Good morning, everyone, and welcome to this press conference of the Q4 and full year 2024 results at ASML. My name is Monique Mols. I am responsible for media relations at ASML. You are here at the auditorium in Veldhoven, our headquarters, and we are here to talk you through our numbers, our developments, what we see in the future. We have a couple of journalists here in the room. Welcome. Thank you for traveling to Veldhoven. Very nice to see you. It's always nice to see people in person. But we also have some people that are watching online, and a very warm welcome to those watching from Asia, and a Happy New Year, and hope the Year of the Snake will bring good luck and fortune to you. So thank you very much for joining in this special day. We appreciate the effort you're making.
The people online, if you have a question, you can actually use the menu on the website. There's a screen that allows you to fill in your question, submit it to us, and we'll take it after the presentations, and for those who are here, just raise your hand when you have a question later on in the Q&A, and we'll get to you. I'm not going to do the presentation for all the numbers, and I realize I forgot my clicker. Wait a minute. I have some help. We've got the CEO of ASML, Roger Dassen, who loves forward-looking statements, so Roger Dassen will talk you through the numbers, and you've seen him before. This year, for the first time, our CEO, Christophe Fouquet, will be here to talk you through the development and our progress in 2024.
So with that, I'd like to welcome Christophe to the stage. Thank you. And I'll hand this over to you.
Thank you very much, Monique. Good morning, everyone. Roger and I are very happy to welcome all of you here at ASML to spend a bit of time to reflect on the 2024, look at what we expect in 2025 and beyond, and of course, take some of your questions. So a bit more than a year ago, we defined 2024 as a transition year, and this has definitely been a transition year, and there's a couple of things I will be stressing in my presentation, which were very, very important in 2024. The first one is AI. I think you have all noticed how AI became important for this industry. I think there's not one day without all of you talking about AI, sometimes in very optimistic terms, sometimes more recently in more fearful terms.
But AI is definitely a major, major change for this industry, and we'll talk about that. And the second transition, you know that ASML is a technology company. I think you know that our success has come from the fact that we have tried to always stay one step ahead when it comes to technology. And 2024 also has been an important year, a very rich year in technology at ASML, preparing also for this technological transition we are going to see happening starting in 2025 with 2-nanometer logic, but also some of the needs of the AI chips, high bandwidth memory, high-power computers. So this has been also a major focus for 2024. Roger, of course, is going to talk about the number, about 24, 25, but I think you already realize that we finished the year very strongly. This is a new record year for ASML.
This is a very strong record Q4 quarter. We are very happy, of course, with those results. And I cannot not use this opportunity to thank the entire ASML team everywhere in the world for all the hard work and the commitment, basically, to help us achieve those numbers. You have noticed also we start 2025 on a good note. Roger will spend some time on that. But again, so far, a pretty good day. So let me go a bit into some of the key trends in this industry. I already mentioned AI. We talked about AI also in our Capital Markets Day in November. And what we explain is that in this industry, we used to talk about semiconductors everywhere, right? Because today, any major application requires semiconductors. Today, more and more, we believe that we will be talking about AI everywhere.
So semiconductors will be everywhere, but AI will be everywhere too. And the reason for that is that the opportunity that AI represents for the entire economy, for the entire humanity, is huge. And we are just at the beginning of that. What you see today is people starting to invest, basically, in hyperscale computers in order to develop their products. And those products will come over time to eat our market. And this is really a very, very important change. What you see here also is the number, the impact potentially on the GDP. We talk about trillions of dollars, which is huge. So this will also mean that we can expect over time that a lot of companies will want to play in this field.
I know some of you maybe were surprised in the last few days that a new company suddenly came and wanted to compete. Well, I think we have to get used to that because this is such an opportunity that there will be more and more players to grab this opportunity. We also explained in November that for AI to succeed, cost has to go down, and energy efficiency has to go down. We also explained that this will come only with innovation. I can promise you that in our industry, any innovation that has ever reduced cost has always created more volume and more opportunity. Again, we will see a lot of that happening, hopefully, in the next few months. What does it mean for us? AI is going to be everywhere, and everywhere means two things.
You will see two types of different products. The very advanced one, the one that we talk about every day, which is going to require very high technology, but also what we call mainstream semiconductors because you need sensors to feed, basically, AI with data. And what we see happening on our market is that both the advanced technology segment is growing. We see more and more appetite, in fact, for advanced technology in logic, in DRAM. This is, of course, a good thing for ASML. But we see also a major increase in the volume of semiconductors overall. So this is mainstream. So this is a bit less sexy. So people talk less about it. But this is, of course, a very, very important part of our business. So this will continue to drive our entire portfolio.
