© 2024 Alpha Spread Limited. All Rights Reserved.
www.alphaspread.com is not operated by a broker, a dealer, or a registered investment adviser. Under no circumstances does any information posted on www.alphaspread.com represent a recommendation to buy or sell a security. The information on this site, and in its related newsletters, is not intended to be, nor does it constitute, investment advice or recommendations. The information on this site is in no way guaranteed for completeness, accuracy or in any other way. In no event shall Alpha Spread Limited be liable to any member, guest or third party for any damages of any kind arising out of the use of any content or other material published or available on www.alphaspread.com, or relating to the use of, or inability to use, www.alphaspread.com or any content, including, without limitation, any investment losses, lost profits, lost opportunity, special, incidental, indirect, consequential or punitive damages. Past performance is a poor indicator of future performance.
You don't have any saved screeners yet
Intrinsic Value is all-important and is the only logical way to evaluate the relative attractiveness of investments and businesses.
Good afternoon. This is the Chorus Call conference operator. Welcome, and thank you for joining the ASM International First Quarter 2024 Earnings Call. [Operator Instructions]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 welcome, everyone, to our Q1 earnings call. I'm joined here today by Gek Lim Loh, CEO; Paulus Verhagen, our CFO; and Hichem M'Saad, CEO and incoming CEO. ASM issued its first 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 accessible on our website, asm.com, along with our latest investor presentation. We remind you, as always, 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, reports, and financial statements, which are available on our website. Please note that the profitability measures mentioned in this call will be primarily based on adjusted non-IFRS figures. For the reported IFRS figures as well as the reconciliation between IFRS and adjusted results, please refer to the quarterly results press release. And with that, I'll now hand the call over to Gek Lim Loh, CEO of ASM.
Thank you, Victor, and thanks to everyone for attending our first quarter 2024 results conference call. I know it is a busy day for all of you. We will follow the usual agenda for today's call. Paul will first review our first quarter financial results. I will then continue with a discussion of the market trends and the outlook followed by Q&A. I will now turn it over to you, Paul.
Thank you, Gek Lim. In the first quarter of 2024, revenue came in at EUR 639 million, which is at the top end of our guidance of EUR 600 million to EUR 640 million. At constant currencies, revenue decreased by 8% compared to the level in the first quarter of last year compared to Q4 revenues up 2%. Equipment revenue was led by ALD and followed by Epi. Further conference revenue dropped year-on-year, mainly reflecting the weaker trends in the power analog segment. Spares and service revenue was up 2% at constant currencies compared to the same period last year. In terms of customer segments, revenue was led by foundry, followed by memory. Sales in the combined Logic/Foundry segment decreased compared to Q1 last year but increased substantially compared to Q4. Year-on-year, leading-edge logic foundry was down and in part offset by a strong increase in the mature logic foundry, mainly in China. We recorded decent gate-all-around related sales in Q1. This followed on the first meaningful orders for gate-all-around that we booked in Q3 and Q4 of last year. Memory sales also rebounded strongly compared to the depressed level in Q1 of last year and also increased significantly compared to Q4. Sales in this segment were for a big part driven by high-K ALD tools for high-performance DRAM and HBM applications. Sales in power/analog/wafer dropped significantly, both compared to the first quarter and compared to the fourth quarter of last year. This is following the weakening order trends in this segment that we already reported in the last couple of quarters. To be sure, this is excluding silicon carbide, where the trend continued to be healthy. Gross margin in the first quarter increased to an exceptional high level of 52.9%, up from 47.9% in the fourth quarter and 51.1% in Q1 of last year. The increase in Q1 gross margin was explained by a generally strong mix, including a continued strong China sales contribution. Sales from the Chinese parks were up both year-on-year and compared to Q4 and reached a quarterly record high. For Q2, we expect China sales to remain strong. For the second half of the year, we expect China to decrease compared to the first half, but to remain at a more resilient level than we anticipated at the start of the year. SG&A expenses increased by a moderate 4% compared to the first quarter of last year and were below the level of Q4 2023, explained by lower variable expenses and reduced spending across various departments. For the full year of '24, we continue to expect a moderate increase in SG&A compared to the full year of 23%. Net R&D increased at a relatively low basis 2% year-on-year. While gross R&D increased by 17%, the increase in net R&D is moderated by the fact that Q1 capitalization of development expenses increased stronger than amortization. In the course of the year, we project this trend to reverse as amortization will start to increase more substantially. This is explained by several newly developed products and applications for which amortization of the related capitalized development expenses will start as of Q2. For the full year, we expect R&D to increase the net R&D to increase by 10% to 20%. This is in line with our previous indication that the net R&D as a percentage of sales in '24 will be somewhat above the target range of high single to low double digits. Below the operating line, financial results included a currency translation gain of EUR 23 million. This compares to a translation loss of EUR 7 million and EUR 25 million in the first and fourth quarters of '23, respectively. As most of you know, these currency gains and losses are explained by the fact that we hold the largest part of our cash in U.S. dollars, and the translation differences are included in our financial results. Let's move to ASMPT. Our share in income from investments, reflecting a stick of approximately 25% in APT, amounted to EUR 5 million in the first quarter, up from EUR 2 million in the fourth quarter and down from EUR 9 million in the first quarter of last year, reflecting the slowdown in the back-end equipment market. Note that the amortization related to the ASM BT stake came largely to an end last year. And