Onto Innovation
Capital-equipment plus recurring. Sells optical metrology/inspection systems (Atlas OCD/thin-film, Dragonfly macro-inspection, Iris films metrology), process-control software, and parts/service to leading-edge chipmakers and OSATs. Cyclical equipment sales tied to fab and advanced-packaging capex, with a growing service + software attach.
The thesis on this name
State of the Memory Supercycle
Metrology/inspection specialist with strong leverage to advanced packaging and HBM — the second name (with Camtek) gating yield in the memory packaging layer. Lithography + metrology systems for bump/panel/HBM inspection put it directly in the TSV/hybrid-bonding content-growth path as stacks get taller. Smaller and less HBM-concentrated than Camtek, so it offers a diversifying second leg on the same packaging-inspection thesis at a more reasonable cyclical multiple. Mispriced as a generic small-cap semicap when a rising share of its mix is the secularly-growing advanced-packaging inspection step.
State of the Memory Supercycle
Metrology + advanced-packaging inspection across both DRAM front-end and HBM packaging — a more diversified, US-listed peer to Camtek on the memory metrology layer.
State of the Memory Supercycle
DRAM front-end metrology + HBM packaging inspection; the more-diversified, lower-torque metrology-layer expression.
Earnings, margins, COGS & capex
Onto is exiting a shallow equipment trough: FY25 revenue was roughly flat at $1.005B (+1.8%), but Q1 FY26 reaccelerated to $291.9M (+9.5% YoY, ~+10% QoQ) on Atlas G6 advanced-node wins, Dragonfly G5 inspection ramp, and surging AI advanced-packaging demand (estimate of inflection; FY26 guide >$1.3B implies >30% growth). The GAAP/non-GAAP margin gap is wide right now because acquisition (Semilab) amortization, integration and one-time items depress GAAP gross margin to 50.1% vs 55.7% non-GAAP. The business is capital-light (capex ~1-2% of revenue) and cash-rich, though the pending Rigaku investment will deploy a large chunk of the balance sheet.
Income statement — where each revenue dollar goes
% of revenueOf every $1 of revenue, ~50¢ is cost of goods and ~39¢ operating expense, leaving ~12¢ of operating profit (~34¢ net).
Revenue trend
Margins
up
down (acquisition/integration drag)
down YoY, up QoQ
down sharply (Semilab amortization)
down YoY
down on lower op cash flow / working capital build
COGS structure
COGS is dominated by purchased components/optics/subsystems and contract-manufactured hardware for metrology and inspection tools, plus field-service/installation labor and warranty. Mix shift toward higher-ASP advanced-node (Atlas G6) and AI-packaging inspection (Dragonfly G5) lifts gross margin, while the Semilab acquisition currently adds inventory step-up and intangible amortization that depress GAAP gross margin (50.1%) well below non-GAAP (55.7%). Software/services carry structurally higher margins than systems.
Capex
Capital-light: $3.6M in Q1 FY26 (~1.2% of revenue) — funds R&D lab equipment, demo/eval tools, and facilities; not a fab. The far larger 2026 capital deployment is financial: the ~$710M Rigaku minority stake (27%, from a Carlyle affiliate), funded partly by a planned $500M bridge loan, closing 2H26 (fact).
Latest earnings
Beat: revenue $291.9M vs ~$285.5M consensus (+2.2%); non-GAAP EPS $1.42 vs ~$1.37 estimate. Stock fell ~6% after hours on a soft-ish Q2 outlook relative to high expectations (fact).
Q2 FY26 revenue $320-330M (midpoint ~+28% YoY); FY26 revenue now seen above $1.3B (>30% growth), 2H growing at least 15% over 1H; advanced packaging +>50% and advanced nodes ~+25% for FY26 (fact).
