The reported 15-year gap between China and leading EUV lithography capability is best understood as an estimate of technological position—not a measured benchmark or a promise that China will need exactly 15 calendar years to catch up. Frank Rohmund, president and CEO of ZEISS Semiconductor Manufacturing Technology, was reported in September 2026 as describing that gap as a reasonable assumption for China’s development of EUV lithography machines. The estimate does not establish a ranking of China’s entire semiconductor industry.
What makes the claim meaningful is the machinery behind it. ZEISS supplies the precision illumination and projection optics used in ASML’s EUV systems, while ASML builds the complete lithography scanners. Those optics are not ordinary lenses with a fancier price tag. They are multilayer mirrors that must work in a vacuum, handle extreme ultraviolet light and preserve an extraordinarily precise image of a chip pattern.
What the 15-year estimate actually measures
The number describes a reported technology gap in extreme ultraviolet lithography, or EUV—the manufacturing method used to project exceptionally small circuit patterns onto silicon wafers. It is a forecast about relative capability, not a laboratory measurement with a start date and an inevitable finish date.
That distinction matters. A country can be behind in one subsystem, close in another and still make progress through a different engineering route. The estimate also does not prove that China is fifteen years away from every advanced chipmaking capability, nor that a future domestic EUV scanner would arrive on a fixed timetable.
The practical comparison is between an established industrial EUV architecture and a Chinese effort whose reported milestones remain less mature. A reported light source or prototype is an important engineering step, but it is not the same thing as a production scanner that can repeatedly print chips at commercial yield.
Why EUV scanners use mirrors instead of lenses
EUV uses light with a wavelength of 13.5 nanometers. At that wavelength, ordinary optical materials absorb the light, and air itself becomes a problem. Conventional refractive lenses therefore cannot guide EUV through a scanner in the familiar way.
The solution is a chain of reflective optics inside a vacuum. EUV systems use multilayer mirrors whose surfaces and coatings are engineered to reflect the required wavelength. ZEISS describes projection optics built around six mirrors, while its EUV architecture also includes separate illumination optics that shape and direct the light toward the mask, or reticle, carrying the chip pattern.
The source is demanding, too. ZEISS describes a system in which tin droplets are struck by two CO2-laser pulses in a vacuum to create a tin plasma that emits EUV light. The light then has to travel through a carefully controlled optical path before reaching the wafer. There is no convenient glass lens waiting to clean up the mistakes.
The precision problem inside an EUV optical system
The difficulty is not just polishing one shiny mirror. It is coordinating a stack of materials, surfaces, coatings, measurements and contamination controls so that the entire system preserves the intended wavefront—the shape and timing of the light wave—as it moves through the scanner.
A technical explanation of EUV optics describes mirrors made from alternating silicon and molybdenum layers, with example layer thicknesses of about 4.1 nanometers and 2.7 nanometers. Those layers act as a Bragg reflector: carefully spaced interfaces reinforce the desired reflected light. A small error in spacing reduces the optical performance.
The same technical material discusses surface-polishing targets in the picometer range. One cited figure is 53 picometers RMS, a video-derived technical value rather than a ZEISS specification. The point is easier to grasp as a rule of thumb: a surface can look perfectly smooth to a human eye and still be wildly inaccurate for EUV imaging.
Contamination adds another headache. The light source produces tin plasma, and tin deposits can degrade the collector mirror. The system therefore needs cleaning and maintenance processes as well as precision optics. In other words, the challenge is a chain of interdependent subsystems—not simply the ability to manufacture a reflective surface.
ZEISS’s own figures show the industrial scale of that chain. Its EUV illumination system contains about 15,000 individual parts and weighs 1.5 tonnes. The projection optics contain roughly 20,000 parts, weigh about 2 tonnes and use six mirrors. Those figures describe the complexity of the established system; they do not by themselves measure the distance to a competing design.
ZEISS and ASML: who supplies what?
ZEISS supplies the optical systems: the illumination and projection assemblies built around precision multilayer mirrors. ASML integrates those optics with the light source, wafer-handling equipment, control systems and other components needed to make a complete EUV lithography machine.
That division is important when interpreting the 15-year claim. Replicating one optical subsystem would not automatically reproduce a full scanner. Conversely, progress in a light source or another subsystem would not demonstrate that the complete machine can print working chips at production quality.
China’s reported EUV and DUV position
The reported Chinese program has two distinct tracks. The first is EUV, where secondary reporting described a prototype capable of generating EUV light but not producing a working chip at the time covered. The second is deep ultraviolet, or DUV, where the reporting places China closer to volume production of immersion systems.
DUV uses a longer wavelength—in the comparison supplied by ZEISS, 193 nanometers—and is a different technological route from EUV. It can remain useful for chip production, including through techniques that require more patterning steps, even when a manufacturer does not have a domestic EUV scanner.
That is why “China is behind in EUV” and “China is advancing in semiconductor manufacturing” are not contradictory statements. They refer to different equipment classes and different levels of demonstrated industrial maturity.
| Technology dimension | ZEISS/ASML EUV position | China’s reported position |
| Optical method | Multilayer reflective mirrors operate in a vacuum at 13.5 nm | The reported domestic effort is still associated with closing the EUV optical and systems gap |
| Complete scanner | ZEISS optics are integrated into ASML EUV lithography systems | No commercial Chinese EUV production system is established by the reported account |
| EUV light source | Tin plasma generated by laser pulses provides the EUV source architecture | A reported prototype generated EUV light, but the account did not describe a working chip made with it |
| DUV progress | DUV is a separate, more mature lithography class | Reported progress toward immersion DUV volume production is closer than the reported EUV effort |
| Production meaning | Industrial EUV is a coordinated scanner ecosystem | A prototype or component milestone is not equivalent to repeatable commercial chip production |
Why 2028 and 2030 remain forecasts
Reports have mentioned 2028 as a target and 2030 as a more cautious outlook for Chinese chip production using domestic EUV technology. Neither date should be read as a confirmed production milestone.
There are several steps between a prototype and commercial output: integrating the source and optics, controlling contamination, measuring wavefront errors, maintaining alignment, processing wafers repeatedly and achieving usable yields. A calendar target can describe an ambition or forecast; it cannot substitute for a demonstrated manufacturing result.
The same caution applies to the reported prototype. A machine that generates EUV light has crossed a meaningful technical hurdle, but chipmaking requires the entire optical and manufacturing chain to work together. That is the difference between “the source works” and “the scanner produces chips reliably.”
What a technology gap does—and does not—tell us
No: a 15-year gap does not automatically mean China will need fifteen years to catch up. The figure describes an estimated distance from an existing capability, not a guaranteed timetable. A latecomer may reuse known solutions, take a different engineering path or accept more manufacturing steps and higher costs while developing the ideal system.
But alternative routes do not erase the EUV challenge. They may reduce dependence on EUV for some products without reproducing the same capability, efficiency or production economics. That is why DUV progress can matter strategically while still leaving a large gap in domestic EUV.
The cleanest reading of Rohmund’s reported estimate is therefore narrow and cautious: China’s domestic EUV effort was assessed as substantially behind the established ZEISS-and-ASML ecosystem, with optics forming one of the hardest technical barriers. The number is useful as a signal of perceived distance. It is not a stopwatch.
Bottom line: EUV lithography is a complete precision-manufacturing system, not a single lens or light source. China’s reported prototype and DUV advances show movement, but they do not turn a forecast into a demonstrated commercial EUV capability.
