Infineon Technologies AG (IFX) Earnings Call Transcript & Summary

Hi, everyone. Thanks for joining us this afternoon. I'm David Mulholland from the technology hardware team here at UBS. It's my pleasure to have 2 gentlemen from Infineon with me today to talk through the power semiconductor side of Infineon's business. So we've got Dr. Peter Wawer, the Division President of the Industrial Power Control business. We've also got Andreas Urschitz, who's the Division President for Power & Sensor Systems business. In terms of the agenda for today, they will both go through a quick overview of the divisions and the strategy of Infineon, then we'll come to a Q&A session. [Operator Instructions] But with that, thanks, everyone, for joining us. And I'll hand over to Andreas to kick us off. Andreas.

Thanks a lot, Dave. Let's go to the next page. If I may bring your attention to the left side of this chart, that one gives you an overview on the overall Infineon Group revenue split by divisions. As you can see, total division revenue, PSS and IPC put together, amounts to 43 percentage points. So that's in line with the Automotive revenue share of Infineon, and that's what we're going to talk about today during this presentation. On the right side of this chart, you can see the revenue split by product category for both the segments IPC and PSS. And if I simplify in a nutshell, roundabout 80% of both the divisions' revenue happens in the area of power semiconductors. Next chart. That one is an overview on Infineon's overall portfolio in the section of discrete power devices. So on the left-hand side, you see the traditional silicon-based technologies and product families, within which everything that goes from 20 to roundabout 900 voltages in breakdown voltage falls more or less under the category of silicon-based MOSFETs. That's handled by PSS as far as non-automotive MOSFETs are concerned. The one above, so 1,200 volt up to 6.5 kV, that's the typical domain of the IPC division. Silicon-based, we are talking mostly about the portfolio of discrete IGBTs and IGBT modules, that's everything on the left-hand side of this chart. On the right-hand side, you can see novel stuff, if you will. When we talk about non-silicon-based materials, so namely wide-bandgap material or compound semiconductors. So in that area, Infineon, in the meantime, has a rich portfolio of silicon carbide MOSFETs, which are starting at the breakdown voltage of roundabout 1,000 voltages, going up. So silicon carbide is a domain of -- and under the leadership of IPC. My colleague Peter Wawer is going to talk about this. Everything which is in the category of 600-volt breakdown voltage or lesser, we do serve with gallium nitride devices. Those are under the regime and ownership of PSS, who is doing the gallium nitrite development on behalf of the entire company on all these areas we're going to talk in the course of the upcoming presentation. Next. That one, I think most of you are pretty well familiar with. So the chart gives an overview on Infineon's market shares by specific categories in the area of power semiconductors. Top left, well known, Infineon in the area of power discretes and modules, is holding 19 percentage point market share, factor 2.5x, roundabout, larger than next-largest competitors, if you will. In the course of the last 5 years, we managed to grow our market share from 18 to 19 percentage points. So that's a strong evidence that our growth strategy is bearing fruit and working out. But that also is a strong indication towards Infineon holding an amount of power semiconductor market share that justifies the building of our famous 300-millimeter wafer fab capacities because they require large chunks of market share in order to operate them at fully loaded volume and reach meaningful cost performance position or pricing points that we can then also do interesting or nice margins at the marketplace. Bottom left, you can see a different split of the entire market, and that picture goes a little bit already towards Infineon's, in principle, strategy in the area of power-power systems. So we are coming from power components, which are power discretes and modules moving towards power systems, where we are trying to amend power discretes with drivers and controllers. Or remember the picture from the last Capital Markets Day where we said that power transistor is a switch, is a muscle, and the muscle needs a brain to tell the muscle when to do and what to do at highest power efficiencies and best-in-class power densities. And that combination, putting both the product categories, discretes and ICs into one bucket, is a base market within which Infineon today is owning already 30 percentage points of market share. So also when you combine ICs and discretes, Infineon's leading ahead of Texas Instruments, as you can see. And also in the combined version, you can see that Infineon in the course of the last 5 years, as we also indicated and told you during Capital Markets Days, we increased our market share from 12 to 13 percentage points. So the entire idea of amending our portfolio in discretes through adjacent drivers and ICs, putting them in 1 single package or selling them discretes, providing power systems to the marketplace, is working out and helps us to: A, fuel our growth going forward; and b, further increase and grow our margins. Next page. And last, not least, a couple of elements that are known as success factors for Infineon's power success story, if you allow me to say. Definitely innovation, innovation in the area of power technologies always has been core and the backbone of the company. Now it is going from silicon into wide-bandgap materials, if you will. Due to size and size considerations and market share, also, economies of scope play an important role because we have over-proportional large amount of R&D force, if you will, or a number of people and experience that allows us to now expand from a discrete area into overall systems as a function of large R&D and customer embeddedness and P2S, building right-fit product for our customers that help us to extract -- a, create and then extract further value of our customers into our P&L. That definitely was important. Quality. Infineon is always known as a quality leader. Portfolio breadth. So in the area of power semiconductors, here to say we are coming close to somewhat like almost 10,000 products. And last, not least, differentiated manufacturing landscape that has been always the source of success. And at this very moment, I want to hand over to my colleague, Peter Wawer.

