SMHN.DE FY2025 Q3 Earnings Call Transcript
Date: 2025-11-08
Source: Financial Modeling Prep

Operator: Dear participants, we warmly welcome you to today's conference call of the SUSS MicroTec SE following the publication of the 9-month results of 2025 earlier this morning. SUSS is represented by the CEO, Burkhardt Frick; CFO, Dr. Cornelia Ballwießer; and COO, Dr. Thomas Rohe. The Management Board will speak shortly and guide us through the presentation followed by a Q&A session. But before we start the presentation, let me hand over to Sven Kopsel from Investor Relations.
Sven Kopsel: Thank you, Sarah. Yes, and many thanks. Welcome to our Q3 conference call. As you probably know from earlier calls, this call is again being recorded and considered as copyright material. It cannot be recorded or rebroadcast without permission and participating in this call implies your consent to this procedure. Please be aware of our safe harbor statement on Page 2 of the slide deck. It applies throughout the conference call. And now I hand over to our CEO, Burkhardt, for some opening remarks, followed by our CFO, Dr. Cornelia Ballwießer, presenting the financial development. Burkhardt, please.
Burkhardt Frick: Thank you, Sven, and many thanks, and welcome, everyone, to this call. I will go a bit faster over the next few slides to have more time to focus on the margin analysis you guys are all interested in, I'm sure. We showed the next page, we showed this exactly this page already 9 days ago in the extraordinary call. So nothing new here. The changes -- there are no changes to the figures since then. We also mentioned the low level of EUR 70 million in orders received in Q3. After various customer meetings in Korea and Taiwan last week, I'm very happy to report that activities are picking up in the fourth quarter. Orders exceeding EUR 100 million are likely. We do see quite some momentum here. We already communicated last week about the pressure on margins and the fact that we had to adjust our guidance for the gross profit and EBIT margins once again. I will go into details of margin development in a moment. However, I would like to state that the current margin pressure does not impact our 2030 ambitions. We will present our new midterm expectations at our CMD on November 17. Last week, the development of our 2 segments was not yet included. So I'd like to highlight a few things here. First, Advanced Backend Solutions. The order intake remains strong for coaters, but this was not quite enough to offset the decline for bonders. The demand for our UV scanners remain intact. Imaging and Coating Systems showed year-on-year sales growth of larger than 50% each. Bonders still showing slight growth after 9 months. Gross profit margin significantly impacted, more on this shortly. Photomask Solutions, we have a very low order intake again. Orders from China now down EUR 32 million versus previous year but more significant orders expected in this Q4. Still high year-on-year sales growth, but Q3 sales was lower than expected. Unfavorable product mix is the main reason for low gross profit margin of 31.7%. Now we have prepared 3 pages where we compare our initial 2025 guidance for sales, gross profit margin and EBIT margin with the actual year-to-date 9 months figures. Firstly, on sales. After 3 quarters, we reached EUR 384 million or 78% of the midpoint of our sales forecast and therefore, are on track and achieved what we expected to do. Q3 sales, as expected, was EUR 118 million, lower than previous quarters. Reason here lower order intake in the first half of 2025. In the fourth quarter, we need sales of EUR 85 million to EUR 125 million to meet our forecast. EUR 105 million would, therefore, leads to the midpoint, which is EUR 490 million. The product mix is different as planned at the beginning of the year with more coaters and fewer bonders based on orders received in the first half of the year. The recent postponement of 2 high-margin projects to 2026 will have a negative impact on gross profit margin in Q4. Now we'd like to provide more transparency on our negative gross profit development. Let me first explain the methodology we applied here. The table on the left shows our actual figures for the first 3 quarters. These are the left columns and a projection of what our gross profit would have been if actual sales had a gross profit margin of 40%, which is the midpoint of our original forecast of 39% to 41%. Our analysis shows we have a gap of EUR 16 million, which we like to explain. On the right-hand side, we allocate these EUR 16 million to special effects, quantify them, specify the timing and if these effects can be considered as one-offs or not. From top to bottom, first, the UV scanner in Taiwan, the ramp we performed there in the first half of the year. We had extra expenses for training and supply chain efforts amounting to EUR 3.2 million, and that's a one-off. Secondly, we had a write-down on discontinued technology projects amounting to EUR 2.2 million that affected Q2. Also, that is a one-off. Expenses for our new site in Zhubei, EUR 1.2 million for double rent relocation and utilities, they affected us only from Q3 onwards. And they will have -- this will have an impact on expenses in Q4 as well as in Q1 2026. Rework during assembly and customer ramp-up support amounting to EUR 2.4 million since Q1 were necessary to support customers to improve performance of recently installed multiple lines and maximize the output and availability of these in the field. This was really important and is an ongoing effort and it also will open the door for follow-up business, which we are, of course, looking forward to expect. The last point is the unexpected product and customer mix changes, which we often use also to explain deviations in our margin. This is for more coaters, less bonders, many low-margin photomask tools for key customers, and that results in also lower fixed cost coverage due to lower sales and overall business activity. That amounts to EUR 7 million in Q3. In total, EUR 16 million of which slightly less