ASML EUV Monopoly: Can the 1,000-Watt Breakthrough Survive?

There is exactly one company on Earth that can build a $380 million machine capable of etching circuits smaller than a virus. On February 23, 2026, that company proved it can make that machine 50% faster. Three challengers think they can break this 30-year ASML EUV monopoly. Here is why they are probably wrong — and why it matters that they are trying.

TL;DR — ASML’s EUV monopoly faces its first real challengers in 30 years.

• ASML demonstrated a 1,000-watt EUV source — a 50% throughput leap toward 330 wafers/hour by 2030.
• Three challengers (xLight, Canon NIL, China’s SMEE) are attacking from different angles, but none threaten the monopoly within 5 years.
• The real story is geopolitical: lithography is now a national security asset, and both superpowers want to either own or destroy the ASML EUV monopoly.

The Machine That Prints the Future

Picture a factory that costs $20 billion to build. Inside, the single most expensive piece of equipment — at $200 million per unit — is a machine the size of a city bus. It fires three laser beams at 100,000 tin droplets per second, generating extreme ultraviolet light at a wavelength of 13.5 nanometers. That light passes through mirrors polished to sub-atomic precision, then etches circuit patterns smaller than a COVID virus onto silicon wafers.

This is EUV lithography. And there is exactly one company that makes these machines: ASML, based in Veldhoven, Netherlands.

EUV lithography is the printing press of the semiconductor industry. If a chip is a city, lithography is the blueprint printer that determines how small the buildings can be and how many fit per block. The smaller and more precise the printing, the more powerful the chip. Every AI chip powering ChatGPT, every smartphone processor, every data center GPU — they all pass through an ASML machine.

ASML does not merely dominate this market. It owns it. 100% market share in EUV lithography. Roughly 94% of all semiconductor lithography equipment revenue (Mordor Intelligence). This is not a duopoly or an oligopoly. It is an ASML EUV monopoly in the purest sense of the word.

And on February 23, 2026, ASML made that monopoly stronger.

The 1,000-Watt ASML EUV Monopoly Proof of Concept

ASML’s 1,000-watt EUV source demonstration was not a product launch. It was a proof of concept — a signal to the industry about where the technology is heading (Bits&Chips, Electronics Weekly).

Here is why watts matter. In EUV lithography, source power directly determines throughput — how many wafers per hour (WpH) a machine can process. More watts means more photons per laser pulse, which means faster exposure times, which means more chips per hour. It is the factory equivalent of upgrading from a single-lane road to a highway.

Current production EUV systems run at around 600 watts, delivering approximately 220 wafers per hour. The 1,000-watt milestone is step one on ASML’s roadmap to 1,500 watts and eventually 2,000 watts by 2030, targeting 330 WpH — a 50% increase over today (Electronics Weekly, ASML SPIE presentation).

For chipmakers, this is not an incremental improvement. A 50% throughput increase means 50% more chips from the same number of machines. When each machine costs $200 million, that math changes entire capital expenditure plans.

ASML’s Financial Fortress: The EUR 38.8 Billion Backlog

If you want to understand why the ASML EUV monopoly is so durable, look at the numbers.

EUR 38.8 billion in backlog means ASML is sold out through 2027 and beyond. This is not a company worried about next quarter. The demand is driven by the AI infrastructure buildout — every major hyperscaler and chipmaker is racing to secure EUV capacity.

Lithography accounts for approximately 31% of total chip cost of goods sold (COGS) — the single largest cost component, ahead of etch (15%) and deposition (14%). When one company controls 100% of the most expensive step in chipmaking, pricing power follows naturally. ASML’s EUV average selling price has climbed from roughly EUR 120 million in 2019 to EUR 200 million in 2025.

The total semiconductor lithography equipment market is valued at $28-30 billion in 2025 (Mordor Intelligence). The EUV segment alone — $9.7 billion in 2024 — is projected to reach $18.4 billion by 2030, growing at a 14.9% CAGR (GM Insights).

High-NA EUV: The $380 Million Bet That Split the Industry

The next frontier of the ASML EUV monopoly is High-NA (Numerical Aperture) EUV — the EXE:5200B system. If standard EUV is a high-definition camera, High-NA is 8K. It pushes resolution limits further, enabling even smaller circuit features.

The price tag: EUR 350-380 million per tool. That is roughly $380-410 million, making it the most expensive piece of manufacturing equipment on the planet.

But here is where it gets interesting. The industry has split on whether to adopt it.

TSMC’s decision to skip High-NA is the most consequential. The world’s largest foundry — responsible for manufacturing chips for Apple, NVIDIA, and AMD — has concluded it does not need ASML’s most advanced tool. “We don’t need it,” TSMC’s CTO stated at IEDM 2025 (Tom’s Hardware).

