EXECUTIVE OVERVIEW:

Onto Innovation is now a concentrated AI process-control growth story rather than a generic small-cap wafer-fab-equipment derivative. The company has moved from a 2025 transition year, when GAAP profitability was depressed by restructuring, Semilab acquisition costs, inventory write-downs, and factory-transition costs, into a 2026 cycle defined by accelerating advanced packaging demand, advanced-node metrology share gains, and a potentially important strategic repositioning into X-ray and hybrid optical/X-ray metrology through Rigaku. The equity has already discounted a meaningful portion of this improvement, with Bloomberg data showing the stock near $284.67, a market capitalization of roughly $14.2 billion, enterprise value of roughly $13.5 billion, a 12-month total return above 220%, and valuation at approximately 40.3x FY2026E adjusted EPS and 29.3x FY2027E adjusted EPS. That valuation is demanding on current earnings, but not obviously excessive if 2027 EPS migrates from Bloomberg consensus near $9.73 toward the more aggressive sell-side bull case of $11.34+ and if Dragonfly G5 share recapture becomes a multi-customer, multi-application ramp rather than a single-customer qualification event.

The central investment debate is no longer whether Onto can grow in 2026. That is now largely de-risked by the Q1 beat, Q2 guide, backlog commentary, and management’s explicit expectation for revenue growth above 30% in 2026. The more important question is whether the company can sustain a multi-year earnings revision cycle through 2027 and 2028 as advanced packaging inspection intensity, HBM 3D metrology, panel-level packaging, gate-all-around OCD metrology, and hybrid optical/X-ray metrology collectively expand Onto’s served addressable market. Q1 2026 revenue was $291.9 million, up 9.5% year-over-year, with non-GAAP gross margin of 55.7%, non-GAAP operating margin of 26.7%, and non-GAAP EPS of $1.42. Q2 guidance of $320 million to $330 million, non-GAAP gross margin of 56.0% to 56.5%, non-GAAP operating margin of 28.0% to 28.6%, and non-GAAP EPS of $1.65 to $1.73 implies a sharp sequential acceleration and supports an H1 2026 revenue base of approximately $617 million at the midpoint. Management’s statement that H2 revenue should be at least 15% above H1 revenue mathematically implies FY2026 revenue above $1.33 billion, almost exactly aligned with Bloomberg consensus and meaningfully above the original 2025 run-rate. (Onto Innovation)

The bull case rests on 4 pillars. The 1st is advanced packaging, where management now expects revenue growth above 50% in 2026, driven by AI logic packaging, HBM, 3D bump metrology, silicon photonics, panel-level packaging, and the commercial ramp of Dragonfly G5. The 2nd is advanced nodes, where Atlas G6 and integrated metrology adoption at gate-all-around logic and advanced DRAM customers support approximately 25% expected growth in 2026. The 3rd is margin expansion, with management guiding quarterly gross-margin progression and Q4 2026 exit operating margin above 30%, while GF Securities forecasts operating margin rising from 25.4% in FY2025 to 30.0% in FY2026 and 34.9% in FY2027. The 4th is strategic optionality from Rigaku, which gives Onto a credible path into X-ray metrology and hybrid optical/X-ray workflows at a point when increasingly 3D structures, buried interfaces, exotic materials, and hybrid bonding may make purely optical techniques less sufficient over time.

The bear case is also substantive. The stock is no longer early. Bloomberg data show the shares are up approximately 80% year-to-date and more than 220% over 12 months, with beta of approximately 1.86 and a recent 52-week high of $316. A 29x FY2027E adjusted P/E multiple leaves limited tolerance for any evidence that G5 adoption is narrower than expected, that Onto’s prior CoWoS share loss was structural rather than cyclical, that KLA/Camtek/Nova/Applied Materials pressure pricing or share, that AI capex momentum slows, or that H2 2026 pull-ins create an air pocket in 2027. A 30x multiple on Bloomberg FY2027E EPS of $9.73 implies a value near $292, close to the current stock price, while 25x implies approximately $243. Conversely, 35x GF Securities’ FY2027E EPS of $11.34 implies roughly $397, which matches the GF target and implies approximately 39% upside from the latest market price. The equity therefore requires either upward estimate revisions, sustained premium multiple support, or both.