- Advanced Packaging & Specialty Device rev (Q1 FY26)
- ~$160M (~2/3 advanced packaging, ~$25M Semilab, rest specialty/power) (fact)
- Advanced Nodes rev (Q1 FY26)
- ~$80M (~60% memory/DRAM, rest logic) (fact)
- Cash + marketable securities
- $654.2M, no funded debt (Mar 31 2026) (fact)
- Non-GAAP gross margin
- 55.7% (Q1 FY26) (fact)
Growth drivers
- AI advanced packaging — HBM + 2.5D logic driving Dragonfly inspection and Iris films metrology; management guides advanced packaging revenue +>50% in FY26 (fact)
- Advanced nodes — Atlas G6 OCD/thin-film metrology adoption at GAA/nanosheet logic and leading-edge DRAM; advanced nodes guided ~+25% FY26 (fact)
- Dragonfly G5 — initial commercial shipments in Q1 FY26 expanding macro-inspection share (fact)
- Software/AI — Ai Diffract analysis software and process-control software attach, central to the Rigaku X-ray/optical hybrid-metrology alliance (fact)
- Semilab integration — added ~$25-27M revenue in Q1 FY26 in specialty/film metrology (fact)
- Memory recovery + HBM capacity — DRAM/HBM capex upcycle benefiting both front-end metrology and packaging inspection (estimate)
Reported financials — SEC EDGAR
Audited GAAP figures pulled from SEC filings · latest filing 2026-02-24. The audited primary-source spine — not financial advice.
Revenue — annual (GAAP)
Margins & balance sheet — FY’24
Bull & bear
Onto is a pure-play on the two highest-growth process-control vectors — AI advanced packaging (HBM/2.5D inspection + films metrology) and GAA/DRAM advanced-node metrology — reaccelerating from a flat FY25 to >30% growth guided for FY26, with a fortress balance sheet and a hybrid-metrology optionality via Rigaku.
- FY26 guided >$1.3B revenue (>30% growth) with advanced packaging +>50% and advanced nodes +~25% — a clear inflection off the FY25 trough (fact)
- Structural HBM/2.5D packaging demand drives Dragonfly inspection and Iris metrology; advanced-packaging metrology is a multi-year share-of-wallet expansion (estimate)
- Atlas G6 + Dragonfly G5 product cycle landing at GAA logic and leading-edge DRAM at exactly the AI-capex peak (fact)
- Capital-light, ~26-28% non-GAAP operating margins, $654M net cash funding both organic growth and the Rigaku optical+X-ray hybrid-metrology play (fact)
- Outgrowing KLA in select OCD/packaging pockets and designed in at TSMC/Samsung/Intel — credible share-gain story (fact/estimate)
A premium-multiple (trailing P/E >140, ~30-35x forward on consensus) cyclical equipment name whose growth is concentrated in HBM/DRAM capex that can digest quickly, sub-scale against KLA and cost-disadvantaged versus Camtek in packaging, with GAAP earnings muddied by the Semilab/Rigaku M&A push.
- Rich valuation leaves little margin for error — the stock fell ~6% post-Q1 on a merely in-line Q2 outlook; any HBM digestion de-rates it fast (fact)
- Memory concentration cuts both ways: a DRAM oversupply or HBM pause hits advanced nodes and advanced packaging at once (estimate)
- Competitive squeeze — KLA (~55% share, far larger R&D) above and Camtek (cost-advantaged 3D microbump metrology) inside the packaging franchise (fact/estimate)
- GAAP profitability is depressed and noisy: GAAP gross margin 50.1% vs 55.7% non-GAAP and GAAP op margin just 11.5% on Semilab amortization; Rigaku adds a $500M bridge loan and equity-method noise (fact)
- Customer concentration (Samsung/TSMC/SK Hynix historically each >10%) means a single capex cut is material (fact, dated disclosure)
What it is worth
Forward P/E and EV/EBITDA cross-check vs process-control peers (KLAC, CAMT, NVMI), anchored to FY26 guidance (>$1.3B revenue, ~$5.50-6.00 non-GAAP EPS estimate).
~$220-250 (HBM digestion / memory capex pause de-rates the premium multiple toward peers) (estimate)
~$350 (in line with ~$352 average analyst target; FY26 guide met, multiple holds) (fact for target; estimate for thesis)
~$400+ (sustained >30% growth
packaging share gains, GAAP margin recovery, Rigaku accretive) — above Street high of ~$371 (estimate)
Premium cyclical: trailing GAAP P/E >140 (depressed GAAP EPS), but ~30-35x forward non-GAAP EPS on the FY26 reacceleration; trades roughly with/above peers on growth, so it is priced for HBM/packaging share gains to keep compounding.
