EUV power drives ASML vulnerabilities

ASML’s biggest weak point is that each extra wafer requires a huge amount of physics and hardware just to make a small amount of usable EUV light. That makes cost reduction harder than in most chip equipment categories, because fabs pay not just for the scanner, but for power, uptime, source maintenance, field upgrades, and specialized service. It also creates an opening for any alternative that can deliver similar patterning with a simpler source and fewer consumables.

• ASML has spent 25 years pushing EUV source power from 1 watt in its 2010 prototype to 250 watts for volume production in 2018, then to a 1,000 watt demo in April 2025. That history shows the bottleneck is not optics alone, but the brute force needed to make enough light for factory throughput.
  1 asml
• The service tail is large because the installed base needs constant upgrades and maintenance. ASML reported net service and field option sales of €8.2B in 2025, up 26.2% year over year, with more NXE field upgrades shifting revenue from tool sales into installed base revenue.
  2 asml
• Nikon’s current lineup is still ArF immersion, ArF dry, KrF, and i line systems, with no commercial EUV offering on its product pages. In practice that leaves mature and non EUV layers as Nikon’s arena, while ASML owns the hardest leading edge layers but also carries the energy and service burden that comes with them.
  3 nikonprecision 4 nikonprecision

The next phase is a race to turn lithography performance into lower cost per wafer, not just smaller features. ASML is improving source power and process simplification, but the more its revenue mix tilts toward upgrades and service, the more attractive a cleaner architecture becomes for challengers trying to break into advanced patterning.