China EUV Lithography Machine — Strategic Breakthrough
For decades, chip design has not been the primary industrial bottleneck in advanced semiconductor manufacturing. Lithography has been the real constraint. In particular, extreme ultraviolet lithography has been one of the hardest technologies to replicate on Earth. Now that monopoly is being contested. China has apparently built a working prototype of an extreme ultraviolet lithography machine in Shenzhen. The system was completed in early 2025 and was producing EUV radiation during testing, Reuters reported.
But it hadn’t yet produced working semiconductor chips when the program went public. Chinese planners were aiming for working chips in 2028, but sources close to the project said a date of around 2030 was more realistic. That makes a difference. China has no alternative to ASML’s production-ready EUV scanners. Meanwhile, demonstrating an integrated EUV prototype is a strategically important step in dismantling one of the West’s most consequential technological chokepoints.
Why EUV Lithography Matters
Modern semiconductor lithography projects circuit patterns onto silicon wafers. Smaller wavelengths enable smaller features. Conventional immersion deep-ultraviolet lithography uses light at a wavelength of 193 nanometers. EUV instead operates at only 13.5 nm. A simplified form of the Rayleigh resolution relationship is
Critical Dimension ≈ k₁ × λ / NA
In this equation, λ is the wavelength and NA is the numerical aperture. The wavelength is shortened from 193 nm to 13.5 nm, radically increasing the resolution potential. ASML’s standard NXE EUV systems feature 0.33 NA optics and can resolve features of about 13 nm. Critical layers at 7 nm, 5 nm, and 3 nm process nodes are manufactured using them. For the 2 nm and smaller process nodes, ASML’s new High-NA platform raises NA to 0.55 and offers about 8 nm resolution for building. So an EUV scanner is much more than just a fancy “chip printer.” It combines plasma physics, nanometer precision mechatronics, ultrahigh vacuum engineering, computational lithography, metrology, materials science, and some of the most accurate optical surfaces in the world.
China’s EUV Engineering Challenge
A normal lens cannot be used in an EUV machine because ordinary materials and even air strongly absorb 13.5-nm radiation. Therefore, the whole optical path has to be in a vacuum. ASML generates the EUV radiation by shooting high-powered laser pulses at microscopic droplets of molten tin. This process occurs 50,000 times per second, creating a plasma that emits light close to the desired wavelength of 13.5 nm. The multilayer reflective mirrors then collect and focus a fraction of this energy onto the photomask and wafer. The engineering tolerances are very close.
ASML says its wafer stage can position wafers to about a quarter of a nanometer and can adjust about 20,000 times per second. The EUV development has taken more than €6 billion of R&D in 17 years before the technology matured to high-volume manufacturing. That is why it is not the same thing to just generate EUV light and to have a commercially available lithography machine. China still needs to learn about source power, collector-mirror life, optical contamination control, mask defects, EUV photoresists, overlay accuracy, vibration isolation, and sustained wafer throughput. A machine in the lab can show patterns. A production scanner needs to do them millions of times with very low defect rates.
Inside China’s Reported EUV Program
The most important revelation from the open source is how Beijing organized the project. Shenzhen’s effort was part of a highly compartmentalized national program that included government labs, semiconductor firms linked to Huawei, and former engineers from ASML, Reuters reported. It was much larger and less sophisticated than the ASML machines of today, the sources said. China is said to have increased its physical size partly to overcome limitations in source and optical performance. The one big bottleneck is reflective optical systems. ASML’s scanners use ultra-precise optics made by Zeiss of Germany. Reuters said China has struggled to reproduce that capability.
The China Academy of Sciences’ Changchun Institute of Optics, Fine Mechanics and Physics helped bring the EUV radiation into the prototype’s optical system, reports say. At the same time, China has been putting its resources into human capital, too. Reuters said teams tried to lure experienced semiconductor experts from abroad. According to rumors, a research team led by a former ASML light-source specialist has applied for eight EUV-source patents in 18 months. It also hired some 100 university graduates to help reverse-engineer parts. Anonymous sources, not official announcements from China, have reported those details. However, they do reveal something of strategic importance: Beijing seems to be treating EUV as a national industrial capability, not a garden-variety commercial product.
Huawei and SMIC’s 7-nm Breakthrough
China had already shown that just blocking lithography equipment was not enough to stop its semiconductor development. In 2023, TechInsights examined Huawei’s Mate 60 Pro and found that SMIC manufactured the Kirin 9000S on its 7-nm second-generation process. China did this without EUV. Instead, SMIC pushed 193-nm DUV lithography with advanced multiple patterning processes. This approach decomposes complex designs into a series of independent exposure, deposition, and etch cycles. It works, but you add complexity, process time, alignment requirements, and opportunities for defects with each cycle.
Thus, the Kirin 9000S received two points. First, China can do much more with older lithography equipment than export control planners originally believed. Second, the economic and manufacturing costs of repeated DUV patterning created a huge incentive to make domestic EUV. That incentive has only increased. In July 2026, Reuters reported that Shanghai Aishengna of China had begun making immersion-DUV machines that it developed in China. Plans reportedly called for five systems in 2026 and 20 in 2027. By comparison, ASML shipped 131 comparable immersion-DUV systems in 2025. The figures are still small, but the trend is important. China is aiming for complete localization of the lithography ladder, rather than leaping straight to EUV.
The Military Significance
Advanced semiconductors are the computational foundation of modern military power. They support synthetic aperture radar processing, electronic warfare receivers, AESA radar processors, encrypted communication, infrared image processing, autonomous systems, signals intelligence, satellite payloads, and increasingly AI-assisted command systems. The US semiconductor restrictions are aimed at limiting China’s ability to develop advanced computing, artificial intelligence, and defense and intelligence applications, according to the Congressional Research Service.
They also discuss China’s military-civil fusion strategy, which allows commercial technologies to assist in military modernization. So a domestic EUV capability would have implications beyond China’s smartphone industry. Over the longer term, it can decrease China’s dependence on foreign sanctions on next-generation processors used in data centers, military simulation, cryptanalysis, electronic warfare, and autonomous platforms. More importantly, it would alter the strategic character of semiconductor export controls. Controls are most effective when there is no indigenous substitute for the technology being controlled.
Prototype Is Not Production
We should neither overestimate nor underestimate China’s achievement. ASML remains well ahead. Its state-of-the-art EUV systems combine decades of accumulated know-how from ASML, Zeiss, Trumpf, Cymer, and hundreds of specialized suppliers. China still has to prove that it can deliver acceptable overlay, uptime, contamination control, source stability, throughput, and production yield. The historical analogy is helpful.
ASML presented EUV technology years before it was economically feasible for mass production. Its first EUV production system went into operation in 2013, but widespread high-volume use did not begin until much later. China could follow a similar path. However, Beijing has one big advantage: it knows EUV works. It does not need to identify the technological end point. It has to figure out how to replicate the industrial ecosystem to achieve that goal there.
Strategic Breakthrough, Not Yet Parity
References
- Reuters — How China built its ‘Manhattan Project’ to rival the West in AI chips, December 2025. (Reuters)
- ASML—EUV Lithography Systems: Technology, NXE, and High-NA EUV. (ASML)
- Reuters—China starts production of homegrown immersion DUV chipmaking tools, July 2026. (Reuters)
- Congressional Research Service — U.S. Export Controls and China: Advanced Semiconductors, August 2025. (Congress.gov)
About the Author
