Good afternoon. This is the Chorus Call Conference Operator. Welcome, and thank you for joining the ASM International's Fourth Quarter 2024 earnings call. As a reminder, all participants are in listen-only mode. After the presentation, there will be an opportunity to ask questions. Should anyone need assistance during the conference call, they may signal an operator by pressing star and zero on their telephone. At this time, I would like to turn the conference over to Mr. Victor Bareño, Head of Investor Relations. Please go ahead, sir.
Thank you, Operator. Good afternoon, and thank you for joining our Q4 earnings call. With me today are our CEO, Hichem M'Saad, and our CFO, Paul Verhagen. ASM issued its Fourth Quarter 2024 results yesterday at 6:00 P.M. Central European Time. For those of you who have not yet seen the press release, it is available on our website, along with our latest investor presentation.
As always, we remind you that this conference call may contain information related to ASM's future business and results, in addition to historical information. For more information on the risk factors related to such forward-looking statements, please refer to our company's press releases and financial reports, which are available on our website. Please further note that during this earnings call, we will reference profitability numbers primarily on an adjusted basis.
Reconciliation with the reported financial numbers can be found in the press release and in the investor presentation, and with that, I will now turn the call over to our CEO, Hichem M'Saad.
Thank you, Victor, and thanks to everyone for attending our earnings call. I'll start with a few of the highlights. We delivered strong results in the fourth quarter, with revenue growth of 27%. For the full year, our sales increased 12% ahead of WFE growth, marking our eighth consecutive year of double-digit growth. Strategically, we booked solid progress in 2024, as we expanded our addressable market, in particular with the transition to Gate-All-Around, and as we further expanded our R&D engagement with key customers.
I want to thank all our people for their relentless commitment and teamwork, which contributed to another successful year for ASM. The agenda for today's call is as usual. Paul will first review our financial results. I will then continue with the discussion of the market trends and outlook, followed by the Q&A session.
Before handing over to Paul, I'm pleased to let you know that we will have our next Investor Day on September 23rd this year in London. Please mark your calendars. We hope to see many of you there. W ith that, over to you, Paul.
Thank you, Hichem. A lso, thank you all for joining this call. Let me start with a review of our fourth quarter results. Revenue increased 27% at constant currencies to EUR 809 million in the fourth quarter, a nd this was at the upper end of our guidance of EUR 770-EUR 810 million. By customer segment, equipment sales were led by foundry, followed by logic, and then memory a nd combined logic-foundry sales accounted for the largest part of the total and increased strongly, both year-on-year and compared to the third quarter.
The increase was driven by Gate-All-Around-related sales. Memory sales in the fourth quarter increased year-on-year, but were lower compared to the elevated level in Q3. Memory was again driven by HBM-related DRAM sales. Power analog and wafer sales remained soft and were sharply lower compared to the fourth quarter of 2023.
Spares and service sales increased 49% year-on-year at constant currencies. The underlying trend continues to be healthy, with steady increase in our outcome-based services. Like in Q3, the above trend growth in Q4 was in part explained by accelerated demand in China. In 2025, we expect growth in spares and services to return to a more normalized level. Please also note that starting in 2025, we've changed the reporting definition of spares and services.
Installation and qualification sales will be classified as spares and service sales instead of equipment sales. This change is to align with the internal organizational structure at ASM, and further details can be found in our press release. Gross margin of 50.3% in the fourth quarter of 2024 was up from 49.4% in Q3 and 47.9% in Q4 of 2023.
The mix in the fourth quarter remained favorable, including a solid, though reduced contribution from China sales, both sequentially and compared to the fourth quarter of 2023. Net R&D expenses in the fourth quarter increased year-on-year by 34% at constant currencies. The increase in net R&D was higher than the increase in gross R&D due to accounting effects.
As explained in previous quarters, several development projects entered the commercial release phase in 2024, triggering the start of amortization of related capitalized development expenses. SG&A expenses at constant currencies decreased 10% from Q4 2023, reflecting ongoing cost control. Our new orders in the fourth quarter amounted to EUR 731 million, up 8% year-on-year. Orders dropped by 12% compared to the level in Q3, and this was in part explained by the effect of order poolings from Q4 to Q3, as already communicated last quarter.
Logic-foundry accounted for the largest part of our Q4 orders and increased significantly compared to Q3. Gate-All-Around orders increased to a very strong level, as leading customers are preparing for the introduction of 2-nanometer volume manufacturing in 2025. Memory orders decreased compared to the relatively strong level in Q3, but they were still up substantially compared to Q4 2023.
Power analog wafer orders dropped to a fairly low level, reflecting the soft market conditions in this segment, including the silicon carbide markets. Let's now have a look at the full-year results. At EUR 2.9 billion, our sales in 2024 increased 12% at constant currencies to a new record high. In terms of customer segments, our total logic-foundry sales increased slightly and continued to account for more than half of our equipment sales.
