Marc Hijink | 'Focus'... ASML, The Dutch Behemoth

✍︎: The Curious Worldview Newsletter - the ultimate compliment to the podcast...

• Jos Benschop – SVP Science ASML - Episode #56
• Chris Miller – Chip War - Episode #81
• ASML Bloody Good Business

Follow me on Instagram - @ryanfhogg

Watch on youtube - https://youtu.be/Kxmf6P-JHa8

ASML is certainly among the most critical businesses to the modern economy. They make the machines that make the microchips which makes modernity go round. 

You’ve likely heard of TSMC, the Taiwanese behemoth at the centre of Chinese geopolitics and how these are in fact the guys that make these amazing microchips. That is true! But they make them on ASML machines, and that’s the critical point to take home. The insane complexity of ASML’s EUV machine is impossible to fully appreciate. Jos Benschop, ASMLS senior vice president of Technology he makes the claim that ASML’s EUV machine is the most complex machine ever made by man.

When words are overused they lose their meaning, but truly ASML’s EUV machine is by definition, unbelievable.

The CTO Martin Van De Brink has recently retired from ASML – and his cultural legacy really forms the narrative for Marc’s Book – with plenty of fun anecdotes.

Marc Hijink is the guest on this episode, he is a Dutch journalist and tech reporter for Handelsblad and recently went down a three year journey with unparalleled access to ASML and produced a wonderful history of this phenomenal company. 

• 00:00 – Who Is Marc Hijink & ASML
• 02:38 – Writing ‘Focus’ A Dream Project
• 07:38 – ASML Origins & Culture
• 17:23 – Culture & Martin Van De Brink
• 24:35 – The Geopolitics Of ASML
• 26:58 – Complexity Of The EUV Machine & ASML’s Appetite For Risk
• 44:55 – What About The Software? + Recruitment Against Meta, Google, OpenAI
• 50:13 – Growing At 800 Employees A Month
• 55:28 – Fragility Of ASML’s Client Base
• 57:17 – Focus
• 58:38 – Serendipity In ASML & Marc’s Life

🍻☕: https://www.buymeacoffee.com/ryanhogg

Transcript

SPEAKER_00 0:00

ASML is certainly among the most critical and important businesses of the modern economy. They make the machines that make the microchips, which makes modernity go round. You've likely heard of TSMC before, the Taiwanese behemoth at the center of Chinese geopolitics, and how these guys are in fact those who make these amazing microchips. It is true, but they make them on ASML machines. And that's the critical point to take home. The insane complexity of ASML's EUV machine is impossible to fully appreciate. Which is something we get into into the podcast. But as well, if you refer to an earlier episode, I actually recorded with ASML's senior vice president of technology, Jos Bensop, who himself makes the claim that ASML's EUV machine is the most complex thing ever made by man. And when words are overused, they really lose their meaning. But truly, ASML's EUV machine is definitionally unbelievable. The CTO, uh Marten van der Brink, has recently retired from ASML, and his cultural legacy really forms the narrative for Mark's book, with plenty of fun anecdotes, of course, along the way. Mark Heijink is the guest on this episode. He's a Dutch journalist and tech reporter for Handelsblood and recently went down a three-year journey with unparalleled access to ASML and produced a wonderful history of this phenomenal company. Among the many things you can expect in this podcast is a lot about the culture of ASML. What does it take to create a good culture? How do you maintain culture as an organization grows so rapidly? The serendipity of the life of ASML, and then of course, all of the rich and wonderful anecdotes that Mark was able to accumulate during his time researching for this book. So refer to the timestamps to jump through to the parts of the episode that matter most to you. Mark is an absolute legend. So also buy his book. It's called Focus and learn about this magnificent company as well. And for the love of microchips, leave a five-star review for this podcast on Spotify and on Apple. Which, of course, the very phones and computers which these applications run on top of rely on ASML's EUV machine. There is another form episode of this very podcast that I would like to draw your attention to as well. It's with Chris Miller, who wrote the definitive book on the geopolitics of semiconductors. It's called Chip War. A link to that is also in the description. But my hope is that with these three episodes consumed together, it would make for a decent intro into the wonderful world of microchips and as well ASML. And here is the great Dutchman himself, Mark Hijink.

SPEAKER_01 2:39

Well, um I I kind of uh uh stumbled upon ASML and discovered that it was a uh company that was like a Dutch uh high-tech fairy tale. And um uh for me it was it's uh great to have the opportunity to uh uh to tell this story and uh to uh research it. And um sometimes I have to like pinch myself to see that I got access to to to such a uh to such uh uh a big and important company. That also um gave me the opportunity to also uh to not only speak to their leadership and to the employees, but also do research uh within their supply chain or talk to criticists or uh external experts so I could create a like a holistic view of the company. Yeah, that they really wanted to uh um they asked me to make an uh like an objective, like a uh to to create a uh a mirror for the company so they uh could look themselves in the eye. And uh the funny thing is um the c the company is so much focused on on just building one machine, it's like a one-trick pony, a very expensive one-trick pony. Uh but uh uh it also kind of loses attention for all the other things happening on the side. Geopolitics was one of them, but also their logistics was somehow uh sometimes a mess. And also the way they uh handled HR uh was uh a huge problem for for the for the company.

SPEAKER_00 4:20

So they wanted you to create a mirror for them. Did anyone not like the reflection that uh came through your book?

SPEAKER_01 4:27

Oh, nice way of putting it. Um I think in the end, uh ASML is a company that uh really appreciates an honest view. Um and uh so they gave me all the freedom just to talk to everybody, and they never tried to intervene or uh to steer uh the conversations I had, um, which was also a thing I had to pinch myself in the arm uh for because um it they really appreciate an honest view and a holistic view, and I guess um that's part of their uh engineering culture as well. Um so it doesn't um uh help the process to uh be polite to each other, you just have to uh to to to to catch it uh to to to make a deadline, you have to be um uh uh fully aimed at uh um being blunt and creating the the least bit of latency, which means you have to be direct and uh to be honest with each other, and that's the thing I wanted to uh uh emphasize in my book as well. And I think uh that quite was perceived well at ASML, so to speak.