Thanks a lot, Andreas. Two general slides, where the first makes statement on our CO2 footprint. Of course, we all know that the CO2 emission topic is becoming ever and ever more relevant and more critical for mankind looking forward. So the good message here, thanks to the product we provide, which, on the other hand, preserve the consumption or the waste of energy, of electrical energy; they also contribute, of course, directly to CO2 avoidance due to the usage of renewables like PV and wind. So we -- our semiconductors empower these kind of power generation, renewable power generation activities. So the net footprint is very positive, reducing according to this elaborated calculation, to more than 54 million tons, it's not enough. So we have set ourselves, on Infineon side, the ambitious target to reduce our own CO2 footprint by minus 70% until 2025, to finally then become also totally CO2-neutral end of this decade. So what are the main levers? Basically, main lever is purchase of green energy, no gray energy. Energy, in our case, means also electricity. Then a variety of measures regarding the exhaust abatements of our own manufacturing facility. And last but not least, for the unavoidable certificates -- certification to be on the safe side. And on the next slide, I would like to share with you how we see, from a company perspective, how we address the business opportunities regarding the electrical energy chain. So we address all those areas, generation, transmission and consumption. While on the consumption side, PSS and IPC play -- IPC, for example, in the area of EV charging or major home appliances. The high-power side, generation and transmission, renewables, wind, PV, et cetera, these are the predominant markets with the focus of IPC. But clear company strategy is that we address all these areas from left to right and deliver and develop leading semiconductor solution for these applications. With that, I would like now to jump into the section of IPC. And starting with the financial overview on the next slide. Here we go. You see that we nicely expanded our segment result in the recent years, coming from 12 percentage regime in '15, '16, also due to the integration of IR, which was, to a certain extent, diluting the margins. We were able now to climb up by more than 5 percentage points, in line, of course, also with revenue increase. As you can see, for the now already ended fiscal year, we -- thanks to the virus, we were not able to expand revenue. Of course, we had an ambitious revenue target which we did not execute due to the overall situation we are all aware of. But I'm very proud that we managed to keep the segment result. And if you look after the comma, we were even able to slightly expand segment result margin. That was basically driven by portfolio management and quite ambitious productivity measures that we took internally despite the fact that we also had to digest idle and underutilization costs. So all in all, I think we managed fiscal '20 quite good, and we are very optimistic now looking forward. And as you can see also on the bottom, our key customer footprint is very wide-ranged through all the areas of applications and all around the globe. So having a very solid base here. And going to the next slide. Brief highlight how we see the development of the main market segments, so-called applications, you see on the left-hand side. The numbers on the right-hand of the symbols indicate the revenue split of IPC. And the comments are related and the errors to the market outlook we see for calendar '21. So our biggest market that we address, being of course, a very fragmented market, is the industrial drives space. It used to be, a couple of years ago, more than 50% of revenue of IPC, has come down significantly due to the fact that other markets, especially the renewables, have outgrown the industrial space. Nevertheless, industrial market continues to grow with a moderate single-digit CAGR, and drives was down already in '19. After a boom phase, '17, '18; '19, drives were going down, and it was predicted to come back and recover second half of '20. Due to obvious reasons, that did not materialize. So basically, we are now 2 years in a row down on the industrial drive side. But here, of course, we were somehow a little bit conservative regarding the forecast for '21. But from today's perspective, I would say, signs are going into the right direction. So if you ask me, I would already give this now, with respect to '21, a rather greenish arrow, being more optimistic than I was when this slide was created only recently. So that goes in the right direction. Renewables, PV and wind basically saved the show for 2020. While drives was down the second consecutive year, wind and renewables were running the show, so to say, in the now ending calendar '20. And also, the forecast for next year is quite positive. I will comment on that a little bit more in detail. Major home appliance was also quite weak in 2020. We also expect recovery here. The only thing which we see quite conservative being only "responsible" for 10% of IPC revenue is the area of traction also related to the virus and also related to quite boom years in '18 and even in the beginning of '19. I would also like to draw your attention to the Others segment because that is already now, so to say, the second biggest segment, the others, the lump sum of the other things. And inside this, there are very attractive growth applications. So I expect that the Other segment, accounting already for 21% of IPC's revenue, will see nice growth looking forward due to the fact that cool applications like CAV, so the industrial construction vehicles are in sight; and also high-power charging, or EV charging, which is being addressed by IPC. Let's jump to the next page. So get a little bit more in detail, the example which we updated also compared to 2 years ago. And here, you nicely see the revenues and the trend for the renewables in those -- both greenish colors, dark green, wind; and light green, PV. So you see that we had nice growth, not enjoyed a nice growth in '20. And that especially the outlook is very positive. And on the right-hand side, you see that we are engaged in all the big players in PV and wind. Of course, not exclusively, competition is out there, no doubt. But the good thing is that the renewables become lesser and lesser dependent on subsidy schemes. In areas with high irradiation, to give you an example. And high irradiation means already south of Europe, you do not have to go into the desert. PV-powered electricity, meanwhile, reaches something like EUR 0.02 per kilowatt hour. So it's very much -- the business case is very much self-sustained if we know. And we know that, for example, coal-fired power plant generates electricity in the range of EUR 0.06 to EUR 0.07 per kilowatt hour. It's totally clear that the growth will accelerate on PV and wind on a wide scale. So forecast is here rather going to the 10% CAGR looking forward. But it's not only for PV and with itself. At the next slides, which is a little more engineering slide, as I already was hinted to, but that's Infineon, right? We see on the next slide, please? [ A little ] goes hand in hand while the renewables, the amount of renewables being responsible for electricity generation increases. We have to handle the issue with fluctuations, right? So