COMPANY OVERVIEW AND BUSINESS MIX: Onto Innovation is a process-control equipment and software company serving semiconductor front-end wafer fabs and advanced packaging facilities. The company’s portfolio includes optical critical dimension metrology, thin-film metrology, macro defect inspection, 3D metrology, wafer and package inspection, advanced packaging lithography, process-control software, factory analytics, and, through recent strategic moves, surface charge metrology and X-ray-enabled hybrid metrology. The company was created by the 2019 merger of Nanometrics and Rudolph Technologies, combining stronger front-end OCD/metrology exposure with advanced packaging inspection, lithography, and software. Needham describes Onto as having #1 positions in OCD and macro defect inspection and ranking #3 in the approximately $6 billion process-control segment of WFE, behind larger incumbents.

Onto’s strategic relevance comes from the fact that AI infrastructure scaling increasingly depends on process-control intensity rather than only lithography, etch, deposition, or test. The AI accelerator supply chain is pushing multiple simultaneous transitions: leading-edge logic at 3nm, 2nm, and sub-2nm gate-all-around; HBM stack scaling; 2.5D logic packaging such as CoWoS-like architectures; silicon photonics and co-packaged optics; larger packages; denser micro-bumps; potential panel-level packaging; and eventually hybrid bonding. Each transition increases the need to measure, inspect, classify, model, and control increasingly small and increasingly complex structures at multiple points across the manufacturing flow. Onto’s products sit in that yield-enablement layer, where tool attach rates are driven by defect sensitivity, throughput, cost of ownership, and qualification status.

Onto reports 1 segment, but the economic exposure can be separated into advanced nodes, specialty devices and advanced packaging, and software/services. In Q1 2026, specialty devices and advanced packaging revenue was approximately $160 million, of which roughly 2/3 was advanced packaging, $25 million related to Semilab, and the balance was specialty devices including power semiconductors. Advanced nodes revenue was approximately $80 million, with 60% tied to memory, primarily DRAM, and the balance tied to logic. Software and services represented the remainder. This mix matters because the equity multiple is being driven by the growth and margin profile of advanced packaging and advanced nodes rather than by lower-growth specialty or power-semiconductor exposure.

MARKET BACKDROP: The external demand backdrop is unusually supportive. SEMI’s April 2026 300mm Fab Outlook projects worldwide 300mm fab equipment spending rising 18% to $133 billion in 2026 and 14% to $151 billion in 2027, with AI chip demand, advanced-node investment, memory demand, regional supply-chain localization, and data-center infrastructure driving the cycle. SEMI also projects 2027 to 2029 logic and micro investments of $228 billion and memory investments of $175 billion, with DRAM spending supported by HBM demand and 3D NAND spending supported by AI inference and storage demand. (SEMI)

TSMC’s 1Q26 commentary reinforces the same point. TSMC stated that 2026 capex is expected to be toward the high end of its $52 billion to $56 billion range, that AI-related demand remains extremely robust, that full-year 2026 revenue is expected to grow above 30% in USD terms, and that strong AI demand is driving incremental N3 capacity additions, including new 3nm fabs in Taiwan and Arizona with production scheduled for 2027. TSMC also characterized capacity as tight into 2027 and said it is working with suppliers to speed equipment pull-ins. This is directly relevant to Onto because Atlas G6, integrated metrology, Dragonfly G5, 3Di, Iris, JetStep, Firefly, and related software are not discretionary in the same way as lower-priority capacity additions; they are tied to yield, ramp speed, and cost-of-ownership in scarce AI capacity. (TSMC)

