Ladies and gentlemen, welcome and thank you for joining the AT&S conference call on results for the second quarter of 2026/2027. This call will be recorded. At this time, all participants have been placed on a listen-only mode. The floor will be open for questions following the presentation. I would now like to turn the conference over to Philipp Gebhardt. Thank you, Parisa. Good morning or afternoon, ladies and gentlemen. Welcome to the AT&S Q2 2026/27 conference call. Today is with us Michael Mertin, CEO, and Gerrit Steen, CFO. Mr. Mertin will give an overview of the key developments, and Mr. Steen will comment on the financial figures and our guidance. As Parisa mentioned, the presentation will be followed by a Q&A session. Now I would like to hand over to Mr. Mertin. The floor is yours. Thank you for the introduction. Ladies and gentlemen, warm welcome here from the center of Austria, from Leoben. From my side, my name is Michael Mertin, CEO of AT&S, and I will guide you through the figures together with my dear colleague, our CFO, Gerrit Steen. Well, how to start this kind of a call with all you from the capital markets. Maybe please join me and think one and a half years back. So what happened. AT&S was a company when I joined, making losses, having negative cash flows, and I gave all of you a promise that I personally see that this company is capable to return to the markets, to fulfill and satisfy customer demands, and stepwise come back to profitability. The first step was the healing, stop the bleeding, going back and coming back to positive cash flows. The next step, last fiscal year, was coming back to positive EBIT. The third step, end of last fiscal year, was coming in principle back to profitability, showing positive net earnings. This fiscal year, we promised significant changes in terms of profitability and growth. After these healing procedures and after we gained trust from the customer side, again, trust into our technology competence and also trust in our production competence, we have been able to sign many favorable contracts, longer-lasting towards the future, and enabling us to become a significant technology and innovation partner for our customers mid and long term. Our positioning, bringing us to the actual profitability and growth, is based on our repositioning towards an innovation partner. We try to create value for both, for our customers as well as for ourselves. Now the result is, as we can see, with more than EUR 680 million in revenue and EUR 244 million EBITDA just in Q2, EUR 150 plus million EBIT, and quite positive earnings per share. This is more than we earned, the earning side, in the entire last fiscal year and in Q1 together. This is actually the way we promised to you. This is sustainable growth with a positive development on our cash position as well. This is what we promised, and this is what we are actually delivering. Just three weeks ago, together with a small management team and our CTO, we had a lot of visits to our top customers on the West Coast and some other parts of the world, meeting their top executives and discussing about the actual demands, the actual technology, and the technology development of our products and the entire environment for the upcoming future. One of the messages definitely was the market is strong. The entire market can't fulfill the demands. Secondly, the market remains strong over the upcoming years. Not just over one or two years, over the foreseeable future until 2030. The fundamentals are not just driven by pure investments into data centers and artificial intelligence products. There are a lot of business models behind granting our customers long-term positive revenues and cash flows. Of course, it's a little bit difficult to see this just in a quarter-from-quarter discussion and having the analytics just on a quarter-on-quarter cash flow side. Nevertheless, we have this discussion with executive managements from our customers. I can't tell you all the details, of course, because this is not to be disclosed. But for us, the entire market fundamentals, they still remain extremely strong. This is also shown in our long-term relations and the long-term contracts in financing, but also in taking our products for certain prices in these contracts we made with our customer base. Next page, please. If you follow me to the next slide, please. What you see on the left side, this data center CapEx 2026. Quite more or less double from the year before, is a number, $1,000 billion. What one of our customers told us, this is an investment he wants to achieve alone within a few years from now. Unbelievable numbers. Even if a part of this will be realized, the market is not able to deliver everything. We are still bullish, we are still strong, but we have to deliver the technology as well. It's not just strong from a volume side, it's not just strong from making more of the same. The substrate market, as we see it, will grow from EUR 25.8 billion to more than EUR 20 billion in 2030. It's not just the volume, but it's the quality of the product, the value add of the product itself. If you are looking back one, two, three years, actual older products for client and server applications, one of these substrates had a value of $8 bucks. The latest generation we are producing today is in the range of the low hundreds of EUR each. The size is bigger, the complexity is significantly higher. First embeddings, bringing components into substrates, already are taking place. What we see for the next and over next generation of products is that the product value will raise up to EUR 1,000 and more, and for the same amount of money, we will embed active and passive electronic components into the substrates. The market growth has two components. It is a strong demand still from the side of our customers into all these data centers with the applications behind it, but it is also growth in the quality and technology of the products. This leads us to the position that we, of course, are in a race for technology and innovation. This race is not a race from year to year. We have to follow joint long-term roadmaps, otherwise these technologies cannot be developed and cannot be realized in production and operations. Interestingly, most of the customers detected, well, substrates are nice, they are high-end, the new gold, but there is a lot of things they need around it. You still need all the server applications, you need quite a lot of client applications. The demand of these PCB-sided stuff or these kind of substrates is also extremely strong. It was a little bit forgotten for a year, one and a half years, and now this is coming up again. There is a second reason for that, and the second reason is, come back to that later, the partly still open technology question on how technology will evolve in the future. Important point is, you need both of the competencies for the future, the actual competence of developing and producing substrates, and the competence of developing and producing advanced PCBs. So, both together is one of our core strengths, and this brings us significant opportunities ahead. Coming back to technology a little bit, maybe you will ask about it a little bit later. To answer the most important question from all of you first, yes, now we are officially allowed to make the statement, we are EMIB-T partners of Intel. We got the official approval and the official allowance to communicate that, and of course, we are proud to be in exactly that position. On the slide, you see different ways of packaging. At the end, the question of how to bring information and power delivery both to memory and, of course, to the CPU, GPU, XPU, whatever it is. There is the classical way you see on the left side. You have a underlaying PCB. You have this new gold, the substrate. You have a typically silicon or silicon-like interposer for extremely tiny structures, and on top of this, you have the silicon chips. This is a package as it is today. The problem here is to become faster and to become bigger