Chinese semiconductor equipment makers are no longer just trying to prove that their tools can work in controlled demos. The harder question is whether those tools can survive the slow, expensive qualification process inside active fabs, where yield, uptime, contamination control, and service support matter as much as headline capability.
That was the practical message behind a recent public appearance by SMIC founder Richard Chang and AMEC chairman and CEO Gerald Yin. They appeared together on CCTV’s Dialogue program on May 17, where they reportedly urged Chinese chipmakers to give domestic equipment more time on production lines. The appearance has not been independently verified in every detail, but the direction of the argument fits the pressure now building across China’s semiconductor supply chain: local tools need real fab exposure if they are going to improve quickly enough.
The push comes as China’s equipment vendors are reporting much larger sales, even while margins appear to be tightening. Export controls have made domestic alternatives more important for Chinese fabs, but they have not removed the basic engineering problem. A deposition, etch, cleaning, or lithography system is not useful at scale simply because it can process wafers. It has to do so repeatedly, predictably, and with enough tool availability to protect production economics.
Domestic tool makers are growing, but pricing pressure is rising
China’s semiconductor equipment sector has been expanding quickly, helped by government policy, constrained access to some foreign tools, and demand from local fabs that need less vulnerable supply chains. AMEC, one of China’s most important etch-tool suppliers, reported 2025 revenue of about 12.38 billion RMB, or roughly $1.74 billion, with net profit of about 2.11 billion RMB. Those figures were presented as a strong year-over-year gain, though readers should treat the exact comparisons as company-reported figures rather than independently audited industry-wide proof.
Other domestic suppliers have also been described as growing fast. Naura Technology, which sells across a broader equipment portfolio, was reported to have posted 27.14 billion RMB in revenue across the first three quarters of 2025. Piotech, a thin-film deposition specialist, was reported to have roughly doubled nine-month revenue to 4.22 billion RMB. ACM Research, a U.S.-listed cleaning-equipment company with major operations in Shanghai, reported full-year revenue of $901.3 million, up 15.2%.
Those numbers point to a larger local market for Chinese-built tools, but they do not automatically mean the industry is becoming more profitable. Reports cited in the source material said gross margins declined at AMEC and ACM Research during 2025, with AMEC’s full-year gross margin described as falling to 39.2% and ACM Research’s gross margin described as declining to 44.4%. Those margin figures should be read as reported indicators rather than independently confirmed sector averages.
The commercial issue is straightforward. Chinese fabs need domestic suppliers more than they did a few years ago, but domestic suppliers are also competing against one another for the same localization budgets. Where Applied Materials, Lam Research, Tokyo Electron, ASML, and other overseas vendors once dominated many orders, Chinese toolmakers now have a larger opening. That opportunity can still turn into a price war if multiple local vendors are chasing the same tool categories.
Where localization is working, and where it is not
The strongest progress appears to be in mature-node equipment categories where process windows are better understood and the technical burden is lower than in leading-edge lithography. Chinese fabs are thought to be buying a higher share of domestic equipment than they were a year earlier, with some estimates putting local sourcing around one-third of equipment purchases. Beijing’s informal target for new fab projects is often described as 50% domestic content, though that target and individual fab compliance reports should be treated as policy signals and industry reporting, not as fully transparent procurement data.
The gains are not evenly spread across the fab. Etch and resist stripping are often described as China’s stronger localization categories at mature nodes. Thin-film deposition has made progress but remains more mixed. Lithography is still the most difficult gap.
| Tool category | Reported localization picture | Why it matters |
|---|---|---|
| Etch | Reportedly stronger at mature nodes, with some estimates around 50% to 60% | Etch is critical across many process steps and has become one of the more credible domestic equipment segments |
| Resist stripping | Reportedly above 80% in some mature-node contexts | This is one of the less constrained areas compared with lithography and advanced deposition |
| Thin-film deposition | Often estimated around 20% to 30%, though figures vary by process and fab | Deposition quality affects film uniformity, reliability, and device performance |
| Lithography | Commonly described as below 5% domestic share | This remains the central bottleneck for advanced manufacturing independence |
These categories are not interchangeable. A fab can localize cleaning, etch, or stripping tools and still remain dependent on imported lithography scanners. It can also qualify a local tool for one mature process and still be years away from using it in a more demanding flow. That distinction matters for buyers, investors, and supply-chain planners watching China’s semiconductor buildout.
