Take a calm look at domestically made lithography machines: 5 units delivered—how far is it from challenging ASML?

Author: Hanya Hu, a Silicon Valley investor + online fiction writer

Today, a piece of news about domestic DUV lithography machines entering mass production has made a lot of people instantly hype up.

Actually, this kind of situation isn’t the first time it has happened. Every time it comes time to hype certain stocks, someone will always bring up “domestic substitution” as part of the story.

No hype, no blackening today—let’s look at what’s really going on.

First, the conclusion. A “DeepSeek moment” for domestic lithography machines is not true at this time.

This article doesn’t hype or smear. Its purpose is to help investors form an understanding grounded in facts.

What exactly is the matter: A state-owned enterprise in Shanghai has begun small-batch production of domestic immersion DUV lithography machines. This year, it plans to ship about 5 units. The delivery targets are three companies: Semiconductor Manufacturing International Corporation (SMIC), Hua Hong Semiconductor, and ChangXin Memory Technologies (CXMT). The 2027 production target has been mentioned as around 20 units. The equipment model corresponds to the 28-nanometer process node. Official data states that the overall domestication rate is above 85%, and mass-production yields are stable at over 90%. Almost at the same time, Shanghai Micro Electronics’ SSA800 series also announced full-scale mass production, with an annual capacity target of 50 units, doubling year-over-year.

These two pieces of news are often mixed together in discussion. Paired with headlines like “China’s lithography-machine DeepSeek moment,” it sounds like China has already conquered the lithography machine challenge. But if you take it apart carefully, things aren’t as optimistic as people imagine.

For investors, this news is worth tracking continuously for now. It’s not yet enough to directly infer that ASML’s moat has already started to loosen, nor enough to prove that the relevant domestic equipment companies are about to realize large-scale profit.

What can be confirmed for now

On July 27, Reuters, citing a report from The Information, said that an unnamed China state-owned enterprise-backed company has begun manufacturing domestic immersion deep-ultraviolet (DUV) lithography machines.

The report states that the first batch of equipment is expected to be delivered in 2026 to SMIC, Hua Hong Semiconductor, and CXMT. This year, it plans to produce about 5 units, and in 2027 it plans to increase to about 20 units.

Meanwhile, the report also clearly points out that this set of equipment still lags behind ASML in performance and reliability—it needs further testing, and it is still some distance away from truly large-scale commercial production. As for the EUV equipment that China is developing, it is still at the prototype stage.

These limiting conditions are very important.

1. A more accurate description right now should be “began manufacturing and entered the customer introduction phase.” Phrases like “already in full-scale mass production” can easily make people think the equipment has already passed customer acceptance and can run long-term, stably, at full load in a wafer fab, while public information has not proven this.

2. On the internet, numbers such as “supports 28-nanometer,” “domestication rate over 85%,” and “mass-production yield over 90%” are also circulating. Unfortunately, at present there is no complete test report published by the equipment manufacturer, the customer wafer fabs, or authoritative institutions, and there’s also no clear statistical basis for these figures.

3. Especially “yield of 90%”: does it refer to equipment factory acceptance rate, single-layer exposure yield, wafer yield, or full chip yield? The differences are huge. Without sufficient analysis inputs—test conditions, product types, number of wafers, and customer acceptance standards—this number currently has not enough analytical value.

Therefore, this article does not treat these unverified data as established facts.

What was actually made this time

Lithography machines mainly fall into two categories: DUV and EUV. This time, what was produced is DUV.

What’s the difference between DUV and EUV? A layperson-friendly analogy is the difference between ordinary light and X-rays.

Of course, the DUV system also includes KrF equipment using a 248-nanometer light source, ArF dry equipment using a 193-nanometer light source, and more advanced 193-nanometer ArF immersion equipment.

“Immersion lithography” means adding a layer of high-purity water between the projection lens and the wafer to improve resolution by increasing the numerical aperture. ASML’s current immersion DUV equipment can be used for mature process nodes, and it can also participate in advanced process production through multi-patterning processes.

EUV uses extreme ultraviolet light with a 13.5-nanometer wavelength. Because most materials absorb EUV light, EUV tools cannot reuse traditional lens systems. Instead, they use multi-layer reflective mirrors in a vacuum environment to transmit the optical path.

This means that moving from DUV to EUV involves a comprehensive change across the light source, reflective optics, vacuum systems, masks, photoresists, contamination control, precision wafer stages, and the entire machine control system. The technical complexity sees a massive leap.

So, when domestic immersion DUV achieves progress, it mainly fills the gap in domestic advanced DUV equipment supply. It can improve supply-chain security for mature nodes and some advanced nodes, but it cannot directly mean that the EUV challenge has already been solved.

What tasks do DUV and EUV each handle

In modern wafer fabs, DUV and EUV will coexist for the long term.

A single chip typically contains dozens of layers, or even more. The line width and accuracy requirements differ across layers. The most critical and finest few layers use EUV, while many comparatively less demanding layers continue to use DUV.

ASML has also stated clearly that EUV mainly handles the most complex and critical patterning layers in advanced chips, while the large number of other patterning layers continue to use different types of DUV equipment.