SPEAKER_00 5:43

Oh, I'm happy for you that uh they received it really well, and it must also be quite cool that as a Dutchman, as a as a proud, honest, blunt Dutchman, you got to uh report on you know the biggest, baddest company that your country has created in in modern history.

SPEAKER_01 6:00

Yeah, and but you know the the funny thing is that a lot of Dutch people actually don't realize how important ASML is for the rest of the world. So that was one of the problems that that the company was quite obscure because it's technology is so complex and right, right.

SPEAKER_00 6:17

It's it's uh well uh uh and people just take for granted that a semiconductor is just another component in their electronics.

SPEAKER_01 6:24

Yeah, and and um but but the ecosystem is uh depending very much on ASML, and also the ecosystem of ASML is growing with uh with the company. So um the the whole supply chain has to accommodate uh um ASML, and ASML has to accommodate the uh growth of the semiconductor uh um uh industry. So kind of the whole manufacturing industry in the Netherlands is uh uh leaping forward, and um that has a uh that that took uh a while for the penny to drop in, for example, in the Haag, where the Dutch government uh uh resides. Um but about uh a company that started out very small in a very small town and now is a huge multinational and and the kingping of the uh semiconductor industry.

SPEAKER_00 7:17

Yeah, little town of uh Vlederhoven.

SPEAKER_01 7:20

Yeah, Veldhoven, it's hard to pronounce, but it's it it's it's Brabant, so that's uh that's a southern part of the Netherlands. The the Netherlands is not a really a huge country, but it does have a north and a s uh south, and uh ASML is just like below the rivers uh in the middle of the country.

SPEAKER_00 7:41

And so they were a little uh spin-off from Phillips, the great Dutch company, and headquartered in this little town that was completely flattened in World War II, and then now the whole entire town is built around ASML, right?

SPEAKER_01 7:57

Yeah, if you uh uh look back on on uh the first 40 years of uh ASML, uh the company now is just uh celebrating its uh its 40th uh anniversary, then you see that um ASML kind of took over the rule that uh the role that uh Phillips used to have uh in the city of Eindhoven, which is very near to uh Veldhoven. So Philips used to be huge, but um uh it has diminished now. Yet its uh spin-off, ASML, um, has has grown into this uh huge company um which is I think the the most valuable uh valuable technology company from Europe and um the the the biggest uh and most important uh company in uh in the Netherlands. And it really has altered uh the city of Eindhoven. Uh the the the they have really have to uh adapt to this uh uh pace of growth that ASML is experiencing right now because the semiconductor industry uh is expanding worldwide, especially in Asia, but also in in Europe and and and uh and the United States. And because ASML is building the machines, the chip machines that are used to well uh uh create uh semiconductors that are in all devices, then you have to uh uh adapt uh its growth pace to uh the worldwide expansion, and that really is uh putting uh a lot of stress on Brabant, uh as as we call the region in the Netherlands, and um they have to invest a lot to uh accommodate uh ASML's uh grow uh growth.

SPEAKER_00 9:46

So, how many new people are joining, say, every quarter uh since uh 2020 for ASML?

SPEAKER_01 9:54

Um well the uh if if you look uh at the pace of growth uh of the number of employees, um you see the expansion that uh ASML is uh uh about doubling every five years in general. That has been the case over the last decade. Um but during uh COVID uh the number of employees has uh uh grown um even more rapidly. I think the uh well per quarter is um you could say that ASML uh in its uh at its uh peak uh during the chip shortage was uh expanding at uh a rate of uh like 800 people per month, which is uh crazy for a company that that size. And and they're also they're also expecting to double within the next five or six years. So the company uh will have a I think a total um uh number of employees of about 70 or 80,000, which of them uh half of them would be uh uh based in the Netherlands.

SPEAKER_00 11:04

And very few of them are actually Dutch themselves. Like it's such a multinational um melting pot.

SPEAKER_01 11:12

Yeah, it it it's it's uh well it it it is a melting pot. And if you look at their number of nationalities, uh I think there's over uh a hundred and forty. Of course, um there's still a lot of Dutch people um who join ASML, but if you look at their uh worldwide expansion, the the company's uh growing to be less and less Dutch um every year, which makes sense because the set semiconductor industry is just um well, it's a a global business, and um ASML is uh acting accordingly, you would you could say?

SPEAKER_00 11:48

So if the company is growing less and less Dutch over time, although it's in very much sort of Dutch heartland, and it's famously a small little Dutch spin-off, and the Dutch culture is quite a strong dominating type of culture. How does ASML maintain a culture, or does it sort of just constantly change over time? And how significant is their culture to sort of their success?

SPEAKER_01 12:14

If you look at these um uh first adventurous years of ASML, they uh had this sort of cowboy attitude, they really wanted to penetrate this uh tough, crowded market of lithography machines, and um the way they uh, for example, that they uh uh joined the beauty contest of one of the semiconductor uh manufacturers in in the US called AMD, which is now a very uh big company that um is is a competitor of Intel, and um to um gain a contract to uh deliver and ship these these machines to AMD, um ASML had to uh join a beauty contest and and compete with uh other lithography machine uh manufacturers to see uh who had the best device. So um that's the way they um so they installed one machine there in the US, and uh when they just booted it up, uh it seemed uh that the machine was was running okay, but then they um the engineer uh saw that the uh engine uh that the the that the linear uh uh motors in the machine weren't uh properly working, weren't working properly. So they had to sneak out uh and uh send a message to Velt over okay, please bring us new spare parts. We really need this. And um without telling AMD, they um uh snuck into the the the factory uh secretly at at night and they uh changed these uh uh motors just to um uh to win this beauty contest. And they never they never told AMD. So that's uh it's a bit of a uh uh an improvising war. Yeah, that's that's just the the way uh well we have this one shot, we have uh we have to take huge risks. Um and uh well we have nothing to lose. And I think the nothing to lose mentality was a thing that uh really uh determines um ASML's DNA in the early days, and you see um that has been growing uh uh throughout the history of the company, and it still is very much part of it uh of its culture right away. And maybe it's typically Dutch as well, because uh well the Dutch are not really fond of hierarchy, so they're really used to of uh uh just doing what they uh want to do, and they're usually very blunt and direct, and which is a good um good match with uh the engineering spirit. So for ASML that that that went quite well.