in size, to combine more of the computer chips on the same size or on a bigger size, you need different ways for high-speed interconnecting of all these silicon components. High-speed interconnecting means very short ways, small lines, and a lot of them. Therefore, typically, the interposer was used. But if you are using a high-end interposer and the substrate is becoming bigger, for example, like your tablet computer, you do not have silicon wafers that big, or you have to use an entire 300 millimeters wafer, being structured, drilled with all the power supply holes in it, and so on. This is not possible anymore. So what you do is, you are embedding so-called bridges into a substrate. This is what I mean, the substrate complexity is increasing by embedding, for example, of these silicon bridges. Additionally, you have active and passive components. Now, if you have up to eight, 12, maybe in the future 16 or 20 CPUs, GPUs on one of these big substrates. You have to have the energy supply. One of these components need 1,000 watts, a kilowatt, at 0.7 or 0.8 volts. This is more than a kiloamp in current. This is huge. It is more than your EV needs for charging. So, what is additionally coming up is small MOSFET transistor-based power supply components, transformers, so that you can come in with low currents, high voltage, for example, 25 amps, 40 volts, and transform it directly close to the silicon CPU, GPU, XPU into the necessary low voltage, high current energy supply. These are embedded components or components which are added left or right on the back surface of the substrates. Additionally, if you have all these components on larger scale substrates, the communication speed have to be enhanced as well. So the communication from and to that structure has to fulfill the demand of all of these silicon components. Now we are facing the limitation of electronical short pulse information transport. So when we are achieving the terabit plus communication ratios, we have to come to optical components. We are talking about optical transceivers, but more and more, these optical connectors are going directly into the PCBs, and maybe in the future also into or onto the substrates as embedded optics, as embedded transceivers. This is going on. So now on the right side, there is an alternative for this kind of packaging for the future, and this alternative is having even more components embedded in a more complex substrate. So having all these energy converters, et cetera, internalized, having partly sub-substrates, so substrates a little bit smaller, embedded in a larger scale substrate. This is something where also the industry is looking for in the future. We are actually working on all these new packaging technologies because we are coming from both sides of technologies. We are coming from the advanced PCB side as well as from the advanced substrate side, and we are maybe leading also in the material science. So we have to simulate all these material behaviors of all these packages before you go into the design, before you go into tests and trials and operations. We are quite good here and working in this kind of pre-development closely together with all of our customers. So hopefully this gave you a first hint on our technology positioning and what we are going to do in the future. So that is the what. Now let us come to the how much, where, why. You see a sketch of our Kulim plant, the entire layout as it is today, with the Kulim 1 plant, we call it the K plant. Two lines are actually under production. Third line is actually ramping. The fourth line is actually coming into ramp. This is the existing structure. Kulim two, the J plant, with five lines, is the outer shell, already was there. Now we are equipping the entire plant and bringing it to ramp. You will see this timetable a little bit later on, how fast this will go on. For you, for calculation purposes, we just put the numbers and the number of production lines which can be placed into these plants. For Kulim 2, it's up to five, and we are actually planning to have three customers in that plant. I showed you the increasing complexity of the substrates. A substrate is consisting on two parts at the end. It's an inner part, this is what we call the core, and it's the layer part where you have all the different layers with all the different interconnects. These are the ABF layers. The core itself is the part where you have to embed all the components, all the electronical components into. So core is becoming more and more core technology, core knowledge. Therefore, it is up to us to produce these new generation cores on our own. Therefore, we already decided, and we are already building two parts of a core building, core production building, to enable these technologies. This is also under construction actually, and will supply the Kulim plant two as well as the orange part, what will supply a possible Kulim 3 plant, what we actually have under discussion, with up to four lines. We have no customers contract signed so far. On the other side, we have significantly more requests than we have production lines in a possible Kulim 3 plant. We expect the final decision and the customer negotiations to happen early in the next year so that we can come to decision points in the first half of next year on how we proceed with this entire campus. Now, we did even more, if you follow me to the next slide. Now you see a satellite picture of the campus. On the upper middle, you see the existing campus. You see a parking lot, picture's a little bit older, parking lot where we are actually building the core production site, where you have these blue buildings in the middle. This is the free space for a possible Kulim Plant 3. Now we bought additional plots. On the right-hand side, upper right, you see what we call Kulim 4. This is a possibility for having both, having more core capacity and additional capacity for substrate lines. Kulim five and six is something where we can imagine to build a new plant for advanced PCB, including the capabilities of any kinds of optical connectors. The latest generation of PCB in terms of accelerator boards, complexity, thickness, and integration of optics. This is what we are thinking about for the future for the plots of Kulim 5 and 6. Just from a size, that would have equal size like our operations in Shanghai. We already thought about the future. We are preparing for the future. We are starting these additional discussions with all of our customers in the same way we did it so far quite successfully, and in the same way of contractual relations and forward development. Now, let's go to the related financials, to the contractual relations. I will hand over to Gerrit now, who will give you an idea how fast we ramp this, and he will translate, of course, this ramp later on in numbers and figures and give you the actuals and the mid and long-term outlook. Gerrit, it's yours. Thank you, Michael, and as well from my side, a warm welcome to our Q2 call. Good afternoon, good morning, good evening, wherever you are. When we look now at this page here, you see exactly how the ramp in more detail of what we discussed as well in our Q1 call looks like when it comes to the expansion projects which we announced now recently for Chongqing and Kulim. That certainly is something which is nicely staggered. If you look at the timing of these different lines, starting in Chongqing, where we are de-bottlenecking, where we are seeing a nice upside already in this year based on some improvements we are doing and some further improvements in our utilization. Then as well in the further installment of equipment to bring up additional volumes, which will then really bring some volume growth in next fiscal year. When you move to Kulim 1, which again is the site dedicated to AMD in the future, the first two lines are operational, are ramping up further, and as well bringing up the yield. We have the third line, which we actually moved over some time ago from China, which is being installed as