SWOT
Strengths
- Leading independent position in standalone OCD/optical metrology and a strong, growing share in AI advanced-packaging inspection (Dragonfly) and films metrology (Iris) — outgrowing KLA in select pockets (fact/estimate)
- Fortress balance sheet: ~$654M cash + securities and no funded debt pre-Rigaku (Mar 31 2026) (fact)
- Designed in at the leading edge — selected by TSMC, Samsung, Intel for GAA/nanosheet metrology; deep memory (DRAM/HBM) exposure at the right moment of the AI cycle (fact)
- Capital-light model (~1-2% capex/rev) with ~26-28% non-GAAP operating margins and high-margin software/service attach (fact)
Weaknesses
- Sub-scale vs KLA — a fraction of KLA's revenue and R&D budget, limiting breadth across the full process-control suite (estimate)
- Wide GAAP-to-non-GAAP gap (GAAP gross margin 50.1% vs 55.7%) from Semilab acquisition amortization/integration muddying reported profitability (fact)
- High customer concentration in a handful of memory/foundry leaders (historically Samsung, TSMC, SK Hynix each >10%) (fact, dated 2021-22 disclosure)
- Cyclical, lumpy equipment orders — FY25 was nearly flat, showing the trough drag before the FY26 reacceleration (fact)
Opportunities
- HBM/2.5D packaging structural growth — advanced-packaging metrology/inspection is a multi-year secular share-of-wallet expansion (estimate)
- Rigaku alliance — combining optical + X-ray (CD-SAXS) into hybrid metrology targets a >$1B process-control TAM within 5 years (fact, management claim)
- Gate-all-around / nanosheet logic transition increasing metrology intensity per wafer at advanced nodes (estimate)
- Software/AI (Ai Diffract) and recurring service growth to lift through-cycle margin floor (estimate)
Threats
- KLA's dominance (~55% of inspection/metrology) and its push deeper into packaging could cap Onto's share gains (fact/estimate)
- Camtek's strong, cost-advantaged position in 3D microbump packaging metrology directly contests Onto's packaging growth (fact/estimate)
- Memory capex is highly cyclical — an HBM digestion or DRAM oversupply would hit both segments simultaneously (estimate)
- Export controls / China demand risk and customer-concentration shocks (a single memory leader cutting capex) (estimate)
Moats, dependencies & bottlenecks
Moats
Iris films) and macro-inspection (Dragonfly) leading-edge designed-in wins are sticky per node, but each node transition reopens competition Selected by TSMC/Samsung/Intel for GAA metrology; Ai Diffract software differentiates (fact)
High within a node — recipes, tool matching and fab qualification make swaps costly mid-node Process-control tools requalify slowly; installed base drives parts/service pull-through (estimate)
recurring parts/service/software revenue grows with shipments Higher-margin recurring stream cushions equipment cyclicality (estimate)
this is a disadvantage vs KLA, not a moat Onto is sub-scale; KLA's R&D budget dwarfs Onto's (estimate)
Dependencies
SK Hynix, Micron, TSMC, Intel) High concentration (historically several customers >10%); a single capex cut is material (fact, dated disclosure)
FY26 growth is largely HBM/2.5D-driven; a digestion phase hits the thesis directly (estimate)
Tool COGS depends on specialized optics and electronics supply; constraints raise cost/lead time (estimate)
Each node must be re-won; metrology intensity rising is a tailwind but competition reopens (estimate)
China equipment-sales restrictions and customer-mix shifts affect TAM (estimate)
Advantages
- Pure-play exposure to the two fastest-growing process-control vectors: AI advanced packaging and GAA/DRAM advanced-node metrology (estimate)
- Strong product cycle in 2026 — Atlas G6 and Dragonfly G5 landing at peak AI capex (fact)
- Fortress balance sheet (~$654M cash, no funded debt) funding organic growth + Rigaku optionality (fact)
- Designed in at TSMC/Samsung/Intel; outgrowing KLA in select OCD and packaging pockets (fact/estimate)