Gate-All-Around-related sales increased substantially in the course of the year and represented the larger part of our logic-foundry sales in 2024. Mature logic-foundry sales decreased in 2024, particularly in the second half of the year. In memory, our sales surged by more than 150%. This is compared to a fairly depressed level in 2023, when our sales dropped around 40%. DRAM accounted for the largest part, driven by very strong HBM-related DRAM demand.
3D NAND sales approximately doubled in 2024, but from a low base and represented a smaller part of our memory sales. As a percentage of equipment sales, total memory jumped from 11% in 2023 - 25% in 2024. Power analog and wafer sales dropped by a significant double-digit percentage, impacted by the ongoing cyclical downturn in this market.
The drop in 2024 is compared to a strong level in 2023, when sales in this segment almost doubled. Our silicon carbide Epi sales increased by a mid-single-digit percentage in 2024. This is below our previous projections of double-digit growth, but still a robust performance in view of the significant slowdown in this market. At constant currencies, our equipment sales increased by 9% in 2024.
We booked solid double-digit growth in ALD, which accounted for clearly more than half of our equipment sales. We've increased the amount in both the logic-foundry and memory markets. Our spares and service sales increased 29% at constant currencies in 2024. Over the last four years, since we started to roll out outcome-based services in 2020, we have increased our spares and service sales at an average annual rate of 19%.
Gross margin for the year further improved from 49.3% in 2023 to 50.5%, mainly reflecting mix, including a continued substantial sales contribution from China. For 2025, with the normalization in China sales, we expect the gross margin to return to the target range of 46%-50%. Net R&D increased by 22% at constant currencies, reflecting a further increase in our customer R&D engagements, as well as the af orementioned accounting impact of higher amortization.
As a percentage of sales, net R&D increased from 11.2% - 12.1% in 2024. For 2025, we expect net R&D to remain at the upper end of our target range of high single-digit to low double-digit at representative sales, or even slightly higher, as earlier communicated. SG&A expenses increased 3% at constant currencies in 2024.
The rate of increase in SG&A slowed down significantly compared to the previous two years, as the additional investments we made to strengthen our organization were largely completed in 2023. As a percentage of sales, SG&A decreased from 11.5% - 10.6% in 2024, and for next year, or for this year, for 2025, we expect SG&A to be at around similar levels and as a percentage of sales to decrease to a high single-digit percentage.
Operating profit increased 17% in 2024, with the operating margin improving from 20.6% - 28.0%. Results from investments, largely reflecting our holding in ASMPT, dropped from EUR 21 million in 2023 to EUR 10 million in 2024, due to the continued downturn in the backend markets. Now, turning to the balance sheet, ASM's financial position remains solid. We ended the year with cash of EUR 927 million and no debt.
Free cash flow increased 23% to EUR 584 million in 2024. This was mainly driven by improved profitability and deferred tax payments, and despite slightly higher working capital and higher CapEx. Working capital increased from EUR 425 million -EUR 447 million, with higher accounts receivables, partially offset by higher contract liabilities and accounts payable.
In terms of number of days, working capital dropped to 50, down from 60 at the end of 2023, and for 2025, we expect to return to the guided range of 55 - 75 days. CapEx increased to EUR 168 million in 2024, up from EUR 154 million in 2023, and was within our annual guidance of EUR 100 -EUR 180 million. In 2024, our CapEx was primarily focused on R&D infrastructure, including a new Korean facility scheduled for completion in 2025 and the start of the investment program for a new R&D facility in Scottsdale, Arizona.
2025 will be a year of continued investments, as we further advance the upgrade plans for our R&D facilities. Regarding our cash flow outlook in 2025, it is important to note that the anticipated earn-out payment of EUR 100 million for LPE has previously disclosed at the time of the acquisition in 2022. In terms of shareholder remuneration, we spent EUR 287 million in cash on dividends and share buybacks in 2024.
With our Q4 press release, we announced a new share buyback program for an amount of EUR 150 million, as well as a proposed dividend of EUR 3 per share, up from EUR 2.75 in the prior year. With that, I would like to hand back to Hichem.
Thank you, Paul. Let's now review the trends in our markets. After the drop in 2023, the semiconductor end market recovered by 18% in 2024. This growth was, however, uneven, with mixed dynamics across the different segments. AI has become the main driver of the semiconductor market, significantly boosting data center growth. In 2024, other high-volume end market segments, such as PCs and smartphones, remained relatively sluggish due to slow customer spending, economic uncertainty, and ongoing geopolitical tensions.
This picture in the end markets was also reflected in the spending patterns of our customers, with good momentum in advanced technologies that enable AI, in particular, Gate-All-Around in logic foundry and High Bandwidth Memory, but softer trends in other parts of the market. In total, Wafer Fab Equipment spending was up by a mid-single-digit percentage in 2024. Our leading-edge logic foundry business accelerated in 2024.
In the first part of the year, sales in the segment were still relatively slow, following the softer market condition in 2023. In the second half of the year, leading-edge logic foundry sales increased significantly, driven by investment in the new Gate-All-Around 2-nanometer technology. That momentum continued in the fourth quarter, with a further rise in Gate-All-Around orders and sales.