The real test is qualification, not prototypes
The public argument attributed to Chang and Yin centered on a practical bottleneck: domestic equipment cannot improve if fabs do not give it controlled but meaningful exposure to production conditions. Chang reportedly suggested that fabs begin with small wafer batches of up to 100 wafers before scaling deployment. That specific recommendation has not been independently verified, but the broader logic is familiar across semiconductor manufacturing.
A working prototype is only an early milestone. To become a production tool, equipment must prove that it can operate over long periods without creating unacceptable defects, drifting out of spec, slowing the line, or requiring service patterns that disrupt output. Qualification also tests whether the vendor can support the tool after installation, supply parts, troubleshoot failures, and tune recipes with the fab’s process engineers.
Industry qualification timelines for new etch or deposition systems are often discussed in the range of 18 to 24 months, though the exact schedule depends on the process, node, customer risk tolerance, and whether the tool is replacing an established platform or entering a new process step. The reason is not bureaucracy. The reason is that semiconductor fabs are built around yield learning and repeatability, and a tool that looks promising in a narrow test can still damage output when inserted into a live flow.
Yin reportedly argued that Chinese fabs still tend to default to foreign tools when they can, partly because established global suppliers have longer track records. He also reportedly noted that even tools from the largest overseas vendors can require years of tuning when introduced at advanced fabs. That claim is plausible as a general description of semiconductor manufacturing, but the exact remarks and comparison should be treated as reported commentary rather than a verified transcript.
AMEC’s display equipment story is an ambitious example
One of the examples attributed to Yin involved AMEC’s move into large flat-panel display equipment. The company reportedly decided in December 2023 to enter a segment that Yin described as previously fully dependent on imports. The tool described in that account is unusually large, weighing around 150 tonnes and measuring about 15 meters by 15 meters.
AMEC reportedly built a working prototype in 12 months, met a customer’s next-generation specifications four months later, and shipped the tool to a production line within 18 months of starting the effort. Those claims are impressive, but they have not been independently verified and should not be treated as proof that similar timelines are available for advanced semiconductor tools.
The display-equipment example is still useful because it shows how Chinese suppliers want fabs and panel makers to think about domestic equipment. The pitch is not that every tool is already equal to foreign alternatives. The pitch is that customers should provide enough production-line access for local vendors to close gaps faster.
That is a different kind of buyer decision. A fab choosing a domestic tool is not only comparing today’s performance. It is also deciding whether to accept qualification risk in exchange for supply-chain control, lower exposure to sanctions, and a vendor relationship that may improve over time.
SMIC’s position shows why the decision is difficult
SMIC sits at the center of this problem. It needs more domestic equipment because export controls have made foreign tool access and long-term service support less predictable. At the same time, SMIC cannot afford to damage yield or throughput by inserting immature systems too aggressively.
AMEC has claimed that SMIC has purchased at least 800 of its tools, a figure Chang reportedly cited during the broadcast. AMEC has also claimed that its etch technology is used in TSMC’s supply chain at nodes ranging from 65nm down to 5nm and 3nm. TSMC has not publicly confirmed the scope of AMEC’s role in its production lines, so those claims should be read as supplier statements rather than confirmed customer disclosures.
Recent reports have also described SMIC as facing yield and production issues tied to equipment maintenance and validation delays in 2025. Those reports have not been fully verified in public detail, but they highlight the same operational risk Chang was reportedly discussing. A fab may acquire a valuable foreign tool and still struggle if spare parts, field service, software support, or supplier engineering access become constrained.
That creates a difficult transition period. Domestic tools may not yet match the most mature imported platforms in every category, but imported platforms may carry service and policy risk. For fab operators, the decision is not simply local versus foreign. It is a tradeoff among process performance, supportability, qualification time, supply security, and cost.
