Infineon Technologies AG (IFX) Earnings Call Transcript & Summary

Yes. Hello, everyone. Welcome to this year's IPC call, and at a fair ground in Nuremberg. A very more welcome to our 2 speakers, Peter Wawer, the Division President of IPC; and Peter Friedrichs, Vice President, SiC, so the true experts in power within Infineon. And with that, I hand over. And yes, please go ahead with your presentation. And not to forget, for all in the webcast, if you want to post a question, please send it via mail to [email protected]. Thank you.

Yes. A warm welcome also from my side. Thanks for the kind introduction, Alexander. We are happy to be back, so to say, right? Happy to be here. Almost unbelievable 3 years ago. The good thing is that after 3 years, you might expect some changes. And we would like to give you an update where we stand today, what the outlook is. As we all know, as you know pretty well, right, very interesting times, very challenging times, but also great opportunities for semiconductor manufacturers. And we want today to focus now on the update for IPC. And as Alex pointed out, I share this presentation with my colleague, also Peter, Peter Friedrichs, because he is our key expert on silicon carbide. And we all know that silicon carbide is at least for power manufacturers of specific interest, right? So we come to that in a bit more details in a minute. Let me start with a big overview, how we see ourselves from Infineon, but especially also from IPC perspective? What is our credo or purpose, as we say? We empower a world of unlimited green energy. Due to the recent events, we have become aware not only that, of course, due to the known CO2 challenges, getting rid of CO2 emission is of primary importance, but also the energy dependence on fossil fuels is of now recently of very, very high interest and very high concern, as we all know, to the ongoing war in Ukraine. And therefore, I think this has a double the importance of this topic, right? So we are proud that according to the analysts, we are the #1 provider of semiconductors in the area of renewable energies. We increased energy efficiency by our products due to improving efficiency, but of course, also being the driving force behind the energy transition from burning fossil fuels to generate electricity and heat towards, of course, using renewables, sun and wind predominantly. Of course, we have also a CO2 footprint, right, as a company. If you take into account the products we manufacture and sell, then we are, of course, already net positive due to the positive benefit of renewables, i.e., wind and solar. But of course, we are working also to become carbon neutral regarding our own footprint. And I think that is also worth to mention, that we have very ambitious targets to go towards green energy, reducing down to -- by 70% until 2025, becoming finally carbon neutral at the end of this decade. Let's have a look into the numbers. So after having kind of challenging time during the corona crisis, flattish revenue in 2020. Again, then healthy increase out of the crisis '21. Underestimating a bit the potential of the market, but now clearly, further increasing demand and also supply. We see a positive trend, right? We were plateauing a little bit shy of the 20% segment result margin. Now for the first half, we are above 20% segment result margin. The fiscal quarter we just announced, record revenue for EUR 30 million, we never have achieved before, so purely organic growth for IPC. And as you might remember, might know, typically, the first half of the fiscal year is the weaker one, having the weakest quarter in fiscal Q1. So usually, following the normal pattern, which we also expect for this fiscal year, the second half will be even stronger than the first half regarding revenue. So of course, we are in a tough environment regarding supply chain, China lockdown due to reoccurrence of the COVID virus might impact our performance here and there, let's see. So far in the first half, I would say we managed decently. Only minor impacts regarding revenue shortcomings for IPC. Key customers, I would say, no change in general. Of course, very nice balance traditional industrial customers and upcoming fast-growing customers in the area of renewable energy. I would say, very nicely balanced portfolios. Of course, these are the big customers. And not to forget that IPC is highly diversified regarding products and also customer base. So we talk about thousands of products and also thousands of customers in total. And here, of course, the well-known big ones are named. Interesting, I would say, also the revenue split. Please note that this is the revenue split '21. I think also historically not a very big surprise, but I think I would like to highlight the topic of SiC, which we put here now and dedicated to signal that this is a meaningful number. And as Peter will show a bit more in detail, we will almost double this amount in the now running fiscal year. So these are the '21 numbers. We would love to double. We will not make it completely due to supply constraints, but very aggressive growth that we were executing on silicon carbide-based products. With that, I would briefly guide you through how we see the applications we serve. On the left-hand side, you see the symbols indicating the application. And the percentage number is the exposure to this application revenue-wise for IPC, meaning roughly 1/3 of revenue of IPC goes into automation and drives. 5 years plus ago, this number was well above 50%. So the message is very clear. While this is still a very important big chunk of our business, that is slow growing, and disproportionate growing. And the high-growth area are renewable energies, and of course, also power infrastructure, which outgrow also for the foreseeable future, for the next years to come, boosting our overall growth looking forward. So the current status, market outlook '22 is, as indicated by the green arrows, Automation and Drives is quite resilient. So being very -- in a very good shape also in '21, continues to be in good shape in '22. Renewable energy generation likewise, especially PV, is very strong, while we saw a certain weakness in the area of wind, it's basically overcompensated by the very, very high demand of PV. Basically, we're sold out for the near term. Even maybe beyond '22, we see this very high demand. Home appliance, very clear. That's a topic which used to be strong in '21, but it's very close to consumer business. So this is something where we think we are over the top regarding high demand and expect declining on the customer side while we continue in '22. And also for '23, major home appliance is expected to be a bit more blurry. Transportation was not booming after COVID appeared. So on the transportation side, investments into infrastructure and high-speed trains, where a big bunch of the products are being used, stalled, really significantly went down. Until today, we digest a certain amount of idle because the business did not recover, but we now see really a silver lining, also increasing book-to-bill. So transportation, I would say, trend recovering in the right direction for the foreseeable future. And the other topic regarding power infrastructure, and I will comment on this in a bit more detail. It's another very strong growth driver, longer term looking forward. Only close to 10% share of our total revenue, but that is very solid double-digit growth. Looking forward, I expect overall, something 15%, 16% CAGR for the next years to come. Others, likewise, positive trend. So now I would like to draw your attention to some selected examples. That slide we also showed at the Capital Market Day a couple of months ago, but I think it's nevertheless very impressive. You see on the left-hand side, the indication to what extent power semiconductors are being used in the area of wind, solar and storage. And obviously, for a semiconductor manufacturer, it's very positive that we see this trend away from fossil, where basically rounded per megawatt, the euro content is basically zero, right? Instead of zero, we see here this kind of range, between 2,000 up to 5,000 in the example, in the given example, semiconductor content per megawatts being installed. And now you see here the additions in 2020 for the given example, storage, solar and wind. '21 number is not yet available. Typically middle of the year, they're expected, so rather soon, but these are the numbers for 2020, 114 gigawatts wind addition; solar, 134; and storage, 5. Now there are a couple of scenarios looking forward regarding the targeted CO2 reduction ambitions. And the so-called sustainable development scenario provided by the EAA, that estimates that carbon dioxide footprint is being reduced to 0 until 2070 on the globe. That is rather a bit longer term, right? 