▶ASML is a critical 'choke point' in the global semiconductor and AI supply chain, with the US AI industry being particularly dependent on its technology [28, 30, 25, 18].Apr 2026
▶The company's EUV technology involves an extremely complex process of using a CO2 laser to strike tin droplets 50,000 times per second to generate 13.5nm wavelength light [2, 23, 24].Apr 2026
▶ASML's production capacity for EUV tools is a significant bottleneck for the industry, with current output around 60-70 units per year, planned to increase to 80 units in the coming years [19, 31].Apr 2026
▶Demand for ASML's technology is outpacing supply, leading customers to accelerate their capacity expansion plans for 2026 and beyond [33].Apr 2026
▶There is a debate over whether ASML's advanced machines are experiencing diminishing returns, costing significantly more for only incremental gains [26, 27], versus being essential for enabling next-generation sub-2 nanometer processes [12, 32].Apr 2026
▶Experts disagree on the severity of the production bottleneck. Dylan Patel views it as a primary constraint on AI compute scaling through 2028-2029 [18], while Jensen Huang considers it a solvable two-to-three-year problem once demand is clear [29].Apr 2026
▶The potential for a competitive alternative is debated. Ben Thompson speculates that China could develop a less reliable but more scalable and cost-effective process [1], which contrasts with the implicit view that ASML's technological moat, built with partners like Zeiss, is nearly insurmountable [14, 28].Apr 2026
▶While CEO Christophe Fouquet points to demand outpacing supply [33], the claims of diminishing returns [26, 27] suggest a potential future friction point where customers may balk at the high cost of next-generation machines like the €350 million High-NA tool [32].Apr 2026
Not enough data for timeline
Sign up free to see the full intelligence report
Get started free