5 domestically produced DUV lithography tools have triggered a broad sell-off in global semiconductor stocks.
Will this be the “DeepSeek moment” for the chip industry, or yet another case of the market overreacting?
According to a report by The Information, a company with state-owned capital backing from Shanghai has started mass production of domestic immersion DUV lithography tools. This year, it plans to deliver 5 units; next year, it plans to expand to 20 units. Target customers include SMIC, HUA HONG Group, and UNISOC Memory.
After the news broke, the global chip sector quickly came under pressure:
📉 ASML fell by more than 8% at one point during the trading session, triggering a volatility trading halt;
📉 Applied Materials dropped 7.7%;
📉 Lam Research fell 8.5%;
📉 SanDisk slid by nearly 13%, while SK hynix, Micron, NVIDIA, and others also pulled back in tandem.
What really makes the market tense isn’t the five devices themselves, but the underlying progress in the industry.
In 2023, Huawei used DUV multi-patterning to introduce the Kirin 9000S; in 2025, SMIC began testing domestic immersion DUV; and now there are reports that it has entered the mass production and delivery stage.
This pace is faster than what many institutions had previously predicted.
That said, it’s still important to stay rational. Five tools are still far from changing the global lithography landscape. Currently, the target process is still mainly 28nm. Multi-patterning theory could continue to advance, but what truly determines competitiveness is long-term stable operation, yield, precision, and reliability—all of which still need time for validation. ASML delivered 131 immersion DUV tools last year, with total system shipments exceeding 500 units. At this stage, the two sides are still not on the same scale.
So, this looks more like a shock to expectations rather than a product revolution that has already been fully validated.
What the market is worried about isn’t that 5 units were delivered today, but that China’s semiconductor industry is moving from “impossible” to “testing,” and then to “mass production,” with the timeline continuously shrinking.
In summary, this is an important milestone for China’s domestic immersion DUV moving from R&D into customer validation. However, replacing ASML across the board still requires a long process of engineering validation and industrialization. Personally, unless there are other negative surprises, I don’t see systemic risk in the near term.
#半导体 #芯片 #DUV #ASML