2070. Basically, everybody agrees that this is not aggressive enough. But if we would pursue this scenario, then we would need wind, 110, somehow fits to the status quo, PV, 240, almost doubling or, let's say, adding 100, and significantly increasing a factor of 4 the amount of storage already in this decade. If we become a bit more aggressive and say, hey, 2070 will not be sufficient because we have the problem right now. And we rather have to act more aggressive and sooner, we come to the net zero emission scenario and the EU, which means that we pull the whole thing in under 2050. That is still many years to go, right? If we want to achieve this, then you see here the numbers doubling on the wind side, again, addition of close to 200 watts for solar, and of course, another significant increase for storage. And on the right-hand side, you see then, exemplary under 2030, what that would mean regarding the cumulated additions in solar power capacity. And I think these are very impressive, very big numbers. I think there are a lot of arguments why it will go into that direction. Because as we see regarding levelized cost of electricity, the renewables are there in many areas, in many regions of the world. They are meanwhile, regarding levelized cost of electricity, the most competitive, also not regarding incentive and subsidy schemes, there in most regions of the world need not be any more subsidy. And the other thing regarding the independence of fossil fuels or fossil fuel suppliers, of coal and gas suppliers, right, has due to the recent political developments, become of very, very high interest, especially in the developed economies. So all good arguments why we believe that will move on. But that's not enough, right? One example now for Germany, but the example, I think, in general holds, right? When we change the energy transmission system towards renewables, that means a massive change in infrastructure. So infrastructure needs to be expanded. It needs to be renewed, needs to be set up. And these are the power connecting grid lines planned for Germany: Ultranet, SuedLink, SuedOstLink. And what is not mentioned here is, of course, also the combiner and transformer stations from the offshore sea parks where typically the energy is collected offshore, 200-kilometer out in the sea, then transformed. We have such kind of HVDC transformer stations brought to the shore and then again fed -- converted and fed into the grid. So for each converter station offshore, you need offshore and then onshore for the conversion, such kind of huge stations. These are a couple of hundred millions which need to be invested here. And they are -- basically, these converter stations are full of IGBT modules of this kind of style. And we have a very good position in this area with our existing technologies. Since those technology are slow switching, they are expected to stay with silicon, by the way, out of cost competitiveness considerations. So this is also a byproduct of this transformation, which is perceived very positive for us short, mid and long-term. Another example which now pops up, becoming a huge growth driver also compared to 3 years ago, we now have a feeling and an estimation to what extent that will fuel our growth is the topic of EV charging infrastructure. And this is a very interesting topic where we see also the benefits of now SiC products. Because with help of SiC products, you significantly increase power densities in fast switching applications. I think Peter, you will also briefly comment on the technical session. And very clear, we are from a semiconductor perspective a system supplier being able to supply microcontroller, gate driver ICs, nicely fitting to each other. And then depending on the power you need, either discrete, silicon carbide-based or easy modules. I can tell the initial designs, of course, of these stations were made still by IGBTs, but that is now really quickly converting towards silicon carbide, because in the same form factor, up to 30% more power can be derived out of the station. So a very clear value proposition. Okay. With that, another example, having the smart grid being connected to your home, right? You have a combination of now many dimensions. You want -- if you're lucky and you have a roof, right, and it's well suited, right, you want to generate your electricity to a certain extent on your own, which I do since quite some time. You now consider it storing it if you can afford and have the ability to set up a storage system. And then, of course, the other thing is that you would like to connect once you have an electric car, you want to elect -- connect your electric car. And the interesting thing is that, of course, you can also use the car as a storage and a buffer system, right? And that is not a vision here, that is already a reality. That now on the system side, these kind of topologies and system solutions are being offered, that you can use basically generation, storage, and of course, also transportation together with the grid connection in your connected -- energy connected home. So again, super value proposition for semiconductor manufacturers. And we are trying to serve each and every dimension which we can address with help of our semiconductors. Now coming to the end of my introductory speak, also a perspective on the customer side. We invested quite a lot in developing our web-based community. Of course, we have the traditional forms of supporting our customers. This is also converting more and more to a digital service, right, due to VR glass capabilities, et cetera, and of course, due to the limitations of COVID, lesser travel, lesser on-site support physically, but a lot of support virtually. And now we meanwhile developed nicely the community, the web-based community where I just wanted to share some numbers here with you. Several thousands of users, a lot of discussions ongoing, particular interest in our SiC Forum and a huge amount of monthly page views as an indication that we are here on the right track. And last but not least, what we also get out of this customer interaction, that we want to expand our offering beyond the hardware, right, also providing services regarding analytics and -- about lifetime, about performance of the system where we now offer a web-based service where the customer can upload his mission profile for the given application and derive certain information via simulation of the specific topology he is using, he or she is using, and then deriving automated information about lifetime, about use cases, being able then online supported to optimize its system. And with that, I would like to hand over to you, Peter. And you give us a more detailed insight into where we stand on silicon carbide.