we speak. We are qualifying, and we will still see actually in this fiscal year, towards the end of this fiscal year, the start of our high-volume manufacturing there, which will then supplement it with a new line which we are installing, buying the equipment step by step, and then which will go into full operations then as well in next fiscal year and bring us some nice earnings and growth potential. When we then move to Kulim 2 as well, there we have several lines now, with three lines with one big technology company who is a long-term partner, plus now the fourth line with Marvell, and the fifth line where we are negotiating as we speak and which will be announced sometime in the future. There, that will, because we will still need the building infrastructure in that building to be installed, will bring us then in the following fiscal year some very significant upside when it comes to volumes with the start of the high-volume manufacturing of the first three lines, the first two lines fully in fiscal 2028/2029, third line towards the end of 2028/2029, and then the fourth line in the beginning of 2029/2030. A nice earnings profile and growth profile, which is nicely staggered over the next couple of years, which give us some good visibility when it comes to our growth trajectory, and again, which is fully backed by customer commitment. When it comes to not only the financing through different parts of the agreements we are having, but as well when it comes to long-term supply agreements over five years, plus take-or-pay agreements, which give us as well the security when things are probably a little bit slower and the customers still are committed to paying what they committed to. Therefore, when you now look beyond as well our this fiscal year, a very, very good earnings upside and visibility for the next several years. When you look at the CapEx, now we included as well in our overall CapEx number the further investment for the Marvell line, for the fourth line in Kulim 2, which then brings the total CapEx now related to the Kulim campus at this point to plus or minus EUR 2 billion. You remember our previous guidance, which was more related only to Kulim 1, fourth line, and Kulim 2, 1, 2, 3. Now including fourth line, we are estimating approximately EUR 2 billion of CapEx, mainly hitting this year and next year. When you move on to the next page, let's dive a little bit further into this quarter number. As Michael Mertin already pointed out, record quarter when it comes to our growth, 52% after the 40% in Q1, now 46% for the first half a year. When you break that down, the biggest part of it is volume-driven. More than 30% of that 52% is really volume-driven, really by strong demand from high-performance computing, and mainly driven by our Microelectronics business unit, with the substrates. As well, there is a sequential quarter-over-quarter growth of 23%, quarter-over-quarter. So nice, very strong acceleration of the growth as well within this fiscal year. Driven really by higher utilization, so moving utilization in our key plants, especially in China, close in the bottleneck processes to 100%. Improved pricing, when you look at the 52% in the quarter, a small double-digit percentage in pricing impact. And as well, continued volume mix upside where we are driving more complex and higher value-add products in our sites. There is as well approximately EUR 35 million of impact from these new customer agreements, as we disclosed previously. There is a smaller element which is hitting our P&L directly, so that has been from these new agreements about EUR 35 million. So, when you just break it down as a growth, it's a high single-digit percentage number of growth contributing from these elements. When you look at the material availability, we are managing it quite nicely, otherwise we wouldn't be able to really report that more than 30% volume growth just in a single quarter. But it is a tight situation, very closely with top management involvement in our global procurement organization, very close collaboration, including as well our customers with our suppliers to continue to manage that. But it continues to be a tight situation, which again, we will monitor and I'll come back to that as well when we talk about our guidance. When we move to the next page, please, and we are having a look first of our Microelectronics division, you see that Microelectronics was the growth driver in the quarter. More than doubling our sales versus previous year Q2, and basically quadrupling, more than quadrupling the EBITDA. Very, very strong demand. We could have sold much more than we did because of the very strong demand we see across the board, unabated. Very, very, very strong, as Michael Mertin pointed out already, ask for our products. We are able again to, with further increase in utilization during the quarter, with some de-bottlenecking, and as well specifically in Malaysia and as well here in Austria, continued production ramps and yield improvements to really enable that growth. That is certainly extremely positive, so that operational leverage, and as well the improved pricing which we saw in the Microelectronics division, certainly helped to drive this significant doubling of the overall sales. When you look at our expansion projects, which we touched on before, they are fully on track, and we are very much working with our teams to enable them with everybody on board to deliver this huge CapEx project and upside and expansion as planned, and really commit and meet our customer expectations when it comes to the readiness of these new production lines in the two, specifically the two sites. But certainly as well, and that is as well something I want to comment on when we move to the next slide, it goes beyond these two sites. As we disclosed before, here in Austria as well, and that is then specifically as well related to our Electronics Solutions division, we are investing into further technology upgrades and as well expansion of capacities in our advanced PCB field with embedding technologies. The same in our Chinese Shanghai site, where we are as well upgrading the technology, where we are enabling more business for optical transceivers. That is as well something which will help us nicely as well in specifically the second half of this year to generate growth. That actually as well was the reason that we are only seeing a small 1% growth in the division in the quarter, because we are getting ready for the second half of the year with these upgrades. And technology improvements to prepare these sites for growth in the second half of this year. There was a small volume increase, but again, looking at the quarter-over-quarter sequential growth, it was significant at 26%, mainly driven by better volumes. That is as well something where we as well expect and to see in the second half an improvement of the margins, because pricing will kick in. As we discussed before, we were a little bit slower there in implementing price increases versus our substrate business, and therefore, we are seeing this now in the second half of this year, which will then help nicely improve EBITDA margins sequentially over the next couple of quarters. We had a smaller couple of, especially from China, FX transaction-related margin declines, which were hitting us with about 3 percentage points in the quarter from FX-related effects in the ES division. As well that, certainly when you look at the quarter-over-quarter versus last year, margin development was one factor which suppressed the margin in the quarter a little bit. But again, when we look at the outlook, we expect for the next couple of quarters a significant improvement to be expected going forward. When you now move on to our balance sheet and financing, again, I think we have a very, very strong liquidity profile. Our leverage now is at two already at the end of the second quarter, and you will see that we expect a further improvement, and therefore, as well updated our guidance for the full year to go significantly