ZEISS SMT: the invisible optics monopoly inside every EUV machine
An independent, fully-cited, deliberately neutral teardown of ZEISS SMT — the Semiconductor Manufacturing Technology segment of Carl Zeiss AG and the sole maker of the projection optics inside every ASML EUV and High-NA lithography system. How it makes money, the deep co-dependency with ASML, the High-NA bet, and what could break the position. Researched in English and German.
Almost no consumer has heard of ZEISS SMT, yet roughly four in five chips on Earth — and effectively every leading-edge processor in a phone, data centre or AI accelerator — are printed with its optics. It is the optics monopoly hidden inside ASML's lithography monopoly. The question this case study turns on is whether being indispensable is the same as being safe.
Fiscal 2024/25 was a record: €5.055B revenue (+23%), the first time past €5bn, making SMT ~42% of the entire ZEISS Group. The same facts feed a bull case (a technical monopoly Nikon and Canon could not crack, ~80% of all chips, the sole supplier of High-NA optics) and a bear case (one customer, a cyclical order book, export-capped markets, and a next-generation bet TSMC is in no hurry to buy). This site lays out each side and leaves the verdict to you.
The revenue behind the debate
ZEISS SMT segment revenue by fiscal year (ZEISS's year ends 30 September). The line has roughly tripled in four years on the EUV ramp and a China-driven DUV surge — but it is lumpy, tracking the timing of ASML scanner shipments rather than smooth end-demand, and the next leg depends on a High-NA adoption curve that is still being negotiated.
FY2024/25 €5.055bn and FY2023/24 €4.122bn are disclosed by ZEISS[30][31]; FY2022/23 ~€3.6bn (+29%)[32]; FY2021/22 ~€2.8bn is derived from the disclosed +29% base and labeled approximate.
The four questions this case study turns on
Is the optics monopoly durable — or a monopoly with one customer?
ZEISS optics are in ~80% of all chips and ~100% of leading-edge chips, and the only firms with comparable heritage — Nikon and Canon — both abandoned EUV. The moat looks technical and non-transferable. But ZEISS sells those optics to essentially a single integrator, ASML, which also owns 24.9% of the business.
Does the multi-billion-euro High-NA bet pay off — and when?
High-NA EUV is the next leg of growth and ZEISS is its sole optics maker. Intel and SK Hynix are early adopters and ASML expects first products within months — yet TSMC, the largest buyer, is deferring volume use to 2029-plus on cost (~$400m a tool). The build is ahead of confirmed demand.
Is foundation ownership patient capital — or a constraint?
Carl Zeiss AG is wholly owned by a foundation that legally cannot sell it, giving a decades-long horizon suited to optics that take a quarter-century to mature. The flip side: no equity currency, no acquirer, and growth that must be self- and partner-funded.
How fragile is the position underneath the strength?
Single-customer concentration, semiconductor cyclicality, export controls that bar EUV from China (and now curb DUV), and an extreme geographic concentration in Oberkochen and Veldhoven all sit beneath the monopoly. Indispensable is not the same as invulnerable.
The balance of evidence, at a glance
Why the bull case holds
- A genuine technical monopoly: sole maker of EUV/High-NA projection optics; ~80% of all chips and ~100% of leading edge run on ZEISS optics[1][14].
- Barriers that defeated Nikon and Canon for 15+ years — accumulated, non-transferable optical know-how and 2,000+ EUV patents[4][16].
- Record €5.055bn revenue (+23%), now ~42% of the ZEISS Group, funded by patient foundation ownership and an aligned, investing customer[30][28][38].
- ZEISS is the binding constraint on High-NA supply, so it gates the pace of every node beyond today's EUV — including Hyper-NA[40][25].
Why the bear case holds
- Extreme single-customer concentration: lithography optics sell to essentially one buyer, ASML, which also owns 24.9% and co-sets the roadmap[17][40].
- The High-NA bet is contested by its biggest customer — TSMC defers volume use to ≥2029 on cost (~$400m a tool)[22][23].
- Revenue is cyclical and lumpy, and management is expanding capacity cautiously amid semiconductor volatility[12][34].
- Export controls cap the market (EUV barred from China since 2019; DUV/metrology curbs from Dec 2024) and value-add is concentrated in two cities[42][41].
A 19th-century optics house that became the heart of the chip
ZEISS SMT is the semiconductor arm of a 180-year-old German optics company — and the segment that now drives the whole group.
ZEISS SMT designs and builds the optical systems at the core of chip lithography — for DUV, EUV and now High-NA. It is a component supplier whose customer is ASML, so the public never sees it, yet ZEISS states ~80% of all microchips and ~100% of high-end (AI) chips are made with its optics[47][1]. From €5.055bn revenue in FY2024/25 it is now the group's largest segment[48].
What it actually makes
The product most people mean by “ZEISS SMT” is the projection and illumination optics that sit inside an ASML lithography scanner — the system of mirrors (for EUV) and lenses (for DUV) that shrinks a photomask pattern onto the silicon wafer with nanometre precision. But the segment is broader than optics alone: it also builds photomask repair and qualification tools and a growing line of inspection, metrology and defect-analysis equipment[10][13]. Its lineage runs from the first circuit-printer optics in 1968 to the Philips wafer stepper in 1983 — the same Philips lineage that produced ASML[2].
“Thirty years ago, EUV technology was considered visionary and technically unfeasible. Today we carry the technology around with us in every smartphone.”
A timeline of the long bet
Carl Zeiss founds the optical workshop in Jena (1846); physicist Ernst Abbe, with glassmaker Otto Schott, establishes the Carl Zeiss Foundation (1889) — the ownership structure that still holds today.
ZEISS supplies its first circuit-printer (lithography) optics — the start of the semiconductor lineage.
ZEISS lithography optics go into a Philips wafer stepper, deepening the link to the company that would become ASML.
The semiconductor business group is formed (1994); the strategic partnership with ASML dates to 1997.
The business is incorporated as Carl Zeiss SMT GmbH, headquartered in Oberkochen.
SMT revenue exceeds €1bn for the first time as immersion DUV scales.
ASML buys a €1bn, 24.9% minority stake in Carl Zeiss SMT to co-fund EUV and High-NA development.
The German Federal President's innovation prize goes to the ZEISS/TRUMPF/Fraunhofer EUV team.
ZEISS ships its first High-NA optics to ASML and SMT revenue passes €5bn (€5.055bn, +23%).
Where ZEISS sits in the most complex machine ever mass-produced
An EUV scanner is a three-party European achievement. ZEISS owns one irreplaceable third of it — the optics.
A leading-edge lithography machine is not one company's product. ASML integrates the system and owns the light-source architecture; TRUMPF builds the high-power drive laser; ZEISS builds the optical system that focuses 13.5 nm light onto the wafer. ASML makes only ~15% of an EUV scanner in-house — the rest, including the optics, comes from partners[7][6].
The division of labour
EUV lithography is often called a European joint project. The 2020 Deutscher Zukunftspreis — Germany's top innovation award — was given jointly to one laureate each from ZEISS SMT, TRUMPF and the Fraunhofer Institute, precisely because no single firm could have built it[6]. The pieces only work together, and ZEISS supplies the one piece — the optics — that no one else on Earth can make.
| Role | Company | What it provides | Substitutable? |
|---|---|---|---|
| System integrator | ASML (Netherlands) | The scanner; system architecture; the EUV source (via its Cymer unit) | No — only EUV scanner maker[7] |
| Light source / drive laser | TRUMPF (Germany) | The high-power CO₂ laser that generates EUV light from tin droplets | No — sole supplier[6] |
| Optics | ZEISS SMT (Germany) | Projection + illumination optics; multilayer mirrors; High-NA anamorphic optics | No — sole supplier[14] |
| Buyers | TSMC, Samsung, Intel, SK Hynix | Operate the scanners in fabs to print chips | Concentrated; few leading-edge fabs[9] |
How big is the prize?