there is night. There are clouds. There are sometimes winds, sometimes no wind. Therefore, the topic of energy buffering, in this case, your battery storage is becoming more and more important. And what you see here is the share of renewable energy on the x-axis and the battery storage capacity, which is being forecasted according to The World Energy Outlook, where these numbers come from. And these dots indicate the years, small dot, 2020; big dot, 2030. And the 4 graphs are for the various regions: Red, European Union; the other one is India, China and United States. And you see here that the prediction is that, for the latter ones, the 3 ones, India, China and United States, the attach rate for battery storage rather increases rapidly with the increased share of usage of renewable energies. So why this difference between Europe? Because simply, Europe is the region being best connected via power grid. So you do not need so much storage capability if you have a very capable grid. And hence, in the area of India, China, United States, the expected growth for these kind of buffer battery storage capacity is significantly higher. That in turn adds another good margin for semiconductor manufacturer like ours because we estimate something like EUR 3,000 worth of content per megawatt for this kind of storage capacity because, of course, you have to have the semiconductor to get the energy in and out of this battery. So another very good market opportunity for us. And with that, I would like to draw your attention to the development on silicon carbide. We all know that this is a hot topic in semiconductor, especially in power. And while we have been a latecomer in the area of the silicon carbide transistor, I think we are running now in a very successful catch-up mode. As most of you might know, the silicon carbide-based semiconductor market from the analysts is expected to grow in the area of 25% to 30%, so in this -- roughly in this area. While we have achieved now a significant higher growth in the recent years. Of course, if you look on the left-hand side, that was the already then business based predominantly on silicon carbide diodes, while we then launched our first transistor-based product. And in the last 3 years, after we were not successful in closing the Wolfspeed acquisition successfully, we started internally with the silicon carbide accelerate program. And from fiscal year '18 onwards, we have brought more than 100 new products based on silicon carbide, diodes, and especially of course MOSFET, into the market. So that, according to our own research, we are, regarding the amount of qualified and available products, meanwhile, market leader. And as you can see also, the growth in revenue on silicon carbide-based products somehow reflects this kind of development. At the same time, we were able to triple the amount of customers being addressed via distribution, having now more, for example, more than 2,500 active customers in distribution area. And also worth to mention is that the revenue growth in the last 2 years mainly came from the industrial space, namely IPC and also a little bit PSS. But the year -- the anticipated growth now looking into '21, the growth will be dominated by ATV. Of course, not being part of this telephone conference today, it's important to note that, of course, growth in industrial space continues, but will be now massively accelerated, thanks to the kicking in of revenue from the automotive space, predominantly main inverter in xEV. And since we all know, cost plays a major role, another 2 important slides. While, as I mentioned, we were a latecomer, we initially bet on the so-called trench technology. And in MOSFETs, they exist also in silicon. Disregarding, if we talk about silicon or silicon carbide, they exist, the planar device and the trench devices. And the trench devices are exclusively used, for example, on our innovation for silicon and silicon carbide since they provide advantage in power per area. So you simply get more power out of a given footprint, might it be on the wafer or might it be in a device. And you can translate this, of course, also into cost advantage, right? If you need a required a certain amount of power, of current, then you'd need less silicon or silicon carbide area. And so here, with this independent research report which was provided by System Plus, it's given in the source, you see here that the Infineon Gen 1 is already the best-performing device regarding this figure of merit, which is the RDS(ON), resistance times gate charge, which is, in the expert area, very well known figure of merit. And not only that, we of course now develop our next generation, the second generation, which we intend to bring to the market next fiscal year where we again target for an improvement regarding 20% to 30% compared to the Gen 1. So we are well set also compared to competition. But as you all know, material cost plays a significant portion of the overall cost picture. We also have addressed this thing, which you're going to see on the next slide. Thank you. And that basically explains a little bit how we're going to use Siltectra. You may be aware that we acquired a start-up named Siltectra 2 years ago, which uses an advanced laser process for splitting silicon carbide material. And that is now being explained and indicated here. And this is also a little complicated picture, but worth to understand. On the other left-hand side, we receive, for example, by our supplier, GTAT a boule, a silicon carbide boule. The CrystX is a trademark of GTAT. And we now, based on our laser, Siltectra process, are able to split this boule into wafers. And of course, we can determine the thickness of the wafers, but the main message #1 is the laser splitting, you avoid a huge amount of waste. While with the traditional wire-source splitting, which is still used in the majority of manufacturing cases for silicon carbide, you remove something like 350 micron per wafer, and you turn it into dust, into waste. While with this laser-based process, we are able to lower this number well above 100 micron. Even, we see a road map down to something like 50 micron. That's the first savings. So savings in material translates, of course, in savings in costs. But we do not stop there, right? While the traditional wafer thickness is 350 micron, you know that we have elaborated competencies in handling thin wafers. In a second step, during device processing, and you see this here indicated with wafer splitting, we see the ability to take a typical 350-micron-thick wafer and then again cut it in half, generating a very thin and also another thin wafer, where we simply process the thin wafer to the very end. And the other half, reclaim it and then put it into the beginning in the starting process of the wafer manufacturing. So simplified, making 2 wafers out of 1. So here, we address cost down significantly of silicon carbide, independent of if we are able to purchase a boule, a crystal, or maybe simply taking a wafer that we purchase from a wafer supplier and then also reusing this, reducing, of course, the material cost significantly. So we're making nice progress, running also this activity in a pilot line mode and having an ambitious target to bring this into qualification and production, the sooner the better. And I think with that one, I'm going to hand over to Andreas.