- Capital-light model with high non-GAAP margins (~56% GM, ~27% OM) and growing software/service attach (fact)
Weaknesses
- Sub-scale versus KLA across breadth and R&D (estimate)
- Wide, noisy GAAP-vs-non-GAAP gap from Semilab amortization (GAAP GM 50.1% vs 55.7%; GAAP OM 11.5%) (fact)
- High customer + end-market (memory) concentration (fact)
- Cyclical, lumpy orders — FY25 was nearly flat before FY26 reacceleration (fact)
- Premium valuation that punishes any guidance wobble (stock -6% post-Q1 on in-line Q2) (fact)
- Cost-disadvantaged vs Camtek in 3D microbump packaging metrology (fact/estimate)
Bottlenecks
- Scale and R&D budget vs KLA — limits how broadly Onto can attack the full process-control stack (estimate)
- Customer concentration in a few memory/foundry leaders caps revenue diversification (fact, dated disclosure)
- Cyclicality of memory capex makes the top line lumpy and hard to forecast quarter to quarter (fact)
- Semilab/Rigaku integration and the $500M bridge loan absorb management bandwidth and balance-sheet capacity in 2026 (fact)
- Cost disadvantage versus Camtek in low-cost 3D microbump packaging metrology (fact/estimate)
Top signals & trends
Top signals
The whole reacceleration thesis hinges on these holding through 2H26 (fact)
Onto's largest swing factor; a digestion phase would be an early warning (estimate)
New product cycle is the proximate growth driver; track shipment milestones (fact)
Closing the 50.1% vs 55.7% gap would lift reported earnings quality (fact)
Management expects accretive by Dec 31 2026; watch leverage and integration (fact)
Competitive encroachment from above (KLA) and inside (Camtek) is the key share risk (estimate)
Trends
Structural demand for packaging inspection/metrology; advanced packaging guided +>50% FY26 (fact)
Raises metrology intensity per wafer, favoring OCD/thin-film tools like Atlas G6 (estimate)
Lifts both advanced-node metrology and packaging inspection — but it is cyclical (fact/estimate)
Rigaku alliance targets a >$1B process-control TAM in 5 years (fact, management claim)
Favors KLA's scale; Onto must win on niche leadership and M&A (estimate)
Caps addressable demand and adds revenue-mix volatility (estimate)
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.
Supply precision optics, lasers and detectors for metrology/inspection tools — core COGS input (estimate)
Japan-listed X-ray (CD-SAXS) partner; Onto taking a 27% (~$710M) stake to co-develop hybrid optical+X-ray metrology (fact)
Acquired film/materials-metrology operation; added ~$25-27M revenue in Q1 FY26 and supplies technology into Onto's portfolio (fact)
Build hardware subassemblies and electronics for tool production (estimate)
Leading-edge foundry; selected Onto for GAA/nanosheet metrology — historically a >10% customer (fact)
Korea-listed foundry + memory leader; historically Onto's largest customer (~25% in 2021) (fact, dated)
Korea-listed HBM/DRAM leader; key memory + packaging customer, historically >10% (fact, dated)
US DRAM/HBM maker; advanced-node + HBM packaging demand customer (estimate)
Selected Onto for advanced-node metrology; foundry + advanced-packaging customer (fact)
Outsourced assembly/test houses buying Dragonfly inspection + films metrology for advanced packaging (estimate; Amkor US-listed, ASE Taiwan-listed)
Dominant ~55% of inspection/metrology with far larger R&D; the gravitational competitor above Onto in both front-end and packaging.
Cost-advantaged leader in 3D microbump/advanced-packaging metrology-inspection; the most direct memory-packaging peer (the comparison the board flags).
Broad process + integrated metrology/eBeam offerings; competes at the edges of Onto's metrology franchise.
Japan-listed; dominant in EUV mask/actinic inspection — adjacent process-control niche, limited direct overlap.
Israel-based optical + materials metrology (OCD/XPS); direct competitor in dimensional/materials metrology.
Japan-based; CD-SEM and eBeam metrology — overlaps in advanced-node measurement.