$ONTO $MKSI $Q $KLAC EXECUTIVE OVERVIEW
The cleanest U.S. public-company beneficiaries of PLP, glass substrates, and larger AI packages are not necessarily the AI chip vendors themselves. The highest-purity exposures are the suppliers that monetize before full high-volume glass-substrate revenue appears: lithography/inspection, plating and wet chemistry, panel-compatible deposition, dielectric and copper materials, package test, probe, and advanced packaging assembly. This matters because TSMC’s current commentary indicates that large-size CoWoS remains the main supply approach today, while CoPoS/panel-level production is still expected “a couple of years” after pilot-line activity.
The most direct U.S.-domiciled beneficiaries are Onto Innovation, MKS Instruments, Qnity Electronics, KLA, Applied Materials, Amkor, Intel, Corning, Lam Research, Entegris, Teradyne, FormFactor, Cohu, Synopsys, and Cadence. The largest dollar beneficiaries, but with lower PLP/glass purity, are NVIDIA, AMD, Broadcom, Marvell, Micron, Alphabet, Amazon, Microsoft, and Meta. TSMC, ASE, Camtek, Advantest, BESI, ASML, Tokyo Electron, Disco, and Samsung Electro-Mechanics are highly relevant to the ecosystem but are not U.S.-domiciled companies; they should be treated as global comps rather than the core U.S. public-company basket.
HIGHEST-PURITY U.S. ENABLERS
Onto Innovation (ONTO) — most direct U.S. public exposure to panel-level lithography, glass-panel inspection, and PLP process control. Onto is one of the cleanest public-equity ways to express the transition from wafer-level packaging to panel-level packaging because its toolset maps directly to the core bottlenecks: large-area lithography, TGV inspection, transparent-panel metrology, data analytics, and process control. Onto opened a U.S.-based Packaging Applications Center of Excellence dedicated to PLP innovation for 2.5D/3D chiplet and AI packages, with collaborators across substrates, equipment, and materials. The facility is working toward sub-1.5µm line/space panel interconnects, while Onto’s Firefly G3 is designed to inspect transparent glass-core substrates for missing TGVs and TGV critical dimensions across full panels. Its JetStep X500 glass-panel lithography system supports sub-1.5µm line/space imaging and large shot sizes, directly aligning with the fine-RDL and panel-overlay challenge.
The investment significance is that Onto should monetize at the pilot-line and pre-HVM stage, ahead of actual glass-substrate unit revenue. PLP adoption raises inspection intensity because defect escape on a large AI package destroys expensive logic die, HBM stacks, interposers, substrates, and assembly value. Onto is therefore positioned as a “picks-and-shovels” beneficiary of qualification activity even if the final substrate winner changes. The key risks are timing slippage, customer concentration in advanced packaging capex, competition from KLA and other process-control vendors, and the possibility that 310mm x 310mm PLP adoption remains narrower than the market currently expects.
MKS Instruments (MKSI) — high-quality exposure to plating chemistry, via drilling, wet processing, and recurring advanced-substrate consumables. MKS is a direct beneficiary because PLP and glass-core substrates increase the importance of copper plating, via fill, surface preparation, SAP/mSAP processing, wet chemistry control, and laser-enabled via formation. The strategic logic behind MKS’s Atotech acquisition was explicit: MKS brought via-drilling expertise, Atotech brought electroplating and chemistry expertise, and the combination was intended to enable higher-yield, higher-throughput next-generation interconnects.
MKS is attractive because it combines capex exposure with consumables exposure. Equipment is ordered into new panel and substrate lines, while chemistry is consumed repeatedly once a process is qualified. That recurring element can create stickier economics than a pure equipment cycle. The relevance to glass substrates is particularly high because TGV metallization, copper fill uniformity, adhesion, void control, and panel-scale plating uniformity are central yield determinants. The risks are balance-sheet leverage from prior M&A, cyclicality in broader electronics/PCB markets, China exposure, and the possibility that substrate makers vertically standardize around competing chemistry stacks.
Qnity Electronics (Q) — the map’s DuPont exposure should now be mapped primarily to Qnity, not legacy DuPont. DuPont completed the separation of its electronics business into Qnity Electronics on November 1, 2025, and Qnity began NYSE trading under the ticker Q. This is important because the ecosystem map shows DuPont in resists/plating chemistry, but the cleaner public vehicle for semiconductor electronics materials is now Qnity rather than DD.
Qnity is strategically relevant because advanced packaging is becoming more materials-limited. The company has introduced enhanced packaging materials for organic interposers and emerging glass-based substrate structures, including Intervia 8540HSP copper for micro-bump and Cu-RDL applications and Cyclotene DF6800M dry-film photo-imageable dielectric for glass-core substrates and glass interposers. These products are designed for advanced interconnect formation, RDL designs, fine-feature patterning, planarization, and multilayer build-up.