Yes. Thank you very much, Peter. . Yes, once more, it's also my pleasure to be able to again share with you our development in the area of wide bandgap. So we have a very long history in this segment, especially in silicon carbide. It's more than 25 years that we are in this technology, more than 20 years on -- in the market. And we are very happy to say that we are finally also able to harvest the -- based on the experience we have gained in this segment. Of course, different ingredients are required to be successful even in the wide bandgap market, even if it's a tight market. And therefore, I would like to make you familiar with our recipe for wide bandgap success in the next few minutes. Yes. First of all, I think this is well known to the community, also to you here, the growth rates we expect in the wide bandgap arena are substantial, by far higher compared to the global power semi market. In case of silicon carbide, expected until the middle of the decade, close to 40% CAGR, with a potential of close to EUR 5 billion revenue. Gallium nitride, even higher growth rate, of course, coming from a pretty low level. So therefore, the expectations are high. The potential is very high. And we as Infineon, we feel pretty well positioned here. Looking into the core topics Peter already mentioned in his part. E-mobility infrastructure, renewable energy, these are the key topics for us in the Industrial division. The e-mobility infrastructure is an enabler for the E-mobility itself. So we have the products here, we have a very broad customer base we have developed. And also in the case of gallium nitride, the products we are having right now out are addressing the mega topics for gallium nitride in the -- on the short or midterm scale chargers and adapters and data center, power supplies. And of course, in the future, more to come even in this segment. I will explain it in a bit more detail when going into the portfolio, the current one and into the outlook. Revenue wise, yes, we are very happy to report, last year, close to 100% increase of the silicon carbide revenue, also for this year. We will come close to that. So at the moment, it looks like that we can increase our revenue by 80%, a little bit also influenced by the fact that we are stating this year in U.S. dollars. And I think you are all familiar with the euro versus dollar development. But in general, again, significant growth rates. So we still believe that the EUR 1 billion revenue for us at Infineon can be possible in the mid-20s, with a healthy share between industrial and automotive applications. And what are the baselines here? Of course, a very strong position already today in the market, Especially, the recent market reports indicate that we are climbing up to #2 with respect to market share in silicon carbide. If you just look into the industrial landscape, that's why where we stand for here from our IPC division, we can say we are clearly the #1 here amongst the players right now out. And of course, this is possible -- became possible because we had the highest year-over-year growth among all our peers here. Very important, very broad portfolio. So you will see it in one of the next slides. So we are able to increase width of our portfolio with respect to product types, to power ratings, blocking voltage classes significantly. And this, of course, enables us to win a lot of sockets since this customization topic. So having a specific product for a specific application became more and more relevant also for silicon carbide. And we are also very happy to state that already first product based on our cold split technology are running in production or basically shipped to customers. So that was a pretty fast development, a bit faster than we expected by ourselves. So you gave us 3 to 5 years to commercialize everything. It took us a little bit less than 3 years, what was really a very great achievement, also highly recognized not just by our management, but also by our customers. Yes. A big debate is always about the technology selections. And so we have the big discussion between the trench technology and planar technologies. Many of our competitors are still with planar, planning for trench in the next -- or over next generation. We decided to skip the planar configuration and the motivation for that or the reason I would like to explain once more here, with this simple graph, which indicates basically what we call a technology curve. So such a technology curve always basically reflects the possibility to optimize your device. With respect to reliability and performance, there's always a trade-off. And then it's, of course, in the decision of the manufacturer where to put the focus, whether more on reliability and performance. If you look into the landscape of the planar devices today, so basically, I'm not giving you numbers. But from a qualitative point of view, if you accept that this is positioned there. So we decided, for us with our trench technology, basically since we have now the ability to increase the performance, that we don't go just completely to the right with respect to performance because the reliability we see in the planar configuration is still not sufficient. Therefore, it was our cautious decision to move up a little bit here on this graph to increase in reliability. So sacrificing basically a part of the performance opportunities we have as a trench compared to planar in order to offer here to our customers' device, which is really comparable to the well-established IGBT technology. And this is highly recognized. Also, basically reflected in the commercial success we have with our first generation and second generation is basically very close to be released. Already sampled to key customers. And what we are going to do here is, of course, we will not jeopardize our very high reliability standards. That means we will hear in this case, then fully utilize the option to increase the performance. So positioning, again, will then be here at this place. So maintaining the reliability we already delivered today, but gaining in performance. As already mentioned, portfolio is key. That is the experience we have made not just for silicon carbide, but also in our silicon world. And for that reason, we significantly increased the number of products, the number of platforms, which we equip with silicon carbide. So just basically today or yesterday here at the PCIM, we released a completely new blocking voltage class, 2 KV, which is a perfect fit to especially the renewables and storage systems, which are based today on higher voltage bus levels. We have a couple of very interesting customer projects in the very high voltage segments, especially addressing traction. So there has been some press releases also over the last few months, which already show the potential of silicon carbide in traction systems. There was one about the tram in Munich, proving basically an energy saving potential of 10%, which is really substantial, enabled by silicon carbide. So this is a very nice opportunity which is going to come. But also in the existing segments of 1,200 volt and 650 volt devices, here, we increased the number of components. We added new package platforms. We upgraded the existing technology to the so-called M1H platform, which gives the user a much higher flexibility with respect to the driver design. So all in all, very nice development. And basically, also the ground for the success we were able to show over the last years. And believe me, we will continue here rolling out more products. New discrete platforms are in development. One is already launched now with the 2 KV technology. And also in the automotive segment, which is not directly displayed here. A lot of new developments is ongoing and will be launched over the next couple of months. Same for gallium nitride. So there are some voices that is pretty quiet about gallium nitride for Infineon. Well, we are pretty successful also here with gaining new designs. So we took our time to develop the technology into an -- from our point of view, very mature and especially very reliable status, because this is basically what differentiates us, in many cases, not just in wide bandgap, but also on silicon from our competition. And you see the lineup which we have already launched, respectively which will come up within the next few months. So what will be new is the segment of 100 volt based components based on Schottky gate HEMT structure. We are also going to add further versions of the IPS solutions. These are small systems in the package, not just the power part and the driver, but also adding here the control of functionality. So it's basically a very easy to design in solution for markets which are very dynamic, where basically also the customers expect, let's say, a product with a lot of functionality already from the beginning. And, we also plan to add then further package platforms here for our high runner 600 and 650-volt technology. So basically new versions of the CSMD platforms, which are then intended to address the target markets for gallium nitride, which we see at the moment predominantly in the already mentioned Power Supply segment, in the Charger segment, with further additions of the portfolio. Also, low power drive segments will be addressed, and to a certain extent, then also charger applications in the cars of volt and onboard for lower power ratings. As already mentioned by Peter before, one additional aspect of our success in the market is that we are not only offering the power part, but also the ecosystem around. And one