China has started making its own immersion DUV lithography machines, and ASML’s market quickly reacted. On Monday, after The Information reported the move, ASML shares fell nearly 6%. By July’s close they were already down almost 17% for the month, a nasty run that put the Dutch company on course for its worst month since October 2024.
I read that sell-off as a market reacting to more than a product update. Lithography is where chipmaking turns from chemistry and engineering into leverage. If China can build even a limited domestic version of the tool that has long sat near the center of ASML’s power, it is trying to rewrite the price of dependence.
What happened
The first machines are expected to go to Semiconductor Manufacturing International Corp, Hua Hong Semiconductor, and ChangXin Memory Technologies this year, according to The Information. That list is revealing. SMIC, Hua Hong and ChangXin are not boutique labs playing with theory. They are the factories that turn advanced process capability into actual output, and they need tools that can survive the violence of production, not just look good in a presentation.
These are immersion DUV systems, meaning the wafer sits under a thin layer of ultra-pure water during exposure. That little trick lets the optics draw finer lines than older dry DUV tools. It is an industrial detail that sounds dull until you realize entire parts of the world economy depend on it. DUV is not the glamour chip technology; EUV gets the headlines, but DUV gets the volume.
China’s version is still young. The report says the machines lag on performance and reliability and still need further testing before mass production. A lithography tool that falters under load is not an industrial asset; it is an expensive paperweight with a power cord.
Why Beijing wants it
The point is not to beat ASML on day one. The point is to stop being trapped inside someone else’s supply chain.
For years, Western export controls have made the Chinese semiconductor industry live with a simple fact: the most advanced lithography gear is not available on demand. ASML has been the choke point, especially because its EUV machines are barred from China and even its advanced DUV sales have faced tighter scrutiny. If you cannot buy the tool freely, building a local substitute becomes a strategic project, not an engineering vanity.
This report matters even though the volumes are tiny. China is said to be aiming for about five DUV machines this year and roughly 20 in 2027. Those numbers do not move the global equipment market on their own. They are, however, enough to show intent. This is how industrial self-reliance begins: one awkward pilot line at a time, then a second, then a production queue that nobody outside the system can easily interrupt.
China is also pushing on EUV, the more advanced form of lithography used for the smallest chip features. That effort remains at the prototype stage, but it is being fast-tracked. I would not confuse “prototype” with “close to commercial.” China has a long way to go before it has a domestic EUV tool that can compete with the Dutch machines. Still, the fact that the work is being accelerated tells you where the state wants the technology stack to end up.
What still gets in the way
China can copy the architecture of a machine faster than it can copy the decade-spanning habits that make one work properly. Precision optics, vibration control, software, light source stability, thermal management, contamination control—all of it has to line up. ASML’s edge is not a single patent; it is the accumulated discipline of making systems that run in factories, not just in a lab.
The gap between “domestic” and “usable” is larger than Beijing’s slogans suggest. If the new tools are slower, less stable, or harder to calibrate, Chinese chipmakers will use them where they can and keep buying foreign kit where they must. The likely first result is not a clean break with ASML. It is a mixed shop floor, part local, part imported, with the local gear improving under the pressure of constant use.
ASML’s near-term advantage still looks intact. The company has scale, a global supplier base, and a track record China does not yet have. But the direction of travel is obvious. A country that can build its own DUV machine, even in limited numbers, is no longer just asking whether it can buy access. It is asking how much of the dependency graph it can tear up before the next round of export controls arrives.
What comes next
The sharp question is not whether China can make a few machines. It already appears to be doing that. The real question is how quickly those machines improve once they are inside actual chip fabs, under actual production stress, making actual money or causing actual headaches.
If the answer is fast enough, ASML’s China business gets squeezed from the inside. If it is slow, the symbolism still bites. Either way, the game has changed. The board is now set for a contest over who controls the tools that make the tools.