below two. So nice deleveraging with this significant EBITDA increase specifically. You remember, it was just at above three even just a half year ago, and above five if you take out the Korea plant divestment proceeds one and a half years ago. So nice deleveraging, nice strengthening of our balance sheet. A very, very balanced profile when you look at our maturities now. We just paid back in October the old hybrid bond early because of our strong liquidity profile, and therefore, as well have some improvement on our interest lines here, so that the remaining maturities for this year are at a low triple-digit number in millions, and as well going forward, a very manageable profile with significant cash and cash equivalents close to EUR 1 billion as of quarter end in our balance sheet. When you move on to the next page, then as well a word in more detail on our cash flow development. Very strong cash flow from the result in the quarter. Certainly driven significantly by the strong result, but as well by client contributions related to our new expansion. We had purposely as well to optimize our financing profile, basically significantly scaled down our factoring program. We didn't see a reason to pay significant money just to get receivables two months early, which are from large tech companies, very, very solvent partners here. Therefore, we decided as well based on our strong cash profile to scale that significantly back. Basically, without that impact, we would have been significantly positive in triple-digit million euros as well on operating free cash flow in the quarter. That is basically behind us. There is a small further impact in Q3, but there's a significant increase as well expected on the operating free cash flow side for the full year and for the second half, so that we will see well into the triple-digit million euros operating free cash flow for the full year, as we always said, as this is one of our key levers and KPIs we are working towards for AT&S. CapEx more than doubled versus Q1. When you look at that sequential development, still certainly when you think about EUR 1 billion-EUR 1.2 billion of total CapEx for the year, we are at EUR 100 million now for the first half, so there's a significant acceleration to be expected for the second half, which again, we are expecting we will see a significant ramp of spending based on the project delivery and progress of our expansion project in the second half. If in the end, something will move into next fiscal, we will see in Q4. That's certainly always something which you cannot time exactly to the week. But generally, we are still expecting some significant spending certainly in Q3 and specifically in Q4 on the CapEx front. Certainly as well when you look at our working capital overall investments with the receivables going up, not only due to the scale-down of the factoring, but due to certainly the significant business growth we have seen. When you look at our balance sheet, we just paid back the hybrid bond, which certainly we've used basically our balance sheet because we took out the hybrid bond on the which was reclassified to debt already as of end of the quarter, and on the other side, the cash. Therefore, on a clean comparable basis, equity ratio now at 25.5% significant improvement versus last year, and as well in absolute numbers, nearly EUR 300 million of additional equity already in our balance sheet. Therefore, based on these strong results, we can fully confirm our guidance for this year. We have been both on the revenue side with 52%, but as well on the EBITDA margin side, fully in our full-year guidance range. Therefore, we can fully confirm that as well for the full year. You can expect continued strong reported year-over-year growth as well in Q3, fully within our full-year guidance range. I think that's important. We are seeing continued very strong demand for our advanced substrates, and as well, as we said already, accelerated growth in our ES division as well, certainly when we look at our comparable prior year quarters. We certainly see some small impact on the substrate side potentially in this quarter on the material availability side, but there's corresponding upside depending as well on the final alignment for the customer contract-related one-timers in our P&L. And looking at the momentum we are having as well, a further acceleration of growth we are seeing right now as well towards then Q4 of this fiscal year. Again, strong quarter and very, very happy about the progress we are making overall, and looking forward now to your questions. Yeah. Thank you, Mr. Mertin. Thank you, Mr. Steen. We will now start the Q&A. In order to give everyone the opportunity to raise questions, we would like to ask you to limit yourselves to two questions. Once we are through, and if there are still questions and still time, we will start another round. Now I would like to hand over to Farisa to handle the session. Thank you very much. Ladies and gentlemen, if you would like to ask a question, please press star, nine, and pound key on your telephone keypad. If you would like to revoke your question, press star, three, and pound key. You can also use the dial-in function in the webcast and raise your hand if you would like to ask a question by phone. I repeat, to ask a question, please press star, nine, and pound key on your telephone keypad. The first question is from Martin Marandon-Carlhian from ODDO BHF. The floor is yours. Thanks for taking my question. My first question is on something you said in the Austrian press this morning, that your planned capacity is four to five times oversubscribed. I think before you were saying a figure closer to three. First, how real do you think that demand is, or do you expect some double ordering there? Second, what are you looking for to sign more capacity for Kulim 3, Kulim 4, et cetera? Considering demand seems to not be the issue, is it about customer diversification first? How favorable the contracts are in terms of financial structure, or just about your own capabilities to ramp? Ramping too soon would be an execution risk. Any clarity would be helpful, and I have a follow-up. As I said also in the interviews this morning, the demand is extremely strong. There are customers which are asking us to build an entire plant for them. What even is maybe not enough. What we are doing is, we are working with proved customers which we can trust each way, and we are looking for customers having an extremely sustainable business model with their investing into data centers and artificial intelligence. The same is valid for everything what is related to optical components and optical transceivers. The market here is significantly bullish as well because of the limited capacities and capabilities of, for example, technologies like mSAP. We own this technology, of course, and we are using it for several applications. Just for you who do not know what it is, there are different methodologies to print conducting lines. There are subtractive methodologies and additive methodologies. With this mSAP process, you can have very narrow lines and narrow spaces, nevertheless, a perfect aspect relation. You can have thick conducting lines quite narrowly spaced. This is necessary for many, many applications and is especially necessary to produce optical transceiver boards. Also here, we are extending our capability as good as possible. We have some further investments which are actually going on in Shanghai as well as in Chongqing. We are extending both of the capabilities in both sides. This will have first impacts in the second half of this year. This is one of the many reasons why we think that also ES, our second business unit, will be significantly grow both in top line as well as in profitability in the second half of the year, besides what Gerrit already stated, that we are through with all the price negotiations and price increases. I just can state For most of the future technologies, all the markets are still bullish. There is a significantly