$ONTO INVESTMENT CONCLUSION
The source material captures a real and investable inflection, but the best interpretation is more nuanced than the claim that optical inspection is structurally obsolete. X-ray metrology and inspection are moving from specialized R&D/reference applications toward broader process-control relevance as logic, memory, and advanced packaging structures become more 3D, more heterogeneous, more buried, and more compositionally complex. Rigaku is a credible beneficiary because its core competence is X-ray measurement, and the Onto-Rigaku transaction provides external validation that major semiconductor process-control vendors view X-ray as strategically necessary. However, the more accurate investment framework is hybrid metrology, not outright optical displacement. Optical inspection and optical critical-dimension metrology retain advantages in speed, areal coverage, cost of ownership, defect discovery, and high-throughput line monitoring, while X-ray is advantaged for buried structures, high-aspect-ratio features, material composition, film thickness, strain, voiding, and dimensional measurements where optical methods hit physical or modeling limits. NIST explicitly frames CD-SAXS as a next-generation in-line dimensional metrology approach because optical scatterometry reaches fundamental limits as devices become smaller and more 3D; Onto and Rigaku similarly frame the opportunity as pairing optical speed and location information with X-ray precision for deeper structures, rather than replacing optical wholesale.
SCOPE OF THE THESIS
The source material uses “advanced packaging” as the headline, but the underlying technical drivers span 2 distinct markets: front-end process control for advanced logic/memory and back-end/advanced-packaging inspection and metrology. Gate-all-around/nanosheet transistors and backside power delivery are primarily front-end and BEOL integration challenges, while hybrid bonding, microbumps, TSVs, chiplets, HBM stacks, and 2.5D/3D packages are advanced-packaging challenges. This distinction matters because Rigaku’s opportunity is not limited to package inspection. The Onto-Rigaku partnership explicitly refers to advanced V-NAND and DRAM metrology, advanced logic and memory, advanced packaging, novel materials, and X-ray opportunities across deeper semiconductor structures. Rigaku’s own advanced-packaging materials cite microbumps, TSVs, hybrid bonding, narrowing process windows, shrinking structures, and non-destructive characterization from roughly 100 nm to 100 µm. The broader investment thesis is therefore that X-ray process control is becoming more relevant across the semiconductor manufacturing stack, not merely that advanced packaging inspection is moving from optical to X-ray.
The node-level framing in the source material is directionally correct but imprecise. The claim links 2 nm, GAA, backside power delivery, and advanced packaging as if they are a single technology transition. The more precise roadmap framing is that TSMC’s N2 uses 1st-generation nanosheet transistor technology and began volume production in Q4 2025, while TSMC’s A16 combines nanosheet transistors with Super Power Rail backside power delivery. That distinction matters because GAA/nanosheets, backside power, and advanced packaging each create different metrology pain points. GAA increases the need to characterize 3D channel geometry, nanosheet dimensions, inner spacers, work-function metals, and conformal films. Backside power introduces wafer thinning, backside alignment, buried power rails, vias, and backside metallization integration. Advanced packaging introduces alignment, voiding, bump geometry, bond-interface quality, TSV integrity, warpage, redistribution-layer defects, and buried interconnect problems. The common denominator is not that all are “packaging,” but that more critical features are 3D, buried, high-aspect-ratio, optically inaccessible, or composition-sensitive.
TECHNICAL VALIDATION
The technical case for more X-ray is strong. As semiconductor structures move vertically, optical systems increasingly rely on indirect inference from surface signals and model-based reconstruction. That remains powerful where structures are periodic, accessible, and well-calibrated, but it becomes less robust when the measurement target is buried under opaque films, embedded in heterogeneous stacks, located at a bond interface, or distributed through a high-aspect-ratio structure. X-ray techniques can probe internal geometry and material properties non-destructively, which is why they have long been used in R&D, failure analysis, and reference metrology. The investable transition occurs when these measurements become sufficiently necessary, repeatable, and high-throughput to justify insertion into production control loops. Rigaku’s framing that accurate non-destructive metrology becomes critical as 2.5D, 3D-IC, chiplet, and heterogeneous-integration architectures evolve is consistent with this technical path. Its disclosed capacity expansion for semiconductor process-control instruments is also consistent with management expecting semiconductor X-ray demand to move beyond niche laboratory usage.