important ingredients, especially for wide bandgap, are driver components, since driving -- operating the part is, to a certain extent, more demanding compared to silicon technologies, especially for gallium nitrate, but also for silicon carbide. And we are very happy to say that for each and every component, we are now able to offer also a perfect fit with respect to driver IC, which meets the requirement of those new fast switching devices. There are aspects like accurate timing behind, aspects like very high voltage, transient immunity. This is all then integrated already in those driver ICs, up to the protection of the new components against failures, which are a little bit different compared to silicon. So this is all done here. It goes from very easy solutions with a low level of functionality, just turning on and turning off the device, up to driver ICs, with the ability to program them, where basically you can eliminate a complete design cycle on board because basically you set up your system and your program your driver IC basically on the fly in the board, which reduces significantly the system design time at the customer. So what are our target markets? We split them a little bit into the ones where we are already very successful, where the penetration rates are substantial. In some of them, like especially EV charging, basically, they are already dominating the landscape. So IGBTs are only -- can only be found in a very small number of those solutions. So here, we already reached the so-called tipping points, which for us is an indication that the majority of the products will switch to the new technology. Power supplies are basically in this transition phase, industrial power supply, same for onboard chargers. And we have a very bright outlook also here in the case of silicon carbide for drivers, the storage systems, commercial air con systems, home clients, very interesting aspect. This is very much cost driven, but still now we have a lot of interesting discussion also in this segment. Wind power and traction, we already talked about. And a similar picture then for the gallium nitride side, charger, Server Data Center Telecom is basically already tick-marked. Motor control for small power motors is a very interesting topic today, increasing more and more in -- the focus for gallium nitride products and all the other things here, of course. In future also, we'll add here. I mentioned already the onboard charger topic. Wireless charging becomes very interesting. Audio amplification is already in a small level present for gallium nitride. Of course, major home clients are target markets for gallium nitride, where the product portfolio will be oriented on in the future. A very nice development for us, I would like also to share with you. And this is basically the width of applications we are now able to address with silicon carbide-based products, for instance. So if you look back in the fiscal year '17, that was all about solar. 80% of the design wins in 2017 were in the segment of solar string inverters, predominantly boosters, but also inverters. So only a small portion of other segments could be addressed by silicon carbide. This has been changed significantly if we look into the last fiscal year. First of all, the sheer number increased by a factor of 5. But if you look to the distribution, that is a very nice development for us because it shows that we can base really on a very wide range of applications, of course, also then, meaning a very high number of customers. So the penetration into the industrial segment is reality. It is no longer concentrated to a few customers and one single application. But basically, all those things here, like UPS, transportation, et cetera, contribute substantially. And if you look to this picture, in 2 or 3 years, we will see again a different distribution, with more focus of other applications and maybe solar dropping with respect to the share. But also, this number here, the 5X will increase in the future. Enabling factors are that we are able to differentiate in technology, chip technology. Trench-based is one key aspect. But on the other hand, we also need to facilitate the interface between the chip and its environment. This is in case of wide bandgap components not so easy because typically we have very small chips compared to silicon. What is a certain hurdle, especially if you want to manage the devices and the application with respect to thermal performance. And here, we introduced recently our so-called .XT technology also for discrete components. That is a technology we have in modules and a certain -- a little bit different realization rate in place for a couple of years. But now also for discrete, that's available. And what is here the major improvement? We replaced the backside contact of the chip in the package, which is traditionally a solder process by a new approach based on a diffusion soldering. So a different metal on the backside forms a very solid, highly performing contact to the lead train. It brings a lot of benefits, especially we can increase the thermal dissipation capability significantly. Meaning, one of the same chip can now handle a much higher power rating. Basically for the user, it means without big changes, he can basically increase the power handling capability by 30%. And it's very simple. It's nothing -- what is a big increase in cost, it is rather the other way around. It's just a new die attach process, even eliminating a certain step in the back-end processing, and [indiscernible]. Very important for silicon carbide and also for the trench technology, we have typically smaller die in SiC and trenches, even smaller than planar. What, as mentioned, is a challenge for the thermal performance. But with this technology, we can compensate for that drawback because it simply significantly reduces the thermal resistance. And therefore, we can still maintain the higher power handling capability even if the chip becomes smaller. And we are very happy that we received for this development just 2 weeks ago Der Deutsche Innovationspreis. Sorry if we're not translating it into English because the German Innovation Award is a little bit different. So therefore, we have to be precise. But we have been the winners for big companies this year in 2022. And yes, this also recognizes basically our way to be a leader, not just in the market, but also in the technology. Yes. And recently, we added our 2 KV technology. We launched it officially here at the PCIM. We have shown it already in the portfolio. So what is behind here? The biggest impact of 2 KV based silicon carbide MOSFETs in power conversion systems, here, an example from PV, is that we are able to increase the so-called power density in the formula kilowatt per kilogram. This is important because in those installations, typically, the weight of the system decides how many installation guys you need in the field. And here, we were able to basically achieve such a big step forward, that customers now are moving away from the multilevel topologies which have been used before to the single stage enabled by 2 kV. And you see an example from all of our lead customers. So the evolution of those systems. Now in 2022, we have a 2 kV system, having done this very nice power density ratio enabled -- and again, enabling technology for that 2 kV silicon carbide MOSFETs. More to come because the same story is also valid for energy storage system for future charging of cars. Also here, we see a trend to higher voltages because higher voltage typically means you have less current, less copper. So again, we are back to the story of saving materials. Solid State transformer in the world of DC, energy transfer became more and more relevant. So also here, you will find the use case. And then, of course, a big topic of PV, which is moving to 1,500 V bus systems. This is going to be entertained with this new technology. And I mentioned the high voltage outlook, we have already in place. So here, we are very happy also to show one example together with a big customer in the traction segment in this case, Siemens. So it's a pleasure to go by tram in Munich, you may be lucky and find this particular street car here, equipped with silicon carbide. So it was already for more than one year in operation, 65,000 kilometers without any problems. And again, 2 substantial improvements. One is the energy saving potential, around 10%, which is basically, in today's time, something which is very much appreciated by the operators of those fleets. And that's also very interesting for people living basically in towns, in the vicinity of those street cars, a significant reduction of motor noise. Not just, by the way, the motor, but also the auxiliary power. So if somewhere a tram is parked in your vicinity today, you always hear some noise due to the auxiliary power supplies running there. If it's coming with silicon carbide, the noise is going to be completely eliminated. So substantial advantage is possible here. And also, the customers really expressed their thanks towards the cooperation we had here in a very, let's say, ambitious project. And as mentioned here by Ingo Wortmann, the Head of Mobility at Stadtweke Munchen, so they are really interested in those solutions. They will pull also towards their suppliers to implement those new technologies, what is also good news for us and for silicon carbide in general. Yes. All in all, we have a big future. And we have made some decisions. So I would like to hand back to Peter to give you a bit of an outlook where we are heading.