higher demand than we can fulfill or deliver, and there are a lot of customers coming to us because of our technology. Please remember, I tried to explain a little bit about the complexity of the upcoming products. Our customers are well selecting their suppliers from which they think they are capable with their R&D and technology resources to even be able to be their partner for this future roadmaps. This is not really given for everyone in the market. Therefore, lots of the customers are directly coming to us, and therefore the demand on our side is extremely high. That was the reason, after we saw this, that we decided to start communication about this additional plant with at least four lines in Kulim. That was the reason we also convinced our supervisory board extremely fast and easily in one of the last supervisory board meetings on site in Kulim to buy the additional plots for further extension of both of substrate capacity as well as for capacity of advanced PCBs, including, as I said, the mSAP technology, et cetera. Hopefully this answers your question. It was a long answer, but at the end, it is a complex mixture between our technology capabilities, R&D capabilities, and the significantly increasing technology demands from the market. This is nothing you can buy off the shelf. Therefore, also the customers like to have these long-term relations. Otherwise, they are not able to have a partner which is capable to co-develop their roadmaps. Yeah. Thank you very much. That's very helpful. The second one was on supply constraints at your customers. Notably, I mean, last week, one of your key customers disappointed on its next quarter guidance due to some supply constraints and lower PC demand. How should we read that? Does it change anything for you that one of your customer ramps a little bit slower, or demand is actually so high that they take all of what they can at the moment, and you see no change on the roadmap whatsoever? You gave the answer already yourself. There are constraints, no doubt about it. The entire market environment, and also we, could grow slightly faster if we would have unlimited access within our supply chain. This is relatively clear. But the constraints we see, and also after all the discussions we had with management of our suppliers, is already built into our prognosis and our plans for the second half of this year and next year. I personally was in Japan meeting with CEOs of our most important suppliers, and believe it or not, sometimes, if they all have a quite good technology background, as my partners partly have, and I have it a little bit as well, even CEOs can find solutions for increased supplies. That's sometimes a funny part of the story. But we take that very serious, and we as top management, Gerrit as well and as myself, we are personally involved here, and everything we know is built in the plan with some security. Otherwise, it could be even a little bit more bullish. I think, Martin- Thank you very much. Just to add on, Martin, it is certainly something, and when you look at their outlook and what they talked about, the client business as well, certainly it is as well something. I mean, we serve both businesses, and certainly if there is potentially probably less demand on client, we are happy about that we get this capacity available to do more server. Therefore, I think from that perspective, there is no concern at all from our side. Okay. Very clear. Thank you. Very much. The next question is from Cantor Fitzgerald, Tim Wunderlich. Floor is yours. Hi, good afternoon. Thank you so much for taking my question. In your press release and also on the call, you talk about power electronics, you talk about optics, advanced testing. I mean, these are the new products and technologies that are going to drive growth and margins in the Electronics Solutions segment starting in the second half of this fiscal year. Can you give us the ballpark number for the size of the growth opportunity and the margin impact that this should have on the Electronics Solutions segment in the second half of this fiscal year, but also for the entire next fiscal year? Then I have a follow-up. Thank you. Sure. Thanks, Tim, for the question. Yeah, again, I think what you have seen is in the first half of the year, certainly, basically zero growth, and that will significantly accelerate as well on lower comparables, certainly from last year. But we expect significant double-digit growth in ES in the next couple of quarters, based specifically on the elements you were mentioning. The businesses have the potential, not in this year, but if you look over the next couple of years, definitely have the potential to all, each of them have triple-digit million sales potential. That's not only something which drives growth in the second half of this year, but for the division as well can drive nice growth in the following years. Yeah, we have a constant growth path here as well for the next foreseeable time, what means for the next three years with all of these technologies you mentioned. The only reason why we can't see it yet is besides the price increase, but the key reason is we have to rebuild most of our plants. We have been a PCB workshop, and now it's going to high-end technology. We are building clean rooms, we are educating people, we need some new equipment. This is everything what is actually going on here as well, together with our partners, with our customers, and we enable ourselves and we enable our customer for new products for the highest end technology what is actually available in the world. Complex product, but the world needs these products. Yeah, we are talking all about these advanced chipsets, et cetera. You have to test all that stuff. To test all that stuff, you need extreme complex Microelectronics. We are supplying it. Everybody is talking about mini nuclear reactors to supply the data centers, but somehow this huge amount of energy has to be brought to a computer chip. This is what we are doing. This is what we are doing with power electronics, for example. This will not end this fiscal year. We are enabling ourselves actually in all of our sites. Optical transceivers is something I already mentioned. We are leveraging on technologies we already had, but not in that volume. This is the big advantage, and this is the difference to other suppliers, and that's the reason why the customers come to us. We can demonstrate that we already have this technology and that we already are capable here, and now we are bringing it to volume. That's the reason why it took some time, but you see it from the second half on. Gerrit already mentioned the numbers. We will see it prospering also for next year. Thank you. Sounds fantastic. What kind of EBITDA margin is your aspiration with the new technologies coming online in the Electronics Solutions segment? I mean, currently we are below 20%. Is it fair to assume you would like to see it move towards 25%, maybe even 30%? Before Gerrit will answer it from the financial side, please remember, I am not so sure if the EBITDA margin is the right measure here. Maybe it is better to go to EBIT here, because overall, the depreciation part of the P&L will be a little bit lower in that area than it is in the substrate area. So the EBITDA margins in terms of net profitability are not really comparable. But we are aiming at more or less similar EBIT margins. But Gerrit, your turf. No, thank you, Michael. Tim, yeah, sure. I mean, we expect certainly a margin expansion and contribution which is above what we are seeing right now from these businesses. That is a little bit as well what we were discussing already when it comes to portfolio optimization and mix, utilizing our existing capacities and as well the ones we are basically enabling right now for higher margin business. That is exactly what we are doing. I do not want to give a specific margin guidance now for the ES. I know then you will always come back and ask me each quarter now for the next margin guidance for the business