Netlist’s Q3 Miss Is the Wrong Fight: The Real Trade Is a 2026 Memory Squeeze With Litigation Optionality

The print was not good versus the Street, and that matters because Netlist is still a small-cap memory participant with little room for narrative slippage: revenue was $42.2 million against $44.0 million, a -4.0% surprise, and EPS was -$0.02 versus -$0.01, off a near-zero estimate base where the percent surprise is not meaningful. What was priced in was a sequential continuation of the Q2 FY2025 recovery, when revenue had stepped to $41.7 million and EPS had narrowed to -$0.02; what actually surprised was that revenue only reached $42.2 million in Q3 FY2025 despite that low bar, while gross margin at 4.4% still sat far below the levels implied by a true memory upcycle. The stock debate after this print should not be whether Q3 was clean. It was not. The actionable question is whether the miss invalidates the larger thesis that a tightening DRAM and NAND market, plus product qualifications into AI infrastructure, can change Netlist’s revenue mix faster than the Street is giving credit for.

That distinction is important because the historical financial path shows a company that has already moved from subscale revenue to a more investable base, but not yet to a structurally profitable model. Revenue was $9.0 million in Q1 FY2023, $16.7 million in Q3 FY2023, $40.2 million in Q3 FY2024, and $42.2 million in Q3 FY2025. That is not a straight line: Q4 FY2024 revenue fell to $34.3 million, Q1 FY2025 fell again to $29.0 million, and the Q2 FY2025 rebound to $41.7 million made the Q3 FY2025 sequential gain of +1.3% look underwhelming. But the market may be missing that the trough-to-recovery pattern has already stabilized above the old revenue base, while EPS losses have narrowed from -$0.07 in Q1 FY2023 and -$0.07 in Q3 FY2023 to -$0.02 in Q2 FY2025 and -$0.02 in Q3 FY2025. The bear case needs Q3 to be evidence that the recovery has stalled; the bull case only needs Q3 to be a holding pattern before supply-driven pricing and qualification-driven volume show up.

The margin line is the reason the debate is still unsettled, because Netlist’s revenue recovery has not yet carried the economics of the broader memory cycle. Gross margin was 4.4% in Q3 FY2025, up from 3.3% in Q2 FY2025 and above 2.8% in Q3 FY2024, but still below the 6.2% achieved back in Q1 FY2023 and nowhere near what peers are printing in the current memory environment. The most important counterpoint in the data pack is the subsequent financial trajectory in the quarterly history: Q4 FY2025 revenue is listed at $75.7 million with 9.1% gross margin, and Q1 FY2026 revenue is listed at $104.9 million with 21.3% gross margin. Those figures cannot be used to excuse the Q3 FY2025 Street miss, but they do show what confirmation of the thesis would look like if the revenue base shifts from resale-heavy, low-margin activity toward products with better scarcity value. The market’s likely error is treating the Q3 FY2025 4.4% gross margin as the steady-state economics rather than as a transition-period number.

The capacity comments on the call make that transition argument more credible, because management did not describe a generic recovery in memory demand; it described a scarcity mechanism with a dated duration. Chuck Hong said, “AI-related demand is consuming the majority of DRAM and NAND capacity, leaving very little availability for the rest of the market.” That wording matters because it frames Netlist’s opportunity as a supply allocation problem rather than simply a cyclical demand pickup. He also said, “And we believe that this will continue for the foreseeable future, all the way through 2027, when additional capacity is brought online.” The phrase “all the way through 2027” is the commitment in the call: if correct, it means the availability constraint should outlast one or two quarterly inventory adjustments and support a multi-quarter opportunity for companies that can source, qualify, and ship into constrained channels. For Netlist, the Q3 miss therefore becomes less important than whether its qualified products can capture scarce supply economics before new capacity arrives.

That is where the AI infrastructure product discussion becomes the real variant-perception lever, because management disclosed both the current scale and the intended step-up without giving a single-quarter revenue guide. Hong said the relevant product line has produced “probably in the 4%, 5% of our overall revenues in the last few quarters,” and that once qualified at larger server manufacturers it could reach “double-digit percentage of the overall revenues” next year. The numbers are small today: 4%, 5% of overall revenues does not rescue a quarter where reported revenue was $42.2 million versus a $44.0 million estimate. But the potential inflection is also concrete: moving from 4%, 5% of revenues to double-digit percentage of overall revenues would change the mix within a revenue base that was already $41.7 million in Q2 FY2025 and $42.2 million in Q3 FY2025. The key point is not that AI is a slogan for Netlist; it is that management put a current contribution level and a next-year mix threshold around the opportunity.

The spending side adds an underappreciated second leg to the thesis, because Netlist does not need revenue to do all the work if legal expense is actually being reset lower. Gail Sasaki said operating expense for Q3 FY2025 declined 17% compared to the prior year’s quarter and declined 38% on a 9-month basis, while net loss declined 25% compared to the prior year’s quarter and declined 45% for the 9-month period. She also said the company is “on track to have reduced our legal fee expense by 50% over last year.” That last phrase is important because Netlist’s litigation program is central to the equity story, but legal spending has historically competed with operating leverage. The income statement still showed -$0.02 EPS in Q3 FY2025, the same as Q2 FY2025, so the expense cuts have not yet converted the company into a profitable model. But a 17% quarterly operating expense decline and 38% 9-month decline mean the downside from litigation burn is at least being actively reduced at the same time the company is trying to scale product revenue.