Yes. Thanks a lot for giving us these technical detailed insights. I think -- I hope this helps you, right, to also understand the value proposition beyond the pure material physics. It's much more than having one MOSFET, right? It's about considering and understanding the requirements from the system and having the appropriate answers based on customer needs. And I think we hopefully convey to you that the future of IPC Industrial business is bright in silicon, but specifically also in wide bandgap silicon carbide. And so the decision has been taken and also officially communicated, right, that we massively continue investments in capacity expansion. So our Kulim site, you see also the discussion we had just before this meeting, right, is the perfect area where we now have already a huge 8-inch, economy of scale wise very competitive facility, which you see on the left-hand side. Two modules. And now we add a third, even larger module, where we will establish from '24 onwards. Manufacturing is expected to start in the second half of '24, where we now focus on ramping up wide bandgap silicon carbide and gallium nitride. So in total, we will invest there more than EUR 2 billion. And also, the revenue potential is significantly above EUR 2 billion. And yes, the key data, I already disclosed. So we significantly believe in the growth perspective of these technologies. I think coming to the very end last slide. We see that we outperformed the market over the past 5 years. On IPC side, we have been specifically resilient also to the COVID topic. Of course, growth was rather flattish in 2020. But we overall think that we are on a very sustainable growth path due to the adverse megatrends we already discussed. As I see portfolio, I think no need to reiterate seen from the numbers that we are perceived as leader in industrial applications. And then another topic I would like to highlight, we are a firm believer that looking into the sheer discrete module product is not enough, right? We have this perspective. And I think also Peter nicely conveyed the message, that we need to understand the application, designing our products and the total product offering consisting out of the semiconductor switch gate driver and microcontrollers in an appropriate way that we help our customers to get the best value proposition out of their application together with our semiconductors, might this be silicon-based, silicon carbide or gallium nitride, doesn't matter. We will find the best solution together. Complementing this with help of reference designs. That is a very important topic for the customers, because, as Peter mentioned, wide bandgap is not adopting replacement. So we want to lower the hurdle for our customers, proposing them reference designs, which in certain areas are quite complex, like DC-DC converter stage for an EV charger, which you can download on Infineon web page, for example, software, and of course, also digital services, which I touched upon very briefly looking forward. With this, we're done. Thanks for your attention for today. And I think now we go for the Q&A session, and feel free. And as Alex indicated, we will have Q&A here in the room and you write an e-mail to Alexander Groschke. Please?

