Phase A — Understand the business
Lens 1 · Company Overview
Nova Ltd. (incorporated 1993, Rehovot, Israel; NASDAQ since April 2000, Tel-Aviv since June 2002) designs and builds metrology and process-control systems for semiconductor manufacturing. In plain terms: when a fab etches/deposits/patterns a wafer, it must measure — at sub-nanometre precision, inline, without destroying the wafer — that each layer is dimensionally, materially and chemically correct. Nova sells the instruments (and software) that do that measuring, plus the service to keep them running. It is a "picks-and-shovels of the picks-and-shovels" play: it does not make chips; it sells the yield-assurance tooling to the people who do.
- Business model / revenue split: FY25 revenue $880.6M, of which products $705.6M (~80%) and services $175.0M (~20%). Products = capital-equipment sales tied to fab capex cycles; services = extended warranty, time-and-materials, service contracts, install-base upgrades — a recurring annuity that grew for a 13th straight sequential quarter into Q1 2026.
- Customers: all leading manufacturers in logic, foundry, memory and packaging — the world's largest IC makers. Highly concentrated (Lens 4). Sold mainly to chipmakers, occasionally to other process-equipment OEMs.
- Product franchises: Dimensional metrology (the legacy scatterometry/OCD franchise), Materials metrology (incl. XPS / Sentronics) and Chemical metrology — Nova's stated three-pillar strategy. Named platforms surfacing in recent results: Metrion (record DRAM sales), WMC, Semdex (HBM bookings).
- Contract structure: capital-sale (not take-or-pay, not subscription); cyclical with fab capex. The service layer is the recurring ballast.
Lens 2 · Supply Chain
Map: specialised optical/electron/X-ray components & subsystems → Nova assembly (Israel, USA, Germany) → chip manufacturer (the fab) → end devices (AI accelerators, HBM, smartphones, etc.).
- Upstream (inputs): Nova integrates lasers, X-ray sources, electron optics & detection, vacuum systems, spectroscopic and electrochemical modules, plus heavy in-house software/algorithms (physical modelling + ML). It is a systems integrator of precision photonics/electron-optics rather than a fab itself — so its upstream chokepoints are specialised component vendors (named suppliers are not disclosed in the 20-F; n/a — not disclosed, a genuine gap).
- Manufacturing nodes: Israel (primary), USA, and Germany via Sentronics (acquired Jan 30 2025). Concentration of inventory/WIP in these "Manufacturing Facilities" is itself flagged as a risk in the filing.
- Downstream (the buyers — names): the customer base is the fab oligopoly — TSMC, Samsung, SK Hynix, Micron, Intel, and the leading Chinese foundries/memory makers (SMIC, CXMT, YMTC-type buyers). Geographically, FY25 sales went China 33% / Taiwan 29% / Korea 16% / USA 9% / Other 13% — i.e. the downstream is literally TSMC (Taiwan), the Korean memory duo (Korea), and the Chinese fab build-out (China).
- Chokepoints / single-source dependencies: the binding constraint is customer concentration, not supplier concentration — Customer A is 23% of revenue (Lens 4). The second chokepoint is geopolitical: a third of revenue routes through China under a tightening US/Dutch export-control regime (Lens 10/13).
Lens 3 · Competitive Advantages (moats)
Nova is the clear #2 in process control behind KLA — it holds roughly 25% of thin-film and critical-dimension metrology, second only to KLA, in a market KLA dominates with ~54-55% overall process-control share. The moat is real but narrower than KLA's:
- Switching costs / recipe lock-in (the core moat): metrology tools are qualified into a customer's process recipe at a specific node; once Nova's tool is the reference measurement for a fab's 2nm or advanced-DRAM line, ripping it out mid-node risks yield. This is why the install base compounds and services grow every quarter. Durable.
- Technical breadth (the widening moat): Nova deliberately diversified from Dimensional-only into Materials + Chemical metrology and bolted ML onto physical modelling — expanding its served market and making it a multi-technique supplier rather than a one-trick OCD vendor. The Sentronics (Materials/XPS) deal extends this.
- Bargaining power — asymmetric. Over customers: moderate — the fabs are giant and concentrated (they need a metrology answer, but KLA is the default, so Nova competes hard on differentiation/price). Over suppliers: high — it's a systems integrator buying components. Net: Nova needs the big fabs more than they need Nova, which caps pricing power vs. a true monopolist.