The tools we are shipping now are predominantly in support to our customers' high-volume manufacturing ramp in 2025. The 2-nanometer node offers significant improvement in device performance and power efficiency. Some of our customers have reported that they are seeing substantial demand for 2-nanometer from their own customers, including AI chip makers, and that capacity for the 2-nanometer node is expected to be larger than for 3-nanometer.
The transition from FinFET to Gate-All-Around has increased our addressable market with $400 million based on 100K wafer starts per month capacity. In addition, we believe we successfully maintained our leading share in ALD in this transition, and we strengthened our position in silicon Epi. While our total silicon Epi sales decreased slightly in 2024 due to the downturn in the power and wafer segments, sales of our reduced-pressure Intrepid ES platform increased substantially, reflecting our growing share in Gate-All-Around applications.
There have been some further shifts in CapEx forecast among the customers in the leading-edge logic foundry segment, but overall, our forecast for a substantial increase in Gate-All-Around-related sales in 2025 is unchanged. During 2024, we also expanded our R&D engagement with key logic foundry customers for the next version of Gate-All-Around, including the 1.4-nanometer node.
We expect innovation, such as backside power delivery, widening adoption of metal ALD, and selective ALD to drive further increases in ALD intensity. Our memory business delivered an exceptional performance in 2024. High bandwidth-related DRAM has become the main driver of the memory market. The high-performance DRAM devices used in these HBM stacks require high-K metal gate ALD technology, in which ASM has a leading position.
In 3D NAND, we benefited from technology-driven demand for our ALD gap-fill solution, but overall, the contribution of 3D NAND continued to be relatively modest. In memory, we expect healthy sales in 2025, supported by continued solid demand for HBM-related DRAM, although it's too early to tell if memory sales would be at the same very strong level of 2024.
Other parts of the DRAM segment, as well as the 3D NAND segment, are likely to be subdued in 2025, in the absence of a more meaningful recovery in consumer-related end markets, such as PCs and smartphones. Longer term, the memory market remains a strategic growth area for ASM. As the DRAM industry transitions to the next nodes, more logic-like technologies will be adopted, which plays to the strengths of ASM.
In the new DRAM node, D1c or 1-gamma, we have increased our share of wallet with new ALD layers. We are working with leading customers on additional ALD and new Epi layers that will be required in the next generation, such as vertical channel access transistor DRAM or for SiC. The power analog wafer market is still in a cyclical correction that started by the end of 2023.
As the automotive and industrial end markets continue to be impacted by soft demand and inventory correction, there are no signs that spending in this space will pick up anytime soon. The moment that it will, ASM is well-positioned to benefit. In the last few years, we have won several new customers and positions in the silicon-based power analog and wafer markets, including in China, especially on the back of new innovative products, such as our Sonora vertical furnace and Intrepid ES Epi platform.
In silicon carbide epitaxy, our sales increased by a mid-single-digit percentage in 2024, as just mentioned by Paul. We believe this is a better performance than the overall silicon carbide epitaxy market, which significantly weakened in the course of the year. Our sales last year were supported by the new customer wins that we announced in 2023 and 2024.
While the outlook for 2025 has further weakened, we believe that the longer-term prospect for the silicon carbide market remains positive. In addition, we are well-placed to further extend our market share with the launch last September of the PE208, our new cluster system for 200-millimeter silicon carbide Epi application, combining best-in-class field performance with a new high-productivity dual-chamber platform. Looking at China, the sales contribution was again strong in 2024.
However, within the year, sales decreased in the second half, and also from Q3 to Q4, down from an exceptional level in the first half. This was in line with our previous indication. This lower demand is mainly due to several customers in logic foundry in China entering a phase of digestion, following the substantial new capacity investment that they made in 2023 and 2024.
We expect this softening trend to continue in 2025, with sales in China down in the full year. We project equipment sales in China to represent low- to- high 20s percentage of ASM's total revenue in 2025. The impact from new U.S. export control announced in December 2024 was in line with our assumption and is reflected in our projection for 2025. Next, I'd like to highlight our progress in sustainability.
In 2024, we published our climate transition plan. As a first milestone on our path to decarbonization, we achieved 100% renewable electricity in our worldwide operations. This contributed to a 52% decrease in our combined Scope 1 and 2 greenhouse gas emissions in 2024. If we then look at the longer term, the prospects for our company continue to be strong.
Third-party research forecasts the semiconductor market to reach $1 trillion by the end of the decade, with AI being the key driver. Next to growth in the data center segment, AI is expected to move in the coming years to edge devices like PCs, smartphones, and industrial applications, supporting growth in broader parts of the semiconductor market. As AI applications demand higher performance and power efficiency, the need for advanced semiconductor devices will continue to rise.
To enable these next-generation devices, more 3D structures and new materials are expected to be introduced. This, in turn, will require more ALD and Epi steps, increasing ASM's addressable markets. Next to driving sales growth in the longer term, we remain focused on improving the efficiency of our organization and operations. Examples are ongoing digitalization of our business processes, as well as further streamlining of our supply chain and manufacturing operations.