for 2024, it will amount to just EUR 0.4 million. Now turning back to ASM operations. Our new orders in the first quarter were EUR 698 million, an increase of 10% at constant currencies compared to Q1 of last year. Looking at the breakdown by customer segment, Foundry was the largest segment followed by Memory and then Logic. Combined, Logic/Foundry orders representing the largest part of total orders were down from Q4, but up from Q1 last year. Orders related to gate-all-around were again healthy. These orders were from multiple customers and included for the largest part ALD tools but also multiple Epi-tools. Mature logic foundry demand mainly from Chinese customers also remained strong in Q1. Memory orders increased strongly in Q1 and DRAM accounted for the largest part, but we also saw a decent uptick in demand for ALD in the 3D [ NAND ] segment. power/analog/wafer bookings, excluding certain carbides, remained relatively low in Q1, reflecting the softer conditions in this market segment.Silicon carbide Epi orders, on the other hand, remained healthy in the first quarter with a meaningful part driven by 200-millimeter tool demand from Chinese customers. Next to the balance sheet. ASM's financial position continues to be strong. We ended the quarter with EUR 720 million in cash, up from EUR 637 million at the end of the fourth quarter. Free cash flow amounted to EUR 62 million in the quarter, down from EUR 155 million last year in Q1. While profitability remains solid, the decrease in free cash flow was mostly explained by increased working capital. Accounts receivables, which tend to fluctuate from quarter to quarter were up in the first quarter. Inventories also increased in the first quarter in preparation of expected higher shipment levels in Q2 and in the second half of the year. Days of working capital increased to 70%, up from 61% in Q1 last year, and remained within our target range of 55 to 75 days. CapEx amounted to EUR 31 million in the first quarter. As a reminder, our target for the annual CapEx continues to be between EUR 100 million to EUR 180 million. A significant part of CapEx would be for infrastructure investments, primarily for R&D, including the expansions in Korea and in Arizona that we announced last year. And with that, I hand the call back over to Gek Lim Loh.
Thank you, Paul. Let's now continue with a review of the market trends. Against the backdrop of macroeconomic uncertainty, sluggish consumer spending, and geopolitical unrest, the recovery in the semiconductor market continued in the first quarter, albeit at a relatively slow pace. AI remains the visor with very robust growth. The smartphone market returned to year-on-year growth and PCs appear to have passed the bottom. Although the recovery in both segments is still early and somewhat slow. The automotive semiconductor market weakened and has been impacted by inventory corrections. In WFE, transitions were still mixed in the first part of the year. Memory picked up somewhat, primarily driven by HBM DRAM. Except for early investments in gate-all-around, the leading-edge logic/foundry segment is still relatively weak. Spending in the Chinese market remains strong. If we look in more detail at our logic/foundry business, capacity spending on the current advanced nodes such as 5-nanometer and 3-nanometer continued to be soft in the first quarter. Momentum in 2-nanometer is, however, picking up speed. Our customers have been talking about significant interest among their own customers for the 2-nanometer node, including for new AI applications in the coming years. Several new 2-nanometer fabs have been announced over the last couple of months by key customers, which suggests that the 2-nanometer node is going to see sizable capacity investments in the next couple of years. In addition, all key customers reconfirm their targets to ramp the 2-nanometer node in high-volume production in the course of 2025. Following the first meaningful orders in the second half of last year, we booked again a healthy level of gate-all-around-related orders in the first quarter, still largely related to pilot line activity. We expect continued orders for 2-nanometer throughout the year with the first orders for high-volume manufacturing in the second half. For the full year of 2024, we continue to expect that gate-all-around will account for the majority of our order intake in leading-edge logic/foundry. As we discussed on previous occasions, Gate-All-Around will be an important driver for ASM. We confirm our estimate that this transition will increase our SAM in launching a foundry by some USD 400 million for each additional 100,000 monthly wafer capacity. Many new ALD layers are required to engage all of our devices, and we believe we successfully maintained our leading market share in 2-nanometer compared to the 3-nanometer loan. Epi is also an enabling technology for gate-all-around. In the transition to 2 nanometers, we believe we will significantly increase our Epi market share, putting us on track toward our market share goal of 30% by the end of 2025. If we then look at the memory business, demand is gradually recovering, even though spending is still mostly related to the conversion of existing capacity. DRAM is the main area of growth on the back of strong AI-related demand for HBM. ASM is well placed in this area as we have a leading position in high-KLD, which is a key technology for high-performance DRAM in HBM stacks. Looking at the full year, we expect memory sales to rebound strongly compared to the lower level in 2023. Next, our business in China. As Paul just mentioned, sales from the Chinese market continued to be exceptionally strong in the first quarter. With the publication of our fourth quarter results, we conservatively expected normalization in the second half of this year in China, following the strong spending levels in recent periods. Based on current visibility, we still expect sales from China to be lower in the second half compared to the first half, but the decrease will be more moderate than we earlier expected. Spending trends in power analog wafer excluding silicon carbide, has been slowing down in China, similar to other geographies. This is, however, offset by continued robust spending in the mature logic foundry segment. This segment, which is the largest part of our sales in logic/foundry in China continues to be dynamic with several new projects being launched. In addition, the size of the existing internal market in China, for instance, for 28-nanometer semiconductor devices is very substantial.If we then look at the power/wafer/analog segment, demand has weakened significantly as we already reported in the last 2 quarters. Revenue in the first quarter was meaningfully lower. And also for the