As I will explain later, that's also why we are continuing to invest, basically, in all those technologies. Advanced one like EUV, less advanced one like DUV. If we look at the number, we still talk about a trillion-dollar opportunity in 2030. There's a lot of debate on how big this could be. Is it bigger than one trillion? How much bigger? I think we are not really entering into this debate. This is already a big number. We're going to drive that. Of course, our business will scale with whatever this number will scale with. We're still looking, of course, at a very significant opportunity. The only thing we don't know exactly today is how this is going to happen, right? We see the investment of everyone in AI. We don't know exactly how this will play out into applications.
So when do you start yourself to buy a mobile phone with AI, a computer with AI, a car with AI, et cetera, et cetera? So that's the part I think we have to still understand. We're going to watch what's happening this year, next year. And this is why we're also still, in our case, sometimes people say a bit cautious on the long-term view because we don't have yet all the understanding, again, on how the AI opportunity will really play out in the coming years. Now, this being said, I think you have seen this slide in the past. This is showing, basically, the EBIT, total EBIT of our ecosystem, all the people that are being playing, investing in this industry. This is huge. And this has been growing more than 13% year- on- year for many, many years. And what does it say?
This says that, basically, this industry has the means to continue to fuel innovation. In fact, you most probably don't have any other industry that has this kind of means to keep driving innovation, so that's why we always show that because this is still good news. There are means for innovation, and as you see it here, the return is very good, and people are investing on not only what's happening next year or the year after. They invest for what's going to happen in five and ten years, and as you know, we do the same in ASML because we have to look, basically, not only to 2025. We have to look even already at what may happen in the next decade, so this is good news. If I summarize a bit the future, it's bright. We're still very bullish.
I think if anything, AI has strengthened the opportunity for this business. So AI is going to be a key driver starting next year. But on the short term, it's also created a shift. And that's what we have also experienced in 2024. Some of our customers have done extremely well with AI, in fact, so well that they cannot produce enough chips to supply to their customers. And those customers have been driving our own demand, and they will continue to do that in 2025. But you have also seen in 2024 that some of our customers were not doing as well because they were not yet, basically, on this AI opportunity. And this has created a bit of this change also of dynamic last year. The second thing, AI has a couple of challenges: cost, energy efficiency. To improve that, innovation is needed.
One of the innovations is Moore's Law, and we expect an acceleration of Moore's Law. We also expect that the market will have a higher product mix towards advanced logic and advanced DRAM. And all of that is good for ASML. All of that will contribute to increase, as we explained in November, the need for the most advanced lithography machine, EUV, DUV. And this will contribute to, basically, increase the size of our business. And this is why when we look at 2030, our expectation in terms of revenue is that we will be somewhere between EUR 44 billion and EUR 60 billion. In terms of gross margin, we are looking at a range of between 56% and 60%. So those are good numbers, again, resulting from this opportunity. Now, you know that I'm an engineer still.
I still spend quite some time on technology because technology, at the end of the day, is what feeds our success. This is what prepares, basically, the opportunity for the next few years. And as I said in my introduction, 2024 has been a great year for technology. We have released many important products on EUV. So we talked quite a bit about the latest Low-NA product, the NXE:3800 , which provided almost 40% improvement on productivity. That tool was released to customers at the end of last year. We've been shipping quite a few of those tools. And those systems are now being qualified. They will be a very important part of the ramp of the advanced technology in the next few months. So this is very important. We have demonstrated the full specification at ASML, a major step, again, in terms of performance.
We are working now with customers to mature the platform. High-NA has been big news. High-NA is a program that we started in 2014. So you can imagine that when at the end of Q4, we get our first two customer acceptances, this is a major validation not only of the value of the technology we bring to the customer, but also of the investment we have made because this is millions of hours of work, a lot of innovation. Therefore, we are very, very happy that the feedback from our customers on High-NA is positive. They like what they see. Now we're going to work with them to define exactly when, how, and in which extent to introduce High-NA in the future. High-NA is going to complement very nicely low-NA to drive advanced technology manufacturing moving forward. Now, we talk always a lot about EUV.
DUV remains a very, very important product. Most of the wafers today are exposed today with DUV. EUV is a very small part of the lithography layers, and you see it here. The number of exposures on DUV is going to continue to grow between now and 2030, so there also, we need innovation, also because DUV being a bit of an older technology, maybe the competition around those products is getting higher, so we're continuing to invest, and here also, we come with major innovation on our KrF product line with the first shipment of the NXT:870B, which allows us to get to more than 400 wafers per hour. So giving productivity. Productivity for customers means lower cost. Lower cost means for ASML that the customer will be more willing to work with those tools because they provide a cost advantage.
And we have also shipped our latest generation of immersion. So immersion, if you want, is the most advanced DUV technology before EUV. There also, the NXT:2150 was shipped to an advanced DRAM customer, opening the era for, again, higher performance, higher productivity also on this tool. Finally, if you remember, a few years ago, we started our application business, our holistic lithography business. We make still a lot of progress there. We grow the size of this business every year. If you look at the year-start install base, we have more than 1,000 systems right now in the field, which is, of course, very significant. But the one thing I'd like to focus a bit more on today is e-beam. If you remember, we acquired HMI a few years ago, a Taiwanese company doing e-beam.