than half can be characterized as one-offs. Now on this page, we focus on the EBIT. We applied the same methodology. Left column shows the actual development of first 3 quarters, right column, the projection with midpoints of initial gross profit and EBIT margin targets, which was 15% to 17%. The gap here is EUR 7.2 million, which means that more than half of the gross profit gap of EUR 16 million was offset by stricter cost management and a positive balance in other operating income expenses. According to the original guidance, we allowed for OpEx of EUR 92.3 million after 3 quarters and would still be on track to achieve the original EBIT margin targets. The actual OpEx, that is expenditure on R&D, sales and administration amounted to EUR 86.8 million. This shows our short-term cost-cutting measures are having an effect, savings of more than EUR 5 million compared to Q2. In Q4, OpEx is expected to be below EUR 30 million. However, most likely above Q2 level based on increased expenses on IT and digitalization projects as well as rising R&D costs also to support scheduled product launches. I think I said above Q2, I should have said above Q3, right? Yes. We will correct this, and you will see it also in the tables. Now after all these numbers, here are a few impressions from last week's opening of our new site in Zhubei, Taiwan. It was an amazing day with a great atmosphere. We welcomed over 100 guests, including Taiwan's Vice Minister of Economic Affairs, a C-level representation from a leading HBM manufacturer and management from the top foundry in Taiwan. We got a broad confirmation that it's important to increase our presence close to the heart of the semi industry sector. We introduced our large clean rooms and made it clear that we are set for future growth. First modules and tools are already being built in Zhubei and will be delivered to our customers in early 2026. Leases for all old locations will terminate by the end of Q1 2026. The financial double burden will also end at this point. And with this, I'd like to hand over to Cornelia to provide some more insights on our financial performance.
Cornelia Ballwießer: Thank you, Burkhardt. After we've already discussed Q3 in detail, I will just summarize some additional developments on the next slides. We already talked about the slow order intake, which leaves us with an order book of EUR 276.1 million as of end of September. This is 35.9% below the level of the first 9 months of last year. Tool orders with roughly EUR 140 million are scheduled for delivery in '26. The visibility for '26 is improving. Our free cash flow from continuing operations came in at minus EUR 0.7 million in the third quarter with operating cash flow of EUR 5.9 million and cash flow from investing of minus EUR 6.6 million. After 3 quarters, free cash flow is now at minus EUR 28.2 million. For the full year, we still see potential to generate around EUR 28 million of free cash flow so that we could end up at end '25 in slightly positive territory. Total CapEx for the 9 months is EUR 17.8 million, mainly driven by our new fab in Taiwan. At the end of the year, we expect to land at CapEx level of EUR 25 million. In '26, we will return to a level of clearly below EUR 20 million. Without additional projects, the level will be approximately at EUR 10 million. On this slide, you see the development of our most important key performance indicators for the last 7 quarters. You can very clearly see the margin development, especially in the last quarter due to the effects we already talked about today. On this slide, you see the two segments. In the Advanced Backend Solutions segment, margins in the third quarter were roughly at the same level as in the previous quarter. Burkhardt already mentioned the most important drivers. In Photomask Solutions, the margin level is in the first 2 quarters of the year higher. Overall, we're still at 38.4% gross profit margin for the 9-month period. However, the third quarter was weak, mainly due to an unfavorable customer mix, as already explained. Here, you see our order intake by segment and regions. The book-to-bill ratio continued to remain at a very low level of 0.62 for the 9-month period. This is, of course, far too low for a company with growth ambitions as we do have. But as already discussed, we expect increasing orders in Q4. Demand from China continues to be very low. The China share of total order intake in the first 9 months of '25 is now 18.5%. In '24, also after the third quarter, the share was at roughly 30%. But generally speaking, we do not have major shifts in the order intake by region. Finally, let's go over the main developments of the balance sheet. Total assets increased by EUR 22 million. For the noncurrent assets, the main driver was the Taiwan expansion with the right-of-use asset for the site and further installations at the site as well as CapEx in Germany, which we already showed in our half year report. Current assets, we have a decrease by EUR 29 million to a total volume of EUR 413.3 million. Inventories declined and are now EUR 12.9 million below the value of end of December '24. Contract assets and trade receivables increased by EUR 22.7 million. Cash and cash equivalent decreased by EUR 41.8 million due to free cash flow in total of minus EUR 31.5 million and the dividend payment as well as repayments of financial debt, including the leasing liabilities. On the liability side, the main changes also happened in the first half of the year with the inclusion of the leasing liability from the Taiwan site. In noncurrent liabilities, the major driver in the 9-month period was also the inclusion of the lease liability for the Zhubei site, which already happened in the second quarter. Current liabilities decreased. Here, the major drivers are still lower advanced payments from our customers who supported last year's ramps and less orders from customers, which have prepayments. After the 9 months, the equity ratio is at 58.2%, which means we improved the equity ratio while we had our ambitious investments. Burkhardt?