- R&D intensity as a moat-maintenance cost: FY25 R&D (net) $143.4M = 16.3% of revenue; ~682 R&D staff incl. ~163 PhDs. This is the price of staying #2 in a market where process-control intensity keeps rising — not optional.
Moat verdict: a strong specialist moat (recipe lock-in + multi-technique breadth) sitting underneath KLA's category-defining one. Nova is the best house on a street KLA owns.
Lens 4 · Segments
Nova reports as one operating segment (CODM = the CEO, who manages on consolidated operating margin and net income) — so there is no GAAP product-line P&L. Disaggregation available:
| Cut | FY2023 | FY2024 | FY2025 | Trend / cause |
|---|
| Total revenue | $517.9M | $672.4M | $880.6M | +30% then +31% — accelerating |
| Products | $405.0M | $538.4M | $705.6M | +31% YoY; ~80% of mix |
| Services | $112.9M | $134.0M | $175.0M | +31% YoY; ~20% of mix; 13 consecutive sequential-growth quarters |
| Product gross margin | 60%¹ | 61% | 60% | −1pt; mix/ramp cost |
| Service gross margin | n/a | 44% | 47% | +3pt; scale economics |
Geographic (% of sales): China 33% (was 39% '24 / 36% '23 — declining share, export-control + local-competition headwind), Taiwan 29% (was 20% / 18% — rising fast, the TSMC 2nm ramp), Korea 16%, USA 9%, Other 13%.
Demand mix (Q1 2026, web): Memory = 34% of product revenue and the standout driver, with advanced DRAM ~two-thirds of the memory business and record Metrion sales; HBM bookings on WMC/Semdex flagged robust. Read: the growth engine has rotated from foundry/logic toward AI-memory (HBM/advanced DRAM), which is the single most important fact in the forward story.
¹ FY23 product GM stated as "60%" in the FY25 20-F MD&A narrative; the consolidated gross profit % was 57% in FY23/FY25 and 58% in FY24.
Phase B — Measure performance
Lens 5 · Earnings Result
Two prints matter: the FY2025 annual (on-disk) and the Q1 2026 (web).
FY2025 (20-F):
- Revenue $880.6M, +31% YoY. Gross profit $505.2M (57% GM). Operating income $253.5M (28.8% op margin). Net income $259.2M (29.4% net margin) — net income > operating income because of $49.8M financial income (interest on a ~$1.65B cash/securities pile).
- EPS: basic $8.61, diluted $7.96 (FY24 diluted $5.75; FY23 $4.28) — diluted EPS +38% YoY.
- Opex discipline: R&D $143.4M, S&M $82.2M, G&A $26.1M; total opex grew 26% vs revenue +31% → operating leverage.
- Cash flow: operating cash flow $245.6M; capex $27.7M ⇒ FCF ≈ $217.9M (~25% FCF margin). Tax rate ~14.5% (Israeli preferred-enterprise regime).
Q1 2026 (reported 2026-05-14):
- Revenue $235.3M (+10% YoY, +6% QoQ) — beat the $222-232M guide. GAAP net income $69.3M, diluted EPS $2.04; non-GAAP EPS $2.33 (also above guide). GM 57.7%; non-GAAP operating margin 34%.
- Q2 2026 guide: revenue $245-255M, GAAP EPS $2.10-2.24, non-GAAP EPS $2.34-2.48.
- Balance-sheet flags (FY25 20-F): inventory $183.7M (+17% YoY, outpacing the +10% Q1 revenue rate — watch), receivables $151.9M (+9%, in line). Net cash ~$914M after the convert (cash+deposits+securities $1,646M − $731.7M notes).
- Market reaction: stock fell ~8% on the Q1 print despite the beat — a classic "priced for more" reaction at a 35-48x multiple. The stock then ran to a $615.99 ATH by mid-June before pulling back to ~$554.
Unusual vs. own history: the financing line. In Sep 2025 Nova issued $750M of 0% convertible senior notes due 2030 and bought ~$53M of capped calls; the older 2025 converts (~$181M) fully converted, adding 2.43M shares. So the share count and the balance sheet both took a structural step.