The strongest use cases are likely not generic “inspection,” but high-value process-control steps where optical ambiguity, destructive sampling, or low-throughput reference techniques are bottlenecks. These include CD-SAXS for high-aspect-ratio 3D NAND and DRAM structures, XRF/XRD-type measurements for film thickness and composition, X-ray inspection or tomography for TSVs and hybrid-bond voids, and X-ray-based process control for advanced interconnect and packaging interfaces. In high-volume manufacturing, insertion will be determined by cost of ownership, throughput, recipe stability, algorithmic reconstruction quality, tool matching, sampling frequency, and ability to correlate measurements to yield excursions. This is why the Onto-Rigaku combination is important: Rigaku contributes X-ray physics and instrumentation, while Onto contributes optical metrology, modeling, software, analytics, customer relationships, and process-control integration. The strategic value is therefore the creation of a hybrid measurement stack capable of correlating fast optical data with more precise X-ray data, not merely the sale of standalone X-ray hardware.
ONTO/RIGAKU TRANSACTION READ-THROUGH
The Onto transaction is a material strategic signal. Onto agreed to acquire a 27% stake in Rigaku for approximately $710M, implying approximately $2.63B of equity value for 100% of Rigaku at the transaction reference point before considering exchange rates, liquidity, governance, timing, and any strategic-premium or discount effects. Onto also received the right to nominate 1 director. The investment is expected to be accounted for under the fair-value option, not consolidated, and Onto stated that it expects the transaction to be accretive beginning December 31, 2026. The same announcement stated that the Onto-Rigaku offering had been selected by 2 key customers and that external analysts estimate the addressed market at >$1B within the next 5 years. This is not a token commercial agreement; it is a material minority investment linked to a specific hybrid-metrology product strategy.
The source material’s inference that Onto’s investment proves Onto is structurally impaired is too aggressive. A more balanced interpretation is that Onto recognized a strategic gap in native X-ray hardware and chose partnership/equity alignment rather than internal replication, which is rational in a market where tool physics, source/detector know-how, precision mechanics, application expertise, and customer qualification cycles create long development timelines. The transaction is defensive because it secures access to a technology vector that could otherwise erode parts of Onto’s optical metrology and inspection relevance. It is also offensive because Onto can attach high-value software, modeling, AI analytics, and hybrid-metrology workflows to Rigaku’s hardware and potentially deepen its account control. Public materials support the narrower statement that Onto’s Ai Diffract software is being integrated with Rigaku CD-SAXS platforms; they do not support the broader blanket claim that “the software that runs on the Rigaku machines is the Onto software” across all Rigaku systems. The distinction is important because the economic split between hardware, analysis software, service, and customer ownership will determine which equity captures the incremental profit pool.
Rigaku is the cleaner direct exposure to X-ray tool adoption, but Onto should not be treated as a simple short solely on this thesis. Onto remains a scaled process-control vendor with advanced packaging, optical metrology, inspection, software, and analytics exposure. In Q1 2026, Onto reported revenue of $291.9M, non-GAAP gross margin of 55.7%, non-GAAP operating income of $77.9M, and cash and short-term investments of $654M. Management guided Q2 2026 revenue to $320M-$330M and highlighted that its process-control suite had expanded through Semilab and Rigaku advanced X-ray. Separately, Onto’s Dragonfly advanced-packaging platform was described as positioned for >50% growth in 2026 versus 2025, with advanced-packaging demand expected by industry analysts to grow approximately 30% annually over the next 2 years. The key investment question for Onto is not whether X-ray matters; it is whether Onto can monetize X-ray through software, hybrid workflows, and customer control sufficiently to offset any optical share loss.
RIGAKU FUNDAMENTALS
Rigaku’s fundamental profile supports the existence of a semiconductor inflection, but it is not a pure semiconductor company. FY2025 revenue was ¥94.1B, and FY2026 guidance calls for ¥101.0B of revenue, ¥19.4B of operating profit, and ¥12.5B of net profit. Semiconductor process-control instruments are the key growth and margin engine: the segment moved from ¥27.8B of FY2025 revenue to FY2026 guidance of ¥33.3B, while segment operating margin is guided to expand from 29.0% to 31.4%. The segment’s expected growth and margin structure are materially stronger than the group average and therefore carry disproportionate equity-story relevance. Rigaku also disclosed that new-product revenue share in semiconductor process control increased 2.2x year over year and that it had 12 joint evaluation projects, which is directionally positive but still requires conversion into repeat production orders.
The latest quarterly data add both validation and caution. In Q1 FY2026, Rigaku reported revenue of ¥17.933B, down 13.0% year over year, operating profit of ¥630M, down 77.8%, and net profit attributable to owners of ¥329M, down 82.8%. Management did not revise full-year guidance and stated that semiconductor process control exceeded expectations despite the year-over-year decline, with growth expected from Q2 onward due to memory and logic demand, project pull-ins, and progressing new-product projects. This creates a classic semicap inflection debate: the bullish interpretation is that Q1 weakness was timing-related and that the forward pipeline is strengthening; the bearish interpretation is that the stock is being valued on a steep adoption curve while reported earnings remain lumpy and highly sensitive to project timing, mix, and operating leverage.