The reason for that was to bring fast e-beam technology to the market, what we call multibeam. This year, the other big news I will say on technology, together with the first customer acceptance for High-NA product, was the first customer acceptance of our multibeam product by several customers, which is, again, a validation of the technology, the investment, but also the need moving forward of this tool. You are most probably a bit less familiar with this, but these are also some of the very important technology, important products we are developing for our customers. We are also very happy to achieve good results on that last year. Now, we have to prepare the future with just more than technology. People is a strong focus in ASML.
I think that when Roger and I look at the assets of the company, I think people always come first because they are capable year after year to create all those technological miracles in R&D, but also bring them to customers and make sure that customers can really use them in high-volume manufacturing, so investing in people remains absolutely a priority for ASML. And we do it differently than we did it in the past because we have a long-term view, and we invest, I would say, earlier, so what you see here are a few examples. We are getting more and more interns in ASML.
So we are opening the doors of ASML to students as early as possible so that they get an idea of a bit of what we are doing and that we make sure that ASML is high on their list when it comes to choosing their next job. We have been focusing on diversity. We will continue to focus on diversity. We are very happy that over time, the percentage of women in the company has grown. It has been growing every year for the last 10 years. It continues to grow. And in order to do that, we have to work very early on the pipeline. So we have to go basically also to school, university, to make sure that right there, women also find an interest in ASML.
ASML used to be seen as a company that is a bit rough, and you only have guys, engineers working there. I think this image is changing for good reason because this is, of course, a very open, very interesting place to be for anyone. And finally, we're reaching out to university. A lot of you, at least the local journalists, are very much aware of the cooperation we have with TU Eindhoven, and many other universities in the Netherlands. But we're also extending that basically to a very global network of universities. There also, it's a way to invest in the future, and I would say a way to support, in a broader term, the entire industry because you understand that not all those students will come to ASML, but they may end up supporting our customers, our suppliers, et cetera, et cetera. And that's also very important.
So more and more focus is paid on that. I think this is really, really a key element of attention for the leadership in ASML in the recent years. In addition, we are growing. I want to show you the number for 2030. This is clearly calling to growth, growth in people, growth in capacity. And as you know, we are a global company. We have a very strong presence in the Netherlands, of course, where we have the core of our R&D, the core of our manufacturing. But we also have activity in the U.S., elsewhere in Europe, Germany, for example, and in Asia. And as we grow our business, those places will grow with us. So you will see, I mean, if you drove around for some of you this morning, you see that there's still a lot of construction happening here in the campus.
You know about our plan of the second campus in Eindhoven. But this also extends, again, to other parts of the world. And this is in preparation to support basically our growth. So still a lot of activity here. When we look at our ESG sustainability agenda, I think we started very early on to drive this agenda long before we had CSRD and all those regulations coming our way. The reason for that is because we believe this was the right thing to do. We believe it was important to reduce our energy consumption, reduce our impact on our environment through some of the logistic activity we have. We believe, I said it already, that our people are at the core of ASML's success. We developed basically a lot of activity to make sure they feel well in ASML.
I think that we continue in terms of governance to have the ambition to be a leader in this industry. So we want to lead by example. So to give you a few examples on environment, we have reduced our EUV consumption per wafer by more than 54% since the first EUV machine, which is enormous. And we're still working to get another 40% off in the next few years. So we're actively making sure that the amount of energy we need on the EUV machine to expose the wafer goes down. We become a bit brave. We first shipped our first DUV system on a boat, which I can tell you 10 years ago was out of the question in ASML. But we are learning how to do things differently and create new opportunities. And we shipped our first DUV tool, our first YieldStar system as well.
In the future, we plan to do a lot more than that in coordination with our customers, suppliers, et cetera. On social, so we have a lot of activity with the community. We are very proud of that. More than 57 projects and more coming. We're spending time, sorry, on housing, on infrastructure projects, of course, here in Brainport, but not only. Everywhere, we have a strong presence. We believe this is important to do that. We're also very happy that our employees are contributing themselves. We have a strong gift matching program, which is getting also quite some success among our employees. We are very thankful that they do that. I will skip governance. We get a pretty high score on sustainability. I think I'm more hot on what we do than the score we get.
But I think the rest of the industry is also recognizing a bit this effort. One last thing to end, because I think it's important. This is about, again, the community. I talked about housing. We try to do more to really, I would say, increase, improve the experience of the people around us. We had many discussions about growth. And one of the things Roger and I believe, the rest of the leadership believe, is that we want to grow, but we never want to be a burden for the community. And in order to do that, you have to find the right balance and, of course, proactively contribute to make sure that housing, school, infrastructure are available. That's, of course, discussions that require strong support from local and global governments. So we have those discussions as well. But that remains important. Education, STEM. We start very early.
