ASML raises sales forecast as AI chip demand surges

ASML, the Dutch technology giant, raised its sales forecast on expectations that AI-driven demand for computing power will persist. The company reported 21% year-on-year sales growth in the second quarter, driven by strong performance from major customers Intel and TSMC.

ASML manufactures extreme ultraviolet lithography machines—specialized equipment that produces the world's most advanced semiconductors, described as perhaps the most complicated objects made by man. No other company produces these eUV machines, giving ASML a singular position in the global semiconductor supply chain. However, the company currently faces a severe production constraint: it can only manufacture a few dozen machines per year. This limitation has made ASML's output a major bottleneck in the AI industry, as chipmakers racing to expand capacity to meet AI demand all depend on the same supplier.

To capitalize on sustained demand, ASML announced an aggressive expansion plan: it will boost production capacity by 30% in 2025 and a similar amount again in 2028. The company's willingness to commit substantial resources to this expansion reflects confidence that the AI-driven surge in semiconductor demand will continue. Success in executing these capacity increases will be critical not only to ASML's financial performance but to the broader semiconductor industry's ability to supply the chips that power AI systems globally.

ASML's raised forecast reflects a fundamental reality of the AI boom: the supply of advanced semiconductors has become the limiting factor. Because the company holds a global monopoly on eUV lithography equipment and operates under production constraints—manufacturing only a few dozen machines annually—it has become a critical chokepoint in the semiconductor supply chain. The 21% year-on-year sales growth in Q2 and the company's willingness to commit to consecutive years of 30% capacity expansion both signal confidence that demand from major chipmakers Intel and TSMC will remain strong enough to justify the investment. However, the capacity expansion itself—scaling from dozens to a higher production rate—remains an engineering challenge that will determine whether the global semiconductor industry can keep pace with AI infrastructure demands.