Lens 6 · Earnings Calls (sentiment trend)
No transcripts on disk (transcripts=0). From web call coverage:
- What management keeps saying: "record," "memory," "advanced DRAM," "HBM," "share gains," and the $1B-revenue-in-2027 target (now described as "on track," with 2027 orders already underway and a new Asia production facility online by end-2026). Tone over the last ~3-4 calls has shifted from cyclical-recovery framing (2024) to confident secular-growth framing (2025-26) as memory inflected.
- What they stopped emphasising: macro/cyclical caution and China softness — China share is quietly declining (39%→33%) but the narrative now leads with Taiwan/memory, not China risk.
- Sentiment read: clearly bullish/expansionary management voice; the risk is that the tone has caught up with (or exceeded) the valuation, leaving little tolerance for an in-line quarter.
Lens 7 · Comps
Peer set = the process-control / metrology-inspection group. Multiples are, dated; where unsourced, n/a. Do not treat as precise — sources conflict and dates vary.
| Company | Ticker | Mkt cap | Fwd P/E | EV/EBITDA | EV/Sales | Div yield | ROE | Note |
|---|
| Nova | NVMI | ~$17.6B [est] | ~34.5x / ~48x (conflict) | ~31.7x | n/a | 0% | 24.2% | #2 process control; net cash ~$914M |
| KLA | KLAC | ~$227.7B | ~71.7x (trailing) | ~38.7x | n/a | ~0.5% (est) | n/a | ~54% share; the monopoly |
| Onto Innovation | ONTO | n/a | ~35.9x | n/a | n/a | 0% | n/a | advanced-packaging-led; Q1'26 rev ~$291.9M |
| Camtek | CAMT | ~$8.86B | ~35x | ~28x | n/a | ~1.6% | 24.3% | Israeli peer; inspection/packaging; LTM rev ~$471.6M |
Read: Nova trades in line with the smaller specialists (Onto/Camtek at ~35x) and at a discount to KLA's ~70x trailing — but KLA's multiple is partly a quality/monopoly premium and partly trailing-vs-forward optics. On forward earnings the group clusters ~35x; on the higher ~48x quote Nova would be the expensive one. Nova's distinguishing features in the comp set: highest revenue growth rate (+31%), zero dividend, strong 24% ROE, and a fortress net-cash balance sheet that none of the smaller peers match. KLAC's ~38.7x EV/EBITDA being "151% above its 10-year median" tells you the entire group is rich, not just Nova.
Lens 8 · Stock-Price Catalysts (what moves NVMI >5%)
From the pattern over the cycle:
- Earnings prints & guidance are the dominant mover — and the reaction is asymmetric to expectations, not absolutes: Q1 2026 fell ~8% on a beat because the bar was higher.
- Memory/HBM datapoints — the rotation to memory (34% of product rev) makes NVMI increasingly a DRAM/HBM-capex proxy; HBM bookings commentary moves it.
- The $1B-2027 target & the new Asia fab — milestone catalysts management has explicitly dangled.
- China export-control headlines — a third of revenue; new BIS/Dutch rules are a recurring down-catalyst.
- The AI-capex macro — as a second-derivative AI-infrastructure name, NVMI swings with the NVDA/TSMC/memory complex and with any "AI-capex-peaking" scare.
- Recent tape: ATH $615.99 (2026-06-15) → ~$554 (2026-06-17), i.e. a ~10% pullback from the high inside a week — momentum-driven, on no disclosed company-specific news.
Phase C — Judge people & books
Lens 9 · Management
- CEO: Gabriel ("Gaby") Waisman — appointed CEO & President Feb 2023; previously Nova's Chief Business Officer since 2016, 20+ years in global public-company roles across Israel/US/Asia. An internal-promotion, commercially-grown CEO, not a founder.
- Executive Chairman: Eitan Oppenhaim — CEO 2013-2023 (joined 2010); ran the transformation from a ~$200M sub-scale OCD vendor into the diversified $880M multi-technique #2. Now Exec Chairman — so the architect of the growth era is still at the top table. Strong, quantified track record: revenue roughly 4x'd over his tenure with margin expansion.
- Skin in the game / insider ownership: insider holdings not on disk (
insider-transactions.csv absent) — n/a precisely; Nova is a widely-held, no-controlling-shareholder Israeli tech (no founder block disclosed in the 20-F cover).