These multi-year projects will help us to scale the organization in a cost-efficient way and to support continued healthy profitability. Lastly, let's have a look at our guidance for this year. Wafer fab equipment is forecasted to be slightly up, and so far, 2025 shows the same mixed picture as last year. Growth in AI-related segments is expected to remain robust, but visibility for any meaningful recovery in other segments is limited.
The main area of growth for ASM in 2025 and for the broader wafer fab equipment market will be the ramp of Gate-All-Around in the leading-edge logic foundry segment, as just discussed. In memory, we expect healthy sales in 2025, supported by solid demand for HBM-related DRAM, although it's too early to tell if memory sales will be at the same very strong level as in 2024.
In power analog, there are no signs of a recovery in the near term, and the outlook for the silicon carbide market in 2025 has further weakened. Our China sales are expected to decrease in 2025. We confirm our targets for total revenue in a range of EUR 3.2-EUR 3.6 billion in 2025, but it's too early to be more specific.
For the first quarter, at constant currency, we expect revenue to increase from the Q4 2024 level to a range of EUR 810-EUR 850 million, with a projected further increase in Q2 compared to Q1. With that, we have finished our prepared remarks. Let's now move on to the Q&A. We'd like to ask you to please limit your questions to not more than two at a time so that everyone has a chance to ask a question. Operator, we are ready for the first questions.
Thank you. This is the Chorus Call Conference Operator. We will now begin the question and answer session. Anyone who wishes to ask a question may press star and one on their touch-tone telephone. To remove yourself from the question queue, please press star and two. Please pick up the receiver and ask questions. Anyone who has a question may press star and one at this time. The first question is from Tammy Qiu from Berenberg. Please go ahead.
Hi. Thank you for taking my question. So firstly, may I ask about China? Have you seen any situation change over the past month, i.e., becoming worse than you previously expected o r do you see the same situation as you have expected previously, i.e., revenue coming down, but it's not actually a Chinese customer or pulling back spending?
So I can take this question. I think our revenue projection in China doesn't change. We didn't see any meaningful change in the projection that we have made in the last few months. So business as usual for us.
Okay. That's clear. S econdly, just want to confirm on GAA momentum for 2025. You mentioned that your second-half visibility remains low. May I ask if your visibility on GAA for the full year is actually much higher than the other areas b ecause it seems to be that some customers may change their plan, but you have other customers who are willing to pick up?
I think, as we mentioned in our press release, we think that Gate-All-Around is strong in 2025. This is the year of HBM for Gate-All-Around, and we see that continuing unabated. So we don't really see any change in the Gate-All-Around, and 2025 is actually very strong in the 2-nanometer technology now.
Okay. Thank you.
The next question is from Didier Scemama, Bank of America. Please go ahead.
Yes. Good afternoon. Thank you for taking my question. Hichem, I've got a question for you with regards to the order intake, perhaps how you feel about it for the first half. In order to reach the upper end of your range, do you feel like you need to secure a substantial increase in orders in Q1 or Q2 at the latest, given your lead times o r are you fairly relaxed to hit the upper end of the range as you've got a funnel of opportunities into the second half? I've got a quick follow-up. Thank you.
Yeah. Let me take that question, Didier. Yeah. We haven't even seen orders in Q4, which came down compared to Q3, as expected. As you know, we already communicated in Q3 that we saw some pull-in orders that we actually expected in Q4. But as you know, timing is always difficult to predict.
In addition to that, what we've seen is, of course, there's still a power/wafer for the analog market that is very slow, as we said, sharply down. So that didn't really help. China is, as expected, but still lower than what we saw in 2024. So also that, of course, did not help. Gate-All-Around , as Hichem already mentioned, continued to be strong. A s you know, your question about Q1, Q2, we don't guide on orders.
We stopped guiding on orders for a good reason, and it is simply because timing of orders sometimes is bulky, sometimes it's not bulky. We burned our fingers a few times, and we prefer not to do that again. So we will not guide on Q1 and Q2.
Okay. Fair enough. My follow-up, it's talking about the opportunities for molybdenum. So clearly, you've got a big play in metallization going forwards with molybdenum. N ow we see effectively all your competitors talking about molybdenum being implemented, not just in foundry logic, but also in NAND and in DRAM.
So I just wondered if you could talk about how you see those opportunities playing out maybe in 2025, 2026, 2027. A lso related to that, for these new materials, we see also cobalt and ruthenium being mentioned in replacement to tungsten and copper. So I just wondered if you think you could also capitalize on those opportunities. Thank you.
So I'll take this question. What I can tell you is that, as we mentioned before, molybdenum is the new metal for logic devices and memory devices. Our focus is actually molybdenum is with the customer logic foundry and DRAM. We engage with many DRAM customers in molybdenum. We have a very active program.
We have actually lots of engagement with many customers. A s a matter of fact, we're actually shipping tools in molybdenum for HBM this year. So we are very much working on molybdenum. We think that molybdenum, like we mentioned, is a new metal.
