ASML, the Dutch manufacturer of extreme ultraviolet lithography machines, controls the only equipment capable of printing the most advanced semiconductor patterns. Its $530 billion market value now faces a test: tech giants have committed over $600 billion to AI infrastructure spending this year alone, and the company cannot build machines fast enough to meet demand.
Each EUV machine costs between $200 million and $400 million, weighs as much as a school bus, and takes months to assemble. The production process relies on a global supply chain of 2,000 companies providing over 100,000 components. ASML employs 44,000 people to manufacture perhaps a dozen of these machines per year.
Christophe Fouquet, who became ASML's CEO in 2024, laid out the math plainly: "The demand is such that the market overall will be supply-limited for quite a bit." He identified chip manufacturing as the current biggest bottleneck in the supply chain—a polite way of saying ASML itself is the constraint.
Fouquet acknowledged he did not foresee the scale of AI's impact when it accelerated after ChatGPT's release. "Sometimes we wake up in the morning and still check that what is happening is really happening," he said in an interview in Beverly Hills.
ASML's near-monopoly has drawn challengers. Substrate, a San Francisco startup founded by a former Peter Thiel protégé, has raised more than $100 million at a valuation exceeding $1 billion and claims it can develop a rival lithography machine. Separately, reports indicate that former ASML engineers in China have partly reverse-engineered the company's technology, presenting geopolitical risks to semiconductor supply chains.
ASML defends its technological lead by spending €4.5 billion annually on research and development—a figure that itself underscores the capital intensity and complexity of the business. This spending supports its position as the sole producer of EUV lithography systems.
The price of entry remains brutal. ASML's largest customers occasionally pause before committing to purchases, Fouquet noted, despite their massive capital budgets for AI infrastructure buildout. The company's assessment: the world will not have enough advanced chips for years.