full year, we expect revenue in this segment to drop. As revised last year, our sales in power/analog/wafer almost doubled. This industry segment is going through a phase of digestion after 2 years of substantial investments in new capacity. In addition, slowing end demand and inventory corrections in parts of the automotive and industrial markets, which have been the main drivers of power analog demand also led to reduced investments. Despite the current slowdown, power/analog/wafer segment is expected to remain an attractive market in the long term for ASM and we are in a good position to benefit. In the last couple of years, we successfully expanded our position, for instance, with our interpret ESA tool for 300-millimeter power and wafer applications. My comment about the slowdown in power/analog/wafer mostly referred to the circuit-based business. In silicon carbide, the trend is more resilient, even though demand has been impacted by a temporary slowdown in electric vehicle growth. some silicon carbide customers recently pushed out new capacity investments. We still expect meaningful double-digit growth in this business in 2024, although not as strong as we anticipated at the start of the year. Our sales growth this year is in part supported by the contribution of the new customer wins that we reported in previous quarters. With a further increase in electric vehicle penetration, growth in the silicon carbide market is forecasted to remain strong in the coming years. In addition, we expect to further expand our position as the industry is going to transition in the next couple of years to 200-millimeter for which our silicon Epi-tools offer substantial benefits. Now turning to the outlook. For the second quarter, we expect sales to increase to between EUR 660 million to EUR 700 million. This is a bit higher than our previous guidance, explained by somewhat higher demand, particularly in the memory segment. For the second quarter, we now expect sales -- sorry, for the second half, we now expect sales to increase by 10% or more compared to the first half. That means 2024 will be another year of revenue growth for ASM. Sales in China are projected to be lower in the second half as just discussed, although at a higher level than previously expected. The sequential decrease in China in the second half will be more than offset by higher sales in leading-edge logic/foundry, mainly in gate-all-around, and to a lesser extent, increases in Memory.We reiterate our revenue targets for 2025 as communicated in our Investor Day last year, supported by the rebound in WFE with 2-nanometer moving into high-volume manufacturing and further strengthening of the memory market. On a personal note, this will be my last earnings call, I'd like to thank you all. Being the CEO of this unique company has been a privilege, and it is great to see what we have achieved as one ASM team over the past 4 years. Our company has never been in a stronger position than today. I'm confident that with Hichem as CEO, ASM has a bright future and continued success ahead. I'd like to thank you, our investors, and analysts for your support and interaction in the past 4 years. I really enjoyed the many investor meetings, conferences, road shows, and even the earnings calls. I wish you all and your families a healthy and possible future.
Maybe Gek Lim before I hand over, I would also like to take this opportunity to thank you on behalf of everyone at ASM. Since you took over as CEO of ASM in 2020, revenue more than doubled strongly outperforming WFE. Our operating profit more than tripled. Our engagement with key customers further strengthens and we stepped up investments in our people in R&D and in sustainability. Over the past 4 years, the total share return amounted to close to 500% or almost 60% annualized, placing ASM among the top performers in the sector. Gek Lim, thank you for your great contributions to ASM's achievements and success.
Thank you, Paul. Let's now move on to the Q&A. We also have Hichem on the call today, but Paul and I will take most of your questions. As you know, Hichem will take over as CEO with the AGM on May 30.
[Operator Instructions]. The first question is from Sandeep Deshpande with JP Morgan.
It's been a great pleasure to work with you, Benjamin, over the last 4 years. Always was very happy chatting with you on industry developments. So maybe in this last Q&A to ask you, how do you see the cadence of the order intake, given that you've won a lot of business in both ALD and in Epi foundries and at other logic customers and they are ramping up this process in the next year. So how do you see this cadence developing? You had a strong order intake in the first quarter? Do you see this progressively continuing through the year? Or is it going to be lumpy? Or is it going to be fourth-quarter-oriented? That's my first question. And my second question is, again, regarding the ALD and Epi business of ASM. Clearly, there are competitors in the market, they are knocking on your customers' doors all the time and some of them say they will take share from you in the future, et cetera. How do you see that environment developing at this point? In particular, with regard to single-wafer ALD versus batch process, whether a single wafer will have its own position, which is very differentiated from the batch process, particularly in the memory side, which tends to be more cost-conscious.
Sandeep, thanks a lot, and it's certainly been a pleasure working with you over the last couple of years. So maybe I will take the first question, and I will have Hisham try to give you more insights on the market share question. So on the cadence of offering, I mean, ordering, I think we will continue to see, especially gate-all-around all this throughout the year. I think what we are seeing in terms of orders since the second half of last year, is primarily targeted at pilot lines. And we have 2 things. One is they are not ordering a one-shop just all the tools and they order, let's say, gradually a time, that's one. Secondly, I think one thing that has changed very significantly is we now have multiple customers. So when we look at, for example, maybe you're looking at 3-nanometer pilot going DRAM, we were essentially looking at one customer. But now it's very different. We have multiple customers going in, trying to build pilot line capacity for 2 nanometer, and this is creating a situation where we are going to get all this constantly also over the next couple of quarters.Now, of course, the big question is, so when do we expect to get high-volume manufacturing orders? We do expect that this will be on track for the second half of this year. And with respect to the ALD and the Epi market share, I will probably pass this over to Hichem.
