning was light enough that even a clean print could not sustain a multi-day rally without follow-on catalysts.

The lack of significant move is not bearish — it reflects appropriate market pricing of a Q2 print that confirmed but did not accelerate the existing thesis. The real catalyst sequence is forward: Q4 DoD acceleration → xSPI HR full production → LEO satellite customer wins → potentially Frontgrade Phase 2 renewal → potentially new licensing deals. Each is a discrete event that could move the stock 10-20%+ when announced; none was disclosed in Q2.

Street Perspective

Debate: Is the $45M cash position the asymmetric defensive feature or just dead capital?

Bull view: At ~30-35% of market cap in cash with no debt, Everspin has multi-year runway to execute the design-win-to-revenue thesis without dilution or distress. The cash provides strategic flexibility — M&A optionality, ability to weather end-market downturns, ability to invest in product development at scale (including the Microchip Foundry Services Agreement disclosed later in 2026). For a microcap with binary execution risk on individual product cycles, $45M is the equity-margin-of-safety that makes the stock investable.

Bear view: $45M of cash earning low-single-digit yield ($1-2M annually) at a company generating $5M of operating cash flow in a good quarter is structurally inefficient capital allocation. Management has not articulated a capital-return framework or a strategic-use-of-cash plan beyond "effective hedge against macroeconomic uncertainty." The cash drag on ROE is real (cash on a $130M market cap company contributes ~30% of book value but generates minimal returns). Either deploy it into accretive M&A or return it to shareholders.

Our take: The bull case is correct at this stage of Everspin's lifecycle. With the design-win ramp 2-3 quarters from material revenue inflection and the broader thesis dependent on multi-year product-development cycles, optionality on the balance sheet is appropriate. We do, however, monitor for any signs that cash is being deployed into low-IRR strategic projects rather than preserved for higher-IRR opportunities.

Debate: Can the licensing line sustain $2M+ per quarter, or is Q2 the peak?

Bull view: The licensing portfolio reflects Everspin's IP value being recognized by tier-1 partners (Purdue, QuickLogic, Frontgrade/AFRL). As Everspin's STT-MRAM technology continues to advance, more partners will enter licensing arrangements to leverage the IP. The Microchip Foundry Services Agreement (later disclosed in Q4 2025) and the $40M defense subcontract (Q1 2026) are examples of the structural compounding of strategic-partnership economics.

Bear view: Each current contract has a finite life: Frontgrade Phase 1 done with Phase 2 contingent; QuickLogic ~2 years remaining; Purdue at steady state with no contractual extension visibility; TMR sensor foundry project ongoing but at modest scale. Without new contract wins replacing expiring ones, licensing line could decline structurally through FY26. The Q2 $2.1M is the peak; expect normalization to $1.0-1.5M quarterly thereafter.

Our take: Both sides have merit. We model licensing at $1.5M per quarter average for FY25 (declining from Q2 $2.1M peak), with potential to re-accelerate in FY26 if (a) Frontgrade Phase 2 renews, (b) Microchip Foundry Services contributes new licensing revenue, or (c) new strategic partnerships convert. The line is structurally volatile but contributes outsized gross-margin uplift relative to dollar value.

Debate: Is microcap liquidity the structural impediment to Everspin's institutional ownership?

Bull view: Everspin's microcap status (~$130-150M market cap; ~80-150K daily volume) is appropriate for the current revenue base ($50M+ FY25 expected). As revenue accelerates to $75-100M+ over 3-5 years, market cap and liquidity will mechanically expand, opening institutional ownership. The pure-play MRAM exposure is genuinely scarce — institutional investors who want MRAM-specific exposure have no other publicly-traded alternative. As the LEO satellite + defense + AI-edge MRAM thesis matures, institutional appetite will follow.

Bear view: Microcap liquidity is the structural ceiling on Everspin's investability. At sub-$200M market cap with $1-3M average daily dollar volume, the stock is excluded from most institutional mandates. The thesis cannot rerate without institutional ownership; institutional ownership cannot expand without liquidity; liquidity cannot expand without market-cap growth; market-cap growth requires a thesis re-rating. The cycle is self-reinforcing and may require either a strategic-buyer M&A event or a major catalyst (e.g., DoD contract scaling to $100M+) to break the cycle.

Our take: Both partly right. The microcap liquidity constraint is real and is one of the central reasons we are initiating at Hold rather than Outperform — the position-sizing flexibility is limited regardless of conviction. As revenue compounds toward $80-100M+ and market cap potentially expands to $300-500M+, the liquidity profile improves. We monitor the upgrade path: revenue + market cap + average daily volume all need to compound roughly together.

Model Update Needed

Valuation impact: At Aug 7 close of $6.03 and ~22.6M diluted shares, market cap is ~$136M; cash $45M, debt $0, EV ~$91M. On FY25E revenue of $55M and non-GAAP net income of $2.3M, MRAM trades at 1.65x EV/Sales and ~40x EV/non-GAAP-NI. On FY26E revenue of $64M and non-GAAP net income of $5-7M, MRAM trades at 1.42x EV/Sales and 13-18x EV/non-GAAP-NI. The valuation is structurally cheap on revenue multiple but reasonable on earnings multiple given the small absolute revenue/profit base. Specialty-semi peers at similar growth profiles trade at 2-4x EV/Sales; high-quality MRAM-specific exposure would justify 2.5-3.5x. Our 12-month fair-value estimate: $7-9 (16-49% above current); the wide range reflects microcap valuation sensitivity to liquidity and revenue trajectory.

Thesis Scorecard Post-Earnings

Overall: The bull case is structurally credible — MRAM as a category has genuine technical advantages (non-volatility, endurance, harsh-environment reliability) that map to expanding addressable markets (LEO satellites, defense systems, edge AI, automotive). The bear case is structurally credible — microcap liquidity, project-based licensing volatility, single-analyst coverage, and design-win-to-revenue cycle delays are all real impediments to multi-year compounding.

Action: Initiating at Hold. The Q2 print is consistent with a "we said we'd execute, here's the execution" quarter — neither thesis-accelerating nor thesis-breaking. We initiate at Hold rather than Outperform because (a) the structural revenue inflection is 2-3 quarters out (Q4 DoD acceleration; xSPI HR full production; LEO ramp), (b) the microcap liquidity profile limits position sizing, and (c) management's deliberately-conservative posture on gross-margin and FY25 guidance signals appropriate caution. Re-rating triggers: Q3-Q4 revenue exceeding $15M with non-GAAP EPS >$0.07, AND DoD contract recognizing >$3M in Q4, AND visible LEO satellite design-win conversion to revenue. Downgrade trigger: Q3-Q4 revenue below $13M consolidated, OR DoD contract Q4 ramp materially undershoots (<$1.5M), OR gross margin compresses below 48%.