Enovix

Phase A — Understand the business

Lens 1 · Company Overview

Enovix Corporation (Fremont, CA; Nasdaq: ENVX; incorporated Delaware; CIK 0001828318) designs and manufactures lithium-ion battery cells built on a proprietary 3D cellular architecture with a 100% active-silicon anode and no graphite. The founders — drawn from "over 25 years in the hard disk drive and semiconductor industries" — solved silicon's swelling/cracking problem not with new chemistry but with physics: instead of winding electrodes into a conventional jelly-roll, Enovix precisely stacks anode, cathode and separator and applies mechanical constraint that "accommodates silicon's swelling and applies pressure that alleviates the cracking problem". The payoff is energy density: the flagship AI-1 platform delivers a 7,350 mAh smartphone cell at >900 Wh/L — pitched explicitly at AI-enabled smartphones and smart eyewear that need more on-device energy.

How it makes money today vs. tomorrow — the central tension. There are effectively two Enovixes:

1. The legacy/acquired business that pays the bills now — conventional Li-ion and silicon-doped-graphite cells, largely for defense, run through Routejade (acquired Oct 2023; two factories in Nonsan City, South Korea; four automated lines + two electrode-coating lines). This is where essentially all of the ~$32M FY2025 revenue comes from.
2. The silicon-anode story the $1.6B valuation is built on — the AI-1/3D-cell consumer business out of Fab2 in Penang, Malaysia, which has not yet shipped a flagship consumer program at volume.

Target end markets: smartphone, smart eyewear (smart glasses / AR), IoT (wearables, health/wellness, AR/VR, power banks, trackers), and defense.

Customers & concentration (critical): revenue is "derived largely from a limited number of key customers, particularly those in the defense sector. One customer, a defense subcontractor in South Korea, accounted for the majority of our total revenue for fiscal year 2025". In the notes: for FY2025, Customers D and E accounted for ~64% and ~13% of total revenue (≈77% combined). Q1 2026 revenue was again "driven largely by batteries supplied to Korean military contractors".

Contract structure: product-revenue sales (not take-or-pay or recurring SaaS-like). Consumer programs require multi-year customer qualification cycles ("can take years to complete") with safety/qualification testing before any volume orders. Deferred revenue is small ($4.3M current at Apr-5-2026).

Suppliers: Enovix "relies on a manufacturing agreement with a Malaysia-based company for some of the facilities, procurement, and personnel needs" of Fab2 — i.e. a third-party contract-manufacturing dependency sits under the supposedly-owned flagship fab.

Lens 2 · Supply Chain

No supply-chain.md exists in the commercial layer (energy wiki is empty) — mapped from the 10-K + web.

Upstream → Enovix → end customer:

• Raw materials / cell inputs: silicon (active-anode material), cathode active materials (NMC-class), separators, electrolyte, copper/aluminium current collectors, packaging. The 10-K flags reliance on "manufacturing operations and the components necessary to build our lithium-ion battery cells" with named single-source / concentration risk. Battery materials supply is China-heavy industry-wide.
• Electrode coating / pack assembly: vertically integrated via Routejade (South Korea) — the acquisition's stated rationale was to "vertically integrate electrode coating and battery pack manufacturing". This is a genuine chokepoint Enovix now owns rather than buys.
• Cell manufacturing (the new story): Fab2, Penang Science Park, Malaysia — three lines: R&D pilot, Agility line (began shipping cells Oct 2024), and the High-Volume Manufacturing (HVM) line (Site Acceptance Testing completed; HVM SAT confirmed Jan 6, 2026 ). Fab1 (Fremont) is now new-product development only.
• Contract-manufacturing layer: the Malaysia-based partner providing facilities/procurement/personnel for Fab2 — a single-source operational dependency.
• Downstream / end customers: the named-but-coded South Korean defense subcontractor(s) (Customers D/E, ≈77% of FY25 rev) today; prospectively Honor + a second smartphone OEM (qualification in progress) and a lead smart-eyewear customer (AI-1 commercial production started); plus a drone/defense pipeline management is working to convert to revenue.