Burkhardt Frick: Now let's turn to the outlook for 2025 as a whole. First, here is a page that was already shown last week with the reduced guidance ranges for gross profit margin and EBIT margin. Everything stays the same as communicated last week. Last week, we already explained that we are discussing possible measures to sustainably improve the cost structure. However, I ask for your understanding that all decisions will be carefully considered. I do not currently expect that we will be able to communicate these possible measures already in 2025. For now, our full attention lies on Q4 to bring in the anticipated new business and set the stage for 2026. We are now opening the floor for your questions. Thank you.
Operator: [Operator Instructions] We will start with the first raised hand with Janardan Menon.
Janardan Menon: I just want to go back to the order increase that you're expecting in Q4. 9 days ago, you had said that you would see an increase in orders. You said above EUR 100 million is possible. But at that point in time, you had also just commented that your Q4 is always typically quite strong. You've seen a very healthy double-digit increase in quarter-on-quarter in your Q4 orders in both 2024 and 2023. So my question is, this increase that you are expecting in Q4, is it purely a seasonal thing? Or do you see an underlying trend of improving orders amongst your customer base? And -- especially, you have been seeing quite low orders on the temporary bonding side. And one of your big customers looks like he's -- they're getting qualified or have got qualified, who knows. And so is there a clear upswing that you see in that market? Also on the UV scanner, are you seeing an upswing? What I'm trying to get at is the sustainability of this order. I mean it may not be huge, but does Q3 mark the bottom and then more than the seasonal, are we getting a more improvement into next year? Whatever your current thoughts are? Second question is just on the margin. Just trying to piece together the whole thing. You'll end up at about 36% gross margin this year based on your guidance. Are you -- do you think that as some of those one-offs go away in the first couple of quarters of next year? You're likely to get to a higher margin than that? Any kind of color on where we could expect based on current expectations? Where you assume your sales are down in line with consensus for next year? Where would your gross margin end up for next year? Any thoughts there would be great.
Burkhardt Frick: Yes. Of course, we have to be careful in forward guidance, but let me start with the order intake. Yes, there has been some seasonality in the past years. But of course, customers order when they really have demand. And so therefore, I would not really call it seasonality at all. I would rather see it as a consequence of activity in the AI space picking up again. And that has been, of course, communicated for the frontline AI players already a quarter earlier, but it takes a while until this goes through the entire equipment chain and also leads to orders. So there's not an immediate effect at the moment a big memory supplier gets qualified or post their future plans, it will not immediately trigger orders. This is more a question of how utilized are your lines, how much throughput can you get on the existing lines and when is the next window to increase? And that seems to now nearer than before. And that's also why we are confident that we get AI-related orders in the first quarter and especially after those discussions we had with our lead customers. Now this will be a mix, of course. So there will be, of course, HBM-related orders, but also CoWoS or packaging-related orders requiring multiple systems, but we see a clear upward trend. How big this one is, as I said, well, I feel confident that it will be larger than EUR 100 million that -- I stick to that number. How large we have to see because we also have to make sure we can also deliver and build these machines on short notice because the demand is required on short notice. On the second question on the margin expectations, I can hand over to Cornelia. But of course, we want to improve our margin performance. There's no doubt in that. But even in line of potentially declining top line, we have to make sure that we do this with good sense.
