China is coaxing advanced-class chips out of lithography machines that were never designed to make them, while laying the first uneven flagstones of a domestic alternative to ASML. Nothing here knocks the Dutch company off its perch: it still holds the only commercial monopoly on extreme ultraviolet lithography, the tools behind the world's most advanced processors and memory. But the route around that monopoly is no longer imaginary, and the arithmetic behind it is worth reading. The mechanism is patient rather than glamorous. Standard deep ultraviolet machines expose silicon at 193 nanometres, against 13.5 for ASML's EUV kit, and Chinese chipmakers stretch the older machines through multi-patterning, splitting intricate circuit designs across several exposure passes. That produces advanced-class chips at the cost of slower production, alignment risk and weaker yields. Chinese firms have also kept older ASML systems such as the Twinscan NXT:1980i alive by sourcing secondary-market components, working around international service restrictions.
ASML Watch analysis
What happened
The full report, published on Friday by ColombiaOne, pulls the strands together. A July 2026 report identified Shanghai Aishengna Electronic Technology Group, a little-known state-owned firm, as leading production of domestically developed immersion DUV machines, with initial output of about five systems this year and roughly twenty in 2027, destined for SMIC, Hua Hong and ChangXin Memory. The caveat matters: the gear needs more testing and remains far from matching ASML.
The bottlenecks stretch beyond lithography. Analysis from the Center for Strategic and International Studies, cited in the report, notes that strict controls on etching, deposition and metrology equipment constrain China's advanced-node ambitions even more than the exposure machines do.
ASML's exposure is real but bounded. Per the company's 2025 Form 20-F, Chinese customers accounted for 29.1% of net sales in 2025, down from 36.1% in 2024, and the filings warn that export controls can affect operations and demand. The near-term share story is sentiment rather than structure: ASML shares fell about 6% on 14 September in a broader tech selloff, with financial coverage tying the drop to worries over the pace of AI infrastructure spending, including Dario Amodei's call to slow frontier AI.
Key findings
- Multi-patterning works, at a price
Splitting a design across several exposure passes yields advanced-class chips on DUV kit, with slower output, alignment risk and weaker yields. - Aishengna is the domestic bet
The state-owned firm targets about five immersion DUV systems in 2026 and roughly twenty in 2027 for SMIC, Hua Hong and ChangXin Memory. - The bigger chokepoint is not the scanner
CSIS analysis notes that controls on etching, deposition and metrology gear constrain China more than lithography alone. - ASML's China cushion is thinning
Chinese customers took 29.1% of 2025 net sales per the Form 20-F, down from 36.1% the year before. - The 2030 remark
Jensen Huang expects China to develop advanced domestic lithography by 2030, a directional forecast that stops short of claiming an EUV equivalent.
Why it matters
Export controls aim to slow China's access to advanced computing, and lithography sits at the head of that pipe. If multi-patterning DUV plus a trickle of domestic machines can hold an advanced-class line open, the controls buy time rather than deliver denial, and every year of practice compounds.
For ASML the structural risk is quieter than any single headline: Chinese buyers drifting towards local systems, and service revenue on installed machines constrained by the very rules meant to protect the technology. The commercial lead in EUV is untouched, and Reuters-reported plans from TSMC, Samsung, SK Hynix and Intel to move High-NA into production use from around 2028 keep the frontier comfortably Dutch for now.
Our read
The measured take is the correct one: China is narrowing selected gaps, not matching the state of the art. The number to watch is not a benchmark but a production schedule. Five Aishengna systems in a year is a pilot; the test is whether domestic machines run consistently in high-volume production and integrate with local etch, deposition and inspection kit. Huang's 2030 expectation is a directional remark from an interested party, and reporting on it has varied, so treat it as a weather forecast rather than a tide table.
Our verdict: genuine progress, no breach. ASML's monopoly stands, but its China revenue cushion is thinning partly by design, and the equipment race just became less predictable.
What to watch
- Aishengna's first deliveries
Whether roughly five systems reach SMIC, Hua Hong or ChangXin Memory and hold up in production. - DUV line economics
Whether multi-patterned advanced chips become economical at scale or remain a costly workaround. - High-NA dates
Whether the production-use plans from around 2028 hold for TSMC, Samsung, SK Hynix and Intel. - ASML's China share
Whether the slide from 36.1% to 29.1% continues as domestic tools mature.
Discussion spark: Are export controls actually slowing China's chip ambitions, or handing its domestic toolmakers a protected market in which to mature until they no longer need ASML?
Sources and evidence
- How China’s Older Chip Machines Challenge ASML – ColombiaOne.com (18 September 2026, 17:10 UTC)
Independent WittyWires tracker for public updates about ASML. Not affiliated with or endorsed by ASML; this is not an official account.