ASML CEO Christophe Fouquet: Nobody comes to us

Every time you use artificial intelligence, you depend, in one way or another, on a 42-year-old Dutch company, which employs 44,000 people, and spends 4.5 billion euros annually to develop its technology.

ASML, headquartered in the Netherlands, makes the machines that make the chips that make artificial intelligence possible. More specifically, it makes the only machines in the world capable of printing the microscopic patterns on silicon wafers that define the most advanced semiconductors — a process called extreme ultraviolet lithography, or EUV. These machines are about the size of a school bus, take several months to assemble, involve hundreds of suppliers, and cost from $200 million to more than $400 million apiece depending on the generation (prices that make even ASML’s largest customers pause from time to time).

This monopoly made ASML the most valuable company in Europe, worth more than $530 billion. With the four largest US tech companies — Microsoft, Meta, Amazon, and Google — committing more than $600 billion in AI infrastructure spending this year alone, demand for ASML hardware has risen to the point where the company has explicitly said the world won’t have enough chips for years.

All this demand also made ASML a target. Substrate, a San Francisco startup founded by a disciple of Peter Thiel, has raised more than $100 million and been valued at more than $1 billion on the claim that it can build a rival lithography machine. Separately, there have been reports that former ASML engineers in China have partially reverse-engineered the technology, a possibility with massive geopolitical implications.

Christophe Fouquet, who becomes ASML’s CEO in 2024 after more than a decade at the company, sat down with this editor on the rooftop of his Beverly Hills hotel Tuesday morning before his appearance at the Milken Institute’s global conference. Wearing a blue suit and white shirt, he was relaxed – even when the conversation turned to competitors.

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 not with the idea that this would come. It went from a concept — something people thought would eventually arrive — to ChatGPT, which was the first really good example of what AI could do. And now I think we’re looking at AI as the next revolution, not just industrial but societal. Did you see it coming? No, we sit in the middle of it every day, and sometimes we wake up in the morning and still check that what’s happening is actually happening.

The big question everyone is asking is whether the supply chain can keep up with demand. Is that possible?

The demand is so great that the overall market will be limited in supply for a long time. Right now, the biggest bottleneck appears to be in chip manufacturing. As an equipment supplier, we follow our customers, and we have followed them well so far – but we know we have to strengthen our entire supply chain and capabilities. If you talked to super scorers, I think they would tell you that in the next two, three, or even five years, they won’t get enough chips.

TSMC recently made news that its latest devices are expensive. How do you respond?

The UV system, if you look at the price, is going to be more expensive than the low-NA system, but the cost of making a chip using this tool on some advanced layers is going to be cheaper. We can get 20% and 30% cost reduction.

[Editor’s note: Both machines Fouquet is referring to here are EUV machines — the same fundamental 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; high-NA EUV is ASML’s newest generation, capable of printing even finer patterns but carrying a price tag of $350 million or more apiece. Fouquet is arguing that even though the new machine costs more, it produces chips more cheaply.]

I have a lot of questions about whether it will be this month, next month, or the month after that. I usually say that it doesn’t really matter, because we designed high NA for the next 10 or 20 years. You can go back to the press from 2016 and 2017, and you’ll find the same quotes – the NA EUV drop was expensive. We know what happened next. The same thing will happen with high NA.

A startup called Substrate, backed by Peter Thiel, claims it can build a rival lithography machine. What do you think about that?

Wanting it and getting it — that’s still a big difference. The challenges of lithography are many. The ability to create an image is the starting point, but you need to create that image in very high quantity, at very low cost, at high speed, and with nanometer precision. I always say that the only reason ASML can build a UV device is because 80% of it already exists, based on past knowledge and products built over time. We had one problem to solve, which was getting the UV light, and that alone took 20 years. When you’re starting from scratch, the challenge is enormous. I’ve seen a lot of claims. I’ve seen some pictures. But we got our first UV image 30 years ago, and we still need another 20 years of hard work to turn it into a manufacturing system.

What about xLight, a laser startup partly backed by the US government that would like to work with you?

xLight focuses on one element of our UV device – the source that generates the light. The resource we have can extend for many years to come, and we know how to scale it. What xLight does is a new resource that still needs to be built and proven. The only question is whether it offers a performance or cost advantage over what we have. I think the jury is still out. We work with them so that they can demonstrate their technology – and we feel that this is our responsibility. But it’s still a very long journey.

There are also reports of former ASML engineers in China reverse-engineering your devices.

To reverse engineer anything, you first have to have the machine. There is no UV machine in China, and we have never shipped any tools there. All the tools we shipped, we know where they are. They are either in use with customers, and we track them, or they have been dismantled and returned to us. The idea of ​​having one of our systems in China is simply false. Because our UV technology has never been exported there, we also don’t have any people in China trained in UV.

Very early on, when the restrictions were imposed, we created a complete separation within the company between those who had access to UV technology, documentation and training, and those who could not. Our team in China is on the other side of that line. The facts indicate very little progress, if any at all. This is difficult for people to accept because access to this technology is very important.

On export controls more broadly – Jensen Huang was here last night arguing that companies should sell globally, and that more corporate revenue means more tax dollars for the company’s home country. He also said the important thing is to keep the best and newest close to home. Do you agree?

I think he is absolutely right. What it adds – and I think this is what Nvidia did – is that you can maintain a technological advantage by maintaining the generation gap in what you sell. Nvidia sells a few generations, and this allows it to find the balance between continuing to do business and not offering a strong competitive advantage to countries where it won’t sell the latest. We believe that the same approach should apply to our products. Today we ship tools to China — which export controls allow — but it’s a tool we first shipped in 2015. If you apply Jensen’s philosophy to our situation, Nvidia is working with a gap of about eight generations. We’re looking at two or three. There is room for rationalization – finding the right balance between not doing business at all, missing out on a great opportunity, and aggressively inviting others to compete with you.

How do you evaluate the current administration’s position on all of this?

There is good dialogue, which is very important. I think there is a real understanding of the needs of the business, but there is still the challenge of finding the right balance between all the different voices and interests. Dialogue exists, and we appreciate that. I’ve been to Washington several times. At least the discussion is taking place. But it is a very complex topic.

You don’t seem to be worried about anyone shortchanging your technology.

People love having the greatest technology, but they tend to forget what it takes to build it. It has been many years of work – not only at ASML but with our suppliers. Many different groups of people solve very difficult problems, and then one company brings it all together using decades of lithography experience to turn it into a manufacturing system. This is by no means easy. I believe this is also the best protection we have. It’s simply what it takes to put it together.