Cornelia Ballwießer: In terms of margin, of course, our ambition is to have a better margin or to achieve a better gross margin in '25. What I can say is it is probably lower as '24. Currently, we are preparing our budget. And as you see and as explained, the margin depends on the customer and product mix, and we are working on this. And that's all I can say for the moment. Regarding your one-offs, yes, there are, of course, one-offs that will not occur again in '26. For example, the write-down of the discontinued technology project, then our double rent relocation and utilities costs in Zhubei, in Taiwan will end, end of the first quarter '26. And yes, the rework, we will see. It depends how we can satisfy our customer or what is needed. But that's what I can say regarding the margin for the next year.
Operator: So -- and then we move on with [indiscernible]. I can see that you're unmuted, unfortunately, we cannot hear you.
Unknown Analyst: Can you hear me now?
Operator: Yes.
Unknown Analyst: Yes. Great. Sorry for the background noise. A few questions. On the order backlog, can you give a little bit the split in ABS segment? What is the CoWoS, the scanner part in the order backlog? And then in the cleaning equipment market, what is the part of the China business in the PS segment? In the backlog, right, not for order entry.
Burkhardt Frick: Yes, we are not being specific on the individual products on our backlog. Please accept that because we do give this granularity. The China portion, of course, is declining, as already previously mentioned. We see it in both in sales, but order intake significantly. We have for China, for example, only 18.4% of the order intake are China bound. For Taiwan, for example, in contrast, it's close to 40%, that's usually what we can disclose. In terms of further information on the backlog, we have, of course, also announced that EUR 140 million of the current backlog is already bound for 2026. And we can also safely state that we have about EUR 20 million in service and upgrade business also for '26 already slated.
Unknown Analyst: Okay. Maybe let me ask a little bit differently. On your CoWoS, I think the scanner is a little bit older technology generation, right, if I understood that correctly. And the question would be, what are your lead times? I mean when the customer places an order with your scanner business until you ship and final acceptance, what is the time lag there for the scanner business?
Burkhardt Frick: Yes. For scanners, of course, it's around 6 months. But of course, as we stated also in previous calls, we tripled our output capability this year. That means also we are pretty full in that sense. So that's also why we concentrate on our main application field, which you rightly state is CoWoS. Now of course, we also get inquiries, how quickly can you top this up. And that's exactly the discussions we are currently having with those lead customers because they expect basically deliveries already as early as in Q1 next year. So right now, we have very active engagements with these customers who also realize that our lead times reduced, but I think they are waiting really until the last second how to place orders. And then we also have to make sure that we can react very quickly, and that keeps us busy. But that's also causing a bit positive momentum of the last days.
Unknown Analyst: Okay. And would it be fair to assume that the gross margin, the product mix impact was also due to this, yes, high volume ramp in scanner business and that this is a little bit more service intense for you in order to have the machines up and running with your lead customer, and that might change with the second generation of the scanner tool you are planning to introduce next year?
Burkhardt Frick: Yes, it definitely will change with the next generation of scanners. But we need to distinguish between product margin and supporting efforts. So I think the supporting efforts of our scanner are not higher than other 2.5D or HBM type products. So you need to account for that. For some of our products, our support efforts were higher than anticipated, which I explained earlier, which caused the extra cost. But I mean, you're absolutely right that the scanner is not our highest margin product.
Unknown Analyst: Got it. And then final one. If you look at your product mix or backlog, what you have right now and the EUR 140 million for 2026, do you expect that the share delivered from your Asia business will be substantially different from this year? I mean that you have much higher shipments in your Asia locations than here in Europe? And if so, what would be the incremental there, the incremental shipments?
Burkhardt Frick: You mean shipments from or to.
Unknown Analyst: No, from your Asian manufacturing footprint, right, your fabs in Asia.
Burkhardt Frick: Yes. Thank you. So first of all, our regional mix will not change, except what we explained, the decline of the China portion. In terms of the products we manufacture out of Asia, they are the same products we are currently manufacturing. But of course, this can change if we are introducing new products. As you know, we are launching up to 5 new products next year. And we have to see also where we will produce those products. So there's a fair assessment, a fair judgment that the amount of products will increase, which we are going to produce in Asia.