SPEAKER_00 15:01

Do you think that culture is something that is created at the top and then trickles down?

SPEAKER_01 15:08

Um if you look at ASML the the the culture was uh always uh an engineering culture, very direct. If you uh build products that are complex like uh lithography machines, then you really have to uh um be very direct and blunt uh and point out failures or um uh in or faults or errors in in the design, otherwise you have a huge problem when a lith lithography machine um uh well somehow uh gets jammed when it's in a uh in a chip uh factory because then it's really expensive is if the machine is paused.

SPEAKER_00 15:52

But do you think that say the leadership, even in an engineering culture, it's the leadership who are making the decisions, who are the ones that drive whatever that culture becomes?

SPEAKER_01 16:03

Um I I think uh if you look at ASML, the the uh the culture was always uh in the engineers, and the engineers uh ended up in leadership. So one of the people at ASML that was really um like a driving um force to um to uh well to push this uh tech technology further and further was Martin van der Brink, and he was the CTO, and he just uh uh recently stepped down because uh well he's um on his uh uh on his pension now. And um Martin van der Brink has this uh really rough style, you could call it uh a really a confronting way of meeting people and and and trying to uh um well you mentioned earlier this uh Dutch culture might be very direct or at least engineering culture, very harsh, kind of direct.

SPEAKER_00 17:01

And it sounds like from uh what you wrote about Martin van der Brink and maybe what his reputation is, is that he exemplifies that uh kind of completely. What are some examples of Martin van der Brink's uh sort of management style that lead to those sort of anecdotes?

SPEAKER_01 17:16

Well, um I I think uh he likes to uh uh uh uh take people out of the comfort zone. So when you uh uh walk up to him, he usually uh um tries to throw you off your feet. I don't know if it's that's the right expression, but um for example, uh I spoke for for this book, I spoke to him five or six times, and uh we went through the history of uh of ASML. But I remember the first time uh he um um we we met, he said, so why are you writing a book about ASML? Um there's already uh been written some books about ASML, which is true, which which is true, uh um because well but it was about the first decade, which was uh quite a long time ago. But it was just his way of challenging me, and this uh this challenging culture is is very much part of his uh strategy uh to um to get people to uh well um to to uh to drive high performance. To drive high performance, yeah. But also uh um for example, he's uh is he's uh one of the people that uh always emphasizes that ASML, even though it's a market leader, should be humble, so um it should be uh a company that doesn't try to uh um make as much profit as possible, but uh the aim of the company uh of ASML and the company is to keep the the ecosystem um alive and prosperous and um to add value to to the few customers they have, these few huge uh chip manufacturers. And because ASML is like almost a monopol monopolist, and especially a monopolist on EUV machines, uh it might be easy to uh think, well, we can ask uh uh or do anything because uh all our customers are very dependent on us. Um and uh Martin van der Brink has always been well some kind of preacher and saying, Well, we need our customers to be successful, otherwise we will fail as a company. So um when I asked him what his uh what his thoughts were about the uh threats that might uh impact uh ASML in in in the long term, he didn't mention geopolitics or uh growth or whatever. He said success is the uh biggest threat of uh of ASML. Well that's prophetic. Yeah, that's uh because he's afraid that people might come become lazy or uh somehow well uh expecting to be uh successful all the time, and then you create an opportunity for for uh um competition, of course.

SPEAKER_00 20:20

Do you suspect that now Martin has retired and his cultural influence is left behind that uh things are gonna change?

SPEAKER_01 20:31

Um well uh of course he's retired, um but he um stays on as a um he'll keep advising the company for about a couple of days a week. So it's very hard for him to leave the uh the company uh as a whole, and his experiences uh widely appreciated, but especially his spirit is widely appreciated. Um I think the fact that Martin is leaving um uh has certain certainly has an impact, but he has a lot of people uh around him that have been trained and are used to his approach, and some people uh they call themselves like mini Martins, so uh they are um um they're cut from the same cloth. So that's interesting.

SPEAKER_00 21:22

Yeah, yeah. So his his blueprint as uh like a cultural leader will maintain in the way that they work.

SPEAKER_01 21:31

Um yeah yeah, my expectation is that that will definitely happen, but uh maybe the biggest risk of uh uh changing the culture uh of ASML is the rapid uh uh growth of the number of employees. So a lot of people joined the company, but they didn't have the experience um of the early days or in uh this cowbook uh cowboy culture. So they they they uh uh might have to be trained to be cowboys again, but also to not Be not too challenging, which is also part of uh ASML's culture. So they also always include well, you have to be challenging to each other, but you also have to collaborate and care about your colleagues. But it's that's like a conundrum because you can't do uh um all these three things uh together.

SPEAKER_00 22:18

You mentioned that Martin van de Brink um sort of prophetically said that success might be the the downfall of ASML, and you also suggested that they're they're monopolistic when it comes to the EV machine, the highest end of uh chip manufacturing. I mean, being that they are essentially a monopoly, it's the ultimate success. You know, how realistic is the threat of competition to them at this stage from your perspective?

SPEAKER_01 22:47

Uh I think uh there's always a chance that uh ASML uh will have a a new competitor. And for example, uh China spending lots of uh um uh capital uh in in creating uh a competitor that might be uh an alternative for ASML machines. Uh since ASML is limited to uh uh ship these most uh uh advanced machines to China, it it becomes uh uh like a uh a push for China to uh uh create its own uh tool set uh to uh it create its own lithography machine. And there's a lithography manufacturer it in China called SMEE SME, and um yet at they're about 10 years behind uh ASML, 10 or to uh to 15 years. But if somehow the Chinese fabs are forced to use uh SMEE equipment, then SMEE uh gets more uh research money and they can uh get the chance to um improve and to and learn uh to learn and to build more advanced machines. So uh restricting ASML in China might actually um uh uh uh create a competitor in China. That's the that that's the thing that ASML is uh always worried about. Yet they have to uh uh they have their uh strategy to uh out-innovate. So they just spent more money on uh RD. If you look at their uh I think uh last year they spent about four billion euros in RD, which is huge, especially for a Dutch company.