The lithography optics ZEISS makes already image features below 10 nm using 13.5 nm light, inside an illumination system of ~15,000 parts[8]. Demand for that capability is set two layers downstream of ZEISS — by how many leading-edge wafers the world's fabs want to print, which in turn rides AI, mobile and HPC demand. That is the upside: ZEISS sits on the narrowest, least substitutable point of one of the most strategic supply chains on the planet. It is also the risk: ZEISS does not control its own demand, and the market for the most advanced tools is a handful of buyers.
Is the market position an asset or an exposure?
Why the position is enviable
- ZEISS owns the single least-substitutable node in the chain — the optics — with no rival at the leading edge[14].
- It is structurally embedded: ASML literally cannot build an EUV scanner without ZEISS optics[7].
- Leading-edge demand (AI, HPC) is structurally rising, pulling more EUV/High-NA tools through the chain[6].
Why the position is exposed
- ZEISS sells primarily to one integrator and has no direct relationship with the chipmakers who set demand[9].
- End-demand is concentrated in a few fabs (TSMC, Samsung, Intel), so a pause by any one ripples up to ZEISS[9].
- The whole leading-edge market is small in unit terms — a few hundred scanners a year — and intensely cyclical[8].
How a component supplier captures value from the chip boom
Three business units, one dominant one: lithography optics. The model is high-skill, capital-heavy and lumpy.
ZEISS SMT sells optical modules and process-control tools, not finished machines. Revenue is dominated by lithography optics shipped into ASML scanners, supplemented by mask-repair and fab-inspection equipment. It is an R&D- and capital-intensive model — the ZEISS Group spent a record 15% of revenue (€1.731bn) on R&D and €1.294bn of capex in FY2024/25 — and the revenue is lumpy, tracking ASML's shipment timing rather than smooth demand[11][12].
The three units
ZEISS SMT runs three strategic business units[10]:
| Unit | What it sells | Role in the model |
|---|---|---|
| Semiconductor Manufacturing Optics | DUV, EUV and High-NA lithography optics | The core — the optics inside ASML scanners; the bulk of revenue and the moat |
| Semiconductor Mask Solutions | Photomask repair, tuning and qualification (e.g. AIMS EUV) | Recurring, fab-facing; tied to every new mask set |
| Semiconductor Fab Solutions | Inspection, metrology and defect analysis (e-beam, X-ray) | Diversifies beyond optics; competes with KLA, Applied Materials, Lasertec |
Where the money comes from
For the optics unit, the economics are straightforward in shape if brutal in execution: ZEISS designs and builds an optics set over many months (a single High-NA mirror takes about a year), sells it into an ASML machine, and books the revenue as the system ships[12]. Because each unit is enormous in value and small in number, revenue moves in steps with ASML's delivery schedule — which is why a strong DUV year (FY2023/24, lifted by a China pull-forward) can look very different from the next[31]. ZEISS does notdisclose SMT's segment profit margin separately, so external margin figures are estimates.
The Mask and Fab Solutions units add a different, more recurring revenue character — mask qualification and fab metrology recur with every node and every new mask — and pull ZEISS into direct competition with the process-control majors rather than the comfortable optics monopoly[13].
Is the model resilient or brittle?
Resilient features
- Monopoly pricing power in optics, plus recurring mask/fab revenue that recurs with every node[10][13].
- An aligned customer co-funding R&D and capex (ASML's ~€760m commitment), de-risking the heaviest spending[38].
- A large installed base of scanners needing service, masks and metrology over a long life[44].
Brittle features
- Lumpy, capital-heavy revenue tied to ASML shipment timing — hard to smooth[12][34].
- Undisclosed segment margin and a front-loaded ~€3.5bn capacity build raise the stakes on demand showing up[33].
- In Fab/Mask Solutions, ZEISS is a challenger against entrenched majors (KLA, Applied, Lasertec), not a monopolist[13].
A monopoly that two optics giants tried and failed to break
ZEISS is the sole maker of EUV projection optics. The most telling evidence of the moat is who couldn't enter it.
At the leading edge ZEISS has no competitor. It is the only firm that makes the multilayer projection mirrors, illumination systems and High-NA anamorphic optics EUV requires — to a roughness below 0.1 nm and picometre stability[14]. The two companies with the closest credentials, Nikon and Canon, both abandoned EUV — strong evidence the barrier is technical, not commercial[15][16]. The catch: it is a monopoly that sells to one customer.
Why nobody else makes these optics
Nikon and Canon are not small: both are precision-optics and lithography veterans with decades of DUV experience. Yet after 15+ years of knowing EUV was the path forward, neither productized an EUV system. Nikon abandoned its program; Canon never shipped one and instead pursues nanoimprint lithography for niche memory and specialty uses[15]. Independent analysis reads their failure as the clearest possible proof that ZEISS's advantage is accumulated tooling, process and tacit human expertise that cannot be externalized or transferred — the kind of moat capital alone repeatedly fails to buy[16].
ZEISS SMT: Sole maker of EUV/High-NA projection optics; ~80% of all chips, ~100% of leading edge [1][14].
ZEISS clusters in the top-right with ASML and TRUMPF — the three irreplaceable leading-edge suppliers. Nikon, Canon and China's SMEE sit in mature-node, more-substitutable territory. Hover a point for the basis.
Five Forces — a monopolist with an unusually powerful buyer
Applied to ZEISS as a merchant supplier of leading-edge lithography optics, the forces are lopsided in an interesting way: rivalry and entry threat are Low (no one else can make the optics), but buyer power is High because there is essentially one customer — ASML — which also owns 24.9% of the business. Click a force for the evidence.
Where the moat could still erode
Why the moat holds
Where it could be chipped
- Canon's nanoimprint and DUV multipatterning chip away at specialty and mature nodes (not the leading edge)[18].
- National programs (notably China's, via SMEE) are pouring capital into a domestic optics path — gated today, but a long-run wildcard[29].
- A monopoly with one buyer can be squeezed on price and order timing in a way a diversified supplier cannot[17].
The hardest optics ever built — and the buyer who's in no hurry
High-NA EUV is ZEISS's next leg of growth and its deepest moat. It is also a bet whose biggest customer is deferring on cost.
High-NA raises the optics' numerical aperture from 0.33 to 0.55 and uses anamorphic mirrors to print ~8 nm features — ~1.7× smaller transistors and ~2.9× higher density than today's EUV[19]. The optics are a step-change in difficulty: the projection module carries 40,000+ parts and weighs ~12 tonnes, and each mirror takes about a year to make[20]. The bull case is that ZEISS alone can build this. The bear case is that TSMC, the largest buyer, is deferring volume use to 2029-plus on cost[22].
What makes the optics so hard
ZEISS calls the High-NA metrology system the most complex machine it has ever built[20]. The mirrors are coated with 100+ atomic-scale molybdenum/silicon layers and figured to atomic precision. ZEISS says it poured roughly ten million R&D work-hoursover more than a quarter-century into the technology, and frames the result as a European achievement — “the heart of microchip manufacturing for the world comes from Europe”[21].
“The mirrors, manufactured to atomic precision and consisting of more than a hundred layers, are the heart of the machine.”
“Our long breath over more than 25 years has now paid off: ASML's new wafer scanner for High-NA EUV lithography is the most modern, most complex and most precise microchip-manufacturing machine in the world.”
▸ The optical mechanism, in one quote (Dr. Peter Kürz)
“Thanks to the larger numerical aperture and the new High-NA mirrors we can capture more light and so expose in even finer detail and with greater precision.”
The demand question: real ramp, or expensive wait?
The case for the bet is that adoption has begun: Intel deployed the industry's first High-NA system for its 14A node, SK Hynix plans High-NA for next-generation DRAM, and ASML's CEO expects the first memory and logic products built on High-NA within months[24]. The case against is that the industry's largest and most disciplined buyer is not in a rush: TSMC has bought High-NA tools only for R&D and says its A13/A12 nodes (targeted 2029) won't need High-NA, citing cost — roughly $350–400m a tool and reports of up to 2.5× higher production cost versus conventional EUV[22][23].