Unknown Analyst: But you cannot quantify like EUR 50 million more sales from your China -- Asia footprint and versus this year, it's not possible right now from your backlog?
Thomas Rohe: No, that's -- I can answer this. We use both sides really pretty flexible in terms of where we do have rich capacity. So we try to leverage our load of factories in both sides as well as in Asia as well as in Germany.
Operator: So by now, we have 4 participants left who raised the virtual hand. So please be patient. And the next one who is able to ask his question is Michael Kuhn.
Michael Kuhn: I'll start with one on the guidance once more. If I just use the midpoint of your sales and gross margin guidance and then combine it with the midpoint of your EBIT margin guidance, I'm ending up at Q4 OpEx of EUR 34 million, which is clearly above the less than EUR 30 million you're envisaging for the final quarter. So let's assume you do midpoint sales, midpoint gross margin, is it fair to assume that you would rather end up at the upper end of the EBIT margin range, excluding obviously any one-offs you might book in the fourth quarter?
Cornelia Ballwießer: We calculated various scenarios over the last past days. And if we achieve the gross profit margin in the middle of the range, let's say, 36%, it is likely that the EBIT margin will end up above the middle. Yes, could be.
Michael Kuhn: That is good to hear. Then one more in the context of OpEx. So we are obviously in the upper 20s run rate-wise right now. This is still including some double costs. At the same time, I guess, IT costs will rise into next year. From today's point of view, what would you think is a realistic OpEx run rate to assume for next year, maybe from the second quarter onwards when you don't incur the double cost in Taiwan anymore?
Cornelia Ballwießer: Yes, good question. Our ambition is that we have a run rate, let's say, EUR 30 million.
Michael Kuhn: Around EUR 30 million. Okay. And last but not least, you mentioned product launches already. Obviously, those include new products in the Photomask area, including the mid-range product. Do you think part of the softness you see from Chinese customers right now is due to those customers waiting for those products? And that said, is there a chance of, let's say, a little China revival at some point next year once the new product range is available for orders?
Burkhardt Frick: China revival sounds like the rolling stones in concert. But I -- obviously, the mid-end range of the mask cleaner is really geared for nodes between 30, 90 nanometers, which are the predominant nodes China is running on. In the past years, they bought very high-end equipment, which was basically overspecced because they don't have EUV equipment in China. So the mid-end range is a better fit for the Chinese market. So yes, we do expect that, that business will pick up once that system is in mass production. And we already have several reservations and quite some are out of China. But also, of course, this mid-end product is interesting enough to replace the aging fleet of old mid-end mask cleaners. Therefore, there is also quite some replacement need lining up.
Operator: And then we will move on with Madeleine Jenkins.
Madeleine Jenkins: I just have one clarification. The customer that is pushing for kind of expedited deliveries in Q1, I think you said. Is that memory or logic?
Burkhardt Frick: It's fair to say both. It's not a single customer who is pushing.
Madeleine Jenkins: Okay. And then in terms of -- on the kind of HBM side specifically, are you still running at like underutilization at your big Korean customer? Or is that kind of back to the levels where you'd expect incremental orders?
Burkhardt Frick: Well, I think one -- we have 2 out of 3 HBM players. And one is really running at full swing. And then, of course, that's also the one which kind of further scales up. The other one, of course, is just about to accelerate again, and they still have, I would say, headroom left. So we don't see short-term excess business coming up there because I think they're not running at peak utilization.
Madeleine Jenkins: Okay. So the kind of Q4 orders isn't necessarily driven out of Korea. Is that fair?
Burkhardt Frick: Correct.
Madeleine Jenkins: Okay. And then I just had a -- you've got a high-NA cleaning tool, Photomask cleaning tool coming out. Could you just give us a sense of kind of when you expect the first orders for that? And also what sort of ASP uplift versus the low-NA version?
Burkhardt Frick: Yes. Madeleine, you're referring to the MaskTrack Smart cleaning platform, which is launching pretty much as we speak. So we are working with some lead customers who want to position this system in kind of -- it's more than just evaluation. It's kind of early production state. So we do expect that we get the first orders still this quarter for this first system. But we are, of course, in the middle of the negotiations, and it's important that we get this first volume customer order for that system, but we anticipate it this quarter.
Madeleine Jenkins: And just on the ASP...
Burkhardt Frick: Sorry, say again?