unit. But again, it is something which certainly is our aspiration to significantly move that up from where we are right now into the 20s. Okay, great. My final Now we mentioned it. My final question is on a potential relisting. There have been some public press articles, some reports that you are considering this potentially in the U.S. Can you give us a bit more color? I mean, is that something that could happen in 2027, or what are your plans here? Yeah, good question. But again, I think it's certainly something we are evaluating when you look at what is the peer environment, what are the valuation multiples we are seeing as well when it comes to some of our peers at different marketplaces, et cetera. It's something we are evaluating. Is it something which is due for the next few months? No, I don't think so. But it's certainly something where we need to further build our visibility, where we need to further build our investor base in other parts of the world. We are being on the road show as well the next couple of weeks as well in Asia. There's some preparatory work, I think, to be done, because it doesn't help if you get listed somewhere, but you don't have really then a lot of turnover in your shares there. Therefore, there is some preparatory work to be done before that, and then we will decide on the maturity, where we are at which point, when a good timing is. Okay. Sounds good. Thank you so much. The next question is from George Brown, Deutsche Bank. Please go ahead. Hi, guys. Well done on the strong Q2, and thanks for taking my questions. Just two on my side. I can take them one by one. Firstly, I think you said that a decision on further capacity is expected to be made in the first half of 2027. Maybe could you just clarify what this is referring to? Is this referring just to Kulim 3 specifically, or should we expect additional announcements beyond the Kulim 3 in the first half of 2027, for example, in mSAP PCBs or optics like you were just talking about earlier? And then I have a follow-up. Yeah, we have a couple of additional projects already going on where we are finalizing details on business cases, where we are finalizing details on contracts. We have some projects ahead of us where we actually are working on, advanced working on, for maxing out some opportunities we have in China, for both for optics as well as for other products in the substrate area. Also, the China for China business is developing extremely well. There is more to come, definitely. Of course, the big single item will be on the campus in Kulim, a new third plant for substrates, including the related second part of the core production. But there is definitely more to come. Maybe you will add on this, Gerrit? Yeah. I mean, certainly for us, it is always important if we do something, we need to have the right terms and conditions and everything in place. We always say we are in no rush. We want to be sure that whatever we do is something which gives us exactly what we need when it comes to the contractor terms, when it comes to the financing, when it comes to take-or-pay, when it comes to supply reliability and earnings visibility. That is something where we take our time to just get to exactly what we need in terms and conditions, et cetera. Then in the end, we need to as well deliver, right? I mean, that is as well something you saw that as well when we talked about how we are now delivering line by line in the existing expansion, nicely staggered. That is as well something. In the end, operationally, you all need to deliver it, and therefore, as well, we want to continue that nice staggered approach. When that comes together, it is the time when it is, but the demand overall profile is so strong that we do not see a need to rush, but certainly we want to do it right. Perfect. Thanks, guys. Then just a second question. It is good to see you guys have approved or have received approval to talk about EMIB-T finally. What is your sort of view on yield progression here? Because I see rumors everywhere about yields coming out of whether it is Japan or Taiwan around yields being low on EMIB-T, which is understandable given it is a new product and it is very complex. Do you see this as a sort of structural issue going into 2028, 2029? If not, what do you sort of imagine the steady state yield being in high volume manufacturing? Can it be comparable to sort of normal EMIB-T yields? First of all, the first part of the answer is the more complex the substrates become, the more time you need previously for not just process development, but also technology development. Such complex products cannot be manufactured in any kind of an existing plant. You need total new machines, total new equipment, new processes, and you have to test it for a certain time. Even in the early times of other substrates, industry had quite low yields, and this is valid for all these new, let me call them super substrates as well. Is it just a matter of time? No, it is not. I think we will see kind of what my prognosis already was in some calls before. We will see a certain kind of narrowing. Not everybody will be enabled to produce everything in the future. One of the reasons why the customers are lining up in front of us is that they believe that we are not even the biggest company in the world in our area, but we are strongly growing, and we have the technology in-house. This is the reason why we are making these kind of tech tours with our CTO and with some experts, showing what we can do, showing some extremely advanced technologies, which are even not communicated in the market so far. We already show with the most complex volume products which are produced today, that we have one of the highest and partly the highest yields in the industry. That is the reason why the customers are approaching us for these new products. In principle, it will be extremely hard to come to the same kind of yields like it was in previous times. The reason is very simple. If you produce a panel what is a little bit more than quarter of a square meter, 50-something centimeters times 50-something centimeters, and you are producing substrate six by 6 centimeters. You have a bunch of these substrates, 80, 100, let us just for the math, 100 substrates on one panel. If I talk about future substrates having significant bigger size and you just have 20 of these substrates on one panel, and you have one defect somewhere, former time it was one out of 100, now it is one out of 20, and the future it is one out of 10. Just the pure math, the probability of having a defect in one part is significantly higher than in the past. Additionally, the complexity of the product, structure sizes, et cetera, is increasing. So, having the same yields is extremely hard. Is it necessary? Maybe not, because the new substrates will deliver extremely better other advantages. If you can combine more chips on the same substrate, it is better for the energy supply, it is better for the entire energy transformation before. It is better, significantly better for the silicon-to-silicon communication. You can apply router chips directly on these boards as well. The compute part of each of these CPUs, GPUs, XPUs could be significantly enhanced. So you have other advantages. But you are right, there will be certain yields, especially also for all the embedded components, where it does not make sense to go down that route anymore. So there will be a kind of a balance between yields and extremely high complexity for the future of the over and over next generation of products. This will be a balance. So I do not expect the same yields. What we expect is reasonable yields, which are in a kind of a balance with these significant advantages of these new substrate generations. There is a limit, therefore I am a little bit It is a little bit too early to finally decide which concepts in detail will be the winning ones. All these CoWoS, CoPoS discussions, et cetera, at the end, there will be a technology narrowing and things will come together. If you