It started in N2 right now with a very small volume, but there are more and more applications as the technology, the Gate-All-Around technology, transitions to the 1.4-nanometer node and eventually also to the 1.0-nanometer technology node. So you can see more and more molybdenum being run especially in logic and DRAM in front-end of line, in mid-end of line, and even in back-end of line.
Talking about your question about different metal, I think that's what you refer here to about cobalt and ruthenium. What's going to happen in back-end of line and metallization in the future is the end is a move from damascene to metallization. So right now, what you do, you deposit a dielectric, and you etch, and after that, you deposit a metal.
In the future, we're actually going back to the 1990s, whereby you deposit a metal, and then you etch the metal, and you deposit the low-k dielectric between those metals. So the new metals that are going to happen, part of it is actually going to be ruthenium is one of the metals that's going to happen in this new metallization scheme. So yeah, this is going to be run with ALD. It's going to happen in the future, and these are areas that we are also aware of and also working on. Thank you.
Thank you. Can I have just a tiny one on NAND? I think you didn't mention NAND. Is that because of the productivity limitation of your ALD or Epi tools?
I think that for 3D NAND, what we think that in that market, there are some also molybdenum applications. We think in that market that the offering there can be either you can make the offering with a single wafer ALD, but also you can do the offering with a batch ALD. We think that that market will most likely transition to batch ALD in the future because of higher productivity and lower cost.
Makes sense. Thank you.
The next question is from Sandeep Deshpande, J.P. Morgan. Please go ahead.
Yeah. Hi. Thank you for letting me on. My first question is regarding the 1.6- and 1.4-nanometer nodes that you highlighted earlier for logic. Are you in a position to say what revenue you can achieve per 100,000 wafers based on your wins in those nodes? Like you've given that the 2-nanometer node, $400 million per 100,000 wafers. So what kind of opportunities you won in those nodes, which will provide you with growth in 2026 and 2027 onwards? T hen I have one follow-up, please.
Okay. Hi. To answer this question, I think what we always say, and that's really what's happening in the industry, as we transition from one technology node to the other, the ALD intensity increases by double digits. So this is going to continue from the 2-nanometer node to the 1.6- and 1.4-nanometer nodes.
So we see that continuing unabated. And as I mentioned earlier, some of the ALD application actually is going to be in the backside power distribution network. A lso, I just talked a few minutes ago that ALD is going to molybdenum ALD, okay, is going to take more and more layers as you transition from N2 to N1.6 and N1.4. So we still see meaningful double-digit increase in ALD intensity.
M y follow-up, Hichem, is about memory. You've had this very strong growth in 2024 in the memory market. When we hear about the CapEx from the memory players this year, there seems to be more focus on building capacity rather than building buildings rather than putting equipment into those buildings. Is that likely to impact you in terms of your HKMG engagement in the DRAM market, for instance o r do you expect to continue to see growth in the memory market in 2025?
We expect to see growth in the high end of the memory market, in the high-bandwidth memory, in the DDR5, in the DDR6. I think the high part of the DRAM memory is actually increasing unabated, and we see there's a strong growth that we see, and that's going to continue. With it, we see more and more high-K metal gate application in that part of the market. So we are really upbeat about the high-bandwidth memory of the DRAM market.
Thank you.
The next question is from Nigel van Putten, Morgan Stanley. Please go ahead.
Hi. Good afternoon. First, I have a follow-up on the services pool. You already addressed it in the prepared remarks, but just to make it a little bit more clear, when you say that growth will moderate, is that from the fourth quarter level, or should we think about that being sort of on a year-on-year basis, which kind of would imply that the next couple of quarters will be down compared to the fourth quarter? That'll be my first question. Thanks.
Yeah. The question was on. The answer was in the prepared remarks, and Nigel was on the full year. So for the full year, we expect a more normalized level of growth. We still expect some growth, but yeah, we've seen pretty exceptional growth. I mean, 49%, you might agree that's not the standard growth that we will be able to achieve.
Y eah, that had two reasons. One, CS&M is continuing to do well, to be honest. The programs that we kicked off really started to get good traction. But also, equally, and maybe even slightly more important, we saw an acceleration of parts in China a nd that will also, in our view, normalize.
Got it.
I'm not in a position now to give you specific how this will evolve over the year. That's a little bit too early to tell.
Understood. Then the next question would be on Advanced Logic Foundry, where the GAA visibility is very strong. Can you provide some context how big a chunk of the total revenue coming out of Advanced Logic Foundry is GAA?
Just to get a sense of the reliability in that segment outside of GAA because in your press release, it does seem to imply that there's changes going on, while at the same time, GAA has not changed. So I mean, yeah, I'll leave it there just to have a sense of how big that potential reliability is.
Yeah. I think for this year, so for 2025, I think we might even have said it. I'm not 100% sure. But if you look at our total logic foundry sales, more than half will come from GAA. So it's a big year-on-year growth. Yeah. As Hichem already said, we're very positive about GAA. A s I said, yeah, we expect more than half of the material logic sorry, of the total logic foundry to come from Gate-All-Around.