Madeleine Jenkins: Just on the ASP, is it kind of significant uplift versus the last generation?
Burkhardt Frick: It is somewhat more expensive than a MaskTrack Pro. But as you know, it highly depends on the configuration. So this is a tool which can be configurated to a larger extent and therefore, will be also more expensive than the existing platform.
Operator: And then we move on with the questions from Johannes Ries.
Johannes Ries: Also some follow-on questions to the cost side first. Maybe first, what -- the leasing cost for the old production side, which will fall away at the end of Q1, how high is this maybe regarding to the full year or for the remaining 9 months. Therefore, what is maybe the positive impact? Then maybe on the bonders, if the bonders recover, will they have the same margin like in the past, there have been maybe some special high prices regarding the shortage or maybe the urgency at the customer side to cut the products in the past. So are you achieving the same pricing at the temporary bonding side like in the past? And on the coaters, is anything possible also to increase the margins there because it seems that they have comparable low margins. I know there is more competition from Tokyo Electron, for example, but maybe also an update there. And you talked a little bit now on mask cleaners. How is the ramp for all the new products with better margins, the scanners? I have also something like you have a new coater coming on the market for next year. That's maybe all impacting a little bit the cost and the margin side. Therefore, I took all these questions in one.
Burkhardt Frick: Yes. Thanks, Johannes. That's a lot of questions. Let me try to start taking them down one by one. So the bonder orders, of course, we had at the very early phase of the ramp, they did have somewhat better margins because we were -- these were rush orders. We had to expedite things. So once we got into real volume phase, also we had more volume prices applied to that. So the initial systems were more profitable than the volume systems. But this has stabilized now, so we don't anticipate unusual things there. So they are above average compared to the rest of the portfolio. On coaters, we have -- we keep getting stable repeat orders from existing OSAT customers. And that is a very stable business and also this customer continues to place these orders. There was also one of the customers I visited early last week. So we can also expect a good solid business there. You are absolutely right. The competitive situation is very strong. But when you're a tool of record, you at least can retain your seat, but you have to price competitively. And that's why coaters usually are more on the average spectrum of our margin. For the Photomask tools, we are launching, so the new systems, they are completely redesigned. They do have a different margin structure, but you cannot just increase margin without offering new features. So it's always a mix of both. Then I think you had a question on the rental cost, right?
Cornelia Ballwießer: Yes. The impact of the additional rental cost for the old site rental cost that turns out in a positive impact next year is EUR 600,000 per quarter.
Johannes Ries: Per quarter, okay. And when will the scanners be launched this new scanner generation, will it happen in the first half next year?
Burkhardt Frick: No, I think that's a bit too early, but we will deliver the first system around, yes, mid next year to the first customer. And that's, of course, we get more feedback. The broad launch of the system is more towards the end of next year.
Johannes Ries: Okay. Super. And also maybe there's definitely much more but not to go in too much details. The wafer cleaning product will also not launch next year or will it come over the next year?
Burkhardt Frick: They will launch next year. And we kind of -- we get the first hardware at the turn of the year. And then, of course, we need to refine the processes. We have one lead customer who will start evaluating. And then we will have not only the volume tool because the first one is a 200-millimeter wafer cleaner, low volume, there will be high-volume tools coming shortly after. And we have -- since we kind of got quite some customer traction, we have now 300-millimeter customers interested in that tool as well. So we are also now checking how fast can we launch a 300-millimeter tool. So wafer cleaning will be a family of tools, the first one coming next year.
Johannes Ries: Super. Great. Maybe also on our calculation for next year, you mentioned you have on top of the EUR 140 million in product backlog for next year, you have also 20-point something on service and spare parts. What is the normal number for service and spare parts for the whole year? I think it's more than EUR 20 million.
Burkhardt Frick: Yes. Johannes, usually, it's about 15% of the total revenue. I think the numbers, I think we stated before were, of course, the first 9 months and then the portion of 26 out of those first 9 months. But I think it's -- you can roughly assume 15% of the total revenue is the service-related part.
Johannes Ries: Only maybe a follow-on. You mentioned it already in the comments. The recovery you see maybe in the pipeline coming on maybe the whole back-end market and also driven also partly by the strong business with AI. It's not only the OEMs, it's also the OSATs you see a recovery.