DUV capability also cannot be simply understood as “only able to make mature chips.”

Through double, quadruple, and even more complex multi-patterning, DUV can extend the process capability to around the 7-nanometer level. In the past, SMIC was able to manufacture 7-nanometer-class chips without EUV equipment. The key path was relying on imported immersion DUV equipment combined with complex multi-patterning processes.

Taiwan Semiconductor Manufacturing Company (TSMC)’s initial N7 process mainly used DUV. It wasn’t until N7+ entered mass production in 2019 that it became TSMC’s first process with large-scale EUV use.

The problem is: the more advanced the node, the more DUV requires in terms of exposure counts, number of masks, overlay precision, and process steps. That lengthens production cycles, and increases defect probability and manufacturing costs.

The main value of EUV is to reduce complex multi-patterning steps of DUV in some key layers, thereby reducing the number of masks, process complexity, and defect risks. ASML’s annual report shows that EUV can help customers transform some complex multi-patterning into relatively simpler single-patterning.

So, even if domestic immersion DUV passes customer verification, it mainly addresses whether China can independently obtain high-performance DUV equipment. It may have strategic significance for 7-nanometer-class processes, but for the economic issues of mainstream advanced processes such as 5 nanometers, 3 nanometers, and 2 nanometers, it is still completely unable to solve them.

What exactly is the scale of “5 units”

Looking only at the number “5,” it’s easy to fall into two extremes.

Some people think the quantity is too small to matter. Others think that if it can be made, it means ASML’s monopoly is already over.

Both judgments are too simplistic.

According to ASML’s annual report, in 2025 ASML confirmed sales of 131 ArF immersion lithography machines. In 2026, ASML announced in the second quarter that its company’s 2026 immersion DUV annual production capacity is about 130 units, and it plans to increase by about 30% in 2027.

If you compare roughly by unit count, the 5 domestic units planned for this year are about 4% of ASML’s immersion DUV sales in a single year.

Moreover, lithography machines cannot be compared one-to-one purely by unit counts.

The real indicators that determine wafer capacity include wafers-per-hour throughput, overlay precision, equipment availability, continuous run time, defect rate, maintenance cycle, and how compatible the equipment is with existing production lines.

Public parameters of ASML’s current immersion DUV equipment show throughput up to 330 wafers per hour, with overlay precision matching around 2.5 nanometers. Domestic equipment has not yet publicly released complete comparable parameters.

Therefore, the biggest short-term significance of these 5 machines is to enter real production lines at SMIC, Hua Hong, and CXMT to verify whether the equipment can run continuously and stably, whether it can smoothly coordinate with etch, thin-film deposition, cleaning, inspection and metrology equipment, and whether it can secure customers’ second batch of orders.

As long as any key metric cannot meet mass-production standards, the equipment may remain in a trial and validation stage for a long time.

From this perspective, 5 units is the starting point of industrialization, but it has not yet formed the capacity to replace imported equipment at large scale.

Don’t combine the rumored SSA800 situation from Shanghai Micro Electronics

There is another claim in the market that Shanghai Micro Electronics’ SSA800 series has already entered full-scale mass production, with annual capacity reaching 50 units and even 120 units.

These figures exist in multiple versions, and there is still a lack of clear company announcements, customer acceptance reports, and procurement disclosures from wafer fabs.

As of May 2025, when Reuters investigated China’s semiconductor equipment industry, it still described Shanghai Micro Electronics’ front-end lithography equipment capability that has been commercially supplied as at the 90-nanometer level. At that time, although SiCarrier had already applied for multiple DUV-related patents, it had not yet publicly demonstrated a complete product.

It’s entirely possible for technical R&D to make progress within one year, but before new formal evidence is obtained, it’s not appropriate to directly add the rumored SSA800 output to the 5 units in this report. And it’s even more inappropriate to draw the conclusion that China can already produce hundreds of immersion DUV units per year.

For lithography machines—highly complex equipment—formal release, manufacturing completion, entry into customer factories, passing acceptance, and stable mass production are several completely different stages.

EUV still has a larger gap

China’s EUV R&D has made some engineering progress.

Reuters’ investigation in December 2025 showed that China built a domestic EUV prototype machine in Shenzhen. The equipment can already generate EUV light and entered the testing stage, but at that time it had not yet manufactured usable chips.

The related project proposes using a domestic prototype machine to manufacture working chips in 2028. People close to the project believe that 2030 may be more realistic. The report also noted that China still faces significant difficulties in core areas such as high-precision reflective optics.

This indicates that domestic EUV has already taken an important step, and engineering feasibility is being verified gradually.

But moving from “able to generate EUV light” to “able to stably manufacture advanced chips” requires solving a series of issues, including light source power and stability, mirror precision, masks, photoresists, contamination control, overlay precision, throughput, and long-term reliability.

ASML took nearly twenty years from the first working EUV prototype to truly entering commercial mass production, and it invested tens of billions of euros in R&D funding.

Domestic EUV may accelerate the catch-up in the future, but currently it is still at the prototype testing stage. It cannot be discussed with the same industry maturity level as immersion DUV that has already begun initial production.