Okay. Thank you, Benjamin. For our market share, if you look into ALD market share, transitioning from 3 nanometer to 2 nanometer, we believe we at least have maintained our market share, which 35% plus in ALD. And in Epi, we think that we have definitely gained market share with the 3 leading silicon in the 3 leading semiconductor logic and foundry customers. In 3-nanometer, for example, we had one customer only in Epi and in 2-nanometer gate-all-around, we have all 3 leading logic foundry customers.
And maybe lastly, Sandeep, on your question on the single wafer butane, I think both will continue to have its own play. Some years ago, we were thinking that perhaps that ALD, some of it will transition to a single wafer ALD. And that is still happening, but it's happening at a very, very gradual pace. As long as ALD can continue to satisfy the kind of performance, they will continue to be used.
The next question is from Alexander Duval with Goldman Sachs.
Thank you so much, Benjamin, for all the discussions over the years. My first question is about AI. A large foundry player talked about very, very strong AI demand. And you in your introductory remarks, alluded to broader AI strength. I wondered to what extent you've seen stronger-than-expected demand for any of your leading-edge tools if you compare with 3 or 6 months ago. And how would you think about that from a memory and from a logic perspective? And secondly, thinking about memory, you talked about strength on higher metal gate orders. Do you see upside risk or expectations next year in the event of a stronger-than-expected rebound in the memory market? And to what extent is an opportunity for HBM orders to provide a boost to the second half of this year?
So on AI, I think we can speculate which country you are talking about. But AI is really the hot topic today. And it's not just giving a boost as far as the processors are concerned. In other words, not just the GPUs are concerned, but you are seeing a significant driver as far as high bandwidth memory or HBM is concerned as well. So we do think that these are going to be very significant drivers, not just for this year but over the next couple of years. And in fact, from some of our customers, we know that they are trying to increase our capacity as far as the HBM is concerned. Now on the processor side, I think today, what you see as the GPUs, they are still being manufactured or, I would say, the relatively less advanced notes of 4 and 5 nanometers. I think we still have not seen processors jumping onto 2 nanometer. So we do expect that, that is going to help us drive the business significantly forward as well. Memory, yes, we are, of course, at this moment, in a good position because of the increase in the HBM, which requires our high-K ALD tool for the high-K metal gate to make the HBM products. And we think that if you look at the kind of forecast or information that we are getting from all of our DRAM customers, we will continue to see a constant flow of orders for the high-care ALD tools throughout this year and potentially also into next year as they try to ramp up as much capacity as possible
The next question is from Tammy Qiu with Berenberg.
Thank you, Ben, for the past 4 years. It has been great to have you. So on the 2-nanometer ramp-up, if I can just follow up with some details.Do you have already visibility? And can you share with us the magnitude and the shape of the high-volume production ramp-up at your 3 customers, i.e., do we have the normal ramp-up as we have in 3-nanometer, half and half in year 1 and year 2 or that will be slower or faster? And does it differ between customers?
Tammy, thanks again. I think each one of the 3 customers has a different, let's say, philosophy as far as -- or let's say, on how they are ramping up. I think we just saw the leading foundry and some announcements. So that is something that is there. But as far as the capacity or how much they're going to ramp, I think that is still not fully decided. For the other 2, of course, some of them are really further down the road in terms of pilots, probably less. But all 3 of them, including the leading foundry have reconfirmed, they want to go into high-volume manufacturing in 2025. So I think what we are going to see is that in 2025, all 3 of them will enter into high-volume manufacturing at some states also maybe at some point in time. And what that means for us is that equipment needs to be already delivered probably by the beginning of 2025 if they are ramping at the end. So we are working together with them on all of these plans that they have. Some of them, I think, are still being finalized. But generally, we believe that 2025 is the year of the start of high-volume manufacturing for gate-all-around.
And Hichem, may I have a question on the market share you have from DRAM, Hichem, please? Can you talk about your market share within the DRAM market? Are you having relationships with all 3 of them? Or it's still at early stage with certain DRAM makers?
I will take that. For the DRAM market share, I would say that we have -- the best way to characterize that is, as far as the high-K metal gate is concerned, we have the majority. We have the majority of the share for IT Metals India.
Is that right to understand all 3 or not really?
No, we will not go into that. But the best way to do this is we have the majority. And it also depends on what product they are producing, but we have the majority as far as the market share is concerned.
The next question is from Robert Sanders of Deutsche Bank.
I want just to say, Benjamin, it's been great. And good luck with your next moves. I just have 2 questions. One on 3D DRAM. It seems like Samsung has been pushing this 4Square DRAM and development seems to be accelerating. I just wondered what your opportunity set could be there. And then secondly, on ALD, it looks like Lam has taken some share in this molybdenum opportunity at gate-all-around. I was just wondering, are you set to -- do you have an ambition to compete heavily in that sort of metal ALD area? Or is it really more kind of land territory?
Rob, thanks a lot. So you're correct. We recently saw Samsung announcing 3D DRAM, which actually led to a little bit of confusion. It's not the 3D DRAM that we know of is 4c, yes. it will definitely have more opportunities as far as ALD. Maybe also Epi is concerned. And of course, we are working with the customers who are working in that area. So we -- in terms of timing, I think we are still maybe a couple of years out, but we are, of course, actively engaged together with them. On the [indiscernible], the metal, of course, we are bending, and we are involved with all the different layers, different customers. It's probably going to be one of the major transitions as far as the technology is concerned, and we are heavily involved in that.