That's why you see a picture of very young kids. We believe that you have to start as early as possible to explain to children why this is a great world, why this is a positive world, why ASML, why this industry is positively contributing to society, to humanity. We help also people who are working on other ESG innovation, locally again, but also beyond that, and well, you may have seen in our social media also this year, we have quite some activity when it comes to employee giving, so we talked about the gift matching program, but we have also a lot of people volunteering basically every day, many hours in order to engage with the community, so this is a bit who we are today. Roger is going to now give you a view of the number. It's also exciting.
So with no further delay, I'd like to introduce Roger.
Thank you, Christophe. It's good to hear the CEO say that numbers are exciting. And they are. And that's what we wanted to, what we want to share with you. So Christophe said it. If we look at Q4, the fourth quarter, it was a very strong quarter. And it was a very strong quarter. It was a quarter where I think the company really ran on all cylinders. It was a strong quarter based on the top line. So if you look at net sales, it was a record quarter for ASML, exceeding our guidance from that vantage point. It was a very strong quarter from a gross margin perspective, particularly held by the fact that we had very strong install-based business.
That was one of the reasons why revenue was up, but it was also an important driver of the gross margin. It was a very strong quarter because we had a very strong order intake of EUR 7.1 billion, which is a very good start, I think, for this year. And it's a very strong quarter because we had very significant cash generation in Q4, if only to make the CFO very happy. So all in all, I think it's a quarter that ticked all the boxes. And it was a quarter that indeed meant that we could end 2024 as a really strong year. If you look at what we said at the beginning of the year, we said, and Christophe made reference to it, a transition year. We said we expect the business to be approximately similar to what we had in 2023.
Indeed, we were able to get a little bit better than 2023 with a top line of EUR 28.3 billion, gross margin of 51.3%, also a little bit better than what we expected at the beginning of the year. Strong net income, as you see here, EUR 7.6 billion, an earnings per share of EUR 19.25. We were able to return EUR 3 billion to shareholders either through dividends or share buybacks. We'll come back to that in a moment. All in all, I think a pretty strong picture for 2024 and more or less in line and actually a little bit better than what we anticipated at the beginning of the year. If you look at the different businesses one already, Christophe highlighted some of the key elements in there.
If you look at the numbers for EUV, EUV sales, a little bit down, 9% down to EUR 8.3 billion with 44 EUV systems, both NXE and EXE. Also there, it was a transition year. It was a transition year because we did introduce after 10 years, as Christophe explained, after 10 years of working on High-NA, we were able to get the first High-NA tool really to the customer, get it installed at the customer, and get a customer to accept the tool. So I think that was an important part of the transition. We were able to bring the 3800 into the field and indeed get customers excited about that tool as well. That is a tool that really drives the performance of EUV Low-NA to the next level. We used to look at an EUV tool with 160 wafers per hour.
This one, a final configuration, gets you to 220 wafers per hour, so also that, I think, is the biggest step that we made in EUV since the inception in terms of improvement of productivity. So from that vantage point, being able to bring that one into the field, I think, is a major step as well. So EUV, a little bit down in terms of total business, but really a transition year that is positioning us very well for the future. DUV grow a bit, 4%, as you see here. And also there, we introduced important new products into the market with the 2150, Christophe showed it, and the 870B into the KrF system on the NXT platform. Major, major breakthroughs from a technology perspective there. Applications grew very nicely with 20%, both YieldStar and Multibeam in particular, contributing to that growth.
There, I think, and Christophe again mentioned that the fact that we were able to get sign-off from multiple customers on Multibeam, again, a project that we've been working on for many, many years, I think, is a milestone for the company. Finally, installed base, growing very, very strongly at 16%. And this is both on the service side and also on the upgrade side. So all in all, as I mentioned, 2024, quite in line with what we expected at the beginning of the year and actually exceeding that a little bit. If we then move to the composition of our sales, a few things that immediately you will recognize. Immersion very, very strong. You saw that in the growth number for DUV, but clearly driven by the immersion business, making up 44% of our system sales. Also clear that memory is very strong.
And there, I think the importance of high bandwidth memory to make AI work, I think, is clearly recognized. And I think that filtered through also in the order. So memory quite well recovered, I would say, from 2023. If you look at the regional mix, obviously the thing that catches your eye is China. China, very strong at 41%. Already quite strong, I would say, in 2023 at 29%. You know the story. We've said it before. All the way through 2022, our order fill rate for China was quite low. We had an order fill rate for China of around 40%. So we could only give to China 40% of what they were asking for. So as a result of that, the backlog for China had been filled all the way through 2022 massively.