It's Francois Bovine from UBS. I would start maybe to talk about Siltectra. You didn't talk too much in the presentation and the progress you made since you acquired the company. Where are you in this process and development for mass production? And which impact should we expect depending on when it comes? That's my first question. And second one, on the trench first generation. We saw maybe like 2 approach. Some are using uni planar. You seem to invest more in trench. The second generation is underway, as you mentioned. Can you talk about the improvement of the second generation versus the first one? And what's the advantage versus planar? Should we expect on the battery? Or anything you can share would be great.

Should I? Okay. Yes. Starting with Siltectra. I think it was just mentioned in one line, but of course, it deserves a bit more attention, you are right. Well, we are very happy to say that we could qualify our Split technology to generate own wafers out of [indiscernible] we are purchasing already last year. So in September last year, this qualification process was successfully finished. So we got the release to use this in-house produced wafers for normal production. On small scale, of course, we have a pilot line investment. So we transferred the start-up like lab we acquired in 2018 into a small pilot line, able to offer let's say a smaller portion of our total demand with respect to weekly wafer starts, but we are operating this line now since last year in a 2-shift system, delivering certain amount of wafers on top of what we have purchased into the manufacturing. But of course, we will not stop there. We have already now made 2 decisions. One is to install a very high volume production side also for the split technology together with our efforts for front end manufacturing in Asia. So this is basically already started. So the technology transfer is initiated and if everything runs well. So in 2 years from now, we will also add material from this new site. And also in Dresden, where we have the pilot line today, we are again moving into a larger facility, again, increasing here our capabilities. This is a primary focus of continuing R&D, but still having also a small volume production capability in place. So all in all, we have commercialized the Cold-split technology, but there's more to come. There's more opportunities based on this technology in the overall device flow, and this is what we are going to pursue over the next years. Yes?

What is the hurdle that you see still to make for really mass production? Is there any hurdle still?

I think it's a question of ramping now. So one aspect is it's a new technology. It's not like in a traditional device technology flow where you can ramp by just adding available tools into clean room. So here, we also have to deal with the fact that there are new tools. So we need somehow to, let's say, transfer also the tool park into automated high-volume production setup. So that is what we are addressing right now. But the rest is just execution, I would say. Yes. Regarding the question trench -- second generation. It's -- a new generation in our understanding. If we define a new generation at Infineon, we always change what we call the cell, either with respect to the cell concept or with respect to the dimensions. Second generation will be further shrink, relying more or less on the first-generation device concept, but improving the power handling capability for a given die area by 25% to 30%, depending a little bit on the application. This will be the major, let's say, added value with a second generation. There are other smaller aspects then added into those developments, so also related to an increased portfolio of new package solutions. There might be new metal schemes available with the second generation. But the major aspect is an increased power handling capability while maintaining -- or most of the other features, because they have been successful, they are successful today. So it's a further step then to, of course, reduce the cost and to increase the penetration rate. More revolutionary things to come, but the second generation is more or less a shrink.

Andrew Gardiner from Citi. I just wanted to come back to a point you made, Peter, about sort of the energy grid installations and how that was a slow switching application. And therefore, you weren't worried about silicon carbide coming in and cannibalizing your silicon IGBT. Can you sort of speak to that more broadly across the IPC portfolio at the moment? You've clearly seen an increase or diversification in the order wins that you're getting for silicon carbide. But just can you give us an idea of perhaps the percentage of your portfolio or revenue base that you're quite comfortable will not be cannibalized in the future and perhaps that may be?