The next question is from Francois-Xavier Bouvignies with UBS.
And first of all, I mean, Ben, I wish you all the best for the future as well. I have 2 quick questions. One is on the HBM. So you see the orders coming through this quarter. Can you talk about when it's going to impact your revenues? And should we expect the material impact of HBM straightaway? Or is it something that is still not going to be very visible because it's too small? Is it already meaningful in terms of orders and revenues? And when it's going to impact the P&L?
So as we have probably shared previously, sometimes the orders don't really correlate to, for example, our average lead time of 6 months. So I think you are going to see HBM orders being converted into revenue, not necessarily based on the lead time of 6 months, but it will definitely occur over the next couple of quarters. That's what we have. Now in terms of what is the material or meaningful impact. For us, memory is still the smaller part of our business, but it's not nothing. So it is helping us boost up our revenue. And depending on the magnitude of how much investment goes into DRAM, I think it will be a meaningful, I would say, addition to our revenue. And especially when you compare to 2023, we do expect a 9th rebound in memory revenue over 2023.
Makes sense. And maybe the follow-up is on China. I'm not very hopeful about this one, but can you tell us the China market share or market share, your China exposure as a percentage of revenues given the strong gross margin, that would be great. And given I don't have a lot of hope on that one, I'm going to squeeze this more the China as a whole. I mean we see Noa and AMEk, I mean growing strongly the revenues. We know that they are investing significantly in terms of localization. I mean, how do you see the competitive landscape for ALD and epitaxy specifically in China? And do you see any impact potentially in the years to come for [indiscernible]?
Sure. Francois, again, solving to this point. We, in fact, have not -- we are small. We only play deposition compared to some of our peers. And if we disclose it could be competitively sensitive information, which we do not want to disclose. You are very spot on again on the competition. So we have seen some of the domestic competitors increasing their revenues significantly over the year. Now, of course, I think that is coming from 2 main areas. One, of course, the size of the market has increased because Chinese customers are investing significantly. But at the same time, there are also certain export destructions that have made imports more difficult. And I think they have no other choice but to adjust to local domestic suppliers. As far as we are concerned, I think we are still in, I would say, a good position. Of course, with the export control regulations, we comply 100% maybe even more than that. We follow the regulations very strictly. But at the same time, if you look at the type of equipment that we are good at, the ALD and epitaxy, I think we are still in terms of performance, still ahead of the local domestic suppliers. And I think as long as we can sell to them and as long as they can buy, we are still the preferred supplier by that.
The next question is from Stephane Houri with ODDO BHF.
Benjamin, thank you for everything and Hichem, looking forward to work with you soon. Two questions indeed. The first one is on the gross margin. because you have hit a record gross margin. I understand it's notably because of China and that China is going to weaken in the second half so we can expect an impact. But is this impact going to be very strong on the gross margin? And is it going to be already in Q2, even though you said that the business should remain strong in the first half? That's the first question. And the second question is more about 2025 as you seem to say that gate-all-around is going to be a significant driver. At the same time, 6 slowdown is temporary, but also HBM DRAM is driving some growth. So why don't you upgrade your target for 2025 or at least shrink into the upper hand? Because to get to the low end of the guidance, it would probably be the sign of a dramatic year of a very bad year, I would say.
So maybe I'll take the first question on margin, Stephane, and then Ben will take a second question. You're right. So we had a record margin in Q1. 2, maybe even 3 factors. One, indeed, record high sales in China have definitely contributed because by now that's on average profitability is higher there. But also the mix in China was good. So also there, we see quite substantial differences, but that was good. And thirdly, also the mix outside of China was good. So actually, everything fell in the right direction, which is also what you saw, 52.9 million. Moving to Q2, we still expect good China sales. I don't think it will be a new record. So there might be some impact, of course, relative to Q1, but still with good sales. That in itself, everything else equal, will, of course, have a positive effect as well. I'm not going to guide now if it's below 50% -- sorry, above 50% or not. It's really too early to tell because there's also still a significant part outside of China, where there's also a mix element as you know. But you can take as a general rule, of course, if China is higher, then typically, that will -- is a positive impact on the overall mix. But you should also not discount the fact that the rest outside of China is also important. And there, we also had a strong mix in Q1. And then for the second half, China's impact is what we expect normalization is not as much as we initially believed. So China is somewhat more resilient than initially thought. So also there, of course, compared to what we thought before, that will have somewhat of a positive impact on margin.
And on your question regarding 2025. So the way that we look at it today is 2025, the significant driver for us, of course, is gate-all-around going into high-volume manufacturing. We do expect a recovery in memory. And maybe just to clarify, today, when you look at memory, a lot of the investments are just to make high bandwidth memory and more advanced 3D NAND products. There is still not many additions in the new capacity. And that still has to happen. And then there's probably a solid missing. We do expect that to recover as well. Silicon carbide, I think the verdict is still open, whether the recovery will be in 2025. China, again, we do -- again, on our cautious view is that there should be some normalization, but let's see how much that is. But based on the biggest drivers of gate-all-around, high-volume manufacturing and recovery in memory, we are very confident of our 3 million to EUR 3.6 billion targets. We don't see anything that necessitates us to change the target at this moment. But going further, if things are really picking up, if we have to change, of course, we will inform the market.