The investment case is based on qualified recurring materials rather than one-time equipment orders. Once a dielectric, copper chemistry, plating material, or interposer material is designed into a customer package, switching costs are high because reliability qualification is slow and expensive. Qnity should benefit if AI packages require more RDL layers, finer pitch, better dielectric performance, higher copper uniformity, and glass-compatible build-up films. The key risks are customer pricing power, qualification delays, lower-than-expected glass adoption, and the possibility that Japanese materials suppliers retain disproportionate share in the most advanced substrates.
KLA (KLAC) — scaled process-control leader with direct glass-core and panel-interposer relevance, but lower purity than Onto. KLA is positioned to benefit because PLP and glass substrates increase the value of inspection, metrology, direct imaging, defect classification, chemistry control, and yield-management software. KLA has introduced an IC-substrate portfolio that includes direct imaging and new Lumina inspection/metrology systems for advanced IC substrates, including glass core and panel-based interposers. KLA explicitly highlights the need for higher yield as package dimensions increase, feature sizes shrink, and glass materials are introduced.
KLA’s advantage is scale, customer incumbency, process-control credibility, and broad reach across wafers, reticles, packaging, substrates, and PCBs. If advanced packages become larger and more expensive, inspection budgets should rise as a percentage of packaging capex because the cost of defect escape increases. The limitation is purity: KLAC remains a large, front-end process-control company, so PLP/glass is unlikely to dominate near-term EPS sensitivity. The stock may benefit more from the broad AI semiconductor capex cycle than from glass substrates alone.
Applied Materials (AMAT) — broad advanced-packaging equipment beneficiary with a strategic move into panel-level deposition through NEXX. Applied is not the purest PLP stock, but it is one of the most strategically important U.S. equipment companies in the transition. Applied announced a definitive agreement in 2026 to acquire ASMPT’s NEXX business, described as a leading supplier of large-area advanced-packaging deposition equipment. Applied stated that the acquisition broadens its panel-level advanced-packaging portfolio and supports larger-body AI accelerators as package architectures transition from 300mm wafers to panel form factors as large as 510mm x 515mm or more.
Applied’s upside comes from its ability to supply enabling deposition, electrochemical deposition, etch, inspection/metrology, and packaging-adjacent processes into AI package ramps. The company also took a 9% stake in BESI in 2025, reinforcing its hybrid-bonding strategy; hybrid bonding is highly relevant to 3D integration and high-density chiplet architectures. The risk is dilution of theme exposure: AMAT’s earnings remain much more tied to front-end wafer-fab equipment, memory, logic/foundry capex, and China restrictions than to PLP alone. It is a high-quality broad-cap beneficiary, not a PLP pure-play.
Amkor Technology (AMKR) — the most direct U.S.-domiciled public OSAT beneficiary. Amkor is strategically important because it provides outsourced advanced packaging and test capacity, and the U.S. ecosystem is trying to localize more back-end capability around TSMC Arizona, AI accelerators, and advanced nodes. Amkor and TSMC signed an MOU under which TSMC will contract turnkey advanced packaging and test services from Amkor’s planned Peoria, Arizona facility, with the companies jointly defining technologies such as InFO and CoWoS for common customers.
Amkor’s Peoria project is a $7 billion, 2-phase campus with more than 750,000 square feet of cleanroom manufacturing space, positioned to become the largest and most advanced outsourced semiconductor packaging plant in the U.S. The investment significance is high: Amkor could capture incremental outsourced packaging/test demand as AI packages exceed foundry-owned packaging capacity and as customers seek geographic redundancy. The risks are also high. Advanced packaging is capital-intensive, customer concentration can be material, utilization must ramp smoothly, and TSMC may retain the highest-margin, most complex integration steps internally.
Intel (INTC) — high strategic optionality through glass substrates, EMIB, Foveros, and foundry packaging, but with significant execution risk. Intel is one of the few U.S. companies with a credible internal glass-substrate roadmap. It announced glass substrates for next-generation advanced packaging, planned for the latter part of the decade, and stated that glass can support higher interconnect density, lower pattern distortion, better thermal/mechanical stability, and larger chiplet complexes than organic substrates.
The investment case is optional rather than linear. If Intel Foundry converts packaging differentiation into external customer wins, INTC could benefit from glass substrates, EMIB, Foveros, U.S.-based assembly/test, and broader demand for a second advanced-packaging ecosystem outside TSMC. However, the stock is dominated by process-node execution, foundry losses, PC/server CPU share, balance-sheet requirements, and manufacturing restructuring. Glass substrates could become strategically valuable but may not be a near-term earnings driver. INTC is therefore an asymmetric, higher-risk packaging option rather than a clean PLP beneficiary.
