Samsung, TSMC and Intel commit to ASML's High-NA EUV lithography and 12-inch photomasks as Huawei drives China's DUV alternative

Samsung and TSMC have reportedly committed to ASML's High-NA EUV lithography machines for high-volume manufacturing, with Samsung starting in 2028 and TSMC in 2030; Intel already uses the tools. The same three chipmakers plan to move photomasks from the current 6-inch standard to 12-inch, with a TSMC-ASML test line by 2031 and production use by 2033, claimed to lift throughput 40%. Separately, Huawei is coordinating a Chinese effort (via Yuliangsheng, spun out of SiCarrier, backed by its Habo venture arm) to build domestic DUV lithography tools, targeting twelve machines by year-end, currently trialled by SMIC and Huawei on simpler chips.

If accurate, this locks ASML's three biggest customers into its most advanced platform for the next decade, reinforcing a near-monopoly on the equipment that sets the ceiling for chip performance. That matters most for investors and supply-chain planners rather than anyone needing to act now, since the earliest commitment (Samsung, 2028) is still two years off and the photomask shift not until 2033. Huawei's DUV push is a genuine, concrete attempt to build an export-control workaround, but it targets older, less complex chips and remains far behind ASML's capability, so it does not change the near-term picture for advanced-node manufacturing.

This is a third-hand aggregation (The Decoder summarising Bloomberg and FT) with no primary sources or links, so the specific dates, the 40% throughput figure and the $400M-per-tool cost should be treated as reported claims, not verified facts. The article stitches together two loosely related stories, ASML's customer lock-in and Huawei's domestic alternative, into one narrative of "grip vs breaking the grip" that overstates how connected they actually are; Huawei's twelve DUV machines by year-end are unrelated in technology and timescale to High-NA EUV.

1. 01Official confirmation from ASML, Samsung or TSMC of High-NA EUV production timelines (2028/2030), ideally in earnings calls or investor filings rather than press summaries
2. 02Whether TSMC and ASML actually stand up the 12-inch photomask test line by 2031, and any independent verification of the claimed 40% throughput gain
3. 03Yuliangsheng's actual DUV machine output by year-end 2026 versus the stated target of twelve units, and any evidence SMIC or Huawei move beyond simple chips onto these tools