Burkhardt Frick: Yes. And they -- of course, they are somewhat connected because the 2.5D players, they are closely linked to OSATs as well. And you have all these new sites evolving based -- driven by CHIPS Act projects, which are also starting ramping. I mean all the big news were, of course, for the front-end fabs, but you also need the back-end operations somewhat close by, and that's starting to evolve as we speak.
Operator: So before we move over to Martin Marandon, who is waiting for such a long time in the queue, please be reminded that it's still possible to ask questions if you may have. And with this, Martin, please go ahead with your questions.
Martin Marandon-Carlhian: The first one is on temporary bonders. I was wondering there if you mentioned the AI demand picking up. There is also the qualification of one of your customers. But I was wondering if the transition to HBM4 is already a factor here because we know that the number of layers are increasing -- the average number of layers. So it should demand more equipment. So do you think it has started now? Or will we see these effects maybe a bit later? And I have some follow-ups.
Burkhardt Frick: Yes, it's a good question. Of course, our -- at least one of our lead customers is in active pursuit of also planning the ramp for HBM4. And we received the good news last week that we are qualified with our temporary bonder for the HBM4 process. And that is good news because the ramp of that will start from late Q1 or starting Q2 next year onwards.
Martin Marandon-Carlhian: Okay. That's very clear. And maybe still on temporary bonders. I mean, Johannes mentioned some competition with Tokyo Electron, but I was wondering about new entrants as well. So like EVG, for instance, if that's something that you see at some point, multi-sourcing in that market or you do not see it at the moment?
Burkhardt Frick: Yes, we do see, of course, our competition. There are no new entrants. They are the same. They have been the same in the past years. And indeed, EVG and TEL are our main contenders there. And yes, they are actively pursuing our base. So yes, so this is happening to some extent. But I think for now, we have the majority of our equipment at those existing customers of ours.
Martin Marandon-Carlhian: Okay. That's clear. And the last one is on the EBIT margin for next year. I mean, I know it's too soon to give a guidance. But I'm just wondering with the backlog that we see at the moment, it probably implies a down year next year, and you have the consensus down by about 15%. And I'm just wondering in that context, let's say, of a double-digit decrease of sales, how much space do you have to reduce cost on the OpEx side next year? Do you think that, for instance, mid-single digit could be a credible scenario if you have such a down year? Or is it too aggressive?
Burkhardt Frick: You mean mid-single digit for what, which...
Martin Marandon-Carlhian: For decrease of OpEx.
Burkhardt Frick: Yes. I think that's a reasonable assumption. I think we need to stay below EUR 30 million. I think this was mentioned before. We also said that we will not reach gross margins of the heydays like '24. So we will be also there, I think, definitely below 40%, but above the numbers we are currently seeing. So because we have to compensate this with a lower top line.
Operator: And now we have a further virtual hand from a person who has dialed in with the phone ending 847. [Operator Instructions] I can see that you are unmuted, but unfortunately, we cannot hear you.
Malte Schaumann: Can you hear me now?
Operator: Yes. Now we can hear you. So if you can please introduce yourself to us.
Malte Schaumann: It's Malte Schaumann, Warburg Research. First question is a follow-up to the former question of -- related to Chinese waiting for the new tools and the environment of the demand. I would broaden that to the overall customer base. Do you see potentially among other customers kind of holding back because you're about to introduce new product generations? I mean you indicated a pickup in activity and in the pipeline. But do you see generally some customers holding back in light of the upcoming product workovers? Or is that not really the case?
Burkhardt Frick: Yes, Malte, that's very hard to say because we cannot judge if they're waiting for new products, but some of these new products are only launching late next year. So if there is a demand and we don't have the right product, I'm pretty sure customers will order elsewhere. So if they wait, of course, good for us. But we -- where we see a kind of more wait behavior that's on the mid-end cleaner because that is the right tool for that market. There, we get a lot of inquiries. But of course, we have to get the first tool out first before we can be bullish about that. But other than that, we, I think, see customers simply wait till the last moment until they order and then they are rushing and then we have to see how we can, even with our reduced lead time to make it happen. That's the current discussions we have also with our -- among our sites.
Malte Schaumann: Okay. Then on the rework on some tools that impacted the gross margin. What caused that basically? I mean that this happens from time to time, but what caused it this time? Was it kind of design flow? Was it new customer demand? Was it the extreme -- potentially extreme ramp? And do you think that you more or less sorted these out? I mean you indicated that this is kind of mixed effect so might reoccur next year. So maybe you can expand a little bit more on that topic.