Chokepoints / single-source dependencies:

• Customer-side concentration is the dominant chokepoint — one Korean defense buyer is the majority of revenue.
• Fab2 contract-manufacturer dependency — operational single point of failure under the flagship fab.
• Laser/"laser dicing" process step — publicly cited as the cause of 2026 volume-production delays, i.e. an internal process chokepoint, not just a supplier one.

Lens 3 · Competitive Advantages (moats)

The real edge — architecture + IP. Enovix's moat is a process/architecture moat: the constrained-stack 3D cell that makes a 100% active-silicon anode survivable. "A majority of our patents relate to battery architecture, secondary batteries, and related structures and materials," supplemented by trade-secret protection on non-patentable know-how. Independent testing has validated the AI-1 as a class-leading energy-density smartphone cell (>900 Wh/L). If the architecture transfers to volume at yield, that is a durable, hard-to-copy advantage in a spec-driven market where OEMs reward the highest mAh in a fixed volume.

Where the moat is weak:

• No volume proof. A moat you can't manufacture at yield isn't yet a moat. The entire bear case is that the lab/pilot advantage hasn't survived the jump to Fab2 HVM (yield, laser dicing).
• Bargaining power is inverted today. With ~77% of revenue from one-to-two defense buyers and zero shipping flagship-smartphone programs, the customer holds all the leverage — qualification cycles "take years," OEMs dictate the test framework (Enovix had to align Honor on a new silicon-specific 0.2C discharge protocol, replacing the legacy 0.7C test). That is the supplier bending to the buyer, not pricing power.
• Credible competitors are advancing commercially faster. Amprius (silicon nanowire, up to 450 Wh/kg) tripled revenue to ~$73M in 2025, hit positive adjusted EBITDA in Q4, and guides $125–135M for 2026 — 4× Enovix's revenue with a shipping product. Different architecture, same "silicon-anode energy density" buyer pitch.

Net: a potentially excellent technical moat that is unproven at the only thing that matters (volume manufacturing) and currently coupled to near-zero customer bargaining power.

Lens 4 · Segments

segments.csv is empty — Enovix does not report disaggregated product/geography segments in a granular table; it operates as a single reportable segment ("Our CEO has been identified as our CODM, who reviews operating results … for the Company as a whole"). So the meaningful "segmentation" is by customer/end-market concentration and geography, not GAAP segments.

• By product line (de facto): essentially all revenue today = conventional Li-ion / silicon-doped-graphite defense cells (Routejade, Korea). The 3D-silicon AI-1 consumer line is pre-volume-revenue.
• By customer (FY2025):
• Customer D ≈ 64% of revenue
• Customer E ≈ 13% of revenue
• "One customer … South Korea defense subcontractor … the majority of total revenue"
• FY2024 had three >10% customers → FY2025 narrowed to two, i.e. concentration increased YoY.
• Receivables concentration (Dec-28-2025): Customers D, E, G ≈ 43% / 12% / 23% of total AR.
• By geography: revenue is "derived … outside the U.S. [for] a significant portion" — predominantly South Korea (defense). Manufacturing is Malaysia (Fab2) + South Korea (Routejade) + India (some activity) + U.S. (HQ/R&D).

Trend & cause: revenue is growing off a tiny base (FY2025 $31.8M, +38% YoY; Q1-26 $7.6M, +49% YoY) but the growth is defense-led and lumpy, not the consumer inflection the stock needs. Q4-25 was a record $11.3M; Q1-26 stepped down to $7.6M; Q2-26 guided $8–9M. That sawtooth is the tell that this is project/order-driven defense revenue, not a ramping consumer franchise.

---

Phase B — Measure performance

Lens 5 · Earnings Result (latest print — Q1 2026, quarter ended April 5, 2026)

All figures `` unless noted; $ in thousands.

