Veeco Instruments Inc.

Veeco manufactures semiconductor process equipment whose highest-value nodes are Laser Spike Annealing (LSA) — "production tool of record at all three Tier 1 logic customers" (TSMC, Samsung, Intel) — and ion beam deposition used to produce EUV mask blanks for extreme-ultraviolet lithography. The quantitative constraint is process qualification at the leading edge: once a thermal/anneal or mask-blank tool is qualified into a Tier-1 logic or memory fab, it is locked in for the life of the node because re-qualifying a competing anneal/ion-beam tool risks yield on the most capital-intensive wafers on earth.

• LSA "production tool of record at all three Tier 1 logic customers, driving repeat business" — a de-facto sole/primary anneal position at TSMC, Samsung, and Intel (Q2 2026 10-Q MD&A, Semiconductor Market section).
• Next-gen Nanosecond Annealing (NSA) — addresses "critical low-thermal budget applications such as contact annealing, 3D device integration and materials modification." Q2 2026: a Tier 1 customer "successfully completed their evaluation and placed a follow-on order for a second system," and the third Tier 1 received an NSA evaluation tool. All three Tier 1 logic customers are now engaged on NSA.
• EUV mask blank production — Veeco's ion beam deposition is used in the mask-blank step of EUV lithography (10-Q: Semiconductor market "includes mask blank production for extreme ultraviolet ('EUV') lithography, as well as Advanced Packaging… high-end servers, and graphical processors"). EUV mask blanks are a single-supplier-adjacent step feeding ASML's EUV scanners.
• Advanced Packaging (wet processing + lithography) — for HBM and high-end GPU/server packaging; management explicitly ties Veeco to "high-performance AI chips, advanced connectivity and High-Bandwidth Memory ('HBM')".
• MOCVD for compound semiconductor (GaN power electronics, photonics/VCSEL, micro-LED) — Q2 2026 compound-semi revenue +45% YoY.
• Moat is qualification + installed-base, not patents: the 10-Q intangible table shows the technology intangible ($355.9M) is essentially fully amortized; the moat is the fab-locked tool-of-record status, not IP.

Veeco's highest-value nodes are Laser Spike Annealing (LSA) — "production tool of record at all three Tier 1 logic customers" (TSMC/Samsung/Intel) — and ion-beam deposition for EUV mask blanks (feeding ASML's scanners). The moat is qualification lock-in, not IP: the tech intangible ($355.9M) is fully amortized (Q2 2026 10-Q).

• Bypass test = 2. LSA has substitutes (Applied Materials, Screen flash-lamp anneal); MOCVD competes with Aixtron. "Tool of record" is real but re-qualification is costly-yet-possible. Veeco's 2024 heat-processing share is only 5–10% global / 10–15% China (SAMR market definition, Dealreporter) — a niche leader, not an unavoidable toll.
• Concentration test = 2. SAMR reviewed two relevant markets: ion implantation (Axcelis 20–25% 2024) and heat processing (Veeco 5–10%). The EUV mask-blank ion-beam niche is near-single-supplier but a narrow slice of a diversified tool portfolio.
• Hyperscaler exposure = 2. Veeco sells to fabs/foundries, not hyperscalers directly — AI-capex exposure is indirect. The pending all-stock merger further caps the standalone toll-road thesis (VECO terminates into ACLS).
• Composite = 2+2+2 = 6 STRONG PROXY (within PA Layer 1 "mostly 6+" band).

• product_service: thin-film process equipment: MOCVD, ion beam, wet processing, laser anneal, PVD/ALD/CVD, plus service/spares. (10-K Business)
• technical function: deposit/etch/treat thin films on wafers/substrates for semiconductor, compound semi, data storage, advanced packaging. (10-K Business)
• why it is hard: process integration, film uniformity, particle/defect control, high-aspect-ratio, customer qualification, IP/patents. (10-K Business/Risk Factors)
• substitutes: competing capital equipment from other semicap OEMs; alternative process flows (thermal anneal, other deposition/etch); older/in-house toolsets. (10-K Competition)
• pricing power: MED — differentiated in selected advanced nodes/packaging, but capital-equipment cycle, customer concentration, and competing OEMs constrain pricing. (10-K Risk Factors/Competition)

Veeco sells physical capital equipment plus associated process technology. The products are not consumer goods; they are high-value tools that perform thin-film deposition, etch, and annealing steps inside semiconductor fabs and device-manufacturing lines.