SPEAKER_00 24:33

Yeah. Um regarding the geopolitics of ASML, um, there was a book by Chris Miller which came out explicitly about this a few years ago, which I'm sure you um Chip War, yeah, excellent book. Exactly. Yeah, um so in your own words, could you explain to the audience and to me uh why ASML is a globally significant company?

SPEAKER_01 24:57

Um you could uh describe ASML as a supplier of the uh most important uh uh machine you need to build these uh advanced semiconductors. And these are the chips that uh you could find in your smartphone, especially the newest one, but also uh in the AI chips that companies like Nvidia make. So that makes these uh devices ASML produces uh the most crucial uh element of uh uh chip technology.

SPEAKER_00 25:28

How you sort of mentioned uh earlier regarding culture in Martin van der Brink that he didn't see this sort of uh geopolitics to be a uh a threat to ASML, but clearly now it is. Just culturally, for this massive company, uh largest company in the Netherlands, largest tech company in Europe, uh huge multinational melting pot in Vlederhoven, how does the culture think about how does ASML think about just how geopolitically significant they are? Like, how does that affect their day-to-day from what you could observe?

SPEAKER_01 26:04

I think they um uh prefer to focus uh uh on their technology, and um of course the geopolitical uh uh attention uh has uh um pushed the company into the spotlights, which is well, you could call that a complement, because the world's now realizing how important ASML is for this whole supply chain of semiconductors. But uh in Veldhoven they're very much uh um inclined to focus on on the technology because they're they're um it's so complex, you don't need the distraction from the outside world. So it's um it's nice to know that everybody now understands what kind uh uh what company you're working on, but the daily work is still very much uh about technology, which makes sense because it's such a complex machine.

SPEAKER_00 26:56

Yeah, I mean, perfect segue. Speaking of the complexity of it, Jos Banshop, who's an executive at ASML, said that in his opinion, it's the most complex machine ever created by human beings. Um, in your own words, could you explain why the EUV machine is so complex and like all the magical sort of components that go into it?

SPEAKER_01 27:19

Well, uh, I think uh we would need a few podcasts uh to explain uh how the machine actually works, but I could just uh Yeah, but I'm just looking for some flagrant language, Mark.

SPEAKER_00 27:30

You don't have to give us the hardcore tech breakdown.

SPEAKER_01 27:34

Okay. Um because uh ASO it took ASML like 15 years and uh over 15 billion euros to um to get this uh machine working, which uh uh uh which is a huge number and it it's a crazy risks. Uh it's a crazy risk. Um so EUV uh uh it's the abbreviation for extreme ultraviolet light. And if you dive into a lithography machine, uh it's basically a copier. So and just like a regular copy copper, it uses light that's been uh um reflected throughout a couple of mirrors, and not onto a piece of paper but onto a uh a silicon wafer, so this uh big shiny disc with uh a photosensitive uh material on it, and um the structure that it's uh been printed upon um has these really fine lines, so you get this uh uh this fine-grained structure that's used to build transistors, uh, the transistors that power your chips. Um the more transistors you cram on the same surface, the more powerful and advanced and energy efficient the chips uh will be. But um for these lines to be very thin and very uh uh fine-grained, you need a well, the typical light source uh would have been uh um the DUV machine, which which is deep ultraviolet light, and it has a uh a wavelength of 193 nanometers. So uh and then one nanometer is one millionth of a of a millimeter. Of a millimeter. Yeah, that's crazy.

SPEAKER_00 29:27

Impossibly small amount. Yeah, if you look at your fingernail for three seconds, I think it grows ten nanometers, something like that.

SPEAKER_01 29:36

Uh three. It's actually one uh it's uh at least my nails, uh they they grow uh one uh nanometer uh per second. Um and um so it used to be 193 uh nanometers, but EUV, so the extreme ultraviolet light, um, is a light source that well it's like a fine liner, so it has a uh dimension of uh 13.5 uh nanometers, which is much finer, so you can uh create um uh more advanced chips. But this uh uh extreme ultraviolet uh light doesn't exist uh uh on this planet, so you have to uh it only exists uh in the corona of the sun. I think that's the nearest place you can find it in its natural uh uh way. So you have to recreate uh create this light, and the way ASML uh is doing that, it's um uh uh using tin drop droplets, so really small droplets, and it um fires a laser, a really heavy laser uh uh onto this droplet. And um this droplet is being uh it's lined uh exploding into a plasma, and the plasma generates a light which is projected into the system, then reflected throughout a couple of mirrors. I'll get into the mirrors later because that's totally crazy. Um and um then it um gets uh projected throughout these mirrors, uh reflected on a mask with the uh uh chip chuck uh structure, and then it hits uh the wafer, which is uh moving very rapidly. Okay, now back to the tin droplets. So to uh create a plasma that's strong enough and to have enough light to actually write the structure on the on the disc, uh you need uh um the tin droplets to be uh hit with a laser 50,000 times a second. So that's uh it's it's not even rapid, it's it's like uh what do you call it? Uh machine gun? Yeah, yeah, yeah. Impossibly fast. It's it impob uh impossibly fast, and also uh the laser that's been used. Uh you could compare this power um with uh with a jet engine. So when a fighter jet uh is taking off, that's the same kind of uh same amount of power or energy that's needed for the for the laser. And they well, it is it's it's not a thing that ASML builds itself. Um so it's being built in uh in a German town called Ditzingen, and um by the by a company named Trumpf. And they really uh built these crazy lasers with uh uh like three amplifiers in a row to uh to have enough power um to hit that uh tin droplet. And that this was the hardest part of the EUV machine, actually building that light source that never uh existed uh before. And if you look back at EUV, um it's it's a technology that was developed or examined and uh experienced with um in Japan and uh US and also in the Netherlands, but to actually end industrialize it to uh to create it um uh in a machine that can work like 24-7 three and uh every day in uh throughout a year, um that that took a quite a long time. And that was the biggest challenge for ASML to succeed. And but uh okay, let's uh go to the next step of the EUV machine, uh which is uh uh the optic system. So uh that that's also uh a German thing, actually, or a co-invention of uh Zeiss, and uh that that's a German company, and ASML. And the optics system uh uses uh lenses that are very flat because the smallest uh what's the right word? Uh detori uh smallest error uh um yeah, yeah, the smallest bump, even an atom high, would um uh create problems uh uh during uh chip manufacturing because the the margins are very uh small um in in this kind of uh uh machine. So Zeiss really has to uh uh do uh to outperform itself to create the uh the flattest mirrors you can imagine.