- Capital-allocation history: disciplined and shareholder-aware: modest, steady buybacks ($25M program '18; $30M '24; $35M '25 at cost) — used to offset SBC dilution, not lever the balance sheet. No dividend (correct for a 31%-grower). The big '25 move — $750M 0% convertible + capped calls — is a sophisticated, low-cost financing (zero coupon, dilution capped) to fund growth/M&A and the Asia fab; not value-destructive, but it did add the 2.43M-share convert overhang. M&A: the Sentronics tuck-in (~$56M) extends Materials metrology — small, strategic, on-thesis.
- Capital-allocation scorecard: ROE 24.2%, ROIC well above cost of capital given ~$253M operating income on a largely cash-funded balance sheet. Reinvest-first, buy-back-to-offset-dilution, opportunistic tuck-in M&A. Grade: A-.
- Red flags: none material. Comp not flagged as excessive; no related-party deals in the filing; the only "promotional" risk is the management tone outrunning results at a rich multiple.
- Archetype: professional operators, founder-grade continuity (the long-tenure Chairman + internally-grown CEO) — the right profile for a scaling specialist that needs execution, not founder-visionary risk-taking.
Lens 10 · Forensic Red Flags
Forensic read of the FY2025 20-F. This is a clean set of books. Specifics:
- Revenue recognition: standard ASC 606 capital-equipment + service; deferred revenue $67.2M (down from $72.9M) — no channel-stuffing tell; the deferred-revenue decline alongside revenue growth is mild and explained by timing, not a red flag.
- Cash vs. earnings: operating cash flow $245.6M vs. net income $259.2M — a ~0.95x cash-conversion. Slightly below 1.0x, driven by inventory build (+$19.0M) and other working-capital uses, not by aggressive accruals. Healthy.
- Receivables/inventory vs. revenue: receivables +9% vs revenue +31% (conservative — collecting faster than growing). Inventory +17% (FY25) and the absolute level keeps rising — the one line to watch; in a capital-equipment up-cycle this is normal pre-positioning, but if memory demand stalls it becomes the first impairment risk.
- SBC / non-GAAP: SBC $26.3M = 3.0% of revenue — modest by semi-equipment standards. The non-GAAP-to-GAAP EPS gap (Q1'26 $2.33 vs $2.04) is ~14%, driven mostly by SBC and intangible amortisation — reasonable, not abusive.
- Goodwill/intangibles: goodwill $90.8M (+$42.5M from Sentronics), intangibles $45.8M — small relative to $2.36B assets; low impairment risk; one reporting unit, tested annually.
- Convertible accounting: the $750M 0% convert sits at $731.7M net (non-current); EPS dilution handled via if-converted; capped calls economically offset upside dilution. Transparent.
- Internal controls: SOX 404(b) attestation clean; Sentronics (acquired Jan '25) excluded from the first-year ICFR assessment (standard, ~2% of assets / ~6% of revenue).
Regulatory findings (required sub-section):
- SEC Litigation Releases / AAERs: None. EDGAR EFTS search (LR + AAER, 2021-06-20 → 2026-06-20) returned 0 findings.
- Non-SEC enforcement (FTC/DOJ/etc.): None found. Targeted web search ("Nova Ltd" + FTC/DOJ/SEC/settlement/fine/penalty/lawsuit/class-action, 2024-2026) surfaced only unrelated "Nova" entities (Fashion Nova FTC; the NovaTech crypto fraud) — no hits on NVMI the semiconductor company.
- Item 8 / Legal Proceedings (the company's own disclosure): the 20-F carries only ordinary-course commitments & contingencies (Note 13); no material litigation disclosed.
- Conclusion: No material regulatory or legal findings — verified via SEC EDGAR EFTS (LR, AAER), targeted web search, and the 20-F's own legal/contingency disclosure as of 2026-06-20. The regulatory risk that matters here is prospective export-control, not enforcement (see Lens 13).
Phase D — Project & stress-test
Lens 11 · Forward Projection
Build bottom-up from FY2025 actuals + Q1 2026 + the Q2 guide + management's $1B-2027 target. All outputs; inputs labelled. Fiscal years = FY2026 / FY2027 / FY2028 (Dec year-ends).