Lithography remains the hardest gap
None of the progress in etch, cleaning, stripping, or deposition solves China’s biggest equipment chokepoint: lithography. Shanghai Micro Electronics Equipment, commonly known as SMEE, is China’s main domestic lithography scanner supplier in volume, but its publicly known systems remain far behind ASML’s most advanced DUV and EUV platforms. SMEE is associated with a 90nm-class ArF system, while reports of a 28nm-class tool have not translated into clearly confirmed high-volume production.
A more closely watched project is the Shanghai Yuliangsheng immersion DUV scanner reportedly being tested at SMIC. The tool has been linked in reports to Huawei-backed SiCarrier and the codename “Mount Everest.” It is often compared with ASML’s older Twinscan NXT:1950i generation rather than the newer DUV immersion systems used in more advanced production flows.
SMIC is believed by some industry watchers to be aiming for use of the Yuliangsheng tool in a 28nm production flow around 2027. That timeline is not publicly confirmed. Even if the tool reaches that target, it would not mean China has solved sub-10nm lithography with domestic equipment. Advanced-node production can use multiple patterning and other workarounds, but the economic and technical burden rises sharply without access to the most capable scanners.
The implication is blunt: China can localize many parts of the fab before it can localize the lithography stack that matters most for advanced logic. That includes not just scanners, but also photoresists, masks, pellicles, metrology, computational lithography, precision stages, light sources, and the supplier know-how needed to keep the system stable in production.
Export controls raise the stakes for fab buyers
Chinese fabs have continued buying DUV tools while rules allow it. ASML’s sales mix has repeatedly shown China as a major market for DUV systems, including a reported 42% of system sales revenue in the third quarter of 2025. That demand reflects both fab expansion and the desire to secure tools before export windows narrow further.
The proposed MATCH Act in the United States would raise the pressure. The bill has named several Chinese semiconductor companies and equipment makers, including AMEC, Naura, Piotech, ACM Research, SiCarrier, SMEE, SMIC, YMTC, Hua Hong, CXMT, and Huawei, as covered entities or facilities. As described in reports, the legislation would tighten restrictions around exports of DUV immersion lithography tools to China.
The House Foreign Affairs Committee advanced the bill in late April, while removing a proposed ban on cryogenic etch tools that would have affected some non-Chinese equipment suppliers. The DUV immersion restriction remained in the bill as it moved forward. Its final outcome depends on the U.S. legislative process, so it should be treated as a live policy risk rather than settled law.
For Chinese fabs, that risk changes the purchase calculus. If future access to DUV immersion tools narrows, qualifying domestic alternatives becomes more urgent even when the tools are not yet drop-in replacements. If restrictions do not tighten as much as expected, fabs may still prefer to keep buying proven foreign systems where allowed. Either way, the pressure to develop a credible local equipment base will not go away.
What this means for the semiconductor supply chain
The Chang-Yin message, if accurately reported, is less a victory lap than a request for patience and risk-sharing. Domestic equipment suppliers want fabs to treat qualification as a national supply-chain investment, not just a narrow procurement decision. Fabs, meanwhile, have to protect output, customer commitments, and yield targets.
That tension explains why the next stage of China’s equipment buildout may be harder than the first. Revenue growth can be driven by policy, capacity expansion, and urgent substitution. Production credibility requires something slower: installed bases, field data, failure analysis, service networks, process recipes, and repeated customer trust.
For buyers and industry watchers, the most important signal is not whether a Chinese supplier announces a prototype or a shipment. The better signal is whether a fab keeps the tool in production, expands it across more process steps, and orders follow-on systems after the initial qualification cycle.
China’s semiconductor equipment makers have momentum, but the market is now moving from substitution to proof. Etch, cleaning, and some mature-node categories show real progress. Lithography remains the unresolved constraint. The fabs that decide how much production-line time to give domestic tools will determine whether local vendors become durable manufacturing partners or remain partial substitutes in a supply chain still shaped by foreign equipment and export policy.