want to bring all the chips directly on a PCB, what kind of PCB will this be without warping and, and? It is a kind of a super substrate. If you integrate everything in a substrate, what is as big as a tablet, what is it? It is a kind of a super PCB. Things are coming together in a certain way, and it is becoming more and more important now that our customers and our industries, companies like us, enter into long-term relations and long-term R&D technology cooperation. Not just R&D on the product side, but also production technology cooperation. We also, vice versa, have to tell our customers what is possible, what are the design optimizations coming from our analytics, coming from our mathematics, coming from our material science know-how. This is becoming more and more important. Exactly that point was the success point, and I was part of it, between ASML and Zeiss in history. They did exactly that together with a few other companies. They designed the roadmaps together, they developed the technology together, they designed the product at the end together based on capabilities, what is possible to be produced. Now exactly that kind of a philosophy is coming to our industry, and we are front end here of this philosophy, because we have all this R&D and technology knowledge. Maybe not all, but a lot. Sorry for this very long answer. Maybe you hear I am a little bit enthusiastic here because I have seen it once, and it is happening here again. I am more than proud that AT&S can be on the front end of this development now. Thank you so much, guys. That was super interesting. The next question is from George Chang from Aletheia Capital. Hi, thanks for taking my question. Most of my questions actually have been asked, but just one thing is that it looks like you have a more concrete plan this time versus three months ago for Kulim 3. I was under the impression that Kulim 3 investment might take some time because you are resource constrained, given all the ongoing projects. Can you talk about what has changed versus three months ago, and how do you tackle the resource issue? Thank you. What has changed? A lot has changed. 18 months ago, the company was suffering from the finances. We had no trust. Three months ago. I know three months ago, but let me tell the story. Our stock price was one-tenth of the value of our balance sheet. You see a very strong development of our company, and we have to prove our capabilities towards customers. This proof is getting better and better, harder and harder from month to month, from quarter to quarter. You also see it in the numbers. From the last quarter to that quarter, these are the three months, we gained additional significant confidence from the customer side. We have more and more requests to significantly invest more, and besides the requests, we get the commitment, "We will do it together with you." This changed in the last three months, and as said before, it's not just the classical, or not 100%, the substrate. It's also everything around advanced server. Everybody's producing substrates for AI, but there also is a necessity for advanced server substrates, which are also quite complex. This was not in the focus a couple of months ago. Now, they are collecting all their real demands by the implementation of these data centers, and they're coming up with new, even increased demands. This is what the studies from Gartner, for example, show as well. We see just over the last quarter a significant increase in the demand just from Gartner over the next three years. Year over year, we see that. This changed also during the last three months. We decided to rethink our option with Kulim Plant 3 to make it faster. Faster means we, of course, and you're right here, we need some space for the buildings we are doing, so layoff areas for all the construction material, et cetera. We have limited resources, but we have excellent resources on the extension, and these resources are already starting to think about a Plant 3. Great. Thank you. Just a follow-up question is that, so the total investment right now for Kulim amounts to EUR 2 billion, including Marvell line. I was under the impression that this would be over EUR 2 billion. Are you scaling back on some investments elsewhere? No, George. We always said it is EUR 1.5 billion-EUR 2 billion without Marvell, and that is solely always depending as well on the final configuration of the production lines and equipment and everything we put in there. We always have some wiggle room there. Now that additional line came on top. If you take the midpoint of what we previously announced, plus one additional line, you are a little bit above EUR 2 billion. That is exactly where we are, around EUR 2 billion. That is what we wanted to indicate. There is no change in overall scope or whatsoever. As more we go forward, certainly the more specific we come as well, we know better exactly how much we need to spend as well for building and infrastructure. You always put some buffers left and right, and now it becomes more specific, and we get the numbers in more detail. Okay, great. Thank you so much. Thank you. Before we move on with the next question, I just wanted to quickly remind you, if you would like to ask a question, please press star, nine, and pound key on your telephone keypad, or use the dial-in function in the webcast and raise your hand. The next question is from Daniel Lion from Erste Group. Please go ahead. Yeah. Hi. Congrats on the development. Impressive. I would have two questions as well. Let's start with the strategic one. It was just announced, the acquisition of Lumentum by Marvell. How would you see this combination, which we might see more often going forward, to benefit your business going forward? Is it accelerating demand for more complex and optical solutions, integration? What does this mean for you and the industry? Marvell actually acquired Lumentum as a laser and optics company, producing all these communication lasers. Our connection to Marvell is on the substrate, on the computing side. Of course, embedding of optics, and also future, the embedding of optic interfaces, is a point here. The big advantage between Marvell and us is, Marvell is a partner with a certain complexity in their product range. And so we are capable, together with them, to cover a broad variety of product ranges, and being fast as a development partner in the implementation of and testing of new products. This is the core benefit between us as companies. Had a couple of meetings with their CEO and COO, extremely intelligent people, and they recognize exactly these synergies between the companies on the fast handling of complexity in the product range, and bringing things to production. This is the business model from Marvell as well. For the longer time, of course, embedded optics instead of just having external transceivers is something where we are thinking about. These are the core benefits between Marvell and us. But this sounds like more distant in the future than, let's say, a topic for the next one, two years, right? Yeah, as I said, the partnerships with companies like Marvell, they are not started just to be a business relation only for one or two years. They are implemented for a long-lasting real partnership. To co-design, to co-develop, to bring joint ideas on the table, to have dedicated teams on both sides, this is what we are discussing with these partners, and this is exactly the change in our business model. In the past, a customer came and said, "Please produce this and that," and we did. Sometimes initially with some help of one of the very big ones, but at the end, this was the business model. Now, we have dedicated contact points both for technology support, for sales, for production, for quality, et cetera, with our partners. We have aligned processes, we have aligned KPIs, we have aligned roadmaps, and this is a long-lasting issue that goes on five years plus. Yeah. Exactly for that reason, we are able to do these investments with that confidence and