Yes. Hello. Andreas speaking again, talking about Power & Sensor Systems, next page. Same picture, like shown by Peter before, in that case, on PSS. You can see PSS growth over the last couple of years, performing at a compound average growth rate of 8 percentage points and a little bit above, up to EUR 2,650 million the last fiscal year in revenue. As far as profits are concerned, we managed to grow the profitability towards 24 percentage points. So constant margin increase and margin growth in the course of the last couple of years. Based on what ingredients, if you will, I'm going to explain to you a little bit in the course of the rest of the presentation. Before to -- going to that, please have a look on the top right part of this chart. So this shows you the PSS revenue split by product segments. To roundabout 75 percentage points of the revenue are made in the area of power semiconductors. That even includes high-rail, which is also power semiconductor-based business. Some 25 percentage points, PSS is doing in radio frequency and sensor devices. These are not part of today's presentation. Today, we are focusing mostly on power. Bottom left, that's a reflection of PSS customer portfolio. So it's the who's who of the consumer industry, telecommunication industry, smartphone industry and also server and compute industry. And on the right-hand side, you see a picture or see a couple of names of our major distribution EMS partners. Let's go to the next page. That one, in line with IPC, gives you also an overview on the major application segments, along which PSS revenue is made up. The top left side, there is the computing and server segment. That accounts for around about 20 percentage points of PSS overall revenue. Strong acceleration in that business happened throughout 2019, '20, and we expect that to continue also into 2021 and following. Major driving force here is definitely the digitalization, work from home and other such aspects. Telecommunication accounts for almost 10 percentage points of PSS revenue. So that's the business that we do have in telecommunication base stations and related equipment. Also that one rather tends north than sidewards, like indicated in this arrow. So also here, over the last couple of weeks, the trend has definitely further accelerated, and there's improved market outlook on the further deployment, in particular, of 5G equipment throughout the world, also as a function of digitalization, work from home and other such aspects. Smartphone makes up 19 percentage points of our business. So they are mostly our radio frequency component business, but also sensor business account into. And also some portion of PSS USB-C business. I'm going to talk about this in the later part of the presentation. Consumer segment, roundabout 20 percentage points, also growing strongly. People working from home also tend to equip their home with new consumer equipment, talking about smart homes and entertainment systems, if you will. So this is a nice growth area. And last, not least, the others and industrial segment. So here, applications such as battery part application fall into. Also, that one shows nice growth momentum going towards 2021. Next chart. This is a kind of a deep dive into PSS server and data center segment. I've been talking about this a little bit already in the course of the Capital Markets Day, but let me just reiterate and build on what has been said in those days. So server and data center is a segment which intrinsically is growing. So according to the analysts that we are using as a source here, the number of servers and server growth is forecasted to be around 9 percentage points CAGR. I tend to say that this is a figure that has probably upwards potential given the latest developments that we do see. But on top of that one, we see that the bill of materials for power semiconductors, in order to manage the power flow within given data centers, is continuously increasing by factor of 1.3 to 1.4 relative to the CAGR in number of units of server systems. Why that is so? In a nutshell, the power hungriness of latest-generation CPUs, but also the power-hungriness of latest-generation machine learning and GPU architecture chips that are determining the calculation power of server systems, the power-hungriness is continuously increasing. This function of an increase of power-hungriness, the number of semiconductors and power semiconductors that are needed to serve the power flow and given efficiency requirements and given power density requirements is increasing quite substantially. That's a benefit, but that's a market and market trends where PSS is taking a lot of benefit from with regards to future growth predictions and perspective on both revenue growth and margin growth. Next one. This one is another chart on the server and data center segment, and it gives a kind of, in principle, idea of PSS' and I believe overall major parts of Infineon's product groups and business lines strategy, namely. We take our world-class power discrete devices, so mostly MOSFETs, around which we continue building adjacent drivers and controller circuitry. So if you will, we grow from components that we observe, in that example, into a single server, and amend these