• Beat: revenue $7.6M cleared the $6.5–7.5M guide and was up 49% YoY. Non-GAAP gross margin 26.3%, the sixth consecutive quarter of positive gross profit (GAAP and non-GAAP). Non-GAAP loss from operations $28.8M beat the $29–32M guide; non-GAAP EPS $(0.14).
• What drove it: Korean military-contractor batteries (the concentrated defense business), not consumer silicon.
• Margin move: GAAP GM nearly 4× the prior-year quarter (20.4% vs 5.1%) — real operating-leverage progress on the defense product, off a tiny base.
• Why net loss widened despite a smaller operating loss inflection: interest expense jumped to $7.0M (+308%) as the full $575M convertible stack carries — financing cost is now a material drag, partially offset by $5.8M interest income on the cash/investments pile.
• Balance-sheet flags (Apr-5-2026):
• Cash $88,751k + short-term investments $439,985k = $528.7M liquidity.
• Inventory $16,451k, up from $13,617k at year-end — building Korea inventory ahead of shipments, a thing to watch (inventory growing faster than revenue can be a demand-timing risk).
• AR $3,943k (down) — clean.
• Long-term debt $520,160k (the converts).
• Accumulated deficit crossed $1.0B: $(1,016,087)k.
• Stockholders' equity ≈ $242.7M (assets $833.9M − liabilities $591.2M).
• Cash burn: operating cash used $(33,072)k for the quarter, nearly double the $(16,907)k prior-year quarter; FCF outflow ~$36.3M including the semiannual 2030-note interest payment and the Korea inventory build. Q2 capex guided $9–13M.
• Market reaction: stock rose on the print ("Q1 2026 beats forecasts, stock rises") — but the gain reversed within ~10 days when, on May 22, 2026, ENVX fell on a smartphone-battery-delay warning. The pattern: the defense numbers beat, the consumer timeline disappoints — and the market only ultimately cares about the latter.

Lens 6 · Earnings Calls (sentiment trend)

No transcripts/ on disk — sentiment read from call summaries/press releases ``.

• Tone has shifted from "validation" to "execution." Q1-26 framing (Talluri): "Enovix is entering a new phase where commercial execution must scale alongside our technology leadership" — explicitly moving the narrative "from validation of technology to disciplined execution against commercialization milestones". That is management acknowledging the credibility problem: the tech is no longer the question, shipping is.
• Recurring phrases: "qualification," "silicon-specific framework," "commercial production," "ramp through the second half," "Korea pipeline." The Korea defense pipeline was quantified at >$130M — a deliberate move to give investors a near-term revenue anchor while the smartphone story slips.
• Things they've (had to) stop saying: firm flagship-smartphone launch dates. The roadmap has repeatedly moved; the latest framing is "qualification over halfway at Honor," "targeted system-level deployments," "initial smart-frame production ramp," "convert the drone pipeline" — milestone language rather than dates. The Street read it as the smartphone inflection being "pushed out multiple quarters".
• Bottleneck honesty: a Q1-26 call was literally titled around "Progress, Cash, and Bottlenecks" — management is being relatively candid about manufacturing/laser-dicing constraints and the cash cost of the build.

Net sentiment: management credibility is the swing factor. Talluri's tone is measured and operator-credible, but the company has spent its "trust me on timing" capital — the call cadence now leans on the defense pipeline and "execution discipline" precisely because the consumer promise keeps slipping.

Lens 7 · Comps

EV/Sales reality check: Enovix at ~$1.6B equity (and ~$1.1B EV after netting ~$0.53B liquidity against ~$0.52B converts — roughly EV ≈ market cap) on $31.8M trailing revenue is ~50× trailing sales; Amprius at ~$2.30B on ~$73M is ~31× and growing into it with positive EBITDA. On a sales-and-traction basis Enovix screens expensive relative to the one peer that is actually shipping. QuantumScape carries a far larger cap on essentially no revenue — but that's a different (solid-state EV) lottery ticket, not a clean comp. P/E, EV/EBIT, dividend yield, 5-yr avg ROE: n/a — not meaningful (all loss-making, no dividends, deeply negative ROE on accumulated $1B deficit).