with the financing of our customers in different methodologies. This makes us stable now and so confident about the upcoming future, about our stability, about our growth, and of course, about the delivery of the bottom line. It is not possible in any other business model any longer. Our industry, the upcoming technologies, they have the demand for other ways of cooperation. You can't run it in independent running parts of a supply chain anymore. It's definitely just not possible. This is what we learned from the classical semiconductor industry, where this happened, this transformation happened almost 20 years ago. You have seen a narrowing in that market during that period of time as well, and there have been clear winners. The winners have been those which cooperated along the supply chain and implemented strong partnerships, in good as well as in bad times. This is what we are doing. This is a transformation in the business model, it's also a transformation within our company. We need a different kind of mentality also within our entire employees. We have to open up to our customers, of course within contractual relations, of course within NDAs, et cetera. But this kind of a corporation is relatively new for us, and it's totally new for the industry. Our approach is, and our demand is, we want to be the leading industry value enhancement partner in that industry. This is the feedback from our customers, that we are sometimes too slow as Europeans, as always, but they have the trust that we can be that partner. We are accurate, we deliver to promise. Sometimes it takes a little bit too long, so customers have to push us, but the trust level is there. That's a big difference. Yeah, fully understood. Thanks, Michael. I would have one more for Gerrit, related to CapEx. Obviously, the EUR 2 billion are laid out and still subject to change once the fourth line is communicated, and maybe even part three. But how should we think about those CapEx and those maintenance CapEx apart from Kulim? How should we model this in? Yeah, again, I think when you look at what we are doing right now in Kulim, that's certainly all part of these long-term customer agreements. That's basically building out the full plant and with the manufacturing line, so there's no growth CapEx on top of it. There's normal maintenance CapEx, where we always had a couple of hundred million per year when you look at our previous guidance. Then as well for this year, we had another EUR 150 million-EUR 200 million on top for some growth CapEx in some of the other plants. That's something as well I would see to continue. Certainly we will continue doing this on our own, debottlenecking, upgrading, optimizing technology, and so on. That you should continue in your model to basically have the same assumptions. Would there be any ballpark figure, as a percentage of revenue, for example, that's- No, I think- we could- Yeah, certainly, I know when it comes to percentage of revenue, I would rather take a Because we are having a set number of sites right now, and certainly with the new sites, they are still basically in ramp, or they will take a couple of years before we ramp and have high-volume manufacturing. Therefore, there's no maintenance CapEx there. So I would not just take a percentage number, but I would just take an absolute number, because otherwise, I think you would go overboard with the significantly increasing revenue we are having. Oh, okay. Understood. Thanks a lot. Thank you very much. We have George Brown from Deutsche Bank back on the line with a follow-up question. Please go ahead. Yeah, thanks, guys. Just two quick follow-ups. Firstly, obviously, you've got a lot of capacity coming online to support your growth and new announcements likely to come. I wonder if you're planning to have an Investor Day or Capital Markets Day to provide a new mid-to-long-term financial outlook, and when could this happen exactly? Secondly, just a quick follow-up on the client payments. I think you said EUR 35 million this quarter. Do you have an expectation you can share with us for the total client payments for the second half of the fiscal year? I think it will have a swing factor potentially on the sales and the margin. Thank you, guys. Yeah. George, let me take both of them. First of all, Investor Day, yes, we are planning something for next fiscal year. We haven't fixed dates and everything yet, but we want to definitely give an update on where we are heading, on where we are doing as well when it comes to our expansion project, et cetera. That's something which will be upcoming in next fiscal year. Second question, when it comes to the one-time payment, as we discussed before, George, we want to give you the update very transparently in each quarter we announce our numbers. But to give a forward guidance, it depends too much on, as we discussed as well, on final alignment with the auditors on one or the other elements. There are several agreements with several elements in these agreements, so that needs to all be discussed and aligned with the auditors. That might as well move one thing into a next quarter or not. Therefore, I don't want to now give you a specific guidance and then tell you, "Ah, we had four weeks later the alignment, and therefore it moved," or whatever. I think that would not be helpful. Okay, perfect. Thank you, guys. The last question goes to Lukas Spang from Tigris Capital GmbH. Yes. Hi, good afternoon, gentlemen. I would like to start or follow on the topic of Kulim 4 to 6, and you already gave some insight to the products which could be behind these further expansion plans. Just an understanding question, should we expect one by one, so four after three, five after four, and six after five? Or could it be also in a different row that, for example, five comes after three or six after four? How should we think about that? Okay, so three definitely will be the first one that we decide on in the first half of the next year. The other two at the end are independent from that. The upper right one is the one where we think about a further extension of cores and substrates. On the lower right side, this longer plot, this is something where we are thinking about advanced PCB plus optics. The PCB and optics part will be handled separately and definitely can start earlier also from a resources point of view. There will be separate different resources being involved. We are actually in the pre-planning phase, thinking about how it could look like, what are the technologies we want to attribute here. We are in the first phases on starting discussions with customers, what they can imagine in a couple of years, what will the demand be, where is the technology going, how layouts should look like. This is five and six. Five and six will be together. First definitely is three, then most probably it's five, six, and four needs to have at least a start of three in advance. Otherwise, it doesn't work. Mm-hmm. Okay, that helps. Then we saw in October that NVIDIA announced further contracts with Unimicron, IBIDEN, and Samsung Electro-Mechanics. How do you see this in general? Specific to this customer, do you see this customer door now closed, or do you see any chance to still get into it? Actually, we are limited by capacity. It's not a question if you want to enter or if they want to enter. Today, we just can't deliver anything. As I said before, due to the fact that we are advanced in the technology of both, of substrates as well as of advanced PCB plus optics, be assured that everyone in the market is talking to us, especially for the mid and long-term future. But actually, we don't have any capacities. Yeah. Sure. Okay. Thanks. Okay. As there are no further questions, we will conclude today's conference call. Thank you for your participation and questions. If you have any further questions, please feel free to contact our IR team, Johannes, and me anytime. Thanks again. Yeah, thank you very much for your attention and for being interested in AT&S. Thank you, everybody. Goodbye.