Indexed to conventional EUV = 1.0×. A High-NA system costs roughly US$350–400m, about 2× a prior-generation EUV tool, and analyst reporting flags per-wafer production cost up to ~2.5× conventional EUV — the cost gap TSMC cites for deferring High-NA volume use to ≥2029[23][22].
Strategy: the bet weighed
Why the High-NA bet looks smart
- ZEISS is the sole supplier — every High-NA system on Earth needs its optics, so it captures all of whatever adoption happens[19][14].
- Early adopters (Intel, SK Hynix) are live and ASML expects first products within months[24][45].
- Gating Hyper-NA extends the monopoly into the 2030s, not just this node[25].
Why it could disappoint for years
- TSMC — the biggest buyer — defers volume use to ≥2029 on cost, so the demand curve may be slow[22].
- At ~$350–400m a tool and up to 2.5× wafer cost, High-NA's value is contested even by sophisticated fabs[23].
- ZEISS is building capacity ahead of confirmed volume orders — a timing mismatch if adoption slips[34].
ZEISS SMT has no true peer — so compare it three ways
As a merchant EUV-optics maker it stands alone. The useful comparisons are its place in the value chain, the lithography landscape, and its own parent group.
There is no second company that sells leading-edge lithography optics, so a like-for-like “peer table” doesn't exist. Three comparisons do the work instead: ZEISS SMT vs. its EUV-ecosystem partners (ASML, TRUMPF), vs. the lithography landscape (Nikon, Canon, SMEE), and vs. the rest of its own group — where SMT is now the largest segment at ~42% of ZEISS revenue[28].
1 · Inside the EUV value chain
ZEISS SMT (€5.06bn) is a component supplier inside the stack of its customer ASML (€28.3bn FY2024 net sales), alongside TRUMPF (€4.3bn group revenue, of which semiconductor lasers are only a part). The scopes differ — ASML sells the whole machine; ZEISS and TRUMPF sell components — so read this as relative scale, not market share[26][27][30].
2 · The lithography landscape
| Company | Leading-edge EUV? | Position |
|---|---|---|
| ASML + ZEISS | Yes — exclusive | Only EUV/High-NA scanners and optics on Earth[7][14] |
| Nikon | No (exited) | Mature-node DUV; historical Intel ties; quit EUV after heavy R&D[15] |
| Canon | No (never) | Mature-node DUV + nanoimprint (NIL) niche in memory/specialty[15][18] |
| SMEE (China) | No | Qualified at older nodes; gated by foreign optics it can't yet replace[29] |
3 · Inside the ZEISS Group
SMT is now the engine of the whole company. Its €5.055bn was about 42% of the group's €11.896bn in FY2024/25, up from ~38% a year earlier — meaning ZEISS's medical, industrial-quality and consumer-optics businesses together are now smaller than its chip-tools arm[28][30].
- ZEISS SMT — 43%
- Rest of ZEISS Group — 57%
A tripling in four years — and a multi-billion-euro build behind it
SMT revenue has roughly tripled since 2020. The question is whether the capacity going in now meets demand on the way out.
ZEISS SMT grew +23% to €5.055bn in FY2024/25 — its first year above €5bn — from €4.122bn, roughly a tripling in four years[30][32]. The group as a whole reached €11.896bn (+9%) with €1.552bn EBIT (13% margin)[30]. To keep up, ZEISS is running the biggest capacity build in SMT's history — but doing so cautiously, given semiconductor volatility[33][34].
The growth trajectory
The run was powered by two waves: the EUV ramp, and a surge in DUV demand — much of it pulled forward by Chinese fabs buying ahead of export controls. ZEISS named “strong demand for semiconductor production equipment in China” as a FY2023/24 tailwind[31]. That cushion is now fading as the curbs bite (see Risks), which is part of why management frames the outlook cautiously even as it invests.
The build behind the numbers
ZEISS says it is investing around €3.5bn in infrastructure over five years — the most ever — much of it in Germany: a 25,000+ m² expansion of the Oberkochen south factory, a new site at Aalen-Ebnat, and expansion at Jena, plus a new 12,000 m² DUV-optics factory in Wetzlar. SMT's workforce has tripled in seven years[33][35]. This is the financial heart of the bull/bear split: the same spend is either building the supply for an AI-era leading-edge boom, or front-running demand that (for High-NA) TSMC is in no rush to confirm.
Strong numbers, contested outlook
The bull read of the financials
The bear read of the financials
- The China DUV tailwind that lifted recent years is fading under export controls[31][42].
- A ~€3.5bn front-loaded build meets a High-NA demand curve TSMC is deferring — timing risk is real[33][22].
- Management itself stresses cost discipline and “capacity adjustments” amid volatility — not a confident, all-clear posture[34].
Owned by a foundation that legally cannot sell it
A 135-year-old foundation owns 100% of Carl Zeiss AG and cannot dispose of the shares — with a customer holding 24.9% of the SMT business one level down.
Carl Zeiss AG is wholly owned by the Carl Zeiss Foundation, which also owns glassmaker SCHOTT and is barred by its own statute from selling either[36]. One level down, ASML owns 24.9% of Carl Zeiss SMT GmbH[38]. The result is an unusual governance pairing: ultra-patient capital at the top, and a customer with a board seat in the most important business at the bottom.
The foundation at the top
The Carl Zeiss Foundation was created in Jena in 1889 by physicist Ernst Abbe (with SCHOTT founder Otto Schott). It is the sole shareholder of Carl Zeiss AG and SCHOTT AG, and the foundation statute forbids selling the shares; the dividends fund STEM research and teaching[52]. In a 2004 reform the operating companies were turned into legally independent joint-stock companies under the foundation as sole shareholder[37].
“The Carl Zeiss Foundation is the sole shareholder of Carl Zeiss AG and SCHOTT AG. The Foundation may not sell the shares of its two companies — this is laid down in the Foundation statute.”
The customer with a stake at the bottom
ASML's €1bn purchase of 24.9% of Carl Zeiss SMT in 2016 came with ~€220m of R&D support and ~€540m of capex and supply-chain investment over six years, explicitly to co-develop High-NA[38]. It aligns the two roadmaps and de-risks ZEISS's heaviest spending — but it also formalizes a single customer's seat inside ZEISS's most valuable business, which is part of why this is a co-dependency rather than a classic monopoly.
People
Leadership reflects SMT's rise: Andreas Pecher, who led ZEISS SMT, became ZEISS Group CEO, and Dr. Frank Rohmund took over as ZEISS SMT President & CEO from January 2025[4][50]. The segment employs 9,349 people, ~2,000 of them on High-NA, concentrated in Oberkochen and a cluster of German sites[48]. EUV has been a major engine of high-skill German employment: the 2020 Deutscher Zukunftspreis citation noted 3,300+ high-tech jobs created at ZEISS and TRUMPF[39].
Governance: strength and constraint
Why the structure is a strength
Why it is also a constraint
The risks hidden under a monopoly
Being irreplaceable masks four real vulnerabilities: one customer, a cyclical market, export controls, and geography.
ZEISS's strength is also the shape of its risk. Customer concentration (essentially one buyer), cyclicality (lumpy, boom-bust equipment demand), export controls (EUV barred from China; DUV now curbed), and geographic concentration(value-add in two European cities) all sit beneath the optics monopoly. None is fatal; together they make “indispensable” less than “invulnerable.”
1 · Single-customer concentration
ZEISS's lithography optics sell to essentially one integrator — ASML, which also owns 24.9% of the business. That is an extraordinary degree of single-customer dependence for a monopolist, and it leaves SMT exposed to ASML's order timing and roadmap choices. The dependence runs both ways (ASML can't ship High-NA faster than ZEISS makes optics[40]), but the revenue concentration is real.
2 · Cyclicality and a contested next node
Semiconductor-equipment demand is boom-bust, and ZEISS's revenue tracks ASML shipments rather than smooth end-demand. If High-NA adoption stalls on cost — with TSMC deferring volume to 2029-plus — the multi-billion-euro capacity build could run ahead of paying demand for years[43][22]. Management's own language (“cost awareness… capacity adjustments”) signals it is alert to this[34].