The next question is from Andrew Gardiner with Citi.
Best wishes from me as well, Benjamin. Two, if I could, one on gate-all-around and then just a clarification for you, Paul, on the OpEx. So first on gate-all-around 2-nanometer, as you guys have pointed out, another healthy quarter of orders and you're saying that's going to continue in terms of pilot line activity into 2Q. I'm just wondering whether you can give us a sense as to how that has shaped up relative to your earlier expectations for 2-nanometer. There's clearly lots of talk in the market about is it slightly delayed and what size is it going to be? And how is the process performing in the early phase of development?I'm just wondering what kind of feedback you're getting on your tools and what you're hearing from the customers in terms of their ramp. Is it going faster than you had expected? And I suppose I come back to a point you made on the last earnings call and a few times in public statements, Benjamin, about a potential acceleration. Are you seeing that? Or is that something that you still are thinking is perhaps later in the year? And then, Paul, could you just clarify the OpEx? You said it's going to be up slightly in 2024 on a net basis and the R&D is clearly up more. And so, therefore, SG&A is going to be down. I just want to make sure I got that clear.
Andrew, thanks a lot. I think our observations and expectations of the 2-nanometer moving into pilot and eventually going into high-volume manufacturing. I don't think that has changed. We are in the stage now of delivering the tools for pilot manufacturing. And we have not been told of any problems by our customers. So we have seen some market rumors that maybe there were some technical issues and so on, but not that we know. And I think in terms of the acceleration, we could still potentially see some of that. As I said earlier, if you look at where the AI chips today are made, they are still made on 4 and 5 nanometers. And that has to be -- some of them, they are thinking of making them more advanced nodes because of the lower power consumption, better performance. And I'm sure that there are some things that are cooking behind that we are not aware of. But I think a lot of the initial RAM that is going to come from 2-nanometer high-volume manufacturing could, in fact, be because of AI chips.
Then Andrew, on OpEx. So SG&A for the full year, so '24 full year compared to '23 full year. I would expect a marginal increase less than 5% year-on-year increase. Net R&D, I mentioned 10% to 20%. Important to note, Q1 net was low growth. You saw a significant increase, and that was still relatively low because we did not yet start amortization of a few significant projects. That will start in the course of Q2, not all of them on day 1. So Q3 will again be higher than Q2 because then you have the full quarterly run rate of the annuity amortized projects. But taking that into account with this increase in amortization, net R&D is expected to be up 10% to 20% year-on-year. Gross R&D will be up somewhat less than that because gross growth there doesn't have the impact of the relative higher increase in amortization relative to capitalization. I hope that's clear.
The next question is from Janardan Menon of Jefferies.
And Benjamin from my side as well. Also great working with you and all the best in your future endeavors. My first question is on the gate-all-around side once again. You've been taking 3 quarters of pilot orders right now, and you're saying that Q4 also is going to -- I mean Q2 is also going to be pilot. So that's about 4 quarters of pilot orders. And you're saying that it's because there's a 3-nanometer was one customer, now it's multiple customers and it's getting spread out a little bit. So when you're talking about high-volume orders on 2-nanometer gate-all-around in the second half, is that all more from one customer? Or do you expect all your customers to be placing those high-volume orders in the second half of this year? And in terms of delivery, you said that you expect gate-all-around to rise quite a bit in the second half and compensate for some of the China declines that you still expect. Is there any chance that you'll take an order in Q3 and ship in Q4 on the high-volume side because there's not a clear match between orders and delivery? Or will your high-volume orders really start being delivered only in Q1 2025 and all the upside on the second half gate-all-around will be coming from pilot order delivery?
Janardan, thanks a lot. It's been a pleasure. So you are correct that we have seen several quarters of [indiscernible]. And I think that's what we are seeing in terms of the spread of a potentially also in Q2. So as you correctly mentioned, over the last 4 quarters. Second half, high-volume manufacturing. We do expect second half that we will be getting the orders. We will not disclose whether it's one customer, 2 customers, or 3 customers, but we will be getting second-half high-volume manufacturing orders. And to your question on is it possible that we get an order in 1 quarter and deliver the next quarter, it is possible. And like I said, the orders, especially from our big customers, sometimes they work together with us based on forecasts, so the orders can come late. And generally, we are expected to build according to the forecast that they give us. So it is possible.
You said you're seeing strength, but you're also seeing some pushouts. Would it be fair to say that the continuing strength because your 200-millimeter orders in Q1 came from China? So is it your Chinese customers which are continuing to invest quite strongly and some of the pushouts are coming from outside China?
I suppose, Janardan, you are referring to silicon carbide?
Yes, silicon carbide.
So the situation, and Jonathan, you are quite correct. We still see some of our Chinese customers investing and they are investing to prepare for the rebound or the recovery and they are also investing for the transition to it. So that is correct. I think the push out, it is happening, I would say, across the board because of the slowdown in electric vehicles. But I think, in our opinion, this is maybe temporary. And when those start coming back and we all believe that the world needs those electric vehicles, I think see the recovery there. For the time being this year, a couple of things that are helping us to achieve a meaningful double-digit increase. One is we ended the year with a nice backlog. Our Chinese customers are still continuing to invest. And last but not least, we are also benefiting from the non-Chinese customer wins that we have had and which we have reported over the last couple of quarters.
