Netlist
Dual engine: (1) design and sale of high-performance memory modules and storage (DDR5 RDIMM/MRDIMM, CXL/HybriDIMM, SSDs) to data-center/OEM customers as a fabless module assembler buying DRAM from the majors; (2) an IP portfolio (LRDIMM, RDIMM, memory power-management) monetized through licensing plus litigation-driven damages against DRAM makers. Revenue is lumpy product sales; IP produces episodic license/settlement/damages cash that is contingent on appellate and PTAB outcomes.
Earnings, margins, COGS & capex
Netlist's product business inflected to profitability in Q1 FY2026 -- $104.9M revenue (+262% YoY), 21.4% gross margin, $8.6M net income and EPS $0.03 -- after years of losses, driven by the AI-led DRAM shortage/price spike and DDR5 ramp. FY2025 net sales were $188.6M (+28% off $147.1M FY2024) with the net loss narrowed to $24.8M (from $53.9M) as IP legal costs eased. The larger, harder-to-value asset is the litigation book: ~$866M in jury verdicts against Samsung ($303M + $118M) and Micron ($445M) -- all won at the jury/district level but under appeal, and materially undercut by PTAB invalidations of underlying patent claims, plus active ITC and district-court actions against Samsung, Micron, Google and Super Micro. Cash is thin ($17.0M ex-restricted) and burned in Q1 on inventory build.
Income statement — where each revenue dollar goes
% of revenueOf every $1 of revenue, ~79¢ is cost of goods and ~13¢ operating expense, leaving ~8¢ of operating profit (~9¢ net).
Revenue trend
Margins
sharply up from ~4.5% (Q1 2025); FY2025 full-year ~6% on memory pricing
+1,622% YoY
first profitable quarter after sustained losses (Q1 2025 was a $9.5M net loss)
swung toward positive; do not conflate with the $8.6M net figure
COGS structure
Dominated by purchased DRAM/NAND components -- Netlist buys memory chips from the same majors it litigates against (Samsung, SK hynix, Micron) and assembles modules, so gross margin tracks the spread between DRAM pricing and module ASPs. The FY2025-26 margin expansion is a pricing windfall from the AI-driven DRAM shortage, not a durable cost advantage.
Capex
Not separately disclosed; light -- fabless module assembly and IP portfolio maintenance rather than fab capex.
Latest earnings
Beat -- EPS $0.03 vs ~-$0.01 consensus; record revenue. Stock nonetheless fell ~13-16% on the print (sell-the-news / cash-burn and legal-cost concern)
No hard numeric guidance disclosed; management framed continued strong demand for DDR5/AI-server memory. Treat forward revenue as cycle-dependent, not guided.
- Revenue
- $104.9M (+262% YoY)
- Gross margin
- 21.4%
- Net income / EPS
- $8.6M / $0.03
- Cash (ex-restricted)
- $17.0M (down $14.8M QoQ; +$10.0M restricted)
- Inventory
- $41.2M (up from $3.4M)
- Deferred revenue
- $44.5M
Growth drivers
- AI/data-center memory demand -- DDR5 RDIMM/MRDIMM and high-capacity modules for AI servers
- DRAM price surge in 2026 lifting module ASPs and gross margin
- Inventory pre-build ($41.2M) and $44.5M deferred revenue signalling forward customer commitments
- IP monetization — license/settlement cash and potential (but contested) collection on the ~$866M in verdicts if appeals and PTAB proceedings resolve favorably
- SK hynix strategic supply + cross-license relationship (2021) supporting product supply and CXL/HybriDIMM cooperation
Reported financials — SEC EDGAR
Audited GAAP figures pulled from SEC filings · latest filing 2026-03-19. The audited primary-source spine — not financial advice.
Revenue — annual (GAAP)
Margins & balance sheet — FY’25
Bull & bear
A newly profitable AI-memory product business provides a real earnings floor while ~$866M in trial-won verdicts plus fresh ITC/DDR5 actions give large -- if contested -- litigation optionality against a ~$1B market cap.
