Ben Pouladian
Founder of BEP Research, AI Compute Economics
ASML's 6% drop on a report of Chinese immersion DUV mass production is no cause for panic. Per a Chinese semiconductor expert, the first-gen tool is five or six generations behind ASML's 1980-class machines, which Chinese fabs already stockpile, and SMEE's six dry DUV tools sold in 2023 never ran production. No Western fab is likely to qualify a Huawei-linked tool, the lost revenue was already exiting under export controls, and lithography isn't China's binding constraint. x.com x.com
The selloff in ASML and BESI on China's new immersion DUV tools is an overreaction with minimal real impact. China plans five units this year and 20 in 2027 for SMIC, Hua Hong and CXMT, under 4% of ASML's roughly 130 annual immersion DUV shipments. DUV can make 7nm chips via multi-patterning, but at higher cost, lower capacity and more alignment errors, and EUV is a fundamentally different technology. Nikon and Canon face the most pressure. x.com
ASML's selloff on China's domestic DUV news is an overreaction because the revenue hit is small and China is nowhere on EUV, ASML's real moat. China is making about five DUV machines this year versus the hundreds ASML delivers. BofA, which also calls the selloff an overreaction, estimates China going domestic cuts ASML's sales by about EUR 1.4B next year, or 2.4%, yet the stock fell about 6%. It helps China's chip independence far down the road but leaves ASML oversold today. x.com x.com
ASML faces moat erosion now that China has a homegrown alternative to its machines. The Information reports a state-backed Shanghai company has started mass-producing DUV lithography tools. Chinese fabs stockpiling older DUV tools had become a major revenue driver, with China at 36% of sales in 2024 and 29% last year. Output is tiny, about five machines this year and maybe twenty next. But once a second supplier exists, the question shifts from "no alternative" to "for how long." x.com
ASML's fears about Chinese competition are noise: China's claim to compete rests on 20-year-old DUV tools, while nobody else has ASML's EUV machines. ASML is a trillion-dollar monopoly with a 20-year head start, so there is no real competition. The broader tech selloff reflects worries such as Nvidia backstopping OpenAI's buildout and open-weight models like Kimi K3, which are more noise than signal; even good open models need massive infrastructure. x.com
ASML's selloff on China's mass production of DUV machines is a huge overreaction that double counts ASML's own guidance. Export controls made Chinese indigenization inevitable, with SMIC trialling domestic tools since Q3'25, and ASML's China share of sales was already falling from ~50% in 2024 to 33% in 2025 and ~20% guided for 2026. ASML is richly valued, but the stock held up after earnings that showed the same China decline. x.com x.com x.com
ASML could break down from here on the China news. x.com
ASML cannot make enough EUV machines to meet demand, which it says will outrun supply through 2028. Its 2026 sales guide rose from €34-39 billion to €44-45 billion, and gross margin hit 57.5%. The combination of an EUV monopoly, demand outrunning supply and growing pricing power is rare. x.com
China's homegrown DUV tool poses little threat to ASML because it remains almost 20 years behind. Per the Financial Times description, the immersion DUV system SMIC is testing resembles ASML's 2008 NXT:1950i for 32nm-class processes, and having a tool doesn't mean matching defect rates or throughput, as Canon and Nikon show. China has stayed roughly 20 years behind for a decade and is not meaningfully accelerating against ASML. x.com x.com x.com x.com x.com x.com
ASML may have topped, since semicap thrives on shortage and would be hit hardest in an overbuild. Then earnings acceleration from memory and fab projects stops or reverses; the labs' leading-edge chase may be at its apex, positioning is bullish and the moat is threatened. Bullishness on Korean, US and Taiwan fab projects makes sense for now. But if the cycle top is confirmed and labs change how they chase the frontier, or token deflation kills some labs, the shortage can end sooner. x.com x.com … Show more
ASML remains the undisputed lithography champion because DUV and EUV scarcity will ultimately be the real bottleneck to scaling AI compute. Wall Street sold semicap on a Shanghai firm building five 2008-vintage DUV tools, which threaten approximately none of ASML's moat this decade. Meanwhile, Situational Awareness put $500 million into Source Foundry, a startup aiming to break ASML's monopoly by redesigning chip manufacturing. Whoever breaks that monopoly wins, but it is a tough race. x.com