Anchors: FY25 revenue $880.6M, GAAP diluted EPS $7.96. Q1'26 $235.3M; Q2'26 guide midpoint $250M. H1'26 run-rate ⇒ ~$485M ⇒ a ~$1.0B FY26 is plausible if H2 holds the memory ramp. Management guides $1B for 2027, implying they may be conservative on 2026.
| Path | FY2026e rev | FY2027e rev | FY2028e rev | FY2028e GAAP dil. EPS | Logic |
|---|
| Bull | ~$1.05B | ~$1.30B | ~$1.55B | ~$13.5 | Memory/HBM super-cycle holds, Taiwan 2nm ramps, share gains continue, op-margin → 32%; |
| Base | ~$0.98B | ~$1.15B | ~$1.32B | ~$10.8 | Hits the $1B in '27 on schedule; op-margin steady ~29-30%; memory normalises but stays the driver; |
| Bear | ~$0.90B | ~$0.92B | ~$0.98B | ~$7.5 | Memory capex digestion in '27, China share keeps bleeding to local tools, margin −2pt; |
What the price implies: at ~$554 and base FY26 GAAP EPS ~$10, the forward P/E ≈ 55x on GAAP / ~48x on non-GAAP — demanding. To "grow into" ~30x by FY28 the company needs the bull EPS (~$13.5), i.e. the market is asking for the super-cycle, not the base. One source pegged the implied ask at a ~35% 5-year earnings CAGR — directionally consistent with "priced for the bull case."
No forecast.ts create is run — per --watchlist rules, the Brier forecast is only logged on genuine conviction commitment, which is a human-gated /thesis step. Candidate forecast to log there: "NVMI FY2027 GAAP diluted EPS ≥ $10.50, p≈0.45, resolves 2027-12-31."
Lens 12 · Bull vs Bear
Bull case. Nova is the #2 in a market whose intensity is structurally rising — process control went from 5.9% → 8.6% of WFE in seven years, and the hard part of the AI buildout (HBM stacking, advanced DRAM, 2nm GAA + backside power) is exactly where metrology demand compounds fastest (sub-3nm = 9.25% CAGR, the fastest sub-segment). Nova has out-grown the process-control market (27.5% product CAGR vs ~16% sector, 2020-25 ) by widening from Dimensional into Materials + Chemical metrology, and it now rides the single best capex theme in tech (AI memory) with record memory bookings. Fortress balance sheet ($914M net cash), 24% ROE, 25% FCF margins, disciplined capital allocation, and a credible $1B-2027 target with orders already in hand. If HBM/advanced-DRAM capex stays elevated through 2027, the bull EPS (~$13.5 by FY28) is reachable and the multiple is justified.
Bear case (permanent-impairment risks).
- Customer concentration → demand cliff. Customer A = 23% of revenue (up from 18%), top-5 = 51%. A single large customer cutting/delaying capex (a TSMC or memory-maker air-pocket) takes a fifth of revenue with it — and metrology is capex, so it's volatile, not annuity.
- China (33% of sales) is a structural one-way risk. Export controls can shut the high-end overnight; local Chinese metrology champions (favoured by policy) erode the rest. China share is already declining 39%→33% — the bear says this continues to single digits.
- Valuation is the risk. At ~35-48x forward the stock has priced in the super-cycle. An in-line quarter (it beat Q1 and still fell 8%) or a memory-capex pause de-rates a 45x multiple violently — a 35% growth-CAGR ask leaves no margin for error.
Pre-mortem (18 months out, thesis broke): It's late 2027. HBM capex over-shot in 2025-26 and the memory makers are digesting — Nova's memory line (the whole growth story) goes from +record to flat. Simultaneously a BIS rule tightens China advanced-node tooling and Chinese locals take the mature-node metrology share. Revenue stalls near ~$950M, the $1B-2027 target slips, and a stock that was 48x forward re-rates to 25x on now-flat earnings — a ~50% drawdown that has nothing to do with the business being "bad" and everything to do with the price.
Contrarian view (what the market refuses to see): the bull tape treats NVMI as a secular compounder, but a third of its revenue is the most politically fragile demand in the industry (China) and a fifth is one customer. The market is paying a monopoly-quality multiple for a #2 with concentration risk. The contrarian flip side: if you strip out the China third and value only the "allied-fab + memory" two-thirds, the durable business is smaller and cheaper than the headline cap implies — but also cleaner, and that's the version that survives a China-decoupling.
Lens 13 · Devil's Advocate (short-seller)
Dismantling the bull case.