TSMC’s rationale: High-NA costs 2.5 times more than standard EUV, delivers lower throughput initially, and the customer base is not ready for the density it enables. Double patterning with existing 0.33 NA tools is good enough for 1.4nm.

The implication for ASML is subtle but important. If the world’s leading foundry can skip a generation, the High-NA market is smaller than projected. Intel and Samsung — both playing catch-up in advanced logic — need High-NA more, precisely because they are behind. This is the semiconductor bottleneck dynamic playing out at the equipment level.

Three Challengers to the ASML EUV Monopoly

For 30 years, no one has seriously threatened ASML’s lithography dominance. “ASML’s monopoly is the result of a moonshot that every competitor abandoned,” as one venture capital analysis put it (Strange VC). Now, three challengers are mounting attempts from fundamentally different directions.

xLight: The American Moonshot

When Pat Gelsinger departed Intel in December 2024, the industry assumed he would retire. Instead, he joined xLight as Executive Chairman — a startup founded in 2021 to build EUV lithography tools based on Free Electron Laser (FEL) technology.

FEL is a fundamentally different approach to generating EUV light. ASML’s method (laser-produced plasma) fires lasers at tin droplets. FEL uses a particle accelerator to generate coherent light directly. Think of ASML’s approach as burning fuel to create light, and FEL as using electricity to create light — cleaner, potentially more powerful, and tunable to different wavelengths.

The numbers: xLight has secured a $150 million CHIPS Act R&D letter of intent from NIST and the Commerce Department (NIST/Commerce Dept), plus over $40 million in venture capital (TechCrunch). The company is setting up at the Albany NanoComplex in New York, chosen for proximity to existing fab R&D infrastructure (Manufacturing Dive).

xLight targets its first wafer exposure by 2028. That is ambitious. FEL facilities have historically been accelerator-sized installations. Shrinking that to fab-compatible dimensions is an engineering challenge with no precedent (TrendForce).

The risk assessment: High technology promise, unproven path to manufacturing scale. But xLight exists for a reason beyond commercial ambition — the US government wants a domestic alternative to the Dutch ASML EUV monopoly.

Canon Nanoimprint: The Niche Player

Canon delivered its first nanoimprint lithography (NIL) system to TIE Semiconductor in October 2024 (Canon USA) and opened its first new lithography factory in 21 years specifically for NIL production (Conevoelec).

NIL works differently from EUV. Instead of projecting light through a mask, it physically stamps a pattern onto a wafer — like a cookie cutter for circuits. No vacuum chamber needed. No extreme ultraviolet light source. Simpler optics, lower cost per tool.

SemiAnalysis captured the reality perfectly: “Stop saying nanoimprint will replace EUV” (SemiAnalysis). NIL is fundamentally a contact-based technology, which creates defect risks that disqualify it from leading-edge logic manufacturing. Overlay accuracy is insufficient for the sub-3nm logic chips where the ASML EUV monopoly reigns.

Where NIL does work: specific layers in NAND flash memory manufacturing. This is a real market, but it is a niche. Canon is not trying to replace ASML across the board. It is carving out a cost-effective alternative for a narrow application.

SMEE: China’s Long March to EUV

Shanghai Micro Electronics Equipment (SMEE) is China’s primary lithography champion, and its trajectory reveals both the ambition and the gap.

Current capability: SMEE has achieved DUV immersion lithography (ArF-i) comparable to ASML’s NXT:2050 generation. For EUV, reports indicate a 100-150 watt laser-produced plasma prototype (CSIS, Asia Times). For context, ASML was at this power level around 2015. The gap is roughly a decade.

A December 2025 assessment from the Center for Strategic and International Studies (CSIS) titled “Breakthroughs or Boasts?” captured the nuance: SMEE’s realistic timeline for limited EUV production is 2030 at the earliest, and the challenge is not just source power. The entire ecosystem — mirrors, pellicles, photoresist — must be built from scratch (CSIS).

SMEE won a RMB 110 million contract from a domestic customer in December 2025. Reports of a “reverse-engineered EUV tool in a covert lab” have circulated (Tom’s Hardware), though these remain unconfirmed. Asia Times reported “Made-in-China EUV targets AI chip output by 2028” — a claim most analysts consider extremely aggressive (Asia Times).

The strategic reality: SMEE does not need to match ASML to matter. If China can produce 2-3 generation-old EUV tools domestically, it gains leverage in a technology war where access itself is the weapon.

The Geopolitical Premium on the ASML EUV Monopoly

The most telling number in ASML’s 2025 results is not revenue or backlog. It is the China revenue decline: from approximately 36% of total revenue in 2024 to roughly 20% in 2025 (ASML Q4 2025 Report).