Y eah. I mean, if I add something to what Paul has mentioned, like we have talked in our previous earnings call, we said that 2025 is going to be prime time for Gate-All-Around a nd it's actually happening. Gate-All-Around 2-nanometer technology node is in HBM tools team ahead. But also, we see that in 2025, that also there are some technology buys for next-generation Gate-All-Around, like 1.4-nanometer. So Gate-All-Around d oes constitute over 50% of our logic foundry business right now.
Got it. Thank you very much.
The next question is from Jakob Bluestone, BNP Paribas. Please g o ahead.
Hi. Good afternoon. Thanks for taking the question. I was hoping you could maybe comment a little bit on how you've seen your market share evolve for ALD, just sort of any thoughts around how you think that will evolve from here. T hen I've got a follow-up as well.
So like I mentioned in the press release right now in my prepared remarks, we think that our market share in ALD is very strong, continues to be strong in the transition to Gate-All-Around technology.
Right. J ust a clarification on China. Is the decline, is that just purely digestion of inventory, or is there anything embedded in your guidance regarding tariff and further restrictions? You mentioned that the latest round of restrictions were as expected, but do you expect anything more? Thanks.
So in the guidance that we've given for China, what is included there is what we call digestion of an exceptional level of revenue that we've seen, in particular, in the first half of 2024, but also still strong sales in the second half, but lower than the first half. That's one. T wo, as you know, I think on I think it was December 2nd, if I'm not mistaken, the new export controls also have an impact somewhat, which we anticipated.
So also that is reflected in this guidance. What is not yet reflected is potential new restrictions because yeah, at this moment in time, real speculation if anything will happen, what will happen if anything will happen. So that's not yet reflected in this guidance, b ut the latest December 24 restrictions have been taken into account in this guidance.
Very clear. Thank you.
The next question is from François-Xavier Bouvignies, UBS. Please go ahead.
Thank you very much. My first question is on the more longer term, beyond 27, a nd ASML, one of your peers, is talking about High-NA insertion in 27 and then for mass production a bit later on. I was wondering if you could detail a bit the impact you think it could have, this High-NA introduction. I assume that now you have a couple of layers using multi-exposures for your ILD, which was not the case before.
So I was wondering if we should expect some moderation of growth as High-NA comes in. Of course, all else equal, you might have some other opportunity somewhere else. But on the advanced logic side, I was wondering, what do you think High-NA could impact your ILD business on the growth rate?
So I mean, what I can tell you here is that we see that as the technology node transitions, that both ALD and Epi are actually enabling the transition to the next generation. So we see High-NA happening, but we'll also see more and more ALD layer actually taking place.
So if there's one application that goes away because of High-NA introduction reduces the number of ALD layer, there are actually the complexity of the devices that are happening right now is actually giving rise to a new selective ALD deposition market, which was not even present, which has actually more ALD than ever. So what I can tell you, again, is with the increased complexity, you need more E pi, and you need more ALD.
Okay. So still net positive, if I read correctly, Hichem.
Yeah. Absolutely.
M aybe the second question I have is a very quick one. You mentioned China's percentage of revenues will come down to low 20% - high 20% in 2025. Could you disclose your 2024? How much is worth China? Just to put into perspective.
Yeah. For sure. I think you know the answer already. We have given qualitative comments on our 2024 exposure in China, exceptional in the first half and strong in the second half. We have also, because of these export controls, given some more guidance for 2025, as you mentioned, the low to high 20s%, a nd that's what I can confirm and share.
Understood. No problem. Thank you very much.
Thank you.
The next question is from Robert Sanders, Deutsche Bank. Please go ahead.
Yeah. Hi. Thanks for taking my question. You mentioned the services number in China. Can you just give us a sense of how much sort of stockpiling of spare parts there is in China and perhaps some stockpiling of your tools as well at Chinese fabs? I have a follow-up. Thanks.
Yeah. If I would know, I might tell you, Robert, but I really don't know. So I don't know if there's a lot of parts piling, but you might expect there is some, given this accelerated demand, of course, that is happening b ecause you can imagine that the Chinese players are concerned, given everything that's happening from a geopolitical point of view.
In terms of equipment, yeah, most of our equipment goes straight to fabs. Sometimes not, but that's simply because fabs might not be fully ready yet or for other reasons. But the build will go straight into fabs. So yeah, I would not be able to give you more information on this at this point, Robert.
Okay. No worries. J ust a question on backside power delivery. So if we see the power delivery network moving to a different way for effectively and scaling down at a much more aggressive feature size, what does that mean for the opportunity, whether it's metalization or anything else, relative to the transition to Gate-All-Around ? I mean, is it a significant increase in opportunity? How should we think about it in terms of significance? Thanks.
I think it's an increase. I don't call it significant if you ask me, but it's an increase, especially in metal ALD deposition. There's also some dielectric ALD there, but it 's an increase, but I don't really call it a significant.