Lens 8 · Stock-Price Catalysts (what actually moves ENVX)

Mostly ``. ENVX is a high-beta, 31%-shorted, narrative-driven name — it moves on milestones and timelines, not earnings dollars.

Pattern of >5% moves:

• Up: manufacturing-readiness milestones (HVM Site Acceptance Testing completed Jan 6 2026), AI-1 launch (Jul 2025), independent-test validation of AI-1 energy density (Jan 2026), short-squeeze dynamics (31% SI), capital raises that de-risk runway (paradoxically).
• Down: commercialization delays — the dominant negative catalyst. March 2026 production delays from laser dicing + extended smartphone qualification → PT cuts (Craig-Hallum $10, Benchmark $25→$15). May 22 2026 drop on smartphone-battery-delay warning. Dilution/convert announcements. Insider selling (CEO/COO sold equity & warrants).
• Structural: stock down ~9% over the trailing 12 months while the S&P 500 rose ~23% — chronic underperformance as the inflection keeps receding.

What the market reacts to: one variable — credible evidence that a flagship consumer OEM will ship a 100%-silicon-anode Enovix battery at volume. Defense revenue beats barely move it; smartphone-timeline news (good or bad) moves it hard. Plus a powerful short-squeeze overlay (31% SI) that amplifies any genuinely positive surprise.

---

Phase C — Judge people & books

Lens 9 · Management

CEO — Dr. Raj Talluri (President & CEO & Director since January 2023; CODM). This is the single strongest non-technology asset in the story. ~30-year career in portable electronics:

• Micron — SVP/GM Mobile Business Unit; that BU "delivered over $7B in worldwide revenue and over $2B in operating profit in FY2022".
• Qualcomm — SVP/GM IoT BU, "drove the division from incubation to over $1B in product revenue"; also led product management for the Snapdragon flagship apps-processor roadmaps (smartphones/tablets/auto).
• Texas Instruments — 16 years, technical staff → business-unit leadership. PhD (UT Austin), 13 patents.
• Why it matters: the entire thesis is landing and shipping a flagship smartphone. Talluri has spent his career inside the smartphone-silicon supply chain selling to exactly these OEMs. If anyone can convert qualification into a design win, his rolodex and credibility are the right ones. He also brought in Steve Bakos as SVP Worldwide Sales (May 2026) to build the commercial muscle.

Chairman — Thurman John (T.J.) Rodgers (founder of Cypress Semiconductor; sponsor of the SPAC, Rodgers Silicon Valley Acquisition Corp, that took Enovix public in 2021). Held ~22.0M shares ≈ ~10% of shares outstanding at Dec-28-2025 — genuine founder-class skin in the game, and a Silicon Valley operator of real stature. But see the related-party flag below.

Capital allocation: the defining moves are financing, not reinvestment returns yet: raised $575M of convertibles ($215M 2028 @ 3.0% + $360M 2030 @ 4.75%) and acquired Routejade (2023) to vertically integrate + buy the defense revenue that now funds the company. Routejade looks like a shrewd deal in hindsight — it gave Enovix a real, gross-margin-positive revenue line and a manufacturing base while the consumer story matures. ROE/ROIC are deeply negative (accumulated deficit >$1B) — meaningless at this stage; judge them on milestones, not returns.

Red flags:

1. Related-party convertible debt. The 2028 "Affiliate Notes" were issued to an entity affiliated with T.J. Rodgers, the Board Chairman, in a concurrent private placement. The Chairman lending to the company he chairs is a textbook related-party item — disclosed and not necessarily abusive (founders backstopping is common), but it warrants scrutiny on terms and independence.
2. Insider selling. CEO/COO have sold substantial equity and warrants even as public messaging stays bullish — a classic confidence-signal mismatch for a pre-revenue-inflection name.
3. SPAC lineage. Rodgers-sponsored SPAC + private-placement warrants (6.0M to Rodgers Capital; 17.5M warrants @ $11.50 strike) — legacy dilution overhang and the usual SPAC-era promote dynamics.
4. Serial timeline misses (see Lens 12/13) — a management-credibility red flag even if the tech is real.