- Q1 FY2026 proves the product model can be profitable at scale (21.4% GM, $8.6M NI) -- no longer a pure litigation shell
- Verdicts won at trial ($303M + $118M Samsung, $445M Micron) exceed a meaningful fraction of market cap; interest accrues while appeals run, and the district court has finalized the $303M award
- Fed Circuit has handed Netlist some wins (rejecting certain Micron PTAB challenges), keeping the door open that portions of the awards survive
- ITC DDR5 power-management case ('366 patent) vs Samsung/Google/Super Micro adds injunction leverage that can force new licenses before the patent expires in 2028
- Direct leverage to the multi-year AI DRAM/HBM demand and pricing supercycle
The equity is a leveraged bet on binary appellate outcomes -- where PTAB has already invalidated most of the underlying claims -- plus a cyclical DRAM price spike, sitting on an OTC listing with $17M cash, negative FCF, and a history of dilution. The verdicts may never be collected in full or even in part.
- PTAB has already invalidated the patent claims behind the bulk of both verdicts (all four Samsung patents-in-suit; a claim underpinning >95% of the Micron award), so 'won at trial' overstates realizable value -- awards can be cut, vacated, or remanded
- Every headline verdict is still under appeal and collection can take years even in the best case
- Product margin is a DRAM-shortage windfall -- a price downcycle compresses the 21.4% GM fast
- $17.0M cash (ex-restricted) with ~$14.8M quarterly burn and a $41.2M inventory build raises dilution/financing risk
- OTC listing and reliance on suing its own suppliers cap institutional ownership and strategic optionality
- ~5x FY2025 sales for a low-margin, cash-burning module maker is rich absent a litigation payoff that is now materially impaired
What it is worth
Sum-of-parts: operating memory-module business (revenue multiple / normalized earnings on cyclical margins) PLUS risk-adjusted litigation book (probability- and time-discounted value of the ~$866M verdicts and pending ITC/DDR5 actions, net of appeal AND realized PTAB-invalidation risk). A pure sales multiple (~5x FY2025 revenue) overstates the low-margin product side and misstates the option value -- which is now materially impaired by PTAB cancellations, not free optionality.
Awards are cut or vacated as the PTAB invalidations are affirmed, the DRAM cycle rolls over compressing product margin, and dilution funds the burn -- valuation resets materially lower toward the operating-business-only floor.
Partial appellate success / negotiated settlements at a fraction of face value (given the PTAB invalidations), combined with a still-profitable but cyclical product business, roughly supports the current ~$1B valuation with high variance.
Enough of the verdicts survive appeal/PTAB and are collected with interest, the ITC DDR5 case forces new licenses, and the product business compounds through the AI memory cycle -- justifies a multiple above current levels.
At ~$1.01B market cap, the equity prices in meaningful litigation success. The product business alone -- low, cyclical gross margin and negative FCF -- would not support ~5x sales; the gap is embedded value of the verdicts and enforcement pipeline. But PTAB has already invalidated most of the underlying claims, so that embedded value is contingent and contested, not a stable earnings stream. Value is appeal-outcome and DRAM-cycle dependent.
SWOT
Strengths
- Foundational memory-module and power-management IP portfolio that has extracted real settlements and jury verdicts from the largest DRAM makers
- ~$866M in jury verdicts won at trial against Samsung ($303M+$118M) and Micron ($445M) -- large relative to market cap, though contested (see threats)
- Product business turned profitable in Q1 FY2026 with rapid margin expansion (21.4% GM, $8.6M NI)
- SK hynix strategic supply + cross-license (2021, $40M license + up to $600M supply) de-risks one major supplier and adds a licensing counterparty
- Leverage to the AI DRAM upcycle -- top line +262% YoY
Weaknesses
- Thin cash ($17.0M ex-restricted) and negative free cash flow — working-capital and legal spend consume liquidity
- OTC listing (OTCQB, not a national exchange) — lower liquidity, weaker institutional eligibility, higher cost of capital
- Structurally low, pricing-dependent product gross margin -- margin is a DRAM-cycle windfall, not a moat
- History of shareholder dilution to fund operations and litigation
- Revenue lumpy and customer/IP-event concentrated
Opportunities
- Collection on the Samsung ($303M+$118M) and Micron ($445M) awards plus interest IF appeals and PTAB proceedings resolve favorably -- currently far from certain
- New DDR5 power-management enforcement (US Patent 12,373,366, expires June 2028) at the ITC vs Samsung, Google, Super Micro -- exclusion-order leverage
- Expansion of licensing to additional DRAM/HBM makers as AI memory scales