components by drivers and controllers, thus always working towards a getting the energy efficiency everlasting improved because that directly translates into lower operating expenditures for a data center operator. And b, we try to shrink the system size of the entire power supply system, be it on the server on board next to the GPU -- CPU, or be it in the power supply that provides the power from the wall, from the plug, towards the server rack. So we try to make the power system smaller and smaller and smaller, thus increase the power density per given system and thus reduce the size needed for a power supply or a server rack to be put in the given space. [ Save for a power boost ], that leads to the effect that more and more servers are fitting into one of the same space on a given circumstances. And that leads to effect that a data center operator comes to the enjoyment of improvement of capacity expenditures. So we work towards both. OpEx reduction and CapEx improvement, and this is pretty much the story behind the Infineon's success in the area of data centers, telecommunication and other such systems that brought us via efficiency and monetization. Next page. A lot said in the server section. So that picture gives an overview on modern 5G, so to say, telecommunication networks. I keep it short. So 5G is growing. The interesting thing for us is that the average silicon content in power semiconductors that goes into a 5G radio is factor 4 higher as compared to 4G radios. So intrinsically is a growth driver that comes along with the deployment of further 5G equipment. And the other topic is, which is here named under the driver #4. That also fog computing equipment is being used along the 5G signal chain. It's just the preprocessing that happens before data collected through 5G connectivity is loaded into cloud compute centers for computers are used. We have the world's best semiconductor power flow solution in order to make that power flow as energy efficient as possible and make the systems as small as possible. Why size matters in 5G, is a very simple example. Most of the operators deploying 5G are using legacy 4G or 3G base stations and upgrade those towards 5G. There's only 1 problem, the cabinet size is limited. And to get enough power into powering the 5G equipment, the limitation of cabinet size leads to everlasting improvement of the highest energy-efficient power semiconductors that are needed in order to get 5G equipment into 4G cabinets, and therefore save, again, CapEx and OpEx as a telecommunication operator. That's where we grow. Next chart. I told you before that the USB comes into the equation at PSS. So USB is new in PSS portfolio. It came through the acquisition and integration of Cypress. So USB is, if you will, a communication protocol, where, for instance, a charger is being connected through USB or USB-C cable to a mobile device like a smartphone or a tablet. You need to have a USB chip in the adapter, and you have to have another one in the mobile device. These are chips and product families that were amended towards PSS portfolio. Why does it make sense in combination with power semiconductors? The equation is pretty easy. Please go to the next page. And this one is an exemplary page that illustrates PSS position going forward, including and applying USB capabilities in the context of 65-watt high-power density chargers for mobile equipment. So if you start on the top left side, what PSS has in its offering today, amongst others, is also compound semiconductors. So gallium nitride 600 volts switches, also 100 volts, which is in our portfolio. Either discrete devices or integrated power stages. So these are gallium nitride switches together with the driver. Those, we combine with PSS and Infineon's unique digital power technology, which we got through the acquisition of International Rectifier, and prior to that, Primarion. So those 2 technology elements, we put together, digital power, gallium nitride and amend it by the USB functionality that we got through Cypress. And all of a sudden, we put ourselves in a position that we can further shrink the entire semiconductors that are used in a 65-watt kind of high-power density charger to a level, that we can manage to increase the power density to 32 watt per cubic inch, which is another 25 percentage point size reduction relative to the next-best alternative, talking about existing 65-watt chargers that are today available in market. The entire bond for gallium nitride, digital power and the USB portion in a charger is around about $4, a very interesting opportunity for us, amongst others, to grow in particular into that segment, in that particular context, leveraging both the possibilities of gallium nitride. And there we've making power systems much, much smaller, and on the other hand, then also leveraging the USB functionality that we got through Cypress, integrating the whole thing into systems, in a package, making the entire solutions on the customer interface much smaller. And customers are willing to kind of pay for this also a little bit of extra premium that then also translates into the bottom line of PSS and Infineon. Having said that, I'm pretty much through, and I want to hand it back to Dave.