This drop is not market-driven. It is policy-driven. US-led export controls, tightened throughout 2024-2025, have systematically restricted ASML’s ability to sell advanced lithography equipment to Chinese customers. ASML has complied — it is a Dutch company operating under Dutch and EU export regulations, but the pressure originates in Washington.

ASML’s guidance is revealing: management reiterated EUR 30-35 billion revenue for 2026, stating that “total revenue won’t fall below 2025 levels” despite the China decline (CNBC). AI demand from non-China customers is more than compensating. The AI infrastructure capital expenditure cycle is filling the gap that geopolitics created.

Here is the irony. xLight exists because the US government does not want to depend on a European company for the most critical technology in semiconductor manufacturing — even an allied European company. The $150 million CHIPS Act investment is national security spending, not commercial R&D.

SMEE exists for the mirror-image reason. China needs domestic lithography because it cannot buy ASML tools. Both superpowers are funding ASML alternatives — one to reduce dependency, the other to circumvent restrictions.

Lithography has become the ultimate dual-use technology. It is the chokepoint where commercial semiconductor competition and great-power competition converge. ASML sits at the center of what analysts increasingly call the “Silicon Curtain.”

What the ASML EUV Monopoly Means for the Industry

The semiconductor value chain is a hierarchy of dependencies, and ASML sits at the top. The HBM memory race between Samsung and SK Hynix requires EUV. NVIDIA’s platform dominance depends on TSMC’s ability to manufacture at scale, which depends on ASML’s machines. Every layer of the AI infrastructure stack eventually traces back to Veldhoven.

For professionals in the semiconductor and AI ecosystem, three dynamics are worth tracking.

First, the throughput roadmap. ASML’s path from 600W to 2,000W by 2030 will determine how quickly chipmakers can expand capacity without buying proportionally more tools. This directly affects the cost curve of AI chips — and by extension, the economics of AI deployment.

Second, the High-NA bifurcation. TSMC’s skip creates a two-track industry: companies that adopt High-NA (Intel, Samsung, SK Hynix) and those that optimize around standard EUV (TSMC). Which approach yields better economics will define competitive positions for the rest of the decade.

Third, the geopolitical premium. ASML’s stock price, China’s semiconductor ambitions, and US industrial policy are now entangled in ways that make pure technology analysis insufficient. Understanding lithography now requires understanding trade policy, export controls, and alliance dynamics.

Bottom Line. Monopolies don’t die from frontal assault. They die from the accumulation of alternatives. ASML’s 1,000-watt breakthrough bought time — the EUR 38.8 billion backlog proves the fortress holds. But for the first time in 30 years, three challengers are chipping away from different angles, and two superpowers are funding the siege. The clock is ticking.

Career Takeaway. If you work anywhere in the semiconductor, AI, or advanced manufacturing ecosystem, lithography literacy is no longer optional. The decisions ASML, TSMC, and Intel are making about EUV today will shape hiring patterns, supply chain structures, and technology roadmaps for the next decade. Understanding who controls the printing press tells you who controls the future.

References

1. Bits&Chips — ASML 1,000W EUV source proof of concept
2. Tom’s Hardware — ASML breakthrough coverage, TSMC skipping High-NA, SMEE reverse engineering reports
3. ASML Investor Relations — Q4 2025 financial results
4. Counterpoint Research — ASML 2025 revenue analysis
5. TrendForce — High-NA adoption, xLight investment, SMEE developments
6. SK Hynix — First commercial High-NA EUV application
7. NIST/Commerce Dept — CHIPS R&D letter of intent with xLight
8. TechCrunch — Gelsinger/xLight profile
9. Manufacturing Dive — xLight Albany facility
10. Canon USA — NIL system delivery to TIE Semiconductor
11. SemiAnalysis — “Stop saying nanoimprint will replace EUV”
12. CSIS — “Breakthroughs or Boasts?” China lithography assessment
13. Asia Times — China EUV ambitions
14. Mordor Intelligence — Semiconductor lithography equipment market
15. GM Insights — EUV lithography equipment market projections
16. CNBC — ASML Q4 earnings analysis
17. Electronics Weekly — 1,000W to 330 WpH roadmap
18. Strange VC — ASML 30-year monopoly analysis
19. Conevoelec — Canon NIL factory

This article is part of Semiconductor Anatomy S1, a series dissecting the hidden infrastructure powering the AI revolution. Part 1 covered the HBM memory war. Next: the foundry duopoly.

Disclaimer: This article is for informational purposes only and does not constitute investment advice. All data sourced from public filings, industry reports, and news publications as cited.