The industry setup is particularly favorable for inspection and metrology because advanced packaging is becoming a system-performance bottleneck rather than a back-end commodity process. Historically, back-end packaging inspection was less strategically important than front-end process control. That distinction is eroding as expensive AI packages combine large logic die, multiple HBM stacks, reticle-size constraints, high-density interconnects, and complex substrates. Onto’s 2025 annual report emphasizes that inspection rates for advanced packages are high throughout assembly because a single defective chip can destroy the value of an expensive package, and that advanced packaging revenue is more tied to assembly volume than to the cyclicality of node shrinks. That distinction is important: advanced-node metrology tends to spike around node introductions and capacity additions, while advanced-packaging inspection can scale with AI package volume and inspection intensity.

OPERATING PERFORMANCE AND NEAR-TERM EARNINGS INFLECTION: FY2025 was optically mixed but strategically important. Revenue was $1.005 billion, up only 2% from FY2024, while GAAP gross margin declined to 49.7% from 52.2% and GAAP operating margin declined to 13.2% from 19.0%. The decline was not primarily a demand problem; it reflected write-downs of excess and obsolete inventory, restructuring costs, contract-manufacturing setup costs, Semilab transaction and amortization costs, R&D project costs, and other transition items. Systems and software revenue was $847.8 million, or 84% of total revenue, while parts and services were each 8%. FY2025 free cash flow remained strong at roughly $300 million in Bloomberg data despite a reported GAAP margin reset, showing that the business retained cash-generation strength even during the transition year.

Q4 2025 marked the beginning of the current inflection. Revenue reached a record $267 million, up 22% sequentially, with advanced packaging and specialty devices contributing approximately $145 million, advanced nodes revenue up slightly more than 30% sequentially to $72 million, and management highlighting that 2.5D packaging orders for AI devices more than doubled in the quarter. The company also disclosed a volume purchase agreement from an HBM customer valued at more than $240 million through 2027, including more than $60 million for 3D bump metrology systems. That VPA is particularly important because it validates Onto’s role in HBM process control, extends visibility, and demonstrates that the 3Di opportunity is not merely an R&D attach but a production-ramp requirement.

Q1 2026 then converted the order narrative into revenue and margin execution. Revenue of $291.9 million exceeded the high end of initial guidance, and Q2 guidance moved to $320 million to $330 million. Non-GAAP gross margin expanded to 55.7% in Q1 and is guided to 56.0% to 56.5% in Q2 despite cost headwinds from memory, fuel, and shipping. Management expects gross margin to improve by at least 50 bps per quarter in Q3 and Q4 and expects Q4 operating margin above 30%. This is an important operating leverage inflection: if revenue approaches $400 million per quarter in late 2026, the model can potentially move from mid-20s non-GAAP operating margin into the low-30s, with incremental margin enhanced by G5 mix, software, factory utilization, and offshore/extended-factory benefits.

The H1/H2 setup is also notable. Q1 actual revenue of approximately $292 million plus Q2 midpoint guidance of $325 million implies H1 revenue of approximately $617 million. Management’s statement that H2 revenue should be at least 15% above H1 implies H2 revenue above $710 million, or FY2026 revenue above $1.326 billion. Bloomberg consensus of $1.330 billion therefore appears aligned with the company’s minimum H2 framework rather than an aggressive bull case. GF Securities’ FY2026 revenue estimate of $1.418 billion requires H2 revenue of roughly $801 million, or an average of approximately $400 million per quarter. That higher forecast requires continued pull-ins, strong G5/3Di shipments, and advanced-node strength, but it is not implausible given the company’s comments that customer pull-ins are not coming at the expense of 2027 demand and that 2027 continues to look stronger than 2026.