- CXL/MRDIMM/HybriDIMM product positioning into AI-server memory disaggregation
Threats
- Appeal + PTAB risk is largely realized, not hypothetical — PTAB has invalidated multiple patent claims underlying both marquee verdicts (all four Samsung patents-in-suit invalidated by spring 2024; >95% of the Micron $445M award tied to a PTAB-invalidated claim), so awards can shrink, reverse, or be delayed for years
- Fed Circuit outcomes are mixed — some Netlist wins (certain Micron PTAB appeals rejected) alongside affirmed patent cancellations; the $303M Samsung appeal was argued March 2026 with a decision pending
- DRAM price cycle reversing -- module margins compress quickly on a downcycle
- Dependence on the very DRAM majors it sues for component supply
- Litigation-cost drag and binary, hard-to-time cash timing
Moats, dependencies & bottlenecks
Moats
Patent portfolio on memory-module and power-management technology Moderate (litigation-tested, but partly invalidated at PTAB) appeal- and PTAB-outcome bound; erodes as patents expire ('366 in 2028) and IPRs cancel claims The genuine asset -- has extracted real settlements/verdicts, but enforcement-dependent, finite-lived, and materially undercut by recent PTAB invalidations
Fabless module assembly is commoditized; margin follows DRAM pricing, not a structural edge
Secures component supply and a licensing counterparty, differentiating from a pure NPE
Dependencies
Buys memory chips from the same firms it litigates -- adversarial supplier relationships
The litigation asset value is entirely a function of appellate, PTAB, and ITC outcomes and timing -- and PTAB has already invalidated key claims
Product revenue and margin are levered to memory pricing and AI-server demand
Thin cash + burn implies episodic dilution to fund operations and legal costs
Advantages
- IP that has extracted real settlements/verdicts from the largest memory makers (SK hynix $40M; Samsung/Micron jury awards)
- Optionality profile -- asymmetric (if contested) litigation payoff on top of an operating business
- Early positioning in memory-module IP (LRDIMM, RDIMM, power management) now central to AI servers
- SK hynix supply/license anchor most peers lack
Weaknesses
- Chronic cash constraint and dilution history
- Commodity, price-dependent product margins
- Binary, hard-to-time value drivers with PTAB invalidation risk realized
- Sub-national-exchange (OTC) listing
Bottlenecks
- Cash runway -- $17.0M ex-restricted against ongoing burn and legal spend
- Appellate + PTAB timelines that defer (and may erase) any verdict cash for years
- Component supply and allocation from adversarial DRAM suppliers during a shortage
- OTC liquidity limiting the investor base and cost of capital
Top signals & trends
Top signals
Product model validated during the AI DRAM upcycle
Market focused on $14.8M cash burn and inventory build over the beat
Undercuts realizable value of the verdicts; Bloomberg Law/Law360 documented cancellations of the patents-in-suit
Some claims survive; the marquee Samsung award outcome is unresolved
Adds exclusion-order leverage; policy scrutiny of ITC remedies is a watch item, and '366 expires 2028
Financing/dilution overhang
Trends
Primary driver of the revenue and margin inflection
Expands both product TAM and the relevance of the power-management IP
Positive for IP leverage · Netlist among the parties pressing memory makers -- but PTAB has been a strong counter-tool for defendants
Windfall now, sharp compression risk on the downcycle
Negative for injunction leverage · Could narrow the ITC threat's teeth
Ecosystem & competitor graph
Suppliers feed the company; customers pull from it. Line thickness shows the strength of each tie (supply-chain dependency, customer earnings contribution). Hover to isolate a tie.
DRAM die supplier -- and litigation counterparty
DRAM/HBM supplier under a strategic supply + cross-license agreement (2021; up to $600M of product)
DRAM/NAND supplier -- and litigation counterparty
Primary product buyers of DDR5/RDIMM/MRDIMM AI-server memory; specific names largely not disclosed. $44.5M deferred revenue implies committed forward demand
Samsung / Micron / Google / Super Micro (as licensees/defendants) IP counterparties whose settlements/damages are the licensing 'revenue' path
DRAM/module maker, direct product competitor AND litigation defendant ($445M verdict, on appeal, key claim PTAB-invalidated)
Largest DRAM maker; product competitor, supplier, and litigation defendant ($303M+$118M). Context only, not a buy/own call
DRAM/HBM leader; supplier + cross-license partner rather than adversary since 2021. Context only, not a buy/own call
Closest US-listed comparator -- memory-interface IP licensing + chips; the 'IP-licensing done at scale' benchmark for valuing Netlist's book
Memory modules and specialty/enterprise memory -- product-side competitor for data-center memory
Private; largest independent module maker -- product-side competition on RDIMM/server memory