What you see here is that in 2023 and 2024, with the capacity expansions that we've been carrying through, we were able to execute on that historically built-up backlog. As we mentioned before, we do expect that on a going-forward basis, China will go back to more normal percentages of our business. And that means lower 20% is what we expect them to contribute to our sales this year. And that is commensurate with the representation of China in our backlog today. So those are just some of the headlines if you look at the composition of our net sales. I think this gives you a similar picture. Again, it shows very clearly the strength of the memory business, which has been nicely building up, I would say, over the past couple of years. But the real spike this year, given the significance of high bandwidth memory for AI.
This gives you the full P&L, if you like, for the consolidated statements of operation, and again, as I mentioned before, I think very much in line with 2023, so most of the numbers are quite comparable. Although you would say that you would see that on the R&D side, we keep on pushing the pedal to the metal, so it's very clear that with all the projects that we have, with the R&D pipeline that we have, with the very strong ambitions that we have also in the medium to long term, and the good ideas that we have, we keep on investing in R&D because I think that is something that we will be able to very firmly capitalize on in the years to come.
In terms of net bookings, I just want to highlight that because net bookings has captured the headlines in a few quarters last year and not always in a positive way. But just if you look at it from an annual perspective, if you put all those quarters together, you see it wasn't too shabby, was it? I mean, you're looking at an order intake for 2024, more or less on par with the order intake that we had in 2023. So as we mentioned at the beginning of the year, a year quite in line with what we had in 2023. A bit of growth and the fact that we were able to sustain the gross margin in spite of the fact that we had the first High-NA tools recognized, I think, is a strong achievement. Cash return to shareholders.
What we proposed to the AGM is to declare a total dividend of EUR 6.40 per ordinary share. That would be a 5% increase over the total dividend for 2023, and I think that's very much in line with the policy that we have, a policy of growing dividends. Whatever cash remains after we've used the cash in the business, whatever excess cash is there, and after distribution of dividend, will be used in terms of share buybacks. Share buybacks this year were fairly low in comparison to some of the previous years, as you see it here. We bought back shares for approximately EUR 500 million. That is related to what I mentioned a bit earlier. Q4 was very, very strong in terms of cash generation, but it really made up like more than 90%, I think, of the total cash generation for the year.
So the cash came in quite late. So at the end of the year, we're actually looking at a very high cash balance. But as a result of that, we chose not to do too much share buybacks in 2024. In terms of outlook, so starting with the quarter, starting with Q1 2025, we expect net sales between 7.5 billion EUR and 8 billion EUR. Installed base at 2.1 billion EUR is what we expect for the first quarter, which is on par with the very strong installed base sales that we already showed in Q4. Gross margin for this first quarter, between 52% and 53%, a little bit higher than the guidance or narrower, if you want to define it in that way, than the guidance that you see there for the full year.
That is also related to the fact that in Q1, we will not see revenue recognition for High-NA. So High-NA will be more skewed towards the second half of the year as a result of that, a slightly better gross margin in Q1. R&D, approximately at the level that we saw for last quarter. And SG&A, I think well managed, even a little bit below the level that we had in last quarter. And then for the full year, we're looking at total net sales between EUR 30 billion and EUR 35 billion. So this is in line with what we discussed in Q3. Gross margin between 51% and 53%. So you do see a good increase there in comparison to previous years. And an annualized effective tax rate of around 17%. So that is the guidance that we have provided at Q3 and that we're reiterating actually today.
I would say all in all, a very strong start of the year. Typically, what you would see is a very strong Q4, a not so strong Q1. I think what you just saw was a pretty strong Q4, pretty strong Q1. I think that's good momentum. Overall, and looking forward a little bit beyond 2025, what do we expect? I think Christophe gave you the foundational comments for how we look at the business. What is very important, and you saw that also in the R&D number, we continue to invest in R&D in a big way. The roadmap that we have is quite broad and quite deep. We do believe that it sets us up for capturing the opportunities that the markets provide.
The market, as we see it and as many people see it, the market by 2030, it could be a market, a semiconductor market that could exceed $1 trillion. So very significant growth in comparison to where we are today. And obviously, as Christophe said, AI is a very important driver of that growth. And we believe we are well positioned to, with our products, capture the opportunities that such a growing semiconductor market actually provides. And in the capital markets today, we talk quite a bit about our roadmap there and about how we see a very significant share of ASML and an increasing share of ASML in the fabs of our customers, both on the logic side and particularly also on the DRAM side.
In terms of what that pans out, how that pans out and what that leads to, reiterating the numbers that Christophe just gave you as well. So we see growth scenarios leading to a market in 2030 for us, or a business for us by 2030, somewhere between EUR 44 billion and EUR 60 billion, and a gross margin between 56% and 60%. I would say both good optics from where we are today. And finally, reiterating our capital allocation and financing policy, which hasn't changed. So again, what we need for the business, we actually use in the business. And you saw some of the projects that were driving. So we have the R&D efforts, but also the CapEx efforts that were very visible and tangible on the slide that Christophe presented.
