ASML CEO Christophe Fouquet talks about his company’s monopoly: no one comes to us

Every time you use AI, you rely in no small part on the 42-year-old, 44,000-employee Dutch company, which spends €4.5 billion every year on technological advancements.

ASML is headquartered in the Netherlands and manufactures the machines that make the chips that enable AI. More specifically, it will manufacture the only machine in the world capable of printing the microscopic patterns that define cutting-edge semiconductors onto silicon wafers, a process called extreme ultraviolet lithography (EUV). The machines are about the size of a school bus, take months to assemble, involve hundreds of suppliers, and cost between $200 million and more than $400 million each, depending on the generation (a price that sometimes deters even ASML’s biggest customers).

This monopoly made ASML the most valuable company in Europe, with a value of more than $530 billion. And with the big four US tech companies (Microsoft, Meta, Amazon, and Google) spending more than $600 billion on AI infrastructure this year alone, demand for ASML’s machines has skyrocketed to the point where the company has publicly stated that the world won’t have enough chips for years.

All of this demand has also made ASML a target. Substrate, a San Francisco startup founded by Peter Thiel’s disciples, has raised more than $100 million to develop a rival lithography machine, valuing it at more than $1 billion. Separately, there are reports that a former ASML engineer in China reverse-engineered some of this technology, which is expected to have a significant geopolitical impact.

Christophe Fouquet, who became ASML’s CEO in 2024 after more than a decade at the company, sat down with this editor on the rooftop deck of a Beverly Hills hotel on Tuesday morning before attending the Milken Institute Global Conference. Wearing a blue suit and white shirt, he looked relaxed, even when the conversation turned to his rivals.

This interview has been lightly edited for length and clarity.

Did you see the AI ​​explosion coming?

No, not at all. We worked hard, but we never expected something like this to happen. You went from a concept that people thought they would eventually arrive at to ChatGPT, which is the first great example of what AI can do. And I think now we see AI as the next revolution, not just in industry but in society. Did you see that coming? No, if you sit in the middle of it every day, you might wake up in the morning and see what’s going on.

The big question on everyone’s mind is whether the supply chain can keep up with demand. Can you do it?

Demand is so high that the overall market supply will be limited for a significant period of time. The biggest bottleneck at the moment appears to be chip manufacturing. We follow our customers as equipment suppliers, and we have followed them pretty well so far, but we know that we need to strengthen our entire supply chain and production capacity. I think if you talk to hyperscalers, they’ll say they won’t have enough chips for the next two, three, even five years.

TSMC recently made headlines for its newest machines being too expensive. How do you answer?

In terms of price, EUV systems are more expensive than low NA systems, but the cost of making wafers with this tool in some advanced layers is lower. Cost reductions of 20% and 30% are possible.

(Editor’s note: Both machines Fouquet is referring to here are EUV machines, the same basic technology. NA stands for numerical aperture, a measure of how finely a machine can focus light onto a chip. Low NA EUV is the current generation, and high NA EUV is the latest generation of ASML, which can print finer patterns, but at a price of $350,000 per machine. Fouquet claims that even though the new machine costs more, it can produce more chips.

There are a lot of questions about whether it will happen this month, next month, or the month after next. And I would say that it doesn’t really matter because I usually design high NA with an eye on the next 10, 20 years. If you go back to news outlets from 2016 and 2017, you’ll find the same quote. Low NA EUV was very expensive. We know what happened after that. The same thing happens with high NA.

There’s a Peter Thiel-backed startup called Substrate that claims it can build a rival lithography machine. What do you think about it?

There is still a big difference between wanting to have it and having it. Lithography presents many challenges. Being able to create images is a starting point, but you need to create those images in very large quantities, at very low cost, very quickly, and with nanometer precision. I always say that the only reason ASML can build an EUV machine is because 80% of it already exists, based on previous knowledge and products built over time. One problem had to be solved: acquiring EUV light, and that alone took 20 years. When starting from scratch, the challenges are enormous. I’ve seen a lot of claims like that. I’ve looked at some photos. But while we created the first EUV image 30 years ago, it took another 20 years of effort to turn it into a manufacturing system.

Does xLight, a laser startup partially supported by the US government, want to work with you?

xLight focuses on one element of the EUV machine: the light source that produces the light. The sources we have can be extended for many years to come, and we know how to extend them. What xLight is doing is a new source that still needs to be built and proven. The only question is whether there is a performance or cost advantage over existing products. I think the jury is still out. We work with them to demonstrate the technology. I feel that it is our responsibility. But it’s still a very long journey.

There are also reports that a former ASML engineer in China reverse engineered your machine.

To reverse engineer something, you first need to have a machine. Also, there are no EUV machines in China. We have never shipped tools to China. All the tools we shipped, we know where they are. These are either in use by our customers and we are tracking them, or they are disassembled and returned to us. The idea that one of our systems is in China is simply wrong. Additionally, our EUV technology has never been exported to China, so there is no EUV-trained personnel in China.

Very early on, when restrictions were introduced, we completely separated within our company who had access to EUV technology, documentation, and training, and who did not. Our team in China is on the other side of that line. The facts show that little, if any, progress has been made. Access to this technology is so important that it is difficult for people to accept it.

On broader export controls — Jensen Huang was here last night to argue that companies should sell globally and that increased corporate profits mean more taxes back home. He also said the important thing is to have the best and latest available. do you agree?

I think he’s absolutely right. He added, and I think this is what Nvidia has done, they can maintain their technological advantage by maintaining a generation gap in the products they sell. Because Nvidia sells products that are several generations old, it can find a balance between staying in business today and not passing a strong competitive advantage to countries that don’t sell the latest products. We believe the same approach should be applied to our products. We currently ship tools to China where export regulations allow, and this is the first tool we shipped in 2015. Applying Jensen’s philosophy to our situation, NVIDIA is working with a gap of approximately eight generations. I’m considering two or three. There is room for rationalization. It’s about finding the right balance between doing no business at all, missing out on big opportunities, and strongly encouraging others to compete.

What is your assessment of the current administration’s current situation regarding all of this?

There is good dialogue and that is very important. I think there is a real understanding of what business needs, but there is still the challenge of finding the right balance between different opinions and interests. The dialogue is there and we appreciate it. I’ve been to Washington many times. At least the discussion is happening. However, it is a very complex subject.

You don’t seem to be worried about someone taking advantage of your technology.

People want the best technology, but tend to forget what it took to build it. This required many years of engagement, not only with ASML but also with our suppliers. Many different groups of people solve very difficult problems, and then one company uses decades of lithography expertise to pull it all together and turn it into a manufacturing system. This is by no means easy. And I think that’s our biggest defense as well. That’s all you need to do to assemble it.