SPEAKER_00 34:19

So that'sn't it the case it's the flattest substance humans have ever created?

SPEAKER_01 34:26

Yeah, yeah, and I I think they um they use the uh uh example of uh if you were to stretch or if you were able to stretch this mirror uh to the south uh to the size of uh Germany, so like uh thousand kilometers uh across. I think the um the the biggest bump on the on the mirror would be uh less than a hair uh high. So it would only stand out a hair high, which is it's incredible. But I even um I like their other example even better. Um that uh if you were to uh shine a laser on it and you you could point it at the moon, you could actually hit a golf ball on the moon. Uh it's so uh so it's accuracy is uh accuracy, yeah. So it's accuracy, yeah. So it's accuracy is is is tremendous. And um so for this book, I traveled uh around the world and also traveled uh around the main suppliers of ASML, uh ASML's machine. And Zeis um had to build this huge lab, laboratory um to measure these uh mirrors. The they have these huge tubes or canisters where they uh create a vacuum that's like in the EUV machine, and that's where they uh measure their mirrors. Because if you if you can't measure it, then you can't build it. That's the the German approach. And it was uh it took I think three or four years to build uh these machines to measure them, and um, it also took uh ASML like 1 billion euros to invest in Zeiss to be able to create these uh mirrors. Yeah, yeah, that's it's such an appetite for risk. Uh yeah, and I think uh the appetite for risk uh is uh mainly uh um a natural thing for uh for for ASML because that's the way the uh the company succeeded um and survived its um its uh first decade when it was on uh on the on the brink of uh collapse uh a couple of times. Right. So yeah, there's a huge appetite for the uh for for risk, but they're not um being uh riskful uh per se. It's just uh the if if you want to build machines that are gonna add value in 10 or 15 years' time, then you have to look at us now, yeah. Yeah, yeah. So it's really uh a thing about focusing, yeah.

SPEAKER_00 37:02

Oh, nice name drop in the book there. Nice one. Um so the purpose as well of all of these extremely complex parts working together is just to print more fine and narrow lines onto a silicon wafer. So they had deep ultraviolet, they had deep ultraviolet, then they had extreme ultraviolet. What is the next adjective for the next machine that they're building, and what is their projection for how accurately they'll be able to print onto silicon?

SPEAKER_01 37:36

Um, the latest machine they're building right now is uh called High N A. So that's uh basically that's an EUV machine, but with an improved set of mirrors uh called High N A. And that machine is uh the first one is uh has just been uh uh put together in Veldhoven. And it was nice because for the book I uh traveled along uh all these suppliers and uh ASML uh factories and which are we're building like these components, these uh modules, and and um uh last month I got the uh opportunity to uh look at it um well uh all the pieces combined, all the pieces of the puzzle puzzle put together, and then you end up with a machine that's 40 meters long, I think it's uh five meters wide, and it's also five meters high. So it's massive, and it it costs uh about 400 million euros. So if you look at what what can it do, then um the um well maybe we we should um talk about the nanometers that chip manufacturers use and the actual nanometers on on the on the chip itself. Nice because there's sounds good. Yeah, because there's a difference. There um if you look at uh, for example, uh companies like Intel or TSMC, they they say they're working on two nanometer chips. Well, but that's just uh a marketing phrase. Um in real life, if you were able to look at these uh lines, um you could tell that they're usually 20 to 30 nanometers meters uh apart from each other. So there's still a lot of room to grow. But if you look at that 20 to 30 nanometers, then the latest machine ASML is producing is capable of uh well improving that by uh by three uh three times or even five times, um which means that there's still uh well room uh for this um uh lithography machine to make even more fine-grained, narrow lines and to create more advanced chips, especially for for AI.

SPEAKER_00 39:49

Just one more uh moment to linger on the complexity of the EUV machine. So you mentioned the the tin that shoots across um hundreds of times per second, and then the thousands of time, thousands of times per second, and then the um the it's it's heated up by that zap from that big laser to, and I don't know if my notes are incorrect here, but to 40 times the degree of the sun's surface, and then the loud and then that light through the tin is bouncing off and printing onto a wafer. Um, and I remember some details as well how they have to keep the temperature inside the machine to such a such a narrow bandwidth, and there are these uh gyro machines that are constantly balancing the silicon wafer so it's remaining perfectly level.

SPEAKER_01 40:44

Yeah, we'll just yeah, yeah, it's um and and what uh what is also impressive is the speed because the wafer is moving rapidly uh in these uh latest generation of machines, and um it's uh like accelerating uh uh uh like a Formula One car, but even uh way, way, way faster. And and all these small movements have to be well predictable, so you can uh adjust uh the uh the the mirrors to it, and uh also the mask, so where the structure is printed on, is moving um at an even faster pace. So uh it's it's like a copier gone wild, and it still produces this these uh uh uh tremendously uh uh impossible uh small uh lines. So from a technology point, it's really uh uh it's the pinnacle of uh of innovation.

SPEAKER_00 41:42

And just to round off all the complexity of it, and this is something Joss Benship was really really trying to emphasize when I spoke to him, it was that creating this machine, yes, is an absolute marvel, but what he's more impressed by is its ability to work day in, day out, 24-7, and so it's sort of uh durability and uh reliability, um, which is something you hear echoed a lot when Elon Musk is talking. You know, creating a prototype of a really nice car is hard, but it's way easier than reproducing that car at scale.