ADVANCED PACKAGING: Advanced packaging is the highest-conviction growth engine. Management originally framed advanced packaging as growing above 30% in 2026 after Q4 2025, but upgraded the outlook to above 50% after Q1 2026. The change is significant because it indicates that incremental order momentum, G5 qualification, HBM demand, panel-level packaging adoption, and broader customer pull-ins accelerated within a short window. The business is being driven by 2.5D logic packaging, HBM, 3D bump metrology, silicon photonics, and panel-level packaging, with the common thread being the need to improve yield in packages whose economic value is far higher than traditional back-end packages.

Dragonfly G5 is the most important individual product cycle. The G5 platform was qualified at a leading 2.5D logic customer and at an HBM customer, establishing a new inspection platform with higher resolution, higher throughput, and multi-sensor flexibility. Management stated that the system can now detect features below 200 nanometers, versus historical limits near 800 nanometers, and that the platform was built from the ground up with new optics, camera, and staging while leveraging Onto’s historical packaging expertise around die warp, wafer warp, rough metal surfaces, and packaging-specific algorithms. This matters because a large portion of the historical bear case was that Onto had lost share in CoWoS-related inspection and might lack the next-generation optical performance to regain it. G5 does not fully eliminate that risk, but the leading 2.5D logic qualification materially improves the probability that the bear thesis has shifted from “structural share loss” to “execution and breadth of share recapture.”

The near-term contribution from G5 is still modest, which is both a risk and an opportunity. Management indicated that G5 was likely less than 10%, potentially closer to 5%, of the 2026 advanced packaging growth contribution at the time of the Q1 call. Shipments are ramping from a handful of Q1 systems into larger Q2, Q3, and Q4 volumes, with requests for pull-ins. The larger impact is expected in 2027, when G5 becomes a higher percentage of inspection revenue and when the platform’s cost-of-ownership and margin advantages versus G3 should be more visible. The 2027 setup is therefore highly sensitive to G5 application-conversion rates across the pipeline of more than 15 applications and more than 10 customers.

3Di metrology is the 2nd major advanced packaging driver. Management highlighted that advanced bump heights have moved from approximately 15 microns to 25 microns 2 years ago to samples below 6 microns today. That progression increases the importance of coplanarity, bump-height measurement, and 3D metrology because smaller and denser interconnects reduce process margin and increase the risk of yield loss. Onto received more than 10 additional 3Di orders in Q1, including OSAT customers, and the HBM VPA included more than $60 million in 3D bump metrology systems. A process-control industry expert with prior sales and product responsibility indicated that HBM inspection/metrology scales with the number of stacked chips for inter-chip bonding, which supports the thesis that higher HBM stack complexity can create a structural increase in process-control intensity rather than merely a 1-time capex event.

Panel-level packaging and silicon photonics are less mature but strategically important. Onto’s JetStep systems were qualified at 2 packaging suppliers to AI device manufacturers, with ramp expectations in 2027. Management has previously discussed a panel-level packaging opportunity of roughly $200 million over several years, including JetStep and Firefly. The company’s annual report describes the transition from 300mm wafers to rectangular panels as a way to increase the number of large-area packages per substrate and improve cost of ownership, with JetStep X500 and Firefly designed to address large-format advanced packaging lithography and process-control needs. If AI packages continue to grow in size and if panel-level approaches move from development into higher-volume production, Onto would have differentiated exposure because lithography, inspection, and process-control data can be linked in a workflow rather than sold as isolated tools.

ADVANCED NODES: The advanced-node business is becoming more important to the thesis because it reduces reliance on a single advanced-packaging product cycle. Onto’s Atlas G6 OCD system has been selected by a 2nd logic customer for gate-all-around metrology, and management expects advanced nodes to grow approximately 25% in 2026. Q1 advanced-node revenue was approximately $80 million, with roughly 60% memory and 40% logic. The logic opportunity is tied to gate-all-around and sub-2nm process complexity, while the memory opportunity is tied to DRAM and HBM4 development. The company also secured a TSV metrology application win using Atlas, with initial shipments expected in H2 2026. (Onto Innovation)