3 · Export controls cap the market
EUV has never been exportable to China (prohibited since 2019), and from December 2024 the Netherlands and US extended controls to advanced DUV immersion systems and metrology. ASML expected China to fall to ~20% of 2025 sales after a 2024 pull-forward — removing a demand cushion ZEISS had directly benefited from[42][31]. The flip side: those same controls entrench ASML+ZEISS by blocking a Chinese leading-edge alternative.
4 · Geographic single point of failure
Most EUV-scanner value-add is concentrated in two European cities — ASML's Veldhoven and ZEISS's Oberkochen. Independent analysis warns a major disruption at either would constrain global leading-edge chip capacity within roughly 12–18 months[41]. For a technology this central, that concentration is a systemic fragility, not just a company risk.
The risk picture as a SWOT
Strengths
Weaknesses
Opportunities
What decides the next chapter
ZEISS's position is secure; its growth rate is not. Four variables decide which way the next few years break.
The monopoly is not seriously in question this decade — Nikon and Canon are gone and no entrant is close. What's contested is the growth rate: it turns on the slope of High-NA adoption versus its cost, the strength of AI-driven leading-edge demand, how far export controls shrink the served market, and whether ZEISS can scale the hardest optics in the world without over-building into a downturn[46].
The four deciding variables
- High-NA economics. Does the cost (~$350–400m a tool, up to ~2.5× wafer cost) fall enough to pull TSMC and others into volume — or does conventional EUV keep winning on value?[23][22]
- Leading-edge demand. Does AI/HPC keep fabs investing through the cycle, sustaining scanner (and therefore optics) volume?[45]
- Export-control trajectory. Do controls tighten further, stay, or ease — and how much of the DUV market do they keep off-limits?[42]
- Execution and capacity. Can ZEISS ramp High-NA optics output (the binding constraint) on schedule without stranding the €3.5bn build if demand slips?[33][40]
Three scenarios
AI-era demand keeps leading-edge fabs full; High-NA adoption broadens from Intel and SK Hynix to the rest, and TSMC eventually follows. ZEISS, as sole optics maker, captures all of it and gates Hyper-NA into the 2030s. The €3.5bn build looks prescient.
EUV stays the workhorse; High-NA grows but slowly while TSMC optimizes cost and design. ZEISS keeps growing on EUV volume, service, masks and metrology, with revenue that is large, lumpy and cyclical rather than smoothly compounding.
High-NA economics keep big buyers cautious for years, China demand stays capped by export controls, and a semiconductor downturn coincides with the capacity build — leaving ZEISS with expensive capacity and a lumpy, deferred order book.
How this was made — and where it may be wrong
A research compilation built to let you weigh the evidence, not to sell a conclusion. Here is how, and what to distrust.
How the research was done
Every source in the bibliography was fetched and read during research. Disclosed financials come from ZEISS Group press releases (fiscal year ending 30 September); deal terms from ASML's own releases; technical claims from ZEISS and ASML material plus German and English trade press; and the contested points (High-NA cost, TSMC's stance, export controls) from named secondary reporting. Claims are tagged by tier (1 primary, 2 reputable secondary, 3 tertiary), confidence, and stance.
Neutrality commitment
This study is a compilation, not an argument. Each section pairs the supporting and the countervailing evidence and ends in a neutral synthesis. The overall source mix is 29 supporting / 16 neutral / 16 critical — deliberately weighted so neither a bull nor a bear case is presented unopposed. Where the evidence leans one way (e.g. the monopoly's technical durability), the strongest counter is shown alongside it.
Native-language research
ZEISS is a German company, so a portion of the research was done in German — including ZEISS's own German press material, the Carl Zeiss Foundation's German pages, and German trade outlets (elektronikPRAXIS, Elektroniknet) — with named-executive quotes from Frank Rohmund, Andreas Pecher and Peter Kürz rendered in the original German with translation[3][50][51]. About 25% of sources are German-language. That share is meaningful but not dominant, for a specific reason: ZEISS and ASML publish their official investor, technical and deal material bilingually or in English, and the most detailed coverage of the High-NA cost debate is English-language trade press. We used German sources for the domestic voice and primary quotes, and English sources where they were the primary record.
- SMT segment profitability is unknown. ZEISS does not disclose SMT margin, pricing to ASML, or the optics-vs-process-control revenue split. Any such figure is an estimate[12].
- FY2021/22 SMT revenue (~€2.8bn) is derived, not disclosed — inferred from the disclosed +29% FY2022/23 base. FY23–25 figures are disclosed[32].
- Market-share claims (~80% / ~100%) are ZEISS's own, widely repeated but company-sourced; we present them as ZEISS's statement, corroborated by the structural fact that no rival makes EUV optics[1][14].
- The €3.5bn five-year investment is a group infrastructure figure, not SMT-only, though much of it serves SMT[33].
- High-NA cost/throughput figures (~$350–400m, up to ~2.5×) are secondary reporting, not vendor-disclosed, and the adoption timeline is moving[23][22].
- Some structural and single-point-of-failure analysis is Tier-3 (independent analysts); we flag it as interpretation, not disclosed fact[41].
- This is a point-in-time snapshot as of June 7, 2026. Export rules, High-NA orders and leadership can change quickly.
Full bibliography
Every source cited in this case study, grouped by section. Each was fetched during research; German-language sources are quoted in the original with translation. Tier and stance are shown for transparency.
Company, Optics Lineage & Timeline
- [1]About us as company & team — ZEISS Semiconductor Manufacturing TechnologyTier 1supportingHigh confidence
ZEISS SMT (Semiconductor Manufacturing Technology) is the business segment of Carl Zeiss AG that develops semiconductor-manufacturing equipment across three units — Manufacturing Optics (DUV/EUV/High-NA lithography optics), Mask Solutions (photomask repair/tuning) and Fab Solutions (inspection, metrology, defect analysis) — with 9,349 employees and €5.055bn revenue in FY2024/25.
“Approximately 80 percent of all microchips worldwide are manufactured using lithography optics from ZEISS SMT.”
https://www.zeiss.com/semiconductor-manufacturing-technology/about-us.html Carl Zeiss SMT was established as a GmbH in 2001 in Oberkochen; the semiconductor business group dates to 1994; ZEISS first supplied circuit-printer optics in 1968 and lithography optics for a Philips wafer stepper in 1983; SMT revenue first passed €1bn in 2010; subsidiaries were merged in 2014; ASML took a minority stake in 2016.
“2016: ASML purchased minority stake to strengthen partnership.”
https://en.wikipedia.org/wiki/Carl_Zeiss_SMT- [3]30 Jahre EUV-Lithographie-Optiken bei ZEISS SMT — ZEISS SMT (Pressemeldung)Tier 1supportingHigh confidence
ZEISS SMT marks 30 years of EUV lithography optics; CEO Frank Rohmund frames the arc from 'visionary and technically unfeasible' to a technology now in every smartphone. A foundational industry workshop at Oberkochen in 1995 shaped the path to optical EUV lithography.
“Thirty years ago, EUV technology was considered visionary and technically unfeasible. Today we carry the technology around with us in every smartphone.”
https://www.zeiss.de/semiconductor-manufacturing-technology/news-und-events/smt-pressemeldung/2025/30-jahre-euv.html - [4]About us as company & team — ZEISS Semiconductor Manufacturing TechnologyTier 1supportingHigh confidence
ZEISS holds more than 2,000 patents in EUV technology and received the Deutscher Zukunftspreis (German Federal President's award for innovation) in 2020 for the development of EUV lithography. Dr. Frank Rohmund has been President & CEO of ZEISS SMT since January 2025.
“For advanced chips, such as those used in artificial intelligence applications, ZEISS SMT holds a 100 percent market share.”
https://www.zeiss.com/semiconductor-manufacturing-technology/about-us.html - [5]ASML Spotlight — EUV Monopoly, High-NA, Zeiss Optics & Export Controls — SemiconductorXTier 3criticalMedium confidence
ZEISS SMT's position is structural but exposed: it is a component supplier whose optics ship only inside ASML machines, with no direct relationship to the chipmakers who set demand — indispensable yet dependent on a single integrator and invisible to the end market.