Archetype: a professional-operator CEO (Talluri) layered over a founder-financier Chairman (Rodgers) — arguably the right combination for this stage (commercialize + fund), but the related-party financing and insider sales temper the alignment story.

Lens 10 · Forensic Red Flags

Income-statement / balance-sheet / cash-flow risks, every figure labelled.

• Cash flow vs. earnings: the divergence runs the right way for once — net loss $(38.3)M but operating cash burn $(33.1)M for Q1-26, with large non-cash adds: SBC $11.8M + D&A $9.4M. So GAAP loss is not being flattered by cash it doesn't have; the loss is broadly real cash out the door.
• Stock-based compensation is heavy: $69.8M for FY2025 and $11.8M in Q1-26 — i.e. SBC alone is >2× total FY2025 revenue. Non-GAAP metrics (the 26.3% "GM," the $28.8M non-GAAP op loss) strip this out; watch the GAAP-to-non-GAAP bridge — the favorable non-GAAP framing leans on excluding very real shareholder dilution.
• Inventory growing faster than revenue: $16.5M vs $13.6M QoQ (+21%) on ~flat-to-modestly-up revenue. Management says it's a deliberate Korea pre-build for shipments — plausible and benign if the orders land, a write-down risk if they slip. Track this every quarter.
• Convertible-debt overhang: $520M long-term debt; interest expense $7.0M/qtr and climbing; the 2030 note's semiannual coupon is a visible cash event (drove the $36.3M Q1 FCF outflow). Converts are dilution by another name — conversion or refinancing risk into 2028/2030.
• Going concern: none asserted. Financials "prepared assuming we will continue as a going concern"; management states cash "will be sufficient to meet our funding requirements over the next twelve months from the date this … Form 10-Q is filed". With ~$529M liquidity against ~$33M/qtr burn, the literal runway is ~4 years on burn alone — but capex for capacity expansion and the 2028 convert maturity compress that, and management explicitly expects to "need to raise additional funds through … equity, equity-related or debt securities" over time. Dilution is a when, not an if.
• Single-customer revenue-recognition exposure: with ~64% of revenue from one defense buyer, any timing/dispute on that account swings the whole P&L — a concentration risk that is also a rev-rec risk.

Regulatory findings (required sub-section):

• SEC Litigation Releases: None. "No LR found for this company" via EDGAR EFTS over 2021-06-20→2026-06-20.
• SEC AAERs: None. "No AAER found for this company" same period.
• Non-SEC enforcement (web): No material FTC/DOJ/FDA/CFPB enforcement, consent decree, or government fine surfaced for Enovix in searches. (Note: Enovix has been the subject of shareholder securities class-action litigation in the past — common for volatile SPAC-era names — but that is private civil litigation, not a regulator enforcement action; no material regulatory penalty found.)
• 10-K Item 3 (Legal Proceedings): the filing carries the standard legal-proceedings framework; nothing rising to a disclosed material adverse regulatory judgment was extracted.
• Conclusion: No material regulatory or accounting-enforcement findings — verified via SEC EDGAR EFTS (LR, AAER), web search, and 10-K review as of 2026-06-20. The genuine forensic watch-items are governance/quality (related-party converts, heavy SBC, inventory build, dilution path), not fraud or enforcement.

---

Phase D — Project & stress-test

Lens 11 · Forward Projection

This is a pre-inflection, loss-making name with no FY revenue guidance — a precise multi-year EPS model would be false precision. I build a directional revenue/EPS path; output ``, inputs labelled. (Per --watchlist rules I do not log a forecast.ts Brier forecast in this loop.)

Anchors +: FY2025 rev $31.8M; Q1-26 $7.6M; Q2-26 guide $8–9M; Korea defense pipeline >$130M; FY2025 net loss $(156.6)M / EPS $(0.75); ~218M shares; ~$33M/qtr operating burn; SBC ~$70M/yr.

