China announces its first EUV prototype, a strategic success with limited range

China is said to have built its own EUV lithography prototype for the first time. This news is less relevant because of its immediate technical implications than because of its strategic dimension. According to a report by Reuters, a functional system exists that generates EUV light and is basically suitable for wafer exposure. Not yet a production tool, not yet a tape-out, but no longer a theoretical thought experiment.

Beijing has been working for years on breaking open the central bottleneck of modern chip production. EUV lithography is the bottleneck, technologically, economically and geopolitically. ASML has a de facto monopoly here, politically flanked by US export controls. China has therefore been trying for a long time to reduce this dependency using familiar methods, reverse engineering, targeted poaching of skilled workers and the use of existing technology. This is also reflected in the current prototype. According to the Reuters report, the system is based on components from older ASML machines. This is not a sign of weakness, but a realistic approach. EUV systems are extremely complex overall constructions consisting of a light source, mirror optics, vacuum technology and control software. Nobody replicates something like this from scratch.

What is particularly interesting is the speed. As recently as April, ASML CEO Christophe Fouquet stated that China would need many years to develop a comparable technology. The fact that a working prototype now exists at least shifts this timeline forward. Not because China has caught up technologically, but because the gap is apparently shrinking faster than expected. Technically, much remains to be seen. There is no reliable data on the stability of the EUV light source, the radiation output or the precision of the optics. It is also crucial that no chip has yet been manufactured with the system. Without a tape-out, the prototype remains a demonstrator. Nevertheless, sources assume that China could be using EUV technology productively by around 2030, significantly earlier than previous estimates.

The pressure behind this is obvious. The AI boom is driving demand for computing power, computing power needs chips and chips need production capacity. Companies like Huawei are forced to develop their own solutions under sanctioned conditions. The Kirin 9000S from SMIC’s 7 nm production was already a signal that technological detours are expensive but feasible. SMIC’s N3 process, functionally close to western 5 nm nodes, is also less an innovation than the result of political isolation. The EUV prototype fits seamlessly into this picture. It is not a breakthrough in the sense of a technological parity, but a strategic step towards reducing external dependencies. Whether this will result in an industrially viable system depends on scalability, reliability and reproducibility. EUV is not a prestige project, but hard industrial work.

The bottom line is this: China has not broken the EUV monopoly, but has demonstrated that it is not unassailable. For the West, this is not an immediate loss of control, but a clear warning signal. Export controls slow down development, they do not stop it.