SPEAKER_01 42:21

If you look at uh the the machines that ASML built in the in the early days, uh I uh I think most of them are still running um after uh two two decades. And um these are machines that are lithography machines that are very useful for example uh creating chips needed for uh EVs for elect uh for for electric cars, and um uh they call it their uh uh thrift shop. So they still uh get these machines from the early days back and they uh upgrade them and uh repair them and um uh send them back to uh uh to the uh uh fabs, and because there are still a lot of demand for these mature systems as they call it. And um because uh uh I think the number ASML uses is that more than 945% of all the machines they ever made are still being used from day to day, uh which is crazy because uh that that makes it uh an investment uh that uh pays off several times. But you could compare to, for example, the uh the uh aviation industry. Uh so the way you build a plane uh is quite similar to the way you build a lithography machine. So ASML uh uh also like Airbus or Boeing, um uh has a huge supply chain where it sources these modules, these uh uh and and then it puts together um uh the the the system in its own factory so it's the designer the architect and it's the assembler so it uh puts it together and it does uh uh it also does the uh well the servicing uh which takes uh uh like daily care because if you just would let a machine run and uh wait uh and and and never service it it will it will uh uh stop working within a couple of weeks and it become useless so it really has to uh well like a like a plane which is attended to every time it lands and then um uh surface engineers uh check the uh check the plane that's uh uh similar to uh uh how the lithography machine is uh is being surfaced yeah yeah that's a neat analogy because I suppose that's so rare what what about the what about the software component of all of this so clearly some of the best hardware engineers in the world are working at ASML but is it also the case that some of the best software engineers are working there to build these systems that are calibrating the machine and constantly balancing the gyro and and um you know all of these smaller details yeah if um without software this machine would be useless because um there's uh so much power and so much uh well you mentioned a temperature for example but it's just one thing um uh there are a lot of circumstances that change um during the process even in a clean room so if like uh uh temperature is one thing but also air pressure or um uh humidity um or um vibrations earthquake for for example especially in Taiwan um so there are a lot of um uh factors uh that are uh um that have have an influence uh on the chip manufacturing uh process and um the only way to handle that is that you need to write massive amounts of code to uh create all the uh uh right um uh adjustments for uh for all these circumstances and if you look at for example at a waiver uh a chip uh manufacturing is like adding layer on layer um and every time uh a layer is added uh you need to um uh of course have the lithography process print uh the the pattern on on the disc but you you also have to well etch the machine and have all these chemical processes and um uh put it uh the wafer in ovens which kind of uh uh changes uh um the the the wafer itself and and it gets warped and all these things have to be measured so the when a wafer enters the machine again the lithography machine again it has to be measured it creates huge amounts of data and the machine is has to be adjusted to to that to keep the focus right so it's very very much uh a a software game and uh the only way to control that is to hire the best uh software people you can find oh man it's incredible what are the salaries of these types of guys say the you know uh not the top executives but you know a guy who's in his early 30s super well educated uh clear performer so so so usually the the salaries at ASML are are like 10 to 20 percent higher than the other companies in bra in Brabant um which makes it a very intra attractive uh uh employer of course uh yet uh you you also mentioned leadership the leaderships um uh uh the leadership of ASML makes quite a uh uh a nice amount of money but compared to for example the uh the peers in the US uh um it's quite modest. Interesting so how then does ASML uh recruit the best software engineers in the world when they might be able to just go I don't know work at Google or Meta or OpenAI and make way way way more um the I think one of the uh um uh one of the uh attractions no that's not the word uh one of the nice parts of ASML no help me out attractions is fine yeah yeah okay one of the appealing factors yeah that's the word I was looking for thanks um uh so I I tr uh also visited uh San Jose uh which is uh like uh ASML's uh uh division in uh in Silicon Valley and there's a lot of competition from big technology companies um um Tesla uh Google Apple you name them they all pay uh extremely well they might uh pay uh better than ASML in in the in the Silicon Valley area but uh the appealing factor for for a lot of engineers uh to join ASML is that they can actually work on uh uh revolutionary technology and not about uh uh building consumer technology or um uh developing a new uh filter for Instagram or for for for TikTok so they um that as as as they use uh their their catchphrase they're changing the world one nanometer at a time nice and so so so for a physicist that it might be uh uh more of a uh appealing factor to work at the uh asml uh compared to uh a consumer facing company yeah no that makes that makes total sense and if you could say on a pie chart put the number of employees into different um different sectors you mentioned I think early 8,000 new employees per month um at one stage eight hundred eight hundred eight hundred eight hundred eight hundred sorry that makes that that makes more sense but nonetheless eight hundred is a huge number that's 200 new employees a week so it's that that's an impossible uh pace because all these new people have to be trained and to uh have to enter the uh the company have to as I describe it in the book uh land on planet ASML so uh that's that's mind boggling they they they they can't manage their growth uh in that way and they're producing less than a thousand units uh per year so on a pie chart each sector where where is all this new growth going to is it the case that the engineering teams are so so huge or are a lot of these people managing say uh marketing sales the hospitality of running the park you know the the the legal side of things etc or is it still the case that most of them are going into the engineering side of the business well if you look at ASML as a company it's it's growing rapidly and um so you you're right they don't produce as uh uh management many machines as you might expect from a almost 30 billion uh uh dollar uh uh revenue company but um these machines are very expensive and uh very hard to uh uh to manufacture so you need a lot of hands to put them together and you need a lot of minds to uh think of the new generation or to solve the problems with the existing generations of machines so that's the way ASML is expanding and um it also um is expanding uh in um uh the the countries that are uh investing in their own uh chip technology now so for example the US and and Europe um they they they uh um have um uh this chips act um the these these chips acts these subsidy plans to support new uh uh fabs in uh the desert of arizona or in in uh in in Magtenburg in Germany and uh these are all like huge factories that need uh surface uh engineers from ASML as well so you see uh a lot of expansion also yeah on on the surfacing part yeah right right so so ASML is expanding worldwide not only in in in the Netherlands when we spoke uh last year you said that offshoring the Taiwanese chip production at the same scale in the US was impossible do you still think that yeah because um uh the TSMC which is the the the biggest foundry in the world based in in Taiwan um is of course it's building uh these uh fabs in in