Got it. Thanks.
The next question is from Stephane Houri, ODDO BHF. Please go ahead.
Yes. Hello. Good afternoon. Thank you for taking my questions. The first one is on the guidance for the full year, the range, the EUR 3.2-EUR 3.6. I wanted to understand what are the assumptions that you took to maintain it. I understand that Gate-All-Around is super strong still, but memory kind of flattish or maybe power analog down, silicon carbide down.
So what would it take for you to have an H2 that would be down compared to H1? I'm asking that because if you do the math with the elements you've given, you should have it just takes a flat H2 to go to the middle of the range. So I'd like to understand how you build this one. Thank you.
So I'll answer the question. I think to answer your question, I think we kept our EUR 3.2-EUR 3.6 range for 2025. I think that part of the uncertainty that we talked about is really that we don't see signs of recovery in the power wafer analog market. We mentioned that previously that the power analog wafer market is very cyclical. It's usually cyclical. It takes about eight quarters cyclicality.
This time, it looks like maybe it's one more quarter or two more quarters. So it's actually a longer cyclicality, but it's really within the range that we see. We don't see a recovery in the second half of the market and second half of the year. You never know. But right now, we don't see any sign of recovery. The other part is really on silicon carbide. I mean, silicon carbide market is weaker than ever before. So that's also another uncertainty from that point of view.
In memory, we have talked about memory earlier that we see that it's too early to tell if memory sales would be at the same very strong level of 2024. We see a very strong HBM memory strength in 2025. But for commodity DRAM market, it's really too early to tell from that point of view. T hat's why we came up with we still continue with a range of EUR 3.2-EUR 3.6.
Okay. Thank you. T hen I have a follow-up. It's about the OpEx in 2025. How do you see them evolving? You grew a lot, R&D, not too much, SG&A. Are we still heading for the same trends in 2025?
I think so. I think that we have a strong program to have better efficiency, to improve our business processes, to incorporate digitalization a nd we're introducing S/4HANA. We're introducing Teamcenter to really streamline our business processes.
So I think that we have a strong focus to keep in our operation to make sure that we continue to run our operation very efficiently. So that impact actually is going to impact. It's a multi-year impact for the future, a nd we are very much committed to improve our top-line revenue growth for many years in the future. But also, we want to make sure that this growth is very profitable a nd that's what we're doing right now.
R&D specifically, sorry, are you talking about double-digit growth still again? Sorry for interrup ting you.
Maybe to add to that, Stéphane, if you look at the growth R&D, maybe also to help you for your modeling, that is likely to expand 10%-15% year-on-year. Net R&D will expand somewhat more, again, because of increased amortization. So the net R&D will increase faster, like you've seen in 2024, than the growth R&D.
So just then you have, I think, a reasonable guidance for your modeling, a nd SG&A, we expect in absolute terms around a similar level as in 2024. A s a percentage of revenue, of course, it should come down to high single digits.
Very clear. Thank you very much.
The next question is from Aditya Methuku, HSBC. Please go ahead.
Yeah. Good afternoon, guys. Thank you for taking my questions. So firstly, just on Gate-All-Around high-volume manufacturing capacities being installed, I just wondered if you might be able to give some color on the tools you're shipping or you plan to ship in 2025.
What level of Gate-All-Around capacities will that lead to in aggregate at the three customers that are meant to be ramping Gate-All-Around in 2025, based on the discussions you've had with your customers so far? That's my first question, and I've got a follow-up.
We believe that our wafer capacity for 2-nanometer will be actually a significant number. But we are not commenting right now on specific numbers. As I mentioned, end markets continue to be actually very strong based on what we're hearing, and leading customers are actually continuing in their HBM. Our view for a substantial increase in Gate-All-Around sales in 2025 has not changed from that point of view.
Okay. Understood. T hen secondly, on China, if I did my math correctly, I think China is close to 40% of your equipment revenue in 2024, similar to what we've seen at your peers. So now, when I look at the news flow out of China, Naura is talking about shipping multiple high-k ALD systems to multiple clients a nd there seems to be a significant focus on developing CVD and furnaces from the likes of AMEC and Naura again.
So I just wondered, how do you see competition in China developing, and should we be concerned about local players potentially gaining share from you? Any thoughts there would be great. Thank you.
So let me start, and then Hichem can add to Aditya. You're right. Naura is growing steeply, also in terms of revenue. You might have seen that. I don't know. Chinese competitors are clearly trying to catch up, which is logical, given everything that is happening. For us, it is just competition that we have also with other players.
I mean, the key is to innovate, to be ahead of them, and to lead. That's what we try. That's what we'll try compared to our Chinese competitors. We do the same with other competitors. So no real difference there. But yeah, they're definitely catching up. But yeah, not to the extent where we are. I mean, they're still far, far, far behind us.
But yeah, we are paranoid. We always are paranoid about competition, also about Chinese competition. So it just motivates us to go even faster with all kinds of innovation that we have in our pipeline.