The next question is from Didier Scemama with Bank of America.
And allow me to thank you also, Ben, for the last 4 years. It's been a fantastic ride and amazing execution. And please also allow me to congratulate Hichem for your promotion, Mabruk. I've got a question, in fact, for Hichem if you can take that. I don't know, Ben, if you want to take it. Now that the 2-nanometer is sort of locked in and we're moving into production. What's your earlier view of the market share you have in ALD and Epi for the 1.8 nanometer process recipes? I mean historically, given that there is not a huge amount of disruption and changes between 2 and 1.8%, I would imagine that your share should be broadly stable, plus or minus. Is that the right way to think about it? And I'll have a quick follow-up.
So if you look into the ALD market share between 2 and 1.8 nanometers, I think what's important to say is that for ALD, we really have maintained our market share going from 3 nanometer to 2-nanometer. And in Epi, we have increased our market share going from 3-nanometer 2 nanometers. But as you transition from 2 nanometers to 1.8 nanometers, you see more and more ALD happening. And for that, you see more and more ALD high-performance ALD layers, which we are working with our customers, and I think we are very well positioned for these applications.
And my follow-up is on the target model for '27 because we all focus on '25, but it feels like it's in the back by now. So '27, just to remind us, when you gave that guidance of EUR 4 billion to EUR 5 billion for 2027. How much of that was motivated by AI, both on the logic and HBM side?
Yes. Maybe I'll take that one. If you go back to our presentation, you see that by 2017, I think we said that 40% or so of investments in gate-all-around is 2 nanometer is leading edge related. And of course, I think you also made a comment on -- I forgot percentage on how many percentage of devices, we believe that they will have one way or the other in AI functionality. So we have made some assumptions. Are they wrong? 100% sure there on. Will it be higher or lower? That is still hard to tell. I mean we have ultimately taken the overall WFE as a leading, let's say, guidance. From there, we calculate backward, of course, on the different nodes and technologies that we believe how this market will be composed. And from there, literally bought them up layer by layer. We do our modeling, and we validate that again, top-down with what we see happening in the various market segments, and that's how we validate that. Will that be different? For sure, there will be pluses, there will be minuses. It's really too early to tell.
The next question is from Adithya Metuku HSBC.
Best wishes to you, Benjamin. So firstly, just on HBM. I just wanted to understand how the tool demand from the high-K DRAM side changes. There's been this talk of 3x the wafer account for bit growth. Does that in any way -- does that mean that you need 3x number of tools when you go from HBM versus non-HBM when you see high-K metal gate cells? That's the first question. And secondly, just on 200-millimeter silicon carbide epitaxy, [indiscernible] made some recent announcements on wins. Would it be fair to assume that batch tools will still be used at 200-millimeter and it won't all be single-wafer Epi-tools when the industry transitions and issues opportunities on silicon carbide epitaxy with single-wafer tools mainly with Chinese customers? Would that be a fair assumption to make? Those are the 2 questions.
Addy, thanks a lot. And on the HBM side, so I think the better way to understand this is when you look at, for example, 2023, HBM was still a very small portion of DRAM total. I think it was maybe just a mid-single digit kind of percentage, and the market generally is projecting that this is going to go up to -- well, based on some of the forecast that I'm reading, it may go up to as much as 20% of DRAM by the end of this year, early 2025. So if you look at that, there has to be quite some capacity that needs to be either built or converted to be able to make a high-bandwidth memory. And so far, I think what is happening is that most of the -- in fact, all of them are converting to some extent, excess capacity and buying some additional tools like our high K/ALVtools and being able -- not converting them to be able to make [indiscernible] memory. Now as to whether it's going to be 3x wafers or whatever, I am not too sure, and I haven't seen the real correlation between a number of wafers, but I think overall, moving from a mid-single-digit to let's say, 20%, that's a fairly nice increase as far as HBM and demand is concerned. We always look at our competitors with respect and that is the same for silicon carbide. I'm actually not aware of which wins they have announced because I think most of the wins are 200 millimeters, we are well aware of that. So maybe that's something that we are not aware. But in general, I think you're going to find that whether it's bad or whether it's single wafer, there's going to be a play for both. It's just a question of eventually what is the cost of ownership. And I think even though it has been kind of mentioned that single wafer has a higher cost of -- or higher costs, I think we have done a lot of work, and we are very, very competitive. And on top of that, we have excellent performance. And this is actually one of the reasons why we have been winning customers outside of China in both Europe and also in North America, and we continue to be confident of our ability to grow our market share even from here.
The next question is from Jan from Nigel van Putten, Morgan Stanely.
I have 2 follow-ups on China. First is if you could discuss what your main exposure is in China at the moment. Is it still more power analog or rather mature logic or even memory? And then also maybe what has changed in the last quarter, resulting in the pretty significant uptick in the gross margin, that will be the first question.
Nigel, thanks. I will take the China question. I think in terms of exposure, as we mentioned, power analog and wafer is generally down. And that includes also China. So the biggest part of our exposure today in China is probably logic/foundry. And we think that to your question on what has changed. We were kind of cautious as to whether they will continue to invest at the type of capacity or, let's say, a level that they have been investing over the last 2 years. But so far, it seems like they are still continuing. And we still do expect some degree of normalization in the second half, but because of more recent information that we have, I think the moderation is much less than what we had anticipated.