in the US but they're very uh well they're they're small or little or producing less wafers compared to uh to all the uh fabs of TSMC in Taiwan and it's very hard to uh uh run a factory and you see that TSMC has postponed um the uh initi initi uh the the the start of their first uh um uh chip fab in in Arizona for quite some time so because they're having trouble just getting along with the American uh work ethos. Right right so it's uh it's a cultural thing uh it's uh it's a cultural thing but it's also uh a thing about uh um having access to talent and if you see um that the uh um the uh educational system in in in the US is well it's not uh uh producing enough uh or uh producing enough technol technology uh talent to um actually uh uh be be working in these uh chip fabs interesting so the yeah so the US a has a has a problem uh so uh aren't the um people of TSMC the employees rather of TSMC aren't they motivated to just take a huge paycheck move to America and then recreate what they did in Taiwan remove all of that geopolitical risk from the table um uh I don't think uh they can be lured uh by just a paycheck it's a it's a cultural fit of course and um because if they were lured uh with paychecks then they probably uh wind up in China which is offering even even even more sometimes five times the amount they can make uh in Taiwan that's amazing um but if you uh of course TSMC is adjusting to uh uh this uh a new era a new era in geopolitics and so it is building um uh advanced uh fabs and even more advanced fabs in in the future um but it's keeping its uh uh its best technology for itself so that's a way of well you could see it as an insurance policy as long as the world is depending on the the fastest chips uh uh on on Taiwan then it's worth protecting uh Taiwan from a uh well a possible uh uh uh invasion from China for uh for example how does ASML think about the um fragility of their small client base in my notes here I've got that 60% of ASML sales come from just three to four companies yeah um this is typically uh a thing of the uh business to business uh uh market and ASML is depending on just a handful of companies very large companies like TSMC Intel um Samsung uh SK Heinex also SMIC from from China and uh of course uh Japanese companies and Micron you name them um these uh are companies that build uh chips and when when um as long as the demand the global demand for semiconductors or chips is is growing at this present pace it doesn't matter for ASML who's actually buying the machines because it's just uh uh well this e uh this economic law that somebody will produce the chips that are necessary and even for example um that might be a risk that uh if Intel some uh somehow falls behind or doesn't uh achieve its uh very uh ambitious goals uh to to lead the market again um then for ASML it's no problem that the uh global demand for chips is being uh fulfilled with uh other by other companies so that even in the end um if there were just only two companies as long as they produce uh all the chips that the world needs ASML can deliver the machines and they're uh well they're still uh commercially viable commercially viable and focus this has been a word you've um very strategically dropped subliminally for the audience throughout the the podcast or you know to call the focus uh okay um so uh there are hundreds of books uh uh called focus but I guess for ASML it makes sense because the machine itself and the lithography system is purely based on uh on focus uh so the the image has to be projected very uh accurately and also ASML has a focus on the long term so it's uh horizon is uh like 10 or 15 years ahead it's road its roadmap and you also see the uh focus of uh of ASML as a company that's just building one machine one type of machine um so there's not a lot of distraction and uh finally there's also this focus of the world around ASML so uh the geo geopolitical uh uh attention um has uh pushed ASML in the spotlight and for ASML that's something they have to get used to and they they kind of pushed that away they pushed away the outside world and now well everybody knows the company yeah and having books written about them are you surprised by the demand for the book uh yes but uh reading it back and I and I think uh also uh creating uh um or uh uh having the English version out and updating it I realized it's a it's a special story uh um because there was a lot of uh serendipity uh if you look at uh ASML how it started how it was lucky several times how it took huge risks uh so things uh could have uh turned out completely different and uh for uh for me uh it was uh really special to uh uh also for myself to learn that uh it helps to uh focus yourself for a long time on one subject and really do a deep dive because then it it pays off and um people uh are somehow uh enticed by the adventure that ASML had for the last uh 40 years and they're excited to to hear what the company is about in the next 40 years. So well uh it it's kind of a surprise but uh it's also a reassurance that I uh well I picked the right subject. And I think you actually just uh mentioned something real interesting that I'd like to follow up with you on you know that I'm gonna ask you about serendipity but maybe if you could just reflect on the role that serendipity has played in the life of ASML um well let's start with uh one of the lead figures uh in uh in in in the book or would you call it protagonist or what's sure yeah yeah protagonist okay so if you look at the protagonist in the book Martin van der Brink uh the the CTO of the company um when he started out at ASML it wasn't it wasn't a sure thing because actually um he went to his job interview at Phillips uh back in 83 and he turned down their offer uh so things had had gotten uh could have gotten uh a completely different uh uh way but um then the uh director of uh Philips uh showed him uh the brochure of the new machine they were making and uh in in a new company wasn't named ASML uh uh back in the day and he kind of Martin van der Brink kind of fell in love with the concept of a machine that's um uh having um that's uh of a machine that's uh combining all physics uh chemistry um megatronics um you name it all all the uh the the the fun uh science uh things are uh in that machine so he fell in love with it and he kind of um uh started on day one and never left uh since sin since then so there was uh kind of a serendipity and you see that throughout uh all the the the the the the crazy stories in the book that ASML is um has been depending on luck and also uh uh huge huge risks and they're yeah well they had nothing to lose and they still feel that way yeah that's really cool really cool um and obviously the huge risks have paid off they're now a monopoly yeah and the outcome doesn't get much better for a business and and if you uh if I refer to uh your your uh earlier question about culture um I think there the biggest challenge and the biggest challenge of the new CEO Christophe Fouquet is keeping that startup uh uh feeling uh in a multi uh national company yeah without uh uh adding all these procedures and rules yeah how how do you how do how on earth is that possible though when you bring in on 800 people a month 200 people a week it's impossible to feel like we're in a startup you know we're the most valuable company in the Netherlands um maybe that's uh uh a bit about um um um I think uh maybe it has to do with the Dutch uh um uh culture so that that which is very much in ingrained into uh ASML that it's it's non-hierarchical a lot of the awful word a non-hierarchical so there's flat it's it's a flat uh organization so you're really supposed uh to challenge everybody even challenge your boss's uh uh boss or manager because um that's the way you uh uh get the uh the the fastest result and really create a super cool like a true meritocracy yeah it's a meritocracy and you also create a like a collective brand that that's uh uh sorry uh you're creating a collective brain that's um kind of working in parallel just like a human brain and um that's that's I think that's the thing that that's driving ASML.