Corning (GLW) — direct glass and carrier exposure, but lower earnings purity. Corning is relevant because glass carriers and panel-sized glass materials can reduce warpage, support temporary bonding/debonding, and enable fan-out and 2.5D/3D packaging flows. Corning states that its glass carriers can reduce in-process warp by up to 40%, support temporary bonding for wafer thinning and fan-out packaging, and be manufactured into panel sizes such as 515mm x 510mm or 600mm x 600mm for panel-level packaging.
Corning’s upside is technical relevance to glass handling, carriers, ultra-flat substrates, and potentially glass-core supply. The limitation is that GLW is a diversified glass and optical-materials company. Near-term AI-related earnings sensitivity may be larger in optical connectivity and data-center infrastructure than in glass substrates. GLW should be treated as a real but diluted beneficiary, with the glass-substrate thesis providing optionality rather than the core EPS driver.
IMPORTANT ADJACENT ENABLERS
Lam Research (LRCX) — advanced-packaging wet processing, ECD, deposition, and etch exposure. Lam is not a panel-level pure-play, but it has meaningful exposure to the metallization and wet-process intensity created by advanced packages. Lam’s advanced-packaging portfolio is positioned for AI accelerators, increased I/O count, memory-wall mitigation, and energy-efficient computing. Its SABRE product family provides electrochemical deposition for copper damascene, while its portfolio also includes wet processing, DRIE/RIE for high-aspect-ratio structures including TSV/TDV/TGV, and deposition for dielectric/metal films in advanced packaging.
The investment logic is that higher HBM count, TSV/TGV complexity, copper interconnect density, and 3D package structures increase demand for precise deposition, etch, clean, and plating steps. The offset is that LRCX is primarily a front-end wafer-fab equipment company with large memory exposure, so PLP/glass is an incremental growth vector rather than the dominant stock driver.
Entegris (ENTG) — contamination control, materials handling, plating integrity, filtration, and substrate management. Entegris benefits from rising yield sensitivity. As packages become larger and more heterogeneous, contamination, microbubbles, substrate scratches, electrostatic discharge, plating impurities, and chemical-delivery defects become more expensive. Entegris positions its advanced-packaging offering around materials purity, materials performance, substrate integrity, yield, substrate handling, high-viscosity materials handling, Cu-plating integrity, CMP, and clean-fab materials delivery.
This is a lower-beta but potentially durable beneficiary. Entegris sells enabling consumables, handling systems, purification products, and contamination-control solutions rather than the headline PLP tools. That can create resilient recurring revenue if customers increase process-control budgets. The risk is that the company’s broad semiconductor exposure may overwhelm the packaging-specific upside in a cyclical downturn.
Teradyne (TER) — HBM and AI package test intensity beneficiary. Teradyne benefits from the fact that larger AI packages require more HBM, more known-good-die validation, and more test coverage before and after stacking. Its Magnum 7H platform targets next-generation HBM test and supports base-die wafer test, pre-singulated HBM test, and post-singulated HBM test. Teradyne also states that HBM has become integral to AI processors and that Magnum 7H supports HBM3/HBM3E and next-generation HBM4/HBM4E speed testing.
The investment thesis is that HBM attach failures and late-stage package failures become increasingly costly as package value rises. More pre-package and post-package test intensity should follow. TER is not a glass-substrate supplier, but it is a beneficiary of the same root cause: AI packages are becoming more complex, more expensive, and more memory-intensive. Key risks are semiconductor test cyclicality, smartphone/consumer exposure, and competition in memory test.
FormFactor (FORM) — probe-card and known-good-die testing beneficiary. FormFactor benefits from the need to test chiplets, HBM components, and advanced package elements before expensive integration. The company explicitly frames known-good-die testing as a core challenge in advanced packaging and offers Altius probe cards supporting 45µm grid-array microbump pitch testing and >3Gb/s at-speed verification.
The investment significance is that FORM is levered to yield protection. As AI accelerators integrate logic chiplets, HBM stacks, bridges/interposers, and advanced substrates, the cost of assembling a bad die rises. More granular pre-integration test should structurally raise demand for high-performance probe cards. The risks are customer concentration, memory-cycle volatility, and the potential for test-cost optimization to reduce incremental intensity in more mature flows.
Cohu (COHU) — smaller-cap test, handler, inspection, and thermal-control beneficiary for ultra-large AI packages. Cohu is a higher-beta, less liquid way to express the increasing difficulty of testing AI datacenter processors. The company supplies semiconductor test automation, interface, test equipment, inspection/metrology, and software analytics for yield/productivity. In 2026, Cohu announced a second multi-unit order for testing next-generation AI datacenter processors, with its Eclipse platform designed for GPUs, CPUs, custom AI accelerators, ASICs, ultra-large form-factor packages, and high power/thermal requirements.
COHU’s attraction is higher proportional exposure to back-end complexity and thermal test requirements. Its risk profile is also higher because of smaller scale, cyclical demand, customer-order lumpiness, and less structural moat than the larger semiconductor equipment leaders.