We have a policy of growing dividends, and then whatever remains will be shared with our shareholders by means of share buybacks, so that's the longer-term perspective. I think it's looking bright, and it's fueled by many of the dynamics in the market that Christophe talked about and many of the very strong ideas and projects and products that we have on our roadmap. With that, I would like to hand it back over again to Monique, who's going to drill us on Q&A.
Yeah, just stay here.
There you go.
I'd like to invite Christophe as well for some Q&As. We are getting questions online, so continue to submit those if you're watching. Just a small reminder, we are only taking questions from journalists, so it is a media conference, and the conference call for analysts will be later today at 3:00 P.M.
I see a lot of hands going up, and Max has the microphone. So Max, can you start there? Yeah. With Max Smolka, can you say your name first and your publication so everyone knows who you are?
Sure. Thank you very much. It's Klaus Max Smolka of German newspaper, Frankfurter Allgemeine Zeitung. The elephant in the room this week is DeepSeek. What is your expectation? What's the effect of DeepSeek-like technology on ASML? And in particular, is DeepSeek technology going to jeopardize your sales in terms of the most advanced high-technology machines? Thank you.
Yeah, Christophe.
Yeah, so I think a lot of this question in the last couple of days. So first, I'd like to say we're not model experts. So I don't have a strong opinion on how good or bad is the new model. So I'll leave that part aside.
From a more general point of view, we believe that AI success, AI expansion has two main challenges moving forward: cost. Say differently, today, AI is too expensive to be able to be rolled out to many, many applications and power consumption, and in order for AI to be everywhere, we need to see major progress on cost and energy consumption. That's the first thing. The second thing is, if you look at this industry, lowering cost has always contributed to more volume. This is basically the fundamental of Moore's Law. Moore's Law is about reducing cost every single generation so that you create more opportunity, so I would say in general that any technology, whatever it is, that will contribute into a cost reduction overall of AI will increase the opportunity. This has been true in this industry for the last 40, 60 years.
I think it's still true moving forward. That's the first thing. So if you ask me for us, anyone that lower cost is in fact good news for ASML because lower cost means AI can be used in more applications. More applications mean more chips. And we are in the business of providing equipment to people who make chips. So that's the first thing. The second thing also mentioned, AI is a huge opportunity. Therefore, I think you should expect to see a few elephants in the rooms in the next few months or few years because everyone will want to be in. And I think the competition, especially when it comes to software, because I think this is where the threshold to access is the lowest, will be very high.
So you cannot have an industry with this amount of opportunity without the key player being, I would say, challenged. So I think you will see a lot more of those discussions happening in the next few months or years. So I think it's good news if I make it short. Again, I don't know exactly what DeepSeek can or cannot do. But I said again, anything that will drive cost down is good news for ASML on the long term. Yeah.
Thank you, Christophe. Kim here at Wall Street Journal. Thank you. Thank you. Kim Mackrael from the Wall Street Journal. Another DeepSeek question. Sorry about that. I'm wondering, over recent years, export controls to China and to particular fabs have been tightened. Does DeepSeek say anything to you about the effectiveness of those controls?
What impact do you think controls have had on DeepSeek or on the ability to do something like that? Thanks.
Yeah, I think that any answer on that would be quite speculative. So I think I don't know. Again, there's a lot of opinion being given on how this relates to export control. I think it's hard to say. I think you look at, we talk about software. We talk about algorithm. We talk about people working to develop better algorithms than others. I think you're still pretty high at the surface of AI, if you know what I mean. I think the software part, the algorithm part, will be the most open field, most probably moving forward. So I will not draw any conclusion. I think you will see a lot of people trying to do that.
I think, of course, the topic being emotional, this is also a reason for that. But I don't know, to be honest. I think that would be highly speculative.
Yeah. Okay. Let's return to, I see you. Let's return to a question online from Le Monde.
Must be for you.
Yeah. No. What about the Beethoven plan of the government? Is everything okay with the promised investments? And will ASML stay definitely in the Netherlands? I think that's also a question, especially there's some speculation and some chat in the news about France. So maybe you want to address that first, and then we go to Beethoven.
Well, I think we've been very clear many, many times, and I'm going to try to be clear again. But this is the place we want to be.
And this is the place we want to be because we have developed our infrastructure, R&D, manufacturing here in Veldhoven. We believe that keeping those two close is very important for success. So we want to continue to grow here around Veldhoven. And this is why we had this basically agreement to look for a second campus, and we are finalizing that. So we want to be here. I cannot be more clear than that. I think we are working as a resource with the different parties to make Beethoven, our second campus, a success. We are deeply involved. I think we have developed a good working relationship with the municipality, but also with the Dutch Government because some of the things we need require attention.