EPS reality: even the bull case stays loss-making through FY2028E — at this share count and opex ($180M+ FY25 operating expense base ), Enovix needs several hundred million dollars of revenue before GAAP breakeven. The investable question is therefore not near-term EPS; it is "does a flagship smartphone design win convert, and does Fab2 yield hold at volume?" — a binary, ~12–24-month question. Cash runway ($529M) reaches the catalysts; the risk is dilution before then and the inflection slipping a third time.

Lens 12 · Bull vs Bear

Bull case. Enovix owns a physically differentiated, IP-protected, independently-validated silicon-anode cell (AI-1, >900 Wh/L) at the exact moment on-device AI is making battery energy density the binding constraint on smartphones and smart glasses. It is run by a smartphone-silicon operator (Talluri) with the rolodex to land OEMs, has de-risked the manufacturing question (HVM SAT done Jan-26), is already gross-margin-positive six quarters running on a real defense business with a >$130M Korea pipeline, has ~$529M of liquidity (multi-year runway), and is 31% shorted — so a single credible flagship design win could trigger both a fundamental re-rate and a violent short squeeze. The smart-eyewear ramp (H2-26) is a nearer, lower-bar consumer beachhead that could prove the platform before the harder smartphone win.

Bear case (permanent-impairment risks).

1. The flagship never ships at volume / ships late and small. Qualification cycles "take years"; the timeline has already slipped repeatedly (laser dicing, May-22 delay warning); OEMs hold all leverage. If on-device-AI demand is met by competitors (Amprius shipping, plus incremental graphite/SiOx improvements from incumbents like ATL/Samsung SDI/LG), Enovix's window closes and the architecture becomes a great paper that never scaled.
2. Manufacturing yield at Fab2 disappoints. SAT ≠ high yield at volume. The bear thesis is precisely that the pilot-to-HVM jump destroys the unit economics; a structural yield problem would impair the whole thesis.
3. Dilution grinds the equity. ~$33M/qtr burn, $575M of converts maturing 2028/2030, explicit intent to raise more equity — even a successful Enovix may dilute holders heavily on the way, and a delayed one almost certainly raises equity at depressed prices.

Pre-mortem (18 months out, thesis broke): It's Dec 2027. The Honor program "completed qualification" but the design win went to a competitor on yield/cost, or shipped in a niche SKU at trivial volume. Fab2 HVM yields stalled below the level needed for OEM cost targets. Enovix did a dilutive equity raise in 2027 at <$5 to bridge to the 2028 converts. Revenue is ~$60M, still defense-led; the stock is a perennial "next year" story and the short thesis won.

Are multiples too high? On traction, yes — ~50× trailing sales vs Amprius ~31× with positive EBITDA and 4× the revenue. ENVX is priced as an option on a flagship smartphone win, not as a business. That's defensible if you underwrite the binary; it's expensive if you don't.

Contrarian view (what the market refuses to see): The crowd is fixated on the smartphone binary and treats defense as boring filler. But the Routejade defense business + >$130M Korea pipeline + smart-eyewear ramp may quietly compound Enovix into a $100–150M-revenue, gross-margin-positive specialty battery maker even if the flagship phone never lands — which would make today's ~$1.6B a different (lower-multiple, but real) valuation rather than a zero. The asymmetric mistake bears may make is assuming "no flagship phone = broken company," when the floor under the equity is harder than the short interest implies. Conversely, the bull mistake is assuming the squeeze is the thesis.