Rigaku’s production-capacity actions reinforce management’s confidence but also raise execution expectations. The company disclosed expanded capacity for semiconductor process-control instruments at Takatsuki and a partner facility in Yamanashi, including doubled assembly and inspection floor space and additional capacity for core components such as X-ray generators and detectors. Rigaku expects total production capacity to be 50% higher in Q4 FY2025 versus the year-earlier level by unit count and to rise another 50% by 2027 compared with Q4 FY2024. Capacity expansion is a meaningful signal because semiconductor customers require supply assurance, service infrastructure, and stable tool matching before moving process-control tools deeper into production. It is also a risk because the valuation will increasingly depend on utilization of that added capacity and on the conversion of evaluation projects into production fleet orders.
VALUATION AND EXPECTATIONS
The “just buy Rigaku” conclusion is too simple after the stock’s rerating. Public quote data indicate Rigaku recently traded around ¥2,620 per share with market capitalization around ¥593B, enterprise value around ¥632B, trailing P/E around 61x, and trailing price/sales around 6.6x. Using company FY2026 guidance, this equates approximately to 5.9x guided sales, 47x guided net income, and 33x guided operating profit on an enterprise-value basis. Those are premium semicap multiples for a company still generating only about 1/3 of revenue from semiconductor process-control instruments. The valuation can be justified if semiconductor X-ray tools move meaningfully into production, if segment revenue compounds at a high rate, if 30%+ segment margins prove durable, and if Rigaku becomes the strategic scarce asset in X-ray process control. The valuation becomes vulnerable if adoption remains R&D/reference-heavy, if tool throughput limits broader insertion, if KLA/Bruker/internal customer solutions cap share, or if order timing produces multiple quarters resembling Q1 FY2026.
The Onto transaction also provides a useful valuation anchor, but it should not be overinterpreted. Onto’s $710M investment for 27% implied approximately $2.63B for 100% of Rigaku at the transaction reference point. The public market capitalization around ¥593B appears materially higher in dollar terms than that implied transaction value, depending on exchange rate and timing. This does not necessarily mean the stock is overvalued, because the public market may be discounting incremental disclosure, stronger demand signals, index inclusion, scarcity value, or control-option value. However, it does mean that the strategic-partnership announcement itself is no longer undiscounted. Incremental upside likely requires evidence of production adoption, rising semiconductor-process-control mix, JEP conversion, repeat orders from leading-edge customers, and sustained segment margin expansion.
COMPETITIVE LANDSCAPE
Rigaku is differentiated but not alone. KLA has launched the Axion T2000 X-ray metrology system for advanced 3D NAND and DRAM, explicitly targeting nanoscale control of complex memory structures. Bruker also markets non-destructive X-ray technologies for thin-film metrology across logic and memory, including specialized systems for substrate defects, FEOL control, metal films, and wafer-level packaging bumps. This matters because leading semiconductor manufacturers rarely rely on a single tool supplier for an emerging control step unless the technical gap is overwhelming. They qualify multiple vendors, segment applications by layer and measurement target, and use existing service infrastructure and process-control software as part of procurement decisions. Rigaku’s X-ray heritage and Onto alliance create a credible path to share gain, but KLA’s installed base, service model, process-control breadth, and customer intimacy remain formidable.
The source material’s description of Camtek and Onto as traditional optical-based advanced-packaging inspection players is broadly reasonable, but the conclusion that optical is structurally “screwed” is not. Camtek continues to show demand momentum, with Q1 2026 revenue of $121.7M and guidance for Q2 2026 revenue of $129M-$131M, alongside commentary that 2H 2026 revenue should grow >25% versus 1H 2026 based on strong demand. Camtek’s market capitalization was recently around $7.6B, implying substantial investor confidence in advanced-packaging inspection demand despite the emergence of X-ray. This is consistent with a world where optical remains critical for high-throughput macro inspection, surface defect detection, overlay, bump inspection, RDL inspection, and broad process monitoring, while X-ray is inserted into more difficult subsurface or compositional control steps. The likely industry outcome is tool proliferation and hybridization, not a clean 1-for-1 replacement cycle.