Why the analogy of a “DeepSeek moment” doesn’t hold

The phrase “DeepSeek moment for lithography machines” is highly shareable, but it easily misleads investors.

The disruption caused by DeepSeek mainly comes from algorithm architectures, training methods, and engineering efficiency. In the software field, once a solution proves effective, other teams can quickly read the paper, run the code, adjust models, and validate the results. The iteration cycle is usually measured in days or weeks.

Lithography machines face precision manufacturing and complex physical systems.

Whether a machine can maintain overlay precision over the long term requires validation through large numbers of wafers and long-time running tests. Light source stability, lens lifetime, vibration control, contamination control, stage fatigue, part consistency, and maintenance systems all need data accumulated from real production.

Occasional success in exposure from a prototype is a completely different engineering requirement from running 24 hours a day with stable operation.

Therefore, this domestic DUV progress is closer to an extended engineering effort that is starting to enter the stage of industrial verification. It relies on sustained investment, talent accumulation, supply-chain harmonization, and joint debugging with customers. It’s difficult to achieve a full industrial turnaround through a single technological release moment.

China’s true turning point in lithography machines in the future may not appear as a sensational press conference. More likely, it will be reflected in continuously increasing customer acceptance counts, steadily improving equipment availability, year-by-year expansion of repeat orders, and a gradual decline in the import proportion of key components.

Where does domestic substitution stand right now?

China’s semiconductor equipment localization has progressed quickly in recent years, but there are big differences across equipment categories.

Reuters’ investigation in 2025 showed that in some equipment such as debonding and cleaning, the domestication rate has already reached about 50%. However, overall self-sufficiency for equipment that can support 7-nanometer and below process nodes remains below 10%. Lithography equipment is one of the most obvious weak spots.

On the other hand, Chinese wafer fabs have accumulated a large number of imported DUV machines.

A report by the U.S. Institute for Research in 2026 estimates that in 2024, Chinese customers purchased about 70% of ASML’s global immersion DUV equipment, possibly close to 90 units. These imported tools have become an important foundation for China to manufacture 7-nanometer-class chips.

It’s important to note that this report comes from a U.S. policy think tank, with a policy stance of pushing for expanded export restrictions. Its production capacity estimates cannot be directly treated as neutral industry statistics.

But it at least shows one point: China’s current mature process and some advanced process capabilities still heavily depend on ASML equipment imported in the past. With domestic equipment planned for 5 units this year, even if all pass verification, it will be difficult in the short term to replace the stock of imported equipment that has already formed at scale.

Its strategic value is mainly in building a second supply source for the future, reducing reliance on imports for incremental capacity, and accumulating engineering experience for more advanced domestic equipment.

What this means for investors

First, it should be clarified that this news is a long-term positive signal for China’s domestic semiconductor equipment industrial chain.

It shows that years of investment in areas such as light sources, projection optics (objective lenses), wafer stages, control systems, metrology systems, and overall machine integration are beginning to show signs of shifting from R&D to customer-site deployment.

But investors should not yet draw conclusions from a single news item that domestic equipment companies are about to scale up shipments massively, or that ASML’s global competitive advantage is about to disappear.

ASML confirmed sales in 2025 of 131 immersion DUV machines and 48 EUV machines. In 2026, the company still plans to expand DUV and EUV capacity, indicating that demand from global advanced logic and memory customers remains strong.

Domestic equipment can gain policy, capital, and customer coordination support in the China market. But once it enters wafer fabs, it still ultimately must be tested by production efficiency and manufacturing cost.

Over the next two to three years, the truly important indicators to track include whether customers complete formal acceptance, how many wafers per hour the equipment can process, whether overlay precision can remain stable long-term, whether equipment availability can meet commercial production requirements, whether defect rates are stable during continuous operation, and whether initial customers are willing to place second and third batch orders.

Also watch whether key optical components, light sources, precision motion systems, and control parts can gradually achieve domestic substitution, and whether after-sales service teams can quickly complete repairs and replace parts.

For specific listed companies, it’s necessary to verify what components they actually supply, whether they have obtained formal orders, how much revenue and profit the related business can contribute, and whether order growth can cover R&D and capacity expansion investments.

Final judgment

Based on the available public information, this progress is better defined as domestic immersion DUV taking a step from R&D and prototype verification toward initial production and customer introduction.

With planned production of about 5 units, the quantity is limited, but it is enough to start real line validation. As long as the equipment can pass customer acceptance and receive repeat orders, its significance will far exceed the five units themselves.

At this stage, it is still not enough to change the global lithography competition landscape, nor enough to show that China has already solved the bottleneck in advanced process lithography equipment.

A real industrial turning point requires seeing several consecutive signals: the first batch of equipment running stably, customers placing follow-up orders, annual output rising from single digits to dozens or even over a hundred units, key parameters approaching international mainstream levels, and continued declines in dependency on imported core components.

Until these metrics are achieved, the most reasonable stance is to affirm progress, respect engineering laws, and avoid pricing a technical news release into “industrial victory” and investment returns too early.

As for the so-called “DeepSeek moment for domestic lithography machines,” it is still nowhere near that point.