On the margin, Nigel a repeat of what I said earlier. I mean, one record high sales. So we never had so high sales in China as in Q1, which contributes positively to the mix in China, believe it or not, but there is also quite some margin differential from product to product case application. And three, also the mix outside of China was actually very strong this quarter. So adding that all up, we got to, I would almost say an exceptional 52.9% grade margin. I wish I could deliver it every quarter, but I'm afraid that will not be the case. But these 3 factors pushed up the margin.
Yes. Understood. And then as a follow-up, you have already hinted about it, your view into the second half for China. I think previously, it was more based on assumptions around normalization. But as the second half now comes into view, you've alluded to new information, but I guess also the order book would now sort of reflect what you can deliver in the next 6 months. So to what extent would you say your view is still conservative? Or is it now more based on actual orders in hand? And there's like less of a chance of that turn out to be conservative, as I think happened in the sort of first half. Sorry about that.
So you're correct. As we -- of course, we are now in April, at the end of April. So we do have a better view. And this is the reason why we have guided that the second half, we still expect to see some degree of normalization. It's just that it is less than what we had expected. But there will be, and China will be in terms of the second half lower than the first half.
Also from my side, thanks for the discussions and all the best in your future endeavors and also Hichem, look forward to working with you. Thank you.
The next question is from Marc Hesselink with ING.
First, on DRAM. So making good progress on HBM. I think in the past, we spoke about that, eventually, it can also move into the less high-performance DRAM market. Can you share anything on -- do you still expect that to happen? And what kind of time frame that also the traditional DRAM market can be more significant for you?
Marc, thanks a lot. I think there is still the possibility. I think the question today is how fast they want to move. I think if you probably speak to our customers, the DRAM customers, today, they are just really focusing on trying to drive high bandwidth memory. That is today the product for them. And how much they want to, for example, also adopt a high-K metal gate for automotive DRAM and so on. I think that's still left to be seen. But over time, as products become technically more complicated, I would not be surprised that they will have to do that. But the focus today by all the DRAM manufacturers is really how much they can turn out in terms of high bandwidth memory.
And my other question is on the market share in Epi. You said that when that is going to go up significantly going to 2-nanometer. Traditionally, you spoke about moving from 15% to 30%. I guess there are a lot of moving parts also with the traditional market being more strong than probably earlier expected. But is that 30% market share, something that you can still expect? Or you will grow into that more into the 2027 period?
We had mentioned last year in our investor meeting in London that we expect to achieve 30% market share in Epi in 2025. So that's really still our goal, and that's our expectation. Going forward, it's really anybody's call from this point of view. But what we can tell you is that we see extremely good momentum for us in our Epi technology and more and more customer wins in the future. So we are confident of our Epi technology, and we think that we have some technology differentiation going forward.
The last question is from Timm Schulze-Melander, Redburn.
I had one sort of quick housekeeping question and one and a quick follow-up. Just maybe, Paul, you talked about amortization and capitalization. Just my sort of housekeeping question is kind of what kind of scale of amortization should we be thinking of for the year. And then also just a final one on the margins. Obviously, a lot a lot of focus there. You've been very clear about the contributors. China is the only one that I hear you sort of suggesting is going to moderate in any way, really. The application mix should really stay very strong and potentially even build further here. So I hear you also on quarter-to-quarter gross margin, but I just want to just really double-check. There's nothing that we should have in our mind that's potentially going to unwind that you can already see going into '25, '26, or 27.
Thanks for the question. Maybe on amortization, although this is not meant to be guidance, but to give you some more clear indication if you take Q1, so which we have disclosed the actual amount of amortization I think it will not fully double towards year-end, but get close to that. So it's a significant increase, and it won't happen gradually. I think Q3, as I said, we should be at a call for our full run rate because, of course, over time, things will most likely further increase given the amount of gross R&D to be spent. But if you take Q4 and Q3 compared to Q1, they're not fully doubling, but let's say, towards around EUR 20 million because we had EUR 12 million in Q1. So it should give you a reasonable indication. Then on margin, yes, it's always difficult because I have to remind, I think you almost if you recall it, if you go back let's say, 2 years where China was much less of a, let's say, dominant play in the overall margin development. We still saw quite some significant changes from quarter to quarter. So there could be up to 2% changes from one quarter to the other simply because of product mix. And likely that is not going to take and some quarter these things for all to the right direction. Some is a mix and some they fall to the wrong direction in terms of margin. So it is really very hard to tell. And it's not because I don't want to give you any more information. It is difficult to guide. And then I just want to repeat, of course, China, given the share of China now in Q1 and most likely also in Q2 and then in H2, something less, yes, that will still have an important impact on the margin, but I don't expect it to be as large as in Q1, simply because I don't think that we have -- we will hit another record level of China sales in the course of this year. I'm not excluding it, but I don't see it at this moment in time at least.
This was the last question. I turn the conference back to Mr. Benjamin now for any closing remarks.
On behalf of Paul, Hichem, and Victor, I want to thank you all for attending our call today. Stay safe and goodbye. Thanks a lot.
Ladies and gentlemen, thank you for joining. The conference is now over, and you may disconnect your telephone.