BESI COMPARISON
The comparison to Besi is directionally relevant because Besi is a key beneficiary of hybrid bonding and advanced assembly, but it is not a direct substitute for the Rigaku thesis. Besi is primarily an assembly equipment and hybrid-bonding exposure, while Rigaku is a metrology/process-control exposure. Besi monetizes the creation of advanced interconnects; Rigaku monetizes measurement and control of structures that become more difficult to inspect as interconnect density and 3D integration increase. Besi has also been heavily rerated on AI packaging, hybrid bonding, and strategic interest, with Reuters reporting that Applied Materials acquired a 9% stake in Besi in 2025 and that Besi Q1 2026 orders rose 104.5% to €269.7M, driven by hybrid bonding. The relative attractiveness of Rigaku versus Besi therefore depends on whether the market has underappreciated the metrology intensity created by the same technology transition that has already rerated hybrid-bonding tool suppliers.
KEY RISKS
The central technical risk is that X-ray remains indispensable for selected process steps but does not scale into a broad enough production-control TAM to justify current expectations. X-ray tools face inherent trade-offs around throughput, source brightness, detector performance, resolution, field of view, reconstruction complexity, sampling strategy, maintenance cost, and integration into fab control loops. Production deployment requires more than measurement superiority; it requires adequate cycle time, matching, uptime, recipe stability, automated interpretation, and measurable yield impact. A tool that is technically superior for R&D can still be too slow or too expensive for broad in-line use. This is the precise transition the source material asserts has already happened; public disclosures support accelerating adoption and customer engagement, but not yet a fully de-risked mass-production revenue curve.
The central commercial risk is that value capture may split across Rigaku, Onto, customers, and competing vendors. If Onto controls key modeling and analysis workflows, software economics may accrue disproportionately to Onto even when Rigaku supplies the hardware. If KLA or Bruker win specific high-volume logic or memory steps, Rigaku’s share of the incremental X-ray TAM could be lower than the market assumes. If leading-edge customers use X-ray primarily as reference metrology to calibrate faster optical or e-beam systems, revenue intensity could disappoint. If optical vendors integrate enough multi-sensor, AI-assisted, or sampling-based workflows to preserve control-loop relevance, displacement will be less severe than the source material implies. The Onto-Rigaku alliance itself suggests that the industry solution is combined measurement intelligence, not a single-modality winner-take-most outcome.
The central stock risk is expectation asymmetry. Rigaku’s semiconductor process-control segment is high-growth and high-margin, but the group remains diversified, and the current valuation appears to capitalize a large portion of the positive X-ray adoption narrative. Q1 FY2026 results showed significant year-over-year revenue and profit declines even as management maintained full-year guidance and described semiconductor process-control demand as better than expected. That combination can work if subsequent quarters validate the backlog and project-timing explanation. It can also create sharp downside if order conversion, shipment timing, or customer adoption slips. The near-term diligence priority is therefore not whether X-ray is technically useful; that appears well supported. The priority is whether Rigaku can convert JEPs into repeat production orders at leading customers, sustain 30%+ segment operating margins, and expand capacity without diluting tool quality or service performance.
INVESTMENT IMPLICATION
The most balanced conclusion is constructive on Rigaku’s strategic position but not unconditionally bullish at any price. The source material is directionally right that a material portion of next-generation process-control demand is migrating toward X-ray-enabled measurement, and Rigaku is a scarce public equity vehicle tied directly to that transition. The source material is also directionally right that Onto’s 27% investment validates Rigaku’s technology and highlights a strategic need that optical-only vendors cannot ignore. However, the argument becomes too simplistic when it treats optical inspection as obsolete, treats Onto’s stake as evidence of defeat, and ignores current valuation. The more investable formulation is that Rigaku offers high-quality optionality on the conversion of X-ray from R&D/reference metrology into production process control, while Onto represents a hybrid-metrology monetization path and Camtek remains exposed to high-throughput optical inspection growth in advanced packaging.
A rigorous investment stance would likely favor Rigaku as the cleaner long exposure on evidence of production-order conversion, customer breadth, and segment-margin durability, while avoiding a simplistic short thesis on Onto or Camtek absent proof of actual optical share loss. The best confirming data would be repeat production tool wins at TSMC, Samsung, Intel, Micron, SK hynix, major HBM suppliers, or leading OSATs; rising semiconductor process-control mix; accelerating new-product revenue contribution; evidence that the Onto-Rigaku tools are moving beyond 2 initial customer selections; and sustained order strength despite broader WFE cyclicality. The best disconfirming data would be continued reliance on R&D deployments, flat or lumpy semiconductor-process-control revenue, delayed JEP conversion, margin pressure from capacity expansion, or competitive wins by KLA/Bruker in the highest-volume steps. The technical thesis is credible and strategically important; the equity thesis now depends on timing, share, margins, and valuation discipline rather than on simple recognition that X-ray matters.