SPEAKER_00 1:03:50

But how do you how do you escape the bureaucratic slip where you end up just working under a professional manager but not someone who can produce anything themselves?

SPEAKER_01 1:04:01

I I think uh the way uh people at ASML usually solve problems is just by shouting louder than your their other colleagues okay yeah so it's really an exclamation culture so there's a lot of but the thing is the which which I which really stands out to me is that it that might work if everyone was Dutch uh or one culture but they're not it's Indians it's Australians it's UK it's Brazilians it's Mexicans it's Spaniards like all these people together I mean you throw a Swede into the mix nothing would make a Swede more uncomfortable than you know sort of screaming your opinion at them yeah well that's uh you make a good point there because uh uh I interviewed um uh one of their former HR bosses which uh and he used to work for Volvo in Sweden and he also mentioned that it was the like uh the biggest difference he ever saw uh at a company um so uh the Swedes really uh are very polite usually not not very direct they don't say what they mean that's yeah okay and and the Dutch uh are the opposite of that so uh entering uh ASML for uh somebody who's not Dutch um is kind of uh like landing on a different planet and um so the challenging atmosphere is is very uh Dutch but the new employees also uh well they they get training of course and they also um um the company has changed uh a bit since the I think uh start of this decade so early uh 2020 they um um uh started using uh the triple c so not only challenge each other but also uh uh collaborate and care and um so these are like common values you would say um but uh in a in Yeah, it's well, it's a tough spot uh to be in. I I I don't think that uh everybody would uh uh thrive at a company like ASML because it's a very very direct culture.

SPEAKER_00 1:06:17

And so do they then have a high turnover since it is such a like direct, strong culture to join into? Do you find that a lot of people just immediately reject the idea of it and therefore there's a higher turnover? No, 800 people a month is crazy growth. There must be people leaving at the other end of that.

SPEAKER_01 1:06:34

Uh yeah, but okay, these 800 people uh that it was during uh their um the steepest uh uh sorry their their fastest uh period of growth um back in 21 and 22. So but then they were hiring like crazy. Yet the company attracts a certain kind of people, so people that are attracted to the uh technology. And um uh I asked this question uh uh uh I also asked uh Martin van der Bink uh if he thinks that uh ASML's culture might change when he when he's leaving, when he as it just stepped down. Um and he said this company attracts a certain kind of people that likes that like to solve a certain kind of problems, the real hard problems. Um so uh if you look at their turnover, I think it's actually pretty low. The latest numbers I heard were that 95% of the people are still working there after a couple of years, so which is uh it which is high compared to uh other technology companies.

SPEAKER_00 1:07:40

Yeah, oh totally. I suppose if you were just in Silicon Valley a very talented software engineer, um and there isn't much implicit meaning in, like you said before, creating a new filter for Instagram, you will just go to where the next highest paycheck is an equity payout. But perhaps if a company like ASML can give the call to action, we will present you with the hardest problem in the world. You'll work with the smartest engineers in the world, and you might produce a pretty cool machine on the other end of it, it's significantly more compelling.

SPEAKER_01 1:08:11

So you add more value to the world. I I think that's the that that's the secret of uh the attraction of ASML. But um also the uh the thing you mentioned about all these other nationalities uh that have to get uh grow accustomed to the Dutch culture. I think um if you look uh at ASML uh abroad, so I visited the the the uh the their divisions in in, for example, Taiwan, um and Asian culture, uh ASML is very refreshing. Um so in Asian culture, it's usually not uh very uh uh considered very polite to uh criticize uh to criticize your manager, but at ASML people are really uh uh experiencing more freedom than then uh their well then local or national culture. So it's it's kind of uh wow, they get room to breathe now. So that's I think that's uh uh really a good thing uh for uh that's attracting new talent to ASML. And if you look at the US culture, um there's this kind of winner mentality uh uh uh in the U in American uh corporate culture that you're not supposed to uh acknowledge your mistakes or um uh present yourself as your most vulnerable. Um and you're and that's something that ASML acquires that that you you need a require. Sorry. Uh ASML requires people to be uh honest of things they don't know of their mistakes, so other people can help. And uh if you compare that to um uh the other companies from the US, then usually uh uh people have a hard time uh admitting they might be wrong or that they might miss a deadline, so there's this a need to seem perfect. Well, that's not uh true, an ASML thing.

SPEAKER_00 1:10:14

That's nice. All right, Mark. Well, last question for you, mate. Uh, the role serendipity is played in your life.

SPEAKER_01 1:10:21

Yeah, okay. Can you uh I think uh this uh whole book was a was an act of serendipity, so I just uh it came across my path. It was never my dream to write a book or to uh uh dive into the semiconductor industry. Uh I just um uh was a financial reporter, technology reporter for NSA, and uh ASML was one of the companies I covered. Then Corona uh just uh happened, and I happened to uh dive into the company even more. There was some kind of fit uh between me uh and and and the company, and then a publisher called me and said, Hey, isn't there a book uh in it? After I published uh a story uh in uh in a newspaper. And uh so so I reached out to ASML said, Okay, are you willing to do a book? Okay, and and just as as as we uh had uh well just negotiated, nothing in writing, just uh uh uh explaining uh what I was uh uh going to do. Then the chip shortage happened, the geopolitical tensions started to happen. Great timing, and uh which was all uh a matter of serendipity. So I never planned this to happen.

SPEAKER_00 1:11:41

Absolutely. And now it's produced uh, you know, a book in itself is a great attractor for more serendipity. Um you do you see already serendipitous things happening to you because you've now published the book and maybe your profile is growing because of it?

SPEAKER_01 1:11:57

Yeah, uh well, if I look into the future, it's uh kind of a blur now because I don't know what's uh what's gonna happen. Um as a journalist, you you're usually uh uh at the other side of the table. So uh I'm the one asking questions and now I have to answer them. And um I uh uh I'm adjusting to that role because I realize that I uh I think it's a nice story to to uh to tell everybody, uh how a small country like uh the Netherlands in a small town build a huge company out of nothing or practically nothing. And um for me it was also there's this serendipity uh if you look back at ASML's history, that all the geopolitical tensions, the first seeds were planted way back uh um in the 90s already. Uh so it's it's not only a history of the company but also a history of um uh how geopolitics has changed over the uh the years, and ASML or its technology has always been in the middle of it. So EUV was uh American thing, and the discussion of uh licensing ASML, the uh uh EUV technology was a very geopolitical decision way back in what was it 1999, even before the uh uh 2000s. And but with it was a different time. Uh we thought the West had won, which is uh turned out quite different.

SPEAKER_00 1:13:34

Mark Dunkevel. Thank you for being so generous with your time.

SPEAKER_01 1:13:38

And thanks for your time and patience.