“100% exclusive supply of all EUV optics — multilayer projection mirrors, illumination system, High-NA anamorphic optics. No alternative supplier exists or is feasible within this decade.”
https://semiconductorx.com/spotlight-asml.html - [47]Über uns als Unternehmen & Team — ZEISS Semiconductor Manufacturing TechnologyTier 1supportingHigh confidence
In its own German materials, ZEISS SMT states that ~80% of all microchips worldwide are made with its lithography optics and 100% of high-end (e.g. AI) chips, and dates its strategic partnership with ASML to 1997.
“Around 80 percent of all microchips worldwide are manufactured with lithography optics from ZEISS SMT. For high-end chips, e.g. for artificial-intelligence applications, it is 100 percent market share.”
https://www.zeiss.de/semiconductor-manufacturing-technology/ueber-uns.html - [48]Über uns als Unternehmen & Team — ZEISS Semiconductor Manufacturing TechnologyTier 1neutralHigh confidence
ZEISS SMT reports 9,349 employees (FTE) and €5.055bn revenue in FY2024/25, operating from Oberkochen (HQ), Jena, Wetzlar, Roßdorf, Aachen and Coswig in Germany plus Zürich, Dublin and Bar Lev (Israel).
“9,349 employees ... €5.055 billion revenue of ZEISS Semiconductor Manufacturing Technology in fiscal year 2024/2025.”
https://www.zeiss.de/semiconductor-manufacturing-technology/ueber-uns.html
The EUV Machine & Where ZEISS Sits
- [6]ZEISS, TRUMPF and Fraunhofer team awarded the Deutscher Zukunftspreis 2020 for EUV lithography — FraunhoferTier 2supportingHigh confidence
An EUV scanner is a three-party European achievement: ASML integrates the system and owns the EUV source architecture, TRUMPF builds the high-power CO2 drive laser, and ZEISS supplies the optical/projection system. The 2020 Deutscher Zukunftspreis honoured one laureate from each of ZEISS SMT (Dr. Peter Kürz), TRUMPF (Dr. Michael Kösters) and Fraunhofer IOF (Dr. Sergiy Yulin).
“The high-power laser from TRUMPF for the EUV light source and the optical system from ZEISS are key components of these machines.”
https://www.fraunhofer.de/en/press/research-news/2020/november/zeiss-tumpf-and-fraunhofer-research-team-awarded-the-deutscher-zukunftspreis-2020-for-the-development-of-euv-litography.html - [7]WFE Lithography: ASML EUV, DUV Scanners, Nikon, Canon, SMEE — SemiconductorXTier 3neutralMedium confidence
ASML builds only ~15% of an EUV system in-house; ~85% comes from a supplier network (notably ZEISS optics, TRUMPF laser, and ASML's own Cymer light-source unit). ASML supplies every EUV and High-NA scanner on earth and ~85–90% of leading-edge DUV immersion scanners.
“ASML supplies every EUV and High-NA EUV scanner manufactured on earth, approximately 85–90% of DUV immersion scanners for leading-edge production.”
https://semiconductorx.com/wfe-lithography.php - [8]30 Jahre EUV-Lithographie-Optiken bei ZEISS SMT — ZEISS SMT (Pressemeldung)Tier 1supportingHigh confidence
An EUV illumination system already comprises about 15,000 individual parts and weighs ~1.5 tonnes; the optics image features below 10 nanometres using 13.5 nm light. Modern chips now carry tens of billions of transistors on a fingernail-sized die.
“With EUV lithography we found a solution for the seemingly unsolvable, and are the heartbeat and metronome of digitalization.”
https://www.zeiss.de/semiconductor-manufacturing-technology/news-und-events/smt-pressemeldung/2025/30-jahre-euv.html - [9]ASML Spotlight — EUV Monopoly, High-NA, Zeiss Optics & Export Controls — SemiconductorXTier 3criticalMedium confidence
ZEISS's position in the value chain is concentrated and exposed: it sells primarily to a single integrator (ASML), whose own revenue concentrates on TSMC, Samsung and Intel. Demand is set two layers downstream from ZEISS, which has no direct lever on it.
“ASML cannot simply choose a different optics supplier ... the result of joint ASML-Zeiss development that has no independent equivalent.”
https://semiconductorx.com/spotlight-asml.html High-NA development was co-funded by the German federal government and the EU (under the IPCEI Microelectronics program), underscoring that ZEISS's optics are treated as a European strategic-technology project, not a purely private effort.
“funding from the German federal government and the European Union.”
https://www.zeiss.de/semiconductor-manufacturing-technology/inspiring-technology/high-na-euv-lithographie.htmlZEISS's German material states that 80% of all microchips worldwide are made with lithography technologies from its strategic partner ASML — whose scanners are built around ZEISS optics — making explicit how ZEISS's fortunes are routed entirely through ASML.
“Today, 80 percent of all microchips worldwide are manufactured using lithography technologies from our strategic partner ASML from the Netherlands.”
https://www.zeiss.de/semiconductor-manufacturing-technology/produkte/maerkte-und-partner.html
Business Model & Segments
- [10]About us as company & team — ZEISS Semiconductor Manufacturing TechnologyTier 1neutralHigh confidence
ZEISS SMT operates three strategic business units: Semiconductor Manufacturing Optics (lithography optics for DUV, EUV and High-NA-EUV), Semiconductor Mask Solutions (photomask repair and tuning, e.g. the AIMS EUV mask-qualification tool), and Semiconductor Fab Solutions (inspection, metrology and defect analysis).
“Semiconductor Manufacturing Optics ... Semiconductor Mask Solutions ... Semiconductor Fab Solutions.”
https://www.zeiss.com/semiconductor-manufacturing-technology/about-us.html The model is R&D- and capital-intensive: the ZEISS Group spent a record 15% of revenue (€1.731bn) on R&D in FY2024/25 and €1.294bn on property, plant and equipment, with SMT the heaviest consumer of growth investment.
“Innovative strength remains the key factor for ZEISS' medium- and long-term success.”
https://www.zeiss.com/corporate/en/about-zeiss/present/newsroom/press-releases/2025/end-of-year.htmlRevenue is lumpy and tied to ASML system shipments: each EUV/High-NA optics set is a multi-month build (about a year per mirror) sold into ASML's machine, so SMT revenue tracks the timing of ASML scanner deliveries rather than smooth end-demand — and ZEISS does not disclose SMT segment profit margin, limiting transparency.
“manufacturing time: approximately one year per mirror.”
https://www.zeiss.com/semiconductor-manufacturing-technology/inspiring-technology/high-na-euv-lithography.htmlZEISS SMT has broadened beyond optics through acquisitions and process-control lines — e.g. taking over Philips Engineering Solutions in Aachen (2024) and a Swiss acquisition — and ships the AIMS EUV 3.0 mask-qualification platform, diversifying away from pure lithography optics.
“Photomask inspection and repair systems ... Wafer inspection and metrology solutions using X-ray analysis and electron beam techniques.”
https://en.wikipedia.org/wiki/Carl_Zeiss_SMTZEISS SMT's German product pages confirm the three-unit structure — Halbleiterfertigungsoptiken (SMO), Photomasken-Lösungen (SMS) and Lösungen für die High-Volume Wafer-Fertigung (SFS) — spanning DUV, EUV and High-NA optics plus mask defect detection/repair and nanometre-scale wafer inspection.
“With our photomask solutions, defects on the photomask can be detected, analysed and repaired, and specific mask properties measured and optimised.”
https://www.zeiss.de/semiconductor-manufacturing-technology/produkte.html
Monopoly, Competition & Moats
- [14]ASML Spotlight — EUV Monopoly, High-NA, Zeiss Optics & Export Controls — SemiconductorXTier 3supportingMedium confidence
ZEISS SMT is effectively the sole supplier of EUV projection optics worldwide — multilayer projection mirrors, illumination systems and High-NA anamorphic optics — with no feasible alternative supplier this decade. The mirrors require Mo/Si multilayers polished to sub-0.1nm RMS roughness and picometre positional stability.