Thomas Rohe: Well, Malte, Thomas speaking here. So the question cannot be easily answered, to be honest, because it's a lot of facts which really come into this point here. On the one side, for sure, our customers are also very demanding with the request for support there because they also ramped up in a pretty short time and really they already have by themselves a very demanding customer. So the support was really requested by customers to be there on site, sometimes even 24/7 to support this ramp-up of our customers, and this was really partially -- only partially anticipated, and we were really a little bit overwhelmed by the request and also the hard request from customers. Nevertheless, we supported them pretty good, I guess, and this is also why we still have really very good relations with these customers because they are taking us into account also for our next-generation HBM4, as Burkhardt already said. And also, if you go really in this steep ramp-up, we see sometimes also some topics which we did not see if we use our tools in a normal way or 2-shift way. So this is some, let's say, improvements, which we also did also because customers changed the process chemistry partially, where we also had some learnings together with our customers. And this is -- these are the main reasons why we had to support more than we anticipated before.
Malte Schaumann: Okay. And the reason why you indicated that this is a mixed effect that you think you're not fully through, so that might reoccur?
Thomas Rohe: I don't think that it might reoccur. We learned a lot and we learned together with customers and they let us learn together with them. So from that point of view, the learning curve also for us should go down so that we really reduce it. It will not go away completely, but it should really be reduced significantly.
Operator: And we have further virtual hand from Nicole Winkler.
Nicole Winkler: So basically, I have one left regarding operating cash flow development. So basically, in Q3, you turned positive again. Can you give us an indication what we should expect for Q4 and where we could end up for full year 2025?
Cornelia Ballwießer: Yes. As you said, in Q3, we turned in terms of operating cash flow into a positive number. And we think that there is a good chance that we can end up at EUR 25 million in a positive territory. So this means in Q4, we will have or there are a good chance to have the EUR 28 million cash inflow that we need to get in a positive number.
Sven Kopsel: That's for free cash flow, Cornelia, right?
Cornelia Ballwießer: Free cash flow, yes.
Sven Kopsel: And for operating cash flow, for sure, this would mean that this number should be a bit higher because we also still have CapEx ongoing.
Cornelia Ballwießer: Yes, that's right. It's around -- yes, I would say, EUR 30 million, EUR 35 million we need in terms of operating cash flow.
Operator: And then we have a follow-up from [indiscernible]. So you should be able to speak now.
Unknown Analyst: A brief question on your next-generation scanner tool. If I remember correctly from your previous calls, this is also enabling panel level packaging, right? If so, if -- can you give a little bit color around -- I mean, what we hear panel level packaging could bring cost advantages to TSMC, et cetera, well above 30%. So the technology seems to make sense. But then can you elaborate a little bit, are you covering different parts of the manufacturing process? And can you give a little bit color on the competition part of the business? So are you working with one lead customer and you're exclusive there? Or are other companies in the qualification process as well? A little bit color would be great.
Burkhardt Frick: Yes, [indiscernible], thanks for the question. I mean, obviously, yes, this is really for panel level packaging. This new UV scanner can handle both wafers and panel-level package applications. There will be several versions of that also with a path to 1 micron resolution. So it's also a more accurate system, but this will not be launched from the get-go. The first focus is indeed panel level packaging for that one lead customer whom we develop this closely together. So this is the launching platform. This will be applied in similar applications as spaces as the current ones, but we have access to more layers and more process layers than before. And also, it will open the door for more other customers because this is a very interesting field to be. So we will be able to broaden our exposure there.
Unknown Analyst: And competition part?
Burkhardt Frick: Well, competition is the same as we have now, which are I-line steppers and scanners, you have already in the market, but we currently have a lead over them in cost of ownership and throughput. And we, of course, want to maintain that lead.
Operator: And in view of the time, we will come to the end of today's earnings call. So thank you to the Management Board for your presentation and the time you took and also to you, dear participants, for joining and your shown interest. So should further questions arise, yes, Sven Kopsel from Investor Relations will be happy to assist you. And on that point, it was -- yes, it was our pleasure to be your host. And Sven, final sentence belongs to you.
Sven Kopsel: Yes. Thank you so much. Just one remark. You know that we are going to have this CMD on Monday, the 17th of November. If you have not registered yet or if you are unsure, maybe please just contact me or Florian Mangold as soon as possible. We are still accepting registrations. So take care. Goodbye.