And of course, it's a bit of work, but we need, as a country, as a region, to be able to do that because, of course, we are competing with many other places in the world where this doesn't sound like a big deal at all. So I think we have to believe that we can do that, convinced we can do that. And so far, the progress is good. I think Roger can say a few more words because he's involved deeply on an almost weekly basis to track the progress. But I think the intention and the ambition are remaining very, very strong. I don't know, Roger, if you want to.
Yeah, sure. We can say a few words on progress because obviously there are challenges, right?
I mean, if it's about nitrogen deposition or it's about the grid or it's about the execution power in municipalities, which recently drew some attention, and it's all true. But what is very, very clear is that there is a very strong commitment of our government and government at very different levels, be it at the national level or be it at the local or regional level. There is a very strong commitment to making this happen, so that is what works for us and then at the end of the day, there's this collaboration and at the end of the day, making sure that the right resources are working on the right projects and getting stuff done in time and we have all the confidence that the things will be executed in time such that we can execute on the expansion as we have planned.
Okay. Thank you.
Çağan Koç had a question from Bloomberg. Thank you. Hi, Çağan.
Does this work? Can you hear me?
No.
Hello?
Yeah. Perfect.
Two questions. One on DeepSeek. Well, the news on Monday sent tech stocks into a tailspin, especially Nvidia, but also ASML's market cap fell by more than $20 billion on Monday. So was this overblown, this reaction, do you think? And what's your kind of response to investors who question the need to spend billions on advanced chips and hardware on AI? And then the second question is, last week, Dutch Prime Minister Dick Schoof said that he expects the Trump administration to keep pushing the Netherlands for more export controls on ASML sales to China. What else can we expect on export controls this year? Do you expect sales of your older DUV machines to be affected?
Do you want to start, Roger?
I'd like to start the second one, and then I'll go back to DeepSeek. You want to? Okay. Yeah, well, on the second one, in terms of export controls, John, we're not going to speculate on that. I mean, let's be clear, export controls is not new. It's something that we've been living for many, many years, and I think in good collaboration with governments, we have been able to work it, and we'll just see what's going to happen there. There is enough speculation. We're not going to add to the speculation there. I would make one comment on DeepSeek, and then I'll give it to Christophe. We're not second-guessing the market. I mean, the market does what it wants to do.
But if you read the analyst reports on the tech stocks, I think there is an increasing voice that echoes what Christophe said, which is, at the end of the day, if you make it cheaper and if you make it more accessible, if you make it more energy efficient, it can lead to a democratization of AI in a way that is necessary in order for this to become prevalent. And that's what you see now in most of the analyst reports that I see. That is what is being echoed there.
Yeah. And I think what we do on AI, we try to do that also in November. This is a long-term trend, long-term opportunity. And when it comes to this topic, we always invite everyone to look at the long term. I think on the short term, we explain that.
So for us, for example, this created a major shift on the market. And this was not all good, as you have seen last year. I mean, some of our customers really struggled with the fact that this huge opportunity was coming. It was changing the game of the market. And not all of them did well, as you noticed. And we also say, I say that in my presentation, today, we don't know exactly how AI will build itself. Right? So there will be a lot of new players coming in. I don't think that you can define today who is the winner in 2030. I think there will be a lot of competition because, I say it again, the opportunity is huge. And to amplify Roger, we welcome in many ways that competition because this will go against cost, which will increase the opportunity.
It will go in favor of innovation. So I think the challenge is to be able to project yourself, not try to dream up what's happening in the next five years. And like I say, you will get a lot of those discussions because I don't think that DeepSeek is the last company that says, "Hey, this is a great opportunity. I'd like to be in." It's not like everyone is going to sit and say, "Well, we declare the winner today. They are X, B, whatever." No, people are going to fight for the opportunity. And I think it's a good thing. And well, now the question is always when it comes to market, how do you look at short versus long term? We run a business. We look at the long term because this is where the success, the opportunity is.
Thank you. Toby Sterling, Reuters.
Hi.
Yeah, Toby from Reuters. I'm not sure. Is there a logical contradiction there if you say it's good if the prices come down, but it's also good if there's huge demand for lots of AI chips? So I'm just not sure if it makes complete logical sense what you say, that lower prices will drive better volumes for ASML in any case. Maybe it's a wash, right? Second question, or I wanted to ask, you had a great comment, I think, three months ago when you said that outside of AI, the chip market looks pretty sad. And I'm wondering whether you see any change there with the rest of the market.
Yeah. So on the first one, on the cost relation to volume, let me try to put a bit of color on that.
Today, the level of AI sales is pretty high, but most of it goes to hyperscale, hyperscale computer servers. If you look at the number of chips related to that, it's pretty small, in fact. This is not the majority of chips. If it stays like this, then you only have people spending money in R&D to develop models, but you don't have, which is a big question in AI, you don't have the application that basically creates the return on this investment. So what we expect to come over the years is those applications. And those applications most probably will end up on a phone, on a car, in industrial application. You see a lot of new expectations on robotics, for example. That's one new area, of course, on PC, etc., etc., on what we call also edge AI. Today, those chips are not yet in demand.