“roughness below 0.1nm RMS ... entirely unique and the product of decades of co-development with ASML that no other company has replicated.”
https://semiconductorx.com/spotlight-asml.html - [15]WFE Lithography: ASML EUV, DUV Scanners, Nikon, Canon, SMEE — SemiconductorXTier 3supportingMedium confidence
The two firms with the closest credentials — Nikon and Canon — both exited EUV. Nikon abandoned its EUV program; Canon never productized EUV and instead pursues nanoimprint lithography (NIL) for niche memory/specialty uses. Both retreat to mature-node DUV.
“The capital, supply chain, and optical-system development required for EUV proved uneconomic for vendors without ASML's scale.”
https://semiconductorx.com/wfe-lithography.php - [16]ASML Spotlight — EUV Monopoly, High-NA, Zeiss Optics & Export Controls — SemiconductorXTier 3supportingMedium confidence
Independent analysis treats Nikon's and Canon's 15+-year failure to enter EUV as evidence the barrier is technical, not commercial — accumulated tooling, processes and human expertise at ZEISS that 'cannot be externalized or transferred.'
“Nikon and Canon ... have been unable to enter EUV after 15+ years ... This is the strongest evidence that the barriers are technical rather than commercial.”
https://semiconductorx.com/spotlight-asml.html - [17]ZEISS and ASML Strengthen Partnership for Next Generation of EUV Lithography — ASMLTier 1criticalHigh confidence
The monopoly cuts both ways: ZEISS sells to essentially one customer (ASML), which holds a 24.9% stake and co-controls the roadmap. Buyer power is unusually high for a monopolist, and the relationship is co-dependent rather than one of pricing dominance.
“Carl Zeiss SMT is described as ASML's most important strategic partner.”
https://www.asml.com/en/news/press-releases/2016/zeiss-and-asml-strengthen-partnership-for-next-generation-of-euv-lithography - [18]WFE Lithography: ASML EUV, DUV Scanners, Nikon, Canon, SMEE — SemiconductorXTier 3criticalMedium confidence
Substitution risk exists at the edges: Canon's nanoimprint lithography offers cost advantages in 3D NAND and specialty markets, and Chinese fabs extend restricted DUV with multipatterning — both nibble at, but do not replace, EUV's leading-edge role.
“nanoimprint lithography (NIL) ... offers cost advantages at specific applications (3D NAND, some specialty markets).”
https://semiconductorx.com/wfe-lithography.php ZEISS's own German material frames its optics as the literal core of ASML's machines and itself as an OEM: 'the heart of the wafer scanners are lithography optics from the ZEISS division', with the ASML partnership dating to 1997 and the first optics delivery to (today's) ASML in 1983.
“The heart of the wafer scanners are lithography optics from the ZEISS division.”
https://www.zeiss.de/semiconductor-manufacturing-technology/produkte/maerkte-und-partner.html- [59]30 Jahre EUV-Lithographie-Optiken bei ZEISS SMT — ZEISS SMT (Pressemeldung)Tier 1supportingHigh confidence
ZEISS describes itself as the outright technology leader in precision EUV optics. Christoph Hensche, then head of the Semiconductor Manufacturing Optics business: 'In this field we are the technology leader.'
“In this field we are the technology leader.”
https://www.zeiss.de/semiconductor-manufacturing-technology/news-und-events/smt-pressemeldung/2025/30-jahre-euv.html
The High-NA Bet (and Hyper-NA)
High-NA EUV is ZEISS's central next bet: raising numerical aperture from 0.33 to 0.55 and using anamorphic optics (4x demagnification in one axis, 8x in the other) to print ~8 nm features, ~1.7x smaller transistors and ~2.9x higher density than current EUV.
“a CD of 8 nm ... print transistors 1.7 times smaller – and therefore achieve transistor densities 2.9 times higher.”
https://www.asml.com/en/news/stories/2024/5-things-high-na-euvThe High-NA optics are a step-change in manufacturing complexity: the projection optics carry more than 40,000 parts and weigh ~12 tonnes (the illumination system ~6 tonnes and >25,000 parts); each large mirror takes about a year to make and is coated with 100+ atomic-scale Mo/Si layers. ZEISS calls the metrology system the most complex machine it has ever built.
“the most complex machine ZEISS has ever built.”
https://www.zeiss.com/semiconductor-manufacturing-technology/inspiring-technology/high-na-euv-lithography.html- [21]High-NA-EUV-Lithographie: Neue Technologie für weltweite Mikrochipproduktion — ZEISS SMT (Pressemeldung)Tier 1supportingHigh confidence
ZEISS says it invested roughly ten million R&D work-hours over more than 25 years into High-NA and frames the result as a European achievement; serial production of High-NA optics began in 2025. CEO Andreas Pecher: 'the heart of microchip manufacturing for the world comes from Europe.'
“The heart of microchip manufacturing for the world comes from Europe.”
https://www.zeiss.de/semiconductor-manufacturing-technology/news-und-events/smt-pressemeldung/2024/high-na-euv-lithographie.html - [22]TSMC Rejects High-NA EUV Investment Concerns, Confirms Purchase for R&D Use — TrendForceTier 2criticalHigh confidence
The bet is contested by its biggest potential buyer. As of 2026, TSMC has bought High-NA tools only for R&D and is delaying volume use on cost grounds; its A13/A12 nodes (targeted 2029) are not expected to need High-NA, with chairman C.C. Wei saying TSMC will keep cutting cost before moving to production.
“TSMC's A13 and A12 processes (both targeted for 2029) are not expected to require High-NA EUV lithography tools.”
https://www.trendforce.com/news/2026/06/04/news-tsmc-rejects-high-na-euv-investment-concerns-confirms-purchase-for-rd-use-flags-cost-driven-production-timing/ - [23]ASML Expects First High-NA EUV Memory, Logic Products Within Months Amid TSMC's Cost-Driven Delay — TrendForceTier 2criticalHigh confidence
High-NA systems cost roughly US$350–400m each — about double prior-generation EUV — and analyst reporting has flagged that using High-NA could raise production cost up to ~2.5x vs conventional EUV, the core of TSMC's hesitation.
“remain too expensive for use at this stage (up to $400 million each).”
https://www.trendforce.com/news/2026/05/20/news-asml-expects-first-high-na-euv-memory-logic-products-within-months-amid-tsmcs-cost-driven-delay/ - [24]ASML Expects First High-NA EUV Memory, Logic Products Within Months — TrendForceTier 2supportingHigh confidence
There is real demand at the front of the curve: Intel deployed the industry's first High-NA system (Twinscan EXE) for its 14A node, SK Hynix plans High-NA for next-gen DRAM, and ASML CEO Christophe Fouquet expects the first memory and logic products on High-NA within months.
“Intel ... Deployed the industry's first High-NA lithography system in December 2025 ... supports development of its 14A process.”
https://www.trendforce.com/news/2026/05/20/news-asml-expects-first-high-na-euv-memory-logic-products-within-months-amid-tsmcs-cost-driven-delay/ - [25]ASML Spotlight — EUV Monopoly, High-NA, Zeiss Optics & Export Controls — SemiconductorXTier 3supportingMedium confidence
Beyond High-NA, the Hyper-NA (toward 0.75 NA) roadmap (~2032) is effectively ZEISS-gated: ASML cannot commit to a Hyper-NA timeline until ZEISS confirms the optics can be designed and built, making the pace of lithography progress dependent on ZEISS's optics capability.
“ASML cannot commit to a Hyper-NA product timeline without Zeiss confirming that the necessary optics can be designed and manufactured.”
https://semiconductorx.com/spotlight-asml.html ZEISS's German High-NA material puts numerical aperture at 0.55 (vs 0.33), enabling sub-10 nm resolution and ~3x more structures per area, and frames the atomic-precision, 100+-layer mirrors as the 'heart of the machine'.