Lens 13 · Devil's Advocate (short-seller)

Dismantling the bull case:

• Revenue concentration is a loaded gun. ~64% from one Korean defense subcontractor. Lose or delay that account and the income statement craters — and defense procurement is exactly the kind of lumpy, politically-exposed, single-buyer revenue that can vanish for a year. The "diversified end markets" story is fiction today.
• The moat may be weaker than bulls think. A constrained-stack architecture that can't be manufactured at yield (laser dicing) is a lab trophy, not a moat. Incumbents (ATL, Samsung SDI, LG Energy, CATL) are pushing SiOx/silicon-blend graphite and have infinitely more manufacturing scale; they don't need 100% silicon to ship "good enough" higher-density cells into the same phones — and they're already in the phones.
• Most dangerous competitor bulls underrate: Amprius. Different architecture, same buyer pitch (highest energy density), but already shipping, $73M revenue tripling, positive adjusted EBITDA. While Enovix qualifies, Amprius sells. The "pure-play silicon-anode leader" crown may already be slipping.
• Worst capital-allocation / governance items: related-party 2028 converts to the Chairman's affiliate, heavy SBC ($70M/yr, >2× revenue), CEO/COO insider selling into the bullish narrative. None is fatal alone; together they read as a team monetizing/financing around a story that keeps slipping.
• What must hold for today's ~$7 price: a credible path to a flagship smartphone design win within ~18 months and acceptable Fab2 yield and no value-destroying dilution before then. All three must break right.
• If growth disappoints 20–30%: there's no earnings to compress — the damage is to the narrative and the multiple. A third smartphone-timeline slip likely takes the stock back toward the $5–6 / Benchmark-cut zone and forces a dilutive raise, which compounds the de-rate. Down 40–60% is on the table on a single bad qualification update.
• Single permanent-impairment scenario (most plausible): Fab2 HVM yields structurally miss OEM cost targets → no flagship win → Enovix becomes a sub-scale defense/eyewear battery maker carrying $575M of converts. Plausibility: moderate — this is genuinely the base-rate outcome for "revolutionary battery architecture transfers to mass production," which historically fails far more often than it succeeds.

Lens 14 · Management Questions (ordered by information value)

1. Honor / lead smartphone OEM: what specifically remains between "qualification over halfway" and a binding volume purchase order, and what is the realistic month for first commercial shipments — committed, not aspirational?
2. Fab2 HVM yield: what is current yield on the HVM line vs. the yield required to hit your lead OEM's cell-cost target, and what is the gap-closure plan/timeline (incl. the laser-dicing fix)?
3. If the flagship smartphone win does not materialize by end-2027, what does the business look like on defense + smart-eyewear + IoT alone — revenue, gross margin, and cash burn?
4. Customer concentration: what is the path to getting any single customer below ~30% of revenue, and over what timeframe?
5. Dilution: given ~$33M/qtr burn and the 2028 converts, what is your explicit financing plan — do you expect to raise equity before the 2028 maturity, and under what stock-price conditions?
6. Smart eyewear: name the magnitude — what unit volumes and revenue does the lead smart-eyewear customer's H2-26 ramp imply for FY2027, and is it a reference platform or a shipping consumer product?
7. Korea >$130M pipeline: what portion is contracted vs. pipeline, and over how many quarters does it convert to recognized revenue?
8. Non-GAAP bridge: SBC is running >2× revenue — what is the multi-year plan to bring SBC down as a % of revenue, and how should investors think about the real (post-dilution) economics?
9. Related-party converts: can you walk through the independence and terms of the 2028 Affiliate Notes issued to the Chairman's affiliate, and the board process that approved them?
10. Insider selling: how should shareholders reconcile management equity/warrant sales with public confidence in the commercialization timeline?
11. Incumbent threat: why won't ATL/Samsung SDI/LG/CATL silicon-blend cells be "good enough" for AI smartphones, eliminating the need for a 100%-silicon architecture?
12. Amprius: they're shipping with positive EBITDA while you qualify — what do you do that they can't, and where do you actually compete head-to-head for the same socket?
13. Capacity / capex: what is the total capex (and over what years) to take Fab2 to flagship-smartphone volume, and how is it funded?
14. Routejade: is the defense business a long-term strategic core or a bridge — and would you consider it for divestiture/separation once consumer scales?
15. Capital-allocation philosophy: at what milestone do you expect to generate positive operating cash flow, and what gates the decision to keep funding the consumer build vs. harvesting the defense base?

---