The litigation asset remains too material to ignore, but it should be treated as option value with dates rather than as current earnings power. Hong said that in the past 2 years Netlist obtained 2 jury verdicts awarding combined total damages of $421 million for willful infringement of patents by Samsung, and he later framed the broader federal court actions as 3 separate jury verdicts awarding combined damages of $866 million for willful infringement by Samsung and Micron. The case-level numbers are large relative to Netlist’s operating base: an Eastern District of Texas case against Samsung had an order finalizing $303 million of damages award in July 2024, another Samsung case involved $118 million in damages with post-trial briefing completed, and the Micron award was $445 million with the case concluded in district court and appeal briefing underway. The timing, however, is staggered: Samsung’s appeal hearing is estimated sometime mid next year, post-trial motions are expected in the coming months, and the Micron Federal Circuit process is expected to take approximately 16 to 18 months. The equity may react to each docket update, but PMs should not underwrite Q4 FY2025 or Q1 FY2026 operating numbers as if $866 million is cash.

Liquidity is the main near-term constraint on waiting for that option value to season, and the Q3 numbers were mixed rather than clean. Sasaki said cash and cash equivalents and restricted cash ended Q3 at $20.8 million compared with $29 million at the end of Q2, with minimal debt. After quarter-end, Netlist raised $10 million through a registered direct offering, and Sasaki also cited a $10 million working capital line of credit and approximately $74 million available on the equity line of credit. The immediate read is that the company has tools to bridge, but the cost of that bridge can include equity issuance, as the post-quarter $10 million registered direct offering demonstrates. That financing reality is why the expense reductions matter: a company with Q3 FY2025 EPS of -$0.02, cash of $20.8 million, and litigation timelines stretching to mid next year and approximately 16 to 18 months cannot rely on verdict headlines alone. It needs product revenue and legal-fee reductions to lower the probability of repeated dilution before appeals resolve.

The delivery of the call itself argues for caution on timing, even while the content argues the setup is improving. The tone history shows Q3 FY2025 sentiment at 0.09, down from 0.35 in Q2 FY2025, while uncertainty rose to 75.6 from 45.9 and qa_evasiveness moved to 34.9 from -3.5. Guidance tone was 0.47 in Q3 FY2025, slightly above 0.44 in Q2 FY2025, which creates a real conflict: management’s forward-looking language was constructive, but the overall transcript became more uncertain and less direct in Q&A. That conflict fits the quarter. The strategic setup is better than the reported miss implies, but the path depends on qualifications, customer timing, supply allocation, and court schedules that management cannot fully control.

The supply-chain read-through is unusually narrow because the data pack lists no named customers of Netlist and no named suppliers to Netlist, so the clean implication is for the named litigants and the unnamed server-manufacturer channel rather than for a disclosed procurement chain. Samsung and Micron remain central in two ways: they are the defendants tied to the combined damages figures of $866 million across the federal court actions, and they are also memory suppliers exposed to the same DRAM and NAND scarcity described on the call. The named customer set is absent, but management did identify “bigger server manufacturers” as the qualification target, with expected contribution next year after qualifications expected by the end of the year. That makes the second-order read-through precise but limited: the print does not identify a customer winner, yet it does imply that any larger server manufacturer qualifying Netlist’s AI infrastructure products could shift a product line currently at 4%, 5% of overall revenues toward double-digit percentage of overall revenues next year.

Relative to memory peers, Netlist is still not monetizing the cycle the way the broader group is, which is exactly why the stock can be mispriced in both directions. In the latest peer table, Micron shows $41,456.0 million of revenue, 84.6% gross margin, and +345.7% revenue YoY, while WDC shows $3,337.0 million of revenue, 50.2% gross margin, and +45.5% revenue YoY, and STX shows $3,112.0 million of revenue, 46.5% gross margin, and +44.1% revenue YoY. Netlist’s latest peer-table entry is $104.9 million of revenue, 21.3% gross margin, and +262.0% revenue YoY. The comparative point is not that Netlist should trade like Micron; it should not, given the revenue scale gap between $104.9 million and $41,456.0 million. The point is that if the memory cycle is producing peer gross margins of 84.6%, 50.2%, and 46.5%, then Netlist’s Q3 FY2025 gross margin of 4.4% was still pre-inflection, while the later 21.3% peer-table margin shows the kind of operating evidence the market will require before paying for the cycle story.

What to watch next is concrete. First, the Q4 FY2025 quarterly history line already sets the confirmation bar at $75.7 million of revenue, 9.1% gross margin, +79.3% revenue QoQ, +120.9% revenue YoY, and -$0.01 EPS; failure to approach those levels would break the view that Q3 FY2025 was merely a bridge quarter. Second, the product-qualification claim needs evidence by the end of the year, because management said some qualifications are expected to be completed by the end of the year and to contribute meaningfully next year; the measurable test is whether the AI infrastructure product line can move beyond 4%, 5% of overall revenues toward double-digit percentage of overall revenues. Third, liquidity should be tracked against $20.8 million of Q3 cash, the post-quarter $10 million registered direct offering, the $10 million working capital line of credit, and approximately $74 million available on the equity line of credit, because further equity use would dilute the upside from operating improvement. Finally, litigation timing needs to be marked to the stated calendar: Samsung’s appeal hearing sometime mid next year, post-trial motions in the coming months for the $118 million Samsung case, and the Micron appeal process of approximately 16 to 18 months. The thesis holds if revenue and margin step toward the Q4 FY2025 levels while legal expense remains on the path to a 50% reduction over last year; it breaks if Q4 looks like another $42.2 million revenue quarter with gross margin near 4.4% and no visible qualification conversion.