We believe that when it comes to the size of our business, the demand will come more from the use of AI than the research on AI. And to get those chips, if you want to have AI chips on your phone, the cost of those chips has to be a fraction of what maybe the hyperscaler can pay for it. You cannot afford 25,000 chips on your phone, right? So this has to happen. And when this happens, which we see again most probably in the second half of the end of this decade, then AI will really drive volume. That's the relationship. If the cost doesn't go down, you may continue to sell a few very expensive chips and make a few people very happy. But you will not basically bring this technology to, well, let's say, to the mass, and the volume will remain small. Right?
So we have to get to that point. And we don't know exactly when and how, but what we expect is most probably this happens between now and the end of the decade.
Toby, it's a classic case of microeconomics, right, where it's all about price sensitivity. And I think anything related to tech in the past has demonstrated to be very, very price sensitive. So to the extent that you're able to drive down cost and prices, it more than compensates for in demand. I think that's what we've seen in the past. And that's the basis why, indeed, Christophe has been saying, starting last year, cost needs to be driven down in order for AI to be successful.
Yeah.
I think on the second question, Toby, I think what Roger and I went into many discussions on that in the last few months. What we really see is we used to look at when the recovery of the normal business happened. I think today we believe that what's really happening in the market is a real shift. So AI is changing quite a few things, whereas some customers, some of our customers are doing extremely well, some of them having to catch up. We see it changing potentially the roadmap of our customers because they have to develop basically processes that will be more targeting high-power computer or high-bandwidth memory. Roger talked about it. So there is a real change. There is a bit of a reorganization of the market, which I think we really started to sense in the second half of last year.
This came with a bit of turbulence. I think this is still happening because everyone is a bit lining up to basically go after this opportunity. This is quite a big change, to be honest. It affects quite a few parts of our business. That's, I think, the way we look at it today.
We still have a few minutes. There were some questions here. Can you start there? Yeah.
Okay.
We'll try to be brief and get to all of you.
Yes. I'm Huug Kiesbers, Dutch news agency ANP. I was wondering, since I hear you mentioning that innovation is so important also for the semiconductor industry to reach volume, is it a bit concerned that the world markets are fragmenting because of trade wars or the threatening of trade wars with the U.S. government mentioning tariffs almost every day also on semiconductors?
And will this in some way impact your long-term outlook?
So I think we can both. So the short answer, I think we've been consistent on that. I think this industry has developed on a very successful open ecosystem. The model where you decouple this ecosystem, where you try to have two, maybe instead of one, makes things more expensive, that's for sure, and slows us down. And I would say this is true almost for any economic matter. So that, I think, is still true. Does it necessarily affect the long term? So there you can get more tactical, where you say, "Well, now if I have two, then that could be more." So there's no clear picture. It depends a bit on how the whole thing evolves. But the biggest impact is on cost and speed of innovation.
If you have an open ecosystem, both will be better.
Okay. We need to speed it up a little bit. Bert? And Donald gets it. Sorry, Don, I'm trying to get to you.
Bert van Dijk, Financieel Dagblad. Just a follow-up question on Toby's question about the other industry than the AI sector. I was just wondering if you could elaborate a little bit on the auto market and the automotive market and the industry market. Do you see some kind of recovery happening there, or are you talking also about a shift in that industry as well? And the short second question is about the order intake. You're going to stop reporting on that from 2026.
I was just wondering, is that because you feel like that the investors are too much focused on that number that leads to swings in your share price, or could you elaborate a bit on why you decided to stop? Because it seems to be a step back in transparency, so.
So let me take both. On the automotive market, the relevance for us for the automotive market has two angles to it. On the one hand, it is the electrification, and on the other hand, it is ADAS, so autonomous driving or assistance and assistance systems.
Yeah. We're counting down quickly.
So those are the two there. I think electrification is clearly in a significant part of the world is not happening as rapidly as was anticipated. I think when it comes to ADAS, it will be heavily dependent on how regulation is going to develop.
But there we are more positive, and we do see that ADAS will break through. So yes, I think the automotive market globally right now is going through a bit of a dip with the electrification. I think long term, I think ADAS is still strong. When it comes to the transparency argument on what we're doing, I think at the end of the day, what matters to us is that we give investors the right information and the information that they can take the right decisions on. We give them guidance. We give you guidance every single year. That guidance is based on ongoing conversations with our customers. And we think that that guidance that we provide gives you a better indication of the momentum of our business than providing on a quarterly basis order intake, which we have been saying for quite a while is pretty lumpy.
So that's why we think in this case, over the longer term, less is more.
Sorry, we have to end this. So if you have any more questions, contact one of the media relations colleagues, and we'll get back to you. Thank you for joining. Thank you for coming to Veldhoven. Thank you for joining online. Hope to see you next year. Thank you.