Thanks, Andreas. And thanks, both Peter and Andreas for their presentations. Very interesting content. As we move to the Q&A, I'll put a couple of questions to both of you myself. And then as mentioned at the start, we'll move over to the phone line. [Operator Instructions]

First question for me, listening to some of the comments that you both made around kind of current progress in some of the key end markets. I wonder if you could both just address whether -- how you feel about the outlook for your division compared to how you messaged with the full year results? I guess it was only 3 or 4 weeks ago, but it sounds like things seem to be a bit more positive in some areas. And I guess, despite everything that we've seen in terms of lockdowns. So I appreciate if you could both comment on that in your respective divisions.

Yes.

Andreas, go ahead.

Peter, maybe you start and I continue, please?

Yes. So I think while we were a little bit lucky, so to say, but of course, luck strikes those who deserve it, right? We were also, to a certain extent, prepared to manage the downturn which we now had to face in 2020. And the lucky part was basically also on our side, accompanied by the flexibility to shift the product portfolio in that direction where the demand was still healthy, like I said, in the area of renewables. What I foresee -- being not now overly optimistic, I foresee now for '21, that the business area of our business, which still was weak, namely the area of industrial drives, starts to come back, plus the continuous demand and growth also from the renewables. If we add this altogether with continuous high demand on Andreas' side regarding power semiconductors for data and server-based applications; plus enormous increase in demand from the automotive side thanks to recovery of the automotive market; plus, of course, also now increase of xEV in general, right, on the car side, but also on the infrastructure side, I expect quite a positive market where I also could imagine allocation situation here and there rather coming soon in '21. Andreas?

Two trends are our friend, Dave. Trend #1 is the Green Deal in a wider sense, so the need for green energy technologies, be it on the energy generation side, or be it also on the energy conversion side. So talking about power management systems and power flows. That plays into our camp, need for more green technology, translating the need for more power semiconductor systems. And the other topic for sure is digitalization, and digitalization now being accelerated big time through current circumstances, which we do simply see continue and translate into more and growing demand for telecommunication infrastructure, computing equipment and everything that goes into consumer equipment and the so-called smart home. Having said that, your question was pretty much, how do you foresee 2021 and how is the growth momentum? I would say the demand continues to accelerate. We see distribution inventories trending down towards, I wouldn't say, historically low levels, but very low levels. That's a strong indicator for our market demand, in particular in South Asia being super strong at the moment. We see that capacities globally, so to say, in combination with lead times increasing, are getting tight and tighter. I think that all of us here in this call heard and read a lot about foundries being allocated already while we speak. And a lot of us have been reading about material foundry prices and other stuff already going north. Putting all that together, I would say, given then also experience that I do have in the power semiconductor segment, it's fair to say that 2021 will be a strong year, and 2021 very likely is going to be a seller's market.

Thank you both. Just moving on, and I'll try to avoid being too short term-ish from now. In terms of the silicon carbide comments you made, Peter, obviously, a very interesting announcement that you made, again, a few weeks ago with GT Advanced Technologies. One comment that I've always had from the supply chain is how difficult it is to produce the material and making sure that the supply you get is of good quality. And obviously, that's something that Cree historically has had a reputation for being differentiated on. How confident are you that GT Advanced can meet the quality requirements, particularly for end markets like automotive? I assume you've gone through good due diligence. So just give us a bit of color on what work you've gone through there to commit to the supply agreement that you did.

Yes, absolutely. So of course, we were in discussions with GTAT for quite some while. And as you correctly pointed out, the topic of quality of material and reproducibility was a major topic. So it took quite a while. But the GTAT folks are, I would say, very skilled, very engaged. They are very knowledgeable. You might be aware that they have a legacy in the area of crystal growth coming from silicon historically, PV silicon. And they turned this knowledge, basically, I would say, to a competitive advantage now also on the silicon carbide side. They follow a different technological approach compared to the approach of Cree regarding the crystal growth itself. So there's -- for the connoisseur and the detailed, knowledgeable colleagues, there's the so-called resistive heating and the inductive heating method. While Cree is using the first one, GTAT is using the latter one. There are, of course, always pros and cons, but we only were able to sign the deal and also make this public announcement because we are now -- I mean, we were very convinced that GTAT will hold up to the quality. And of course, we already have already tested the material, and we're quite convinced that they are capable of providing consistently high-quality material.