- Where the money structurally breaks: Nova sells capital equipment into a cyclical capex budget controlled by ~5 customers. Strip the cycle and you have a great business; mis-time the cycle and a 31%-grower is a -20% decliner — and the multiple assumes the up-cycle is permanent.
- Revenue concentration / the shift: 23% in one customer, 33% in one country. If Customer A in-sources or dual-sources metrology, or if China is export-control-fenced, the growth story (not just a quarter) is impaired. The China share has fallen three straight years — that's not noise, it's a trend the bulls are narrating around.
- Why the moat is weaker than bulls think: Nova is #2, not #1. KLA (~54%) sets the reference and out-spends Nova ~3-4x on R&D in absolute dollars. In any segment where KLA chooses to push, Nova defends with price. The "multi-technique breadth" is real but is also KLA's strategy, executed with a far bigger chequebook.
- Most dangerous competitor bulls underestimate: not KLA — Onto Innovation in advanced packaging, the fastest-growing adjacency (Onto guiding >30-50% advanced-packaging growth, a >$240M HBM purchase agreement). If packaging metrology becomes the prize and Onto owns it, Nova's memory-led growth narrative has a hole. Plus policy-backed Chinese local metrology for the mature-node third.
- Worst capital-allocation reads: none egregious — but the $750M convert + 2.43M-share conversion quietly diluted holders, and a 0% convert is cheap only if the stock cooperates; at a 48x multiple that's a bet.
- Assumptions that must hold for ~$554: (a) memory/HBM capex stays at super-cycle levels through 2027; (b) China doesn't fall below ~25% of sales; (c) op-margin holds ~30%+; (d) no large-customer air-pocket. Break any one and the multiple, not the business, is the loss.
- If growth disappoints 20-30%: FY28 EPS drops from base ~$10.8 toward bear ~$7.5; a sympathetic de-rate to ~25-28x ⇒ ~$190-210 fair value, i.e. ~60% downside from ~$554. That is the asymmetry the short presses.
- Single permanent-impairment scenario & plausibility: full US/allied export-control fence on China advanced-node + Chinese-local displacement of the mature-node third ⇒ permanent loss of ~25-30% of revenue with no near-term replacement. Plausibility: moderate (20-35% over 3 years) — high enough to demand a margin of safety the current price doesn't offer.
Lens 14 · Management Questions (ordered by information value)
- Of FY25 revenue, how much is HBM/advanced-DRAM-specific, and what is your assumed memory-capex trajectory underpinning the $1B-2027 target — what happens to that number if memory capex is flat in 2027?
- Customer A is now 23% of revenue (up from 18%). Is this concentration rising further, and what is your single-customer exposure ceiling before it changes how you run the business?
- China fell from 39%→33% of sales in two years. Decompose that: how much is export-control-driven vs. local-competitor share loss, and where do you see China settling?
- Which Chinese domestic metrology competitors do you now lose deals to at mature nodes, and on what dimension (price, policy preference, "good-enough" spec)?
- In advanced-packaging metrology — the fastest-growing adjacency where Onto is aggressive — what is Nova's share today and your right-to-win versus Onto and KLA?
- KLA outspends you several-fold on R&D in absolute dollars. In which specific segments are you gaining share on KLA, and where are you structurally ceding?
- The $750M 0% convert + capped calls: walk through the dilution math at various share prices, and what the proceeds actually fund (Asia fab? M&A? buffer?).
- Inventory rose +17% in FY25 while Q1'26 revenue grew +10%. How much is HBM/2nm pre-positioning vs. demand you may not realise, and what's the write-down risk if memory pauses?
- What is the service-revenue ceiling as a % of mix, and its steady-state gross margin — how much of the model can become recurring annuity?
- The new Asia production facility (online end-2026): capacity added, capex, gross-margin impact, and which customers/geos it de-risks?
- Sentronics (Materials/XPS): integration status, revenue contribution trajectory, and is more Materials/Chemical M&A coming?
- At ~30%+ operating margin, where is the incremental operating leverage — can margins expand from here, or is R&D intensity a structural cap?
- Capital-return policy: at ~$914M net cash and 25% FCF margins, why no dividend, and at what cash level does buyback/dividend scale up?
- What is your win-rate at the 2nm/18A node versus prior nodes — are you holding, gaining, or losing reference-tool position at the leading edge?
- If forced to name the one variable that most determines whether Nova is a $2B or a $1B revenue company by 2030, what is it?