Yeah. I mean, if I add that to what Paul has just mentioned, yeah, there's competition and so on, but also we continue in our innovation. We continue to compete in the Chinese market. We have developed a new product, actually, for the China market, like in Epi.
We have introduced the Intrepid ES, atmospheric Epi , for power analog devices. We also have released a very successful product for the China market, which is the Sonora vertical furnace, running many furnace applications in actually 300 millimeter, even in 200 millimeter. We have our dual tool. So yeah, I think we're competing in China. It's a market that's very competitive. But as long as we continue to innovate, we'll be able to be successful there too.
Understood. Thank you, guys.
The next question is from Timm Schulze-Melander, Redburn Atlantic. Please go ahead.
Hi. Good afternoon, guys, and thanks for taking my questions. Maybe the first one for Hichem, please, on molybdenum. I appreciate that it is very, very early. But maybe could you provide some qualitative commentary around sort of win rates and differentiators, and then I had a follow-up. Thank you.
I think molybdenum, as I mentioned, is a new metal. It's used in both memory and logic and foundry. There are many, I mean, in molybdenum, there are many innovations that's happening in the process technology, in the chemistry, in the precursor, in productivity, in how you deliver the chemical s, etc., etc.
So it's an ALD process, which we have lots of experience on how to make the process repeatable with very good performance. We really have a few wins in molybdenum, which we're very excited about. You're going to hear more about it in our investor meeting in September of this year. We have a very strong engagement, and we're really excited about our engagement in molybdenum. Even in DRAM, we are very happy about some of the interaction we have with our customers.
So it's a new metal. It's a new metal. It's a new thing. This is disruptive. I think this opportunity or never to come in and with the innovative solution. We are excited in the engagement that we have so far.
Awesome. Okay. Look forward to hearing more about that at the CMD. Maybe just for Paul, a lot of color there on the operating expense lines, which is really helpful. Just in terms of modeling, China in 4Q, are we close to the 25% corridor?
What do you mean? Sorry. Are we close to the 25% corridor? What do you mean, Timm?
You guided a range for 25% in terms of China as a percentage of sales. I'm just wondering, are we close to that in Q4?
Yeah. Nice try. I can repeat the same answer I gave right before, Timm, but I think you know the answer. It was strong, good, but below Q3 and also below the same year as it's coming down. A s you know, also 25 compared to 24 is coming down. So I leave it to you to try to make sense of that.
I think what is more important is that going forward for this year, yeah, we guided low 20s to high 20s. It should give you a reasonable good indication, which means, of course, that 24 was higher than that.
Got it. Worth a try. Thank you.
Thank you.
T he last question comes from Marc Hesselink, ING. Please go ahead.
Yes. Thanks for taking the question. The first one is just reading through your press release and also the comments during the call today. I have the feeling that maybe the opportunities even accelerated a bit faster than what you have communicated before, given the step-up in the R&D, also especially in the fourth quarter, and also the evaluation tools, again, moving up in the second half of the year, which typically is a good leading indicator. Is that a true observation? Is that true?
I think what you can say is that I think for us that we are looking into all new applications that's happening right now. I think the R&D spending has accelerated. Like we mentioned, there's more and more ALD layer from one generation to the other. So by definition, you have to do more R&D to get those layers.
I mean, if you do R&D right now, I mean, this is going to happen four years from now. So it's not something that really is immediate from that point of view because it takes a long time to develop new processes and new equipment for ALD. But that's an indication of our really strong belief in the ALD market growth into the future.
L ike we mentioned, ALD is a new CVD. We see that happening, replacing PVD layers, replacing even some CVD layers. But also, ALD is happening for the high aspect ratio growth. It's happening in backside power distribution. So it's just normal. As the ALD expands, the market expands, our R&D ne eds to expand with it.
Okay. Great. T hen my final second question is on your cash returns. Given your sort of target cash level that you earlier communicated and the buyback that you now started, it seems a bit small, given buyback for the next year. Maybe your thoughts about that?
I think my thoughts about that, I think that we're looking into our cash allocation strategy going forward. So it's something that we're looking into a nd right now, I think that we didn't make any decision from that point of view. But what we really focus on, okay, how can we provide the best return to our shareholders a nd that's really our focus going forward.
Yeah, a nd maybe to add to that, Mark, I mean, we have also the EUR 100 million earn-out payment in 2025. You should add that as well. We are expanding rapidly R&D infrastructure. As you know, we finished Korea. We started in Scottsdale. So there's a lot going on. So yeah, also for that part, we are maybe a little bit conservative. You might be right. I don't know. But you should take these three things also into account.
Okay. Okay. Very clear. Thanks.
Gentlemen, there are no more questions registered. I turn the conference back to you for any closing remarks.
Last, I would like to thank you, everyone, for attending today's call. Also, on behalf of Paul and Victor, we hope to meet with many of you at the upcoming investor events. Thanks again, and goodbye.
Ladies and gentlemen, thank you for joining. The conference is now over, and you may disconnect your telephone.