Yes. Thanks for the question. I think we learn while we walk, right? And if you look now back over a longer time frame, I would say, from today's perspective, we see that silicon carbide can go further than we thought maybe 4, 5 years ago. And that is always going in line, of course, with the understanding on the system side of our customers, right? So we together running through this kind of learning curve. Generally speaking, you can make a statement that we say, the faster the switching speed and the application, the more likely it's silicon carbide, either today or tomorrow, right? And that is true for the EV charging system. That is true for the PV inverter. So if you go to the grid, that's another extreme, right, then you are switching very slowly today, right? And also, the innovation cycle on those grid infrastructure is extremely slow. And therefore, the confidence level that this stay in silicon for the value proposition, cost performance aspects for the foreseeable future, the confidence level is very high. And then we come into the market of drives, which is our big market, as you have seen, right? Something like 1/3. And here, it's also a mixed bag. I think there are certain applications like server drives, where out of performance considerations and it's also volume floor space considerations and the applications, there is a value proposition for silicon carbide as well. But for general purpose drives, right, a certain very cost-sensitive, not fast-switching application because the switching speed is determined by the rounds per minute, right, of the drive. For the foreseeable future, we do not see any interest in silicon carbide as we speak. I don't want to exclude that there might be innovations coming up also in the drives area with -- where this picture would change, let's say, in 5-plus years from now. I would not exclude this one. But for the time being, that is also demanding and requesting for the next generation of silicon IGBT, that we are also bringing to the market. So IGBT 7 is in the market. We just launched also the 1,700 volt, and we also will have an IGBT 8. IGBT 9, I don't know yet. So there's still also a future in silicon that should not be underestimated, right? So we have economy of scale, 12-inch, will for sure not be in this decade in silicon carbide. That one I make here, I make a bold statement. And so far, I think the race is on, while definitely more and more applications are being eaten up by silicon carbide. That's the dynamic.

And just another one sort of specifically on the silicon carbide growth. You've kept your target vague though it may be in terms of the timing of $1 billion of revenues sort of middle of the decade. Some of your competitors that are currently behind you in terms of the market share that you showed on the slide are saying they're going to get to a $1 billion run rate much quicker. I'm just wondering what you make of the competitive dynamics. How you feel about your sort of your win rate relative to peers? And does all of that sort of mesh?

Yes. I think it's exciting, I would say. There are others who are more aggressive than we are, that is true. I think now showing the growth rate that we have shown, right, now growing year-on-year 80%, I think it's something, right? The point is you always have a very high growth rate if the absolute numbers are small, right? The higher they get, the more challenging it will be. I think I would say we take it from a competitive perspective. And we will try the best to keep our share and defend our share. And I think, to be honest, everyone has to make his own judgment regarding also investments. You need to have the clean room space. You need to have the specific tools and equipment. We know today, the lead times are huge for equipment, right? If you have ordered them already 2 years ago in big numbers, then you're lucky. If not, we might see here and there a delay on the execution side. Let's see.

Olivia Honey from Jefferies. Picking up on the $1 billion of revenue point by 2025. How can we think about the split between industrial and automotive by that point? Clearly, industrial is the larger part of revenues at the moment. But I think Peter said it would be a healthy split between the two by the mid-20s. Can you give perhaps a more numerical pointers on that? And I suppose, how can we think about the growth rates between automotive and industrial between now and mid-20s?

I think giving a rough indication, right, I would say we had definitely a head start on the industrial space because we were from the beginning very much focusing on selected applications, understanding the dynamics on the customer side, and then making -- providing tailor-made products, right, for the given reasons having also key customers. I would say the strategy that ST pursued on the automotive side, more or less, we pursued a bit on the industrial side, if this is a fair comparison, 80-20, like this one. Looking forward, and of course, as the market numbers indicate, the sheer amount of market on the automotive side is, of course, much bigger compared to the industrial space. So -- and I think the -- your numbers are not too wrong regarding the comparison of market size. So looking forward, the bigger chunk of growth at Infineon will come from the automotive segment. Very clear, right? Where we will try everything on the industrial side to defend our market position in certain segments where we are definitely, from today's perspective, in a leadership position. So I think very clearly, automotive, I expect that over the next couple of years, automotive will outgrow us, yes? But also not to be underestimated, right? The very fragmented area of industrial products is also a source of profit. So while we are in short supply, generally speaking, right, having higher demand, we optimize, of course, profitability also addressing those sockets where we see the best margin. And that's not always on the very high-volume opportunities.

Great. One more, if I may. I think you've given a EUR 1 billion order book size for GaN. Can you give us a rough estimate of what the time frame on those bookings coming through would be? And I guess, similarly, can you give an order book figure on the silicon carbide side?

So I'm not sure that I got the question completely. Maybe you mean regarding the slide we showed on gallium nitride growth rate in general? Or let's jump back to it, wait a second. Here we are? So what was the...

No. Not the market as a whole, your bookings?

Our bookings?

Yes.

No, I can because bookings at IPC are zero. So gallium nitride is addressed by PSS. So it's -- as Peter pointed out, it's investing in the area of 650 and below, right? We have an overlap in the area of 650. So IPC definitely looks into gallium nitride opportunistically, because at 650 silicon, silicon carbide and gallium nitride meet, right? But as of today, the applications in the industrial space are completely on silicon either or silicon carbide. So no gallium nitride, sorry. And I don't know by how the numbers of our PSS folks. But clear priority and focus of the application is here, chargers, adapters and SMP as in data centers which are not being served by IPC.

If I may add, the PSS colleagues didn't give out a concrete number. Gallium nitride is way more early in the development. We now have a competitive portfolio in the market. We have high design wins. And now it's about ramping. And then yes, we will come up with revenue numbers as well. But as of today, it's a bit early days.

I see that some vendors are already starting to offer silicon carbide products based on 8-inch wafers. Can you please comment on the Infineon plans for going to 8-inch for silicon carbide?