“The mirrors, manufactured to atomic precision and consisting of more than a hundred layers, are the heart of the machine.”
https://www.zeiss.de/semiconductor-manufacturing-technology/inspiring-technology/high-na-euv-lithographie.htmlZEISS frames High-NA as a 25-year bet now paying off. Andreas Pecher (then SMT President/CEO, now ZEISS Group CEO) called ASML's High-NA scanner the most advanced, most complex and most precise microchip-making machine in the world.
“Our long breath over more than 25 years has now paid off: ASML's new wafer scanner for High-NA EUV lithography is the most modern, most complex and most precise microchip-manufacturing machine in the world.”
https://www.elektroniknet.de/elektronikfertigung/fertigungstechnik/zeiss-baut-das-herzstueck-fuer-die-ic-fertigung.213853.htmlDr. Peter Kürz, head of the High-NA-EUV business field at ZEISS SMT, explains the optical mechanism: the larger numerical aperture and new mirrors capture more light and allow finer, more precise exposure.
“Thanks to the larger numerical aperture and the new High-NA mirrors we can capture more light and so expose in even finer detail and with greater precision.”
https://www.elektroniknet.de/elektronikfertigung/fertigungstechnik/zeiss-baut-das-herzstueck-fuer-die-ic-fertigung.213853.htmlThe physical scale-up of High-NA optics is large in German terms too: a High-NA mirror is about twice as big and ten times as heavy as a previous-generation EUV mirror.
“A High-NA EUV mirror is about twice as large and ten times as heavy as previous EUV mirrors.”
https://www.zeiss.de/semiconductor-manufacturing-technology/inspiring-technology/high-na-euv-lithographie.html- [62]ASML Expects First High-NA EUV Memory, Logic Products Within Months Amid TSMC's Cost-Driven Delay — TrendForceTier 2criticalHigh confidence
High-NA's value proposition is contested even by sophisticated buyers: rather than adopt it early, TSMC says it will keep extracting gains through chip design and conventional EUV — i.e. the optics ZEISS builds may face a slower, narrower market than the technology's capability implies.
“Strategy focuses on extracting gains through innovative chip design rather than relying solely on new ASML systems.”
https://www.trendforce.com/news/2026/05/20/news-asml-expects-first-high-na-euv-memory-logic-products-within-months-amid-tsmcs-cost-driven-delay/
Peer Comparison & Benchmarking
- [26]ASML reports €28.3 billion total net sales and €7.6 billion net income in 2024 — GlobeNewswireTier 2neutralHigh confidence
ZEISS SMT (€5.055bn FY2024/25 revenue) is a component supplier inside the stack of its customer ASML, whose FY2024 net sales were €28.3bn with €7.6bn net income and a 51.3% gross margin; ASML guided 2025 sales to €30–35bn.
“Total net sales: €28.3 billion ... Net income: €7.6 billion ... Gross margin: 51.3%.”
https://www.globenewswire.com/news-release/2025/01/29/3016895/0/en/ASML-reports-28-3-billion-total-net-sales-and-7-6-billion-net-income-in-2024.html The other core European EUV supplier, TRUMPF (which builds the drive laser), posted €4.3bn group revenue in FY2024/25 (down from €5.2bn) across all its laser/machine-tool businesses — of which semiconductor lasers are only a part — illustrating how exposed even the EUV winners are to the broader industrial cycle.
“FY 2024/25 Sales: €4.3 billion ... Order Intake: €4.2 billion.”
https://www.trumpf.com/en_US/newsroom/fiscal-year-figures-2024-25/Within the ZEISS Group, SMT is now the largest and fastest-growing segment: its €5.055bn was ~42% of the group's €11.896bn FY2024/25 revenue, up from ~38% (€4.122bn of €10.894bn) a year earlier.
“Semiconductor Manufacturing Technology ... Revenue: 5.055 billion euros (prior year: 4.122 billion euros).”
https://www.zeiss.com/corporate/en/about-zeiss/present/newsroom/press-releases/2025/end-of-year.html- [29]WFE Lithography: ASML EUV, DUV Scanners, Nikon, Canon, SMEE — SemiconductorXTier 3neutralMedium confidence
On lithography capability, ASML+ZEISS occupy the leading edge alone; Nikon and Canon hold mature-node DUV and (for Canon) nanoimprint niches; China's SMEE is qualified only at older nodes and is gated by foreign optical-system dependencies it cannot yet replace.
“Chinese progress is gated by foreign optical system dependencies that remain difficult to replace domestically.”
https://semiconductorx.com/wfe-lithography.php - [56]ASML Spotlight — EUV Monopoly, High-NA, Zeiss Optics & Export Controls — SemiconductorXTier 3criticalMedium confidence
The decisive way ZEISS SMT differs from any peer is concentration, not scale: unlike a diversified equipment maker, it routes nearly all lithography-optics revenue through one customer and one cyclical end-market, so the most relevant 'benchmark' is its single-point dependency rather than a revenue league table.
“the two companies have co-developed EUV optics technology for over 25 years, sharing process knowledge, aligning roadmaps, and building production capacity in coordinated lockstep.”
https://semiconductorx.com/spotlight-asml.html
Financials, Growth & Investment
ZEISS SMT revenue rose 23% (currency-adjusted) to €5.055bn in FY2024/25 (ended 30 Sept 2025) — over €5bn for the first time — from €4.122bn the prior year. The ZEISS Group as a whole grew 9% to €11.896bn with EBIT of €1.552bn (13% margin) and 46,622 employees.
“Revenue: 5.055 billion euros (prior year: 4.122 billion euros) ... Growth: +23%.”
https://www.zeiss.com/corporate/en/about-zeiss/present/newsroom/press-releases/2025/end-of-year.html- [31]Good overall result and revenue growth for the ZEISS Group (FY2023/24) — ZEISSTier 1neutralHigh confidence
In FY2023/24 SMT grew 16% (currency-adjusted) to €4.122bn with incoming orders up 11%, helped by strong DUV demand and a pull-forward of China orders; ZEISS supplied ASML with its first High-NA optics during that year. Group revenue was €10.894bn (+8%), EBIT €1.444bn (down from €1.686bn).
“special circumstances such as strong demand for semiconductor production equipment in China.”
https://www.zeiss.com/corporate/en/about-zeiss/present/newsroom/press-releases/2024/end-of-year.html SMT's run has been steep: revenue was ~€3.6bn in FY2022/23 (+29% YoY, nearly double the 2020 total), then €4.122bn (FY2023/24) and €5.055bn (FY2024/25) — roughly a tripling in four years.
“The semiconductor manufacturing technology (SMT) division achieved €3.6 billion in sales, representing a 29% increase ... nearly double the total recorded as recently as 2020.”
https://optics.org/news/14/12/22ZEISS is funding the biggest capacity ramp in SMT's history — the group has said it will invest around €3.5bn in infrastructure over five years (much of it in Germany: Oberkochen, Aalen-Ebnat, Jena), with the Oberkochen south factory expanding by 25,000+ m²; SMT's workforce has tripled in seven years.
“We are currently in the biggest ramp-up in ZEISS SMT's history and will expand the south factory in Oberkochen by more than 25,000 square meters of production space.”
https://www.zeiss.com/semiconductor-manufacturing-technology/smt-magazine/ramp-up-at-zeiss-smt-sites.htmlManagement is expanding cautiously: ZEISS frames FY2024/25 as a 'dynamic and challenging' environment and stresses cost discipline, efficient processes and capacity adjustments amid semiconductor volatility — the capacity build coexists with explicit caution.
“strict cost awareness, efficient processes, and ... capacity adjustments.”
https://www.zeiss.com/corporate/en/about-zeiss/present/newsroom/press-releases/2025/end-of-year.html- [35]New multifunctional factory for ZEISS in Wetzlar celebrates topping-out ceremony — ZEISS SMTTier 1neutralHigh confidence
A new multifunctional factory in Wetzlar (12,000 m², ~150 jobs) for DUV lithography optics topped out in June 2024, part of distributing the optics build across more German sites.