1. Laser spike annealing and laser thermal processing

Veeco’s laser annealing systems heat a wafer’s surface for very short durations to activate dopants and modify material properties. This is used in advanced logic transistor formation, including finFET and gate-all-around-class structures.

• Technical function: millisecond-scale thermal annealing, which reduces dopant diffusion relative to conventional rapid thermal annealing.
• Why it matters: As transistor dimensions shrink, controlling temperature and time is critical for yield and performance.
• Substitutes: conventional rapid thermal processing, furnace annealing, microwave annealing, and competing millisecond-anneal tools.
• Difficulty to replicate: deep process integration, optical/laser system know-how, and long customer qualification times.

Source: Veeco 10-K business discussion, SEC EDGAR company record, CIK 0000103145; https://www.sec.gov/cgi-bin/browse-edgar?action=getcompany&CIK=0000103145&type=10-K&dateb=&owner=include&count=40; accessed 2026-05-09. Specific product line definitions should be rechecked against the latest 10-K.

2. Ion-beam etch and deposition

Veeco’s ion-beam systems direct focused ion beams at substrates or targets to etch or deposit thin films.

• Used in data storage: manufacturing thin-film magnetic read/write heads and hard-disk media.
• Used in semiconductor and advanced packaging: directional etch, surface planarization, and physical-vapor-deposition-style film formation.
• Why ion-beam versus conventional etch/deposition: ion-beam etch is directional and can achieve controlled etching of materials that are difficult to etch with plasma chemistries. It is a niche but hard-to-replace technique.
• Substitutes: plasma etch, reactive ion etch, chemical-mechanical planarization for some steps, ion-beam tools from other vendors.

Source: same SEC EDGAR issuer record and 10-K product discussion; no exact accession number verified in this run.

3. Molecular beam epitaxy

MBE grows crystalline films with atomic-layer-level control under ultra-high vacuum.

• Used in: RF devices, lasers, photonics, infrared sensors, quantum devices, and advanced research.
• Technical function: evaporating high-purity source materials onto a heated crystalline substrate in ultra-high vacuum.
• Why it is hard to replicate: UHV chamber design, source-cell stability, contamination control, and process recipes.

4. MOCVD

Metal-organic chemical vapor deposition grows compound semiconductor epitaxial layers such as GaN, GaAs, and InP.

• Used in: gallium-nitride power electronics, RF devices, micro-LED, laser diodes, photonics.
• Why it matters: GaN power and RF are growing end markets because of electrification, data-center power, and RF front-end demand.
• Substitutes: MBE for some applications, HVPE for thick GaN, or competing MOCVD tools from Aixtron and others.

5. Other process equipment

Veeco has historically been involved in:

• Wet-process/surface-preparation equipment, depending on generation and acquired product lines;
• Maskless lithography or direct-write lithography for packaging and research;
• Precision surface processing and legacy equipment.

The exact current product list should be read from the latest 10-K Item 1. Current product-line revenue split: needs primary source validation.

Veeco’s moat is not a single patent or product; it is an integration and qualification moat:

1. Process qualification is expensive and slow. A fab may require years to qualify a tool for production. Once a Veeco tool is integrated into a manufacturing process, replacing it risks yield loss and requalification cost.
2. Installed base creates aftermarket lock-in. Service, spare parts, and upgrades follow the installed tool base.
3. Niche technical expertise. Ion-beam and MBE tooling require vacuum, beam, and process-domain knowledge that is not trivial to replicate.
4. Low-volume, high-mix market. The total market is not large enough to attract many competitors, which supports pricing discipline but also limits growth.

Veeco likely has moderate pricing power in its niches, but not across the board:

• Strong pricing power: ion-beam and MBE tools where alternatives are limited and qualification costs are high.
• Weaker pricing power: MOCVD and some data-storage equipment, where Aixtron and other competitors maintain a strong presence.
• Cyclical pricing pressure: In downturns, semiconductor equipment makers sometimes discount to maintain factory utilization and customer relationships.

Whether Veeco can raise prices without losing share depends on the specific product segment. A single corporate-level “pricing power” score is less useful than a segment-level analysis. Quantitative price/volume trend: needs primary source validation.

• What is the status/timing of the China SAMR decision, and is the "shipments held by U.S. Customs" issue export-control-related and deal-relevant?
• How much of the +67% contract-liability build is pre-payment tied to the pending merger vs. organic demand?
• Post-close, does the combined Axcelis+Veeco entity get re-rated as a full-suite semi-cap bottleneck, and does ACLS (not VECO) become the machine's physical-AI capital-equipment vehicle?