Yes. We have a project set up development and transferring the 6-inch technology to 8-inch. It will be put in place in our new facility in Kulim in high-volume manufacturing. The initial activities already start at our R&D side in [indiscernible]. And the -- timing wise, we will optimize this regarding availability, and of course, also cost competitiveness of 8-inch material. So since we rely on raw material supply, not mandatorily wafer supply, but bulk supply, as explained by Peter, for us, it's prerequisite that from a cost perspective, we come from a business case to a cost crossover, right? Not to forget that today, on an area comparison, still the 6-inch wafer is more costly than the 4-inch wafer. But of course, to the other productivity advantages, 6-inch makes sense, right? For us, now it still takes some time until we have enough supply secured on 8-inch, and then go for the transition. So technically, we have first wafers, and we are starting the activity. But the decision for going for high-volume manufacturing is still to be made depending on the cost and availability. One important point to mention, the tools we now invest and order will have the ability to switch to 8-inch. So the reusability of the equipment in Kulim will guarantee also a smooth transition to 8-inch. Then we go to the web.

Okay. Maybe a question from the web. This one is from Sandeep. What is your percentage of your [indiscernible] approximately trench? I think most of it's 100%, but diodes?

Diodes. The trench technology is, of course, something specific for the transistor for the MOSFET there. The diode technologies are all based on our proprietary NPS system, which is not a trench. So it's an -- but also simple Schottky barrier diode, but in the case of trench, it's 100% trench.

Okay. And yes, is it Infineon's intention to lead the SiC market? And what will enable that? Does it need IFX to get the big U.S. auto customers? Or yes, what would be the way to. Yes, first question, is it our intention to become #1 in SiC, as in power in general, and how can we do it?

So definitely, our self-understanding was and is and will be that we are a leader in silicon carbide. And now we can debate on #1, #2, #3, I don't care to be honest. The way to get there is definitely to provide the best products in a cost competitive manner. And I think we have proven that we are capable of doing so in the industrial space. I think the colleagues on the automotive side making great progress regarding design wins. I think I would leave it up to them to publish and announce the recent successes there. But I think portfolio matters besides manufacturing capability and, of course, then understanding the customer needs. So I'm very confident that having those growth rates shown on the industrial side, we are definitely capable to make this happen in the automotive space as well. So I'm super confident that we will reach ourselves the targets.

Okay. Fine. Another one from Rob Santos from Deutsche Bank. What is your latest view about Soltec Smart-Tech technology? Soltec is sounding very confident about gaining 30% share by 2028, with a value proposition mainly to driving significantly yield improvements.

Do you want to say something?

Yes, of course, since we are open to any new development in the segment, we have a very detailed look into this opportunity as well. For us, it is important to understand what kind of potential value this technology from Soltec can bring. We are on the way to do that. It's not yet finished. So we received material. We look at it and we make our conclusions out of that, whether the value promised by this new technology. And it is really about value. So I think this is important to understand the Soltec strategy is to sell value and not to offer something less expensive. So therefore, we need to see what kind of value this material can bring in our technology. And based on the results, we are agnostic there. So if this is a better approach, and why not switching to it? But first of all, we need to prove it. That's ongoing.

Maybe additional comment on that one. Just in principle, right? The point is about the material cost, which today is of rather higher impact for silicon carbide compared to silicon. And then, of course, the yield. And the yield is related to the material quality. And the material quality is determined by the wafer itself. And then, of course, the epitaxy, we do on top of it. Epitaxy, we do in-house, we can also purchase of comparison purposes. We, of course, do both. But it's the complete set, the combination, cost, and then performance, i.e., defect density, which translates into yield. And so the question is, in the end, what is the best value proposition? What is the best package? If I go the standard way and provide standard bulk silicon carbide at efficient costs, and if I do a very sophisticated epitaxy regarding lowering the defect density, right, then I might achieve be able to achieve very high yields. And we should not underestimate also the following. The die size impacts the yield. Because the larger the die, the more negative impact defects have, because the probability is high that this kind of defect sits into your device. And therefore, also the trench technology, being able to shrink the devices systematically generates higher yields. So being able to have a tranche means smaller device at the same performance, meaning higher yields. So that is exactly the thing we look at and try to optimize in all dimensions, right? And if there's a value proposition, we go for it. And if not, we don't.

So anybody from the room? Okay. Then I have another one from David O'Connor. Can Peter talk about margins across SiC discrete modules? And do you need 8-inch to get to group margins for SiC? And SiC margin profile industrial versus automotive.

Okay. I comment on Industrial. So Industrial is margin accretive, gross margin-wise, on IPC level. Therefore, we love those kind of products. And margins are -- we are able, of course, to allocate the volume, optimizing and increasing the margin to the extent possible, meaning modules and discretes must be on the same level. If not, we will allocate more volume to those who generate more margin. Because as we speak, we are sold out for the foreseeable future in the industrial space. So it's up to us together with the customer, of course, to optimize also the margin.

Okay. I can only add on the ATV margin. In the last earnings call, the Board mentioned that it's accretive, also in automotive. So we've seen a great improvement. And yes, that's all I can repeat from the latest earnings call. Okay. I do not have any more questions from the web. Once again, anybody here in the room? Any follow-ups? Okay.

Yes. With that, thanks a lot. Thanks to all of you in the room for the interesting discussion. And of course, thanks a lot for all of you on the web interface. Okay. Bye-bye. Thanks. Bye-bye.

Thank you. Bye.