That's great. And coming back to you, Andreas, in terms of the commentary around the server market. Obviously, we've seen quite a big acceleration there. And I think you've been mentioning at various occasions we've met you about the content opportunity in servers. Is there -- the figures you gave are pretty helpful. But just in terms of the return on investment that an operator might be able to get from putting a more efficient solution in, is there -- I think in the past, you've given some figures to help us understand why this does make sense to spend more on the content from the semi side for the power supply. So I just wonder if you could run through that for us.

Yes. So it's still in line. So what I'm going to tell you now is still in line with what I showed you at the Capital Markets Day. So at the very end, if you look into next-best alternative power flow systems that are in the marketplace and take that as a comparison base, by the use of Infineon's latest and greatest power flow solution, which comprises both the dimensions, so AC-to-DC power supply solutions, also incorporating gallium nitride in those for efficiency and density reasons, plus incorporating the latest and greatest, DC-to-DC converting solutions, including then also point of load DC-DC conversion systems. With those save-plus-value, If you apply a complete Infineon power flow, in a novel data center architecture, roundabout the operational expenditure saving amounts to 10 to 15 percentage points. And that directly translates into lesser cost per mouse click, if you will. So somebody who offers a search engine and has therefore to bear and face certain costs and energy bill per search request, that's, of course, a big business case for such operators. And the other topic is simply your save-plus-value will also get the CapEx improved by somewhat like 10 percentage points, plus/minus. So this depends a little bit on the building construction, the cooling system that's in the data center and so on and so forth. But also CapEx can be further optimized because you get more servers into the same given space. Thus within a given space, you increase the calculation and storage capacity of a given data center farm or server farm, and that's the deal that we are providing our customers, too. And that's why customers are buying to us, also allowing us to extract the fair share in form of margins that we can extract over the buying or sales prices.

That's great. So at this point, we'll go to the phone line for any questions, for anyone that's in the queue. [Operator Instructions]. But with that, I'll hand over to Rita to conduct the Q&A. Thanks, Rita.

The first question is coming from Adithya Metuku from Bank of America.

Yes. So firstly, I just wondered if you could give us a little bit of commentary around how you view your boule-cutting technology compared to Soitec's Smart Cut technology. What are the key differences? Do you have any advantages that we need to be aware of? And secondly, a question. So the first one is for Peter and the second one for Andreas. Just on the time-of-flight sensors, I just wondered if you could give us some color on any demand trends you're seeing there. Are you -- and if you could comment a bit on whether this demand is coming from the front side of the plane, rear side of the plane, automotive end market? Any color you can give on the time-of-flight demand for 3D sensing, that would be much appreciated.

Yes. Thanks for the question. So I'm going to start with the question regarding Soitec. So we are continuously looking into, of course, also competitive approaches regarding material splitting. So far, the process I'm aware of regarding Soitec is used for the split technology and wafer bonding technology used for silicon. So far, we see also a complete different approach regarding the technology because the Soitec process is not laser-based. Ours is laser-based and with a certain cold-split specifics, which are technological-wise very, very different. As said, we are always looking left and right to be sure that we are betting on the right horse. We are not yet in production. We are preparing ourselves, still in R&D mode, to become productive. And of course, also to qualify the first wafers and hence of course devices. But as I speak now today, we do not see that the Soitec process and the Soitec approach is being used or usable or competitive in the areas of silicon carbide. Question answered, first part?

Yes, thank you.

And hand over to Andreas.

Talking about time of flight, couple of things. So point number one, time of flight used to get popular in some [ airplanes ], in particular on the user-facing site, mostly for authentication purposes. So there, we do see time-of-flight cameras being in use to a certain degree for 2 or 3 years in the meantime, point one. Point number two, in addition to this one, we have observed the market and smartphone makers also trending towards using 3D time-of-flight cameras, also on the world-facing side. So mostly, those were facing cameras having 3D functionality use those for improved photography features. So if for instance, you look into a Huawei P40 device, so the photography and the photography effect that you can get with the help of the depth sensor, which is the 3D time-of-flight camera that has been improving photographic quality massively. Beyond this one, we see further proliferation in the marketplace since use cases around augmented reality are more or less finally coming. That was always the prediction, that augmented reality may drive the proliferation of time-of-flight sensors. That's now coming, and we see that mostly to be populated through the iPhone 12 in particular, where amongst, so to say, 3D camera-ing the one or the other model, on the world-facing side has also launched the latest system or ecosystem around augmented reality development landscape for ecosystem that they do offer to AR developers who are doing games and other stuff on this one. So all these factors together, I think, bring the market and form the market towards a direction that 3D time-of-flight cameras are used on both sides: On the user facing for authentication mostly, but then also on the world-facing side for AR and photography reasons, mostly. As far as Infineon is concerned, I think that's not a secret. So Infineon has a bunch of design wins which we also officially reported on. So one of the bigger ones was reported last year throughout the CES in the context of then also LG as a smartphone maker. And then more recently, you can find Infineon's time-of-flight camera also within certain models of Chinese smartphone makers, one of the bigger ones and the bigger designs that we have is then also the Huawei P40. So if you open that up, there is Infineon inside. Further growth in that segment expected, but not yet being part of the growth predictions that we are communicating on the capital market, such as we did already 2 years ago. So it's a kind of an ongoing wildcard that is, in the meantime, starting to fill up with life and revenue.

