ASML Orders Keep AI Boom Alive

ASML’s backlog is one of the clearest real-world checks on AI spending, and it still points upward. The constraint is shifting from demand to how quickly the semiconductor supply chain can build more advanced manufacturing equipment.

• –AI demand is driving spending across advanced logic and memory, not just Nvidia-style accelerator production.
• –ASML’s 30% planned capacity increases for 2027 and 2028 may still fall short, signaling unusually strong multi-year visibility.
• –More lithography capacity should eventually expand GPU, accelerator, and high-bandwidth-memory supply.
• –Developers should expect AI infrastructure investment to remain aggressive, though chip availability—not model demand—may determine deployment timelines.
• –The main risk is that customer forecasts reflect defensive capacity reservations and could still be revised if AI revenue fails to justify hyperscaler capex.