“The state-of-the-art multifunctional factory will offer us excellent conditions for the production of DUV lithography optics in the future.”
https://www.zeiss.com/semiconductor-manufacturing-technology/news-and-events/smt-press-releases/2024/topping-out-ceremony-in-wetzlar.html
Ownership, Governance & People
Carl Zeiss AG is wholly owned by the Carl-Zeiss-Stiftung (Carl Zeiss Foundation), which is also sole owner of glassmaker SCHOTT AG. The Foundation is barred by its statute from selling the shares — an 'inalienable' ownership in place for over a hundred years, rooted in Ernst Abbe's 19th-century foundation deed.
“The Carl-Zeiss-Stiftung is the sole shareholder of the two foundation companies Carl Zeiss AG and SCHOTT AG.”
https://www.carl-zeiss-stiftung.de/en/foundation/structureFoundation ownership is also a constraint: because the foundation cannot sell the companies, ZEISS has no equity currency, cannot be acquired or merged, and SMT's billion-euro growth must be self- and partner-funded rather than raised on public markets.
“the foundation companies were demerged from the Foundation in 2004 and transformed into legally independent joint-stock companies.”
https://www.carl-zeiss-stiftung.de/en/foundation/structure- [38]ZEISS and ASML Strengthen Partnership for Next Generation of EUV Lithography — ASMLTier 1neutralHigh confidence
ASML's 24.9% stake in Carl Zeiss SMT GmbH (acquired for €1bn in 2016) aligns the two roadmaps and brought ~€220m of R&D support plus ~€540m of capex/supply-chain investment over six years — but it also formalises a customer's seat inside ZEISS's most important business.
“ASML will support Carl Zeiss SMT's research and development for approximately €220 million as well as capital expenditures and other supply chain investments for approximately €540 million over the next six years.”
https://www.asml.com/en/news/press-releases/2016/zeiss-and-asml-strengthen-partnership-for-next-generation-of-euv-lithography - [39]ZEISS, TRUMPF and Fraunhofer team awarded the Deutscher Zukunftspreis 2020 — FraunhoferTier 2supportingHigh confidence
EUV created thousands of high-skill German jobs: the 2020 Deutscher Zukunftspreis citation noted over 3,300 high-tech jobs created at ZEISS and TRUMPF and EUV turnover above €1bn already in 2019, underscoring the talent and regional-economy stakes around Oberkochen and Jena.
“over 3,300 high-tech jobs have been created at ZEISS and TRUMPF to date and an annual turnover of more than one billion euros was generated in 2019.”
https://www.fraunhofer.de/en/press/research-news/2020/november/zeiss-tumpf-and-fraunhofer-research-team-awarded-the-deutscher-zukunftspreis-2020-for-the-development-of-euv-litography.html - [52]Carl-Zeiss-Stiftung: Förderung von Wissenschaft und Forschung — ZEISSTier 1supportingHigh confidence
The Carl Zeiss Foundation, founded in Jena in 1889 by Ernst Abbe (with SCHOTT founder Otto Schott), is the sole shareholder of Carl Zeiss AG and SCHOTT AG and is barred by its statute from selling the shares; its dividends fund STEM research and teaching.
“The Carl Zeiss Foundation is the sole shareholder of Carl Zeiss AG and SCHOTT AG. The Foundation may not sell the shares of its two companies — this is laid down in the Foundation statute.”
https://www.zeiss.de/corporate/ueber-zeiss/gegenwart/daten-fakten/carl-zeiss-stiftung.html
Risks & Challenges
- [40]ASML Spotlight — EUV Monopoly, High-NA, Zeiss Optics & Export Controls — SemiconductorXTier 3criticalMedium confidence
Customer concentration is the defining risk: ZEISS SMT's lithography optics sell to essentially one integrator, ASML, which holds a 24.9% stake — a degree of single-customer dependence rare for a monopolist, leaving SMT exposed to ASML's order timing and roadmap choices.
“ASML cannot ship High-NA EUV scanners faster than Zeiss produces High-NA optics sets.”
https://semiconductorx.com/spotlight-asml.html - [41]ASML Spotlight — EUV Monopoly, High-NA, Zeiss Optics & Export Controls — SemiconductorXTier 3criticalMedium confidence
Geographic concentration is a systemic fragility: most EUV-scanner value-add sits in two European cities (ASML's Veldhoven and ZEISS's Oberkochen); a major disruption at either would constrain global leading-edge chip capacity within roughly 12–18 months.
“A major disruption at either Veldhoven or Oberkochen would directly constrain global leading-edge semiconductor manufacturing capacity within 12-18 months.”
https://semiconductorx.com/spotlight-asml.html - [42]ASML expects impact of updated export restrictions to fall within outlook for 2025 — ASMLTier 1criticalHigh confidence
Export controls cap the addressable market: EUV has never been exportable to China (prohibited since 2019), and from December 2024 the Netherlands/US extended controls to advanced DUV immersion systems and metrology; ASML expected China to fall to ~20% of 2025 sales after a 2024 pull-forward, removing a demand cushion ZEISS had benefited from.
“exports of DUV immersion lithography systems to these specific locations could also be affected ... China business ... to be around 20% of our total net sales for that full year.”
https://www.asml.com/en/news/press-releases/2024/asml-expects-impact-of-updated-export-restrictions-to-fall-within-outlook-for-2025 - [43]TSMC Rejects High-NA EUV Investment Concerns, Confirms Purchase for R&D Use — TrendForceTier 2criticalHigh confidence
Cyclicality and a contested next node compound the risk: semiconductor-equipment demand is boom-bust, and if High-NA adoption stalls on cost (TSMC deferring volume to ≥2029) the multi-billion-euro High-NA capacity build could run ahead of paying demand for years.
“the main reason the High-NA EUV system has not yet been introduced into mass production is purely cost-related.”
https://www.trendforce.com/news/2026/06/04/news-tsmc-rejects-high-na-euv-investment-concerns-confirms-purchase-for-rd-use-flags-cost-driven-production-timing/ - [44]About us as company & team — ZEISS Semiconductor Manufacturing TechnologyTier 1supportingMedium confidence
Mitigants exist: a vast installed base needing service and mask/process-control tools, diversification into Fab and Mask Solutions, ASML's large multi-year backlog, and the fact that ZEISS is the binding constraint on High-NA supply all soften the single-customer and cyclical risks.
“Semiconductor Mask Solutions ... Semiconductor Fab Solutions: Inspection, metrology, and defect analysis tools.”
https://www.zeiss.com/semiconductor-manufacturing-technology/about-us.html
Forward View
- [45]ASML Expects First High-NA EUV Memory, Logic Products Within Months — TrendForceTier 2supportingMedium confidence
ASML expects the first memory and logic products built on High-NA within months — an early signal that, even with TSMC deferring, the technology ZEISS supplies is entering production at Intel, Samsung and SK Hynix.
“the first memory and logic products manufactured on High-NA EUV systems within the next few months.”
https://www.trendforce.com/news/2026/05/20/news-asml-expects-first-high-na-euv-memory-logic-products-within-months-amid-tsmcs-cost-driven-delay/ The decisive variables ahead: the slope of High-NA (and later Hyper-NA) adoption versus its cost; whether AI-driven leading-edge demand stays strong; how far export controls shrink the served market; and whether ZEISS can scale the hardest optics in the world fast enough without over-building into a downturn.
“Approximately 2,000 of ZEISS SMT's 9,349+ employees work on High-NA-EUV.”
https://www.zeiss.com/semiconductor-manufacturing-technology/inspiring-technology/high-na-euv-lithography.html- [57]TSMC Rejects High-NA EUV Investment Concerns, Confirms Purchase for R&D Use — TrendForceTier 2criticalHigh confidence
The bear path is concrete: if High-NA economics keep big buyers cautious (TSMC deferring volume to ≥2029) and a cyclical downturn coincides with the multi-billion-euro capacity build, ZEISS could be left with expensive capacity ahead of a deferred order book.
“TSMC will continue improving efficiency and reducing costs before moving the technology into production.”
https://www.trendforce.com/news/2026/06/04/news-tsmc-rejects-high-na-euv-investment-concerns-confirms-purchase-for-rd-use-flags-cost-driven-production-timing/