ASML and Aixtron both raised their 2026 revenue outlooks in April as AI-driven capital spending reshaped fab equipment order books. ASML reported increased orders for extreme ultraviolet lithography systems -- the machines you need to print sub-5nm logic -- and stated that 'AI is the primary growth driver in advanced logic.' Aixtron, which makes the epitaxial deposition equipment used for silicon carbide and gallium nitride wafers, upgraded its outlook on demand for SiC in electric vehicle powertrains and GaN in power conversion.
The lead time signal is the one that matters operationally. EUV systems are already running more than 12 months from order to delivery, with customers booking slots into 2027. That is not a supply shortage in the traditional sense -- ASML is capacity-constrained, not will-constrained. They cannot build systems faster than they can source the subsystems that go into them. For any fab planning a capacity expansion tied to AI inference workloads, the equipment procurement window is now longer than most program planning cycles assume.
The Aixtron SiC demand is a different story with a different driver: EV powertrain efficiency, not AI directly. But the indirect AI connection is real -- the power systems feeding AI training clusters are increasingly SiC-based because the efficiency gains at 1200V operation are substantial. Both markets are pulling on the same Aixtron capacity.
What this means for chip buyers: advanced logic capacity coming online in 2027 and beyond is being booked now. Fabless companies and chiplet integrators that have not secured process technology agreements for N2 and below are behind. Equipment lead times are a leading indicator for wafer capacity tightness -- and the indicator is pointing at an extended squeeze.