The next question is coming from Janardan Menon from Liberum.

My first question is actually on profitability, especially from a point of view of the automotive division's strong demand for power semiconductors, especially from EVs. What impact does that have on the profitability of your division because of capacity utilization levels rising or that we need to add additional capacity? Is there any indirect impact that you have? And specifically, on IPC, you said that quite a large part of your revenue is from modules. And the Automotive division has always said that modules is a low-margin product for them. And they have got a JV in China. Is there anything that you could do on the module side to improve overall profitability? Or are you satisfied with your current levels of profitability? And then a quick follow-up on the PSS side. You've said 19% of revenue from smartphones. And I'm sort of assuming that quite a big chunk, close to EUR 300 million, would be from the MEMS microphone side, which has been a very successful product for you. But we sort of lost a little bit of track of what is happening to Infineon on the RF side, where you lost a socket at a major smartphone manufacturer some years ago, et cetera. Can you give us just an update on what is happening there? How you see that part of your business progressing into 2021 and beyond?

Okay. So let me start. First 2 questions, I would like to take. So margin spillover effect from Automotive business, yes and no. So basically, the underloading cost. Usually, if you do not load, manufacturing capacity is allocated to the division not loading it. And that means, of course, that the major chunk of the low demand that ATV was specifically facing in last fiscal year was on the P&L and went into the bottom -- went into the bottom line of ATV. But there, due to the accounting system, of course, kind of spillover effect, definitely. So while now in comeback of revenue and demand is primarily beneficial for ATV itself, it also has a spillover -- positive spillover effect because certain corporates or central costs also on the operations side are then diluted. So also here, we will benefit disregarding how our own business will develop. And I think, as we said, the outlook is also positive on PSS and IPC side, we will benefit from it. So all in all, whole Infineon will see a positive effect. Will be not so pronounced, as I tried to explain, but will be a positive benefit. So to your second question that's a very interesting one. There's a fundamental difference in module business between ATV and IPC. While ATV is focusing on a very few models of modules which have to be run in very high volume at a very competitive cost and hence price, on the IPC side, it's very different because we have a huge variety of products. Each and every of the very distributed, fragmented of the industrial applications of IPC requires specific modules. Quite some of them are really custom-made for the application. Or what we in recent years enforced is our capability to even customize modules on specific customer requests. That's a topic that we do not highlight too much to the outside world because we don't want to make our competitors aware of it. But the reason for that very fragmented and customized approach by application, or even in certain cases, if the business case allows by customer, leads to the result that the module business at IPC is margin-accretive. So it's above the average of IPC's margin. So it's a totally different picture. And here, I hand over to Andreas for the second.

Yes. Coming to the question related to the sensor and radio frequency segment within PSS and the correlation towards smartphone-s, smartphone markets and so on and so forth. So point number one. Within our entire sensor business, you're correct. 2/3 of the revenue is made up a very successful silicon microphone business. But already 1/3 of the sensor revenue is made up by radar, 3D sensors, pressure sensors and also our gas sensors. So these are all accelerating segments, if you will, to some degree, wildcard, as I've mentioned before, where we put investment into with regards to fueling a growth story to come and to be reported when timing is mature, if you will. Coming to the other part of the question that related towards radio frequency business. And it has become a bit silent, I think I heard you say before. So the point related to radio frequency is as follows. In recent months, the RF business sees a very nice renaissance, if I may say so, driven out of demand coming from China. And what happens here is pretty much that Chinese smartphone makers are trying to put themselves into, so to say, sources and suppliers where they have, in their perception, more supply security, if you will. And Infineon is a European-based radio frequency experts, so mostly on radiofrequency component is seen a very valuable supplier to those. So long story short, we benefit from this trend, so to say. And America first, making Chinese companies now also in the RF section, go away from the one or the other module supplier towards RF discrete suppliers such as us. And we are going again and entertaining more long-term partnerships in the section of radio frequency components for smartphone makers, in particular in China and Asia-Pac.

The next question is coming from Pierre Ferragu from New Street Research.

I wanted to come back on a couple of slides you presented on your position in silicon carbide. And I was wondering if you could give us a perspective on how you see your competitive landscape today. So you have about EUR 150 million of revenues in this year. I think maybe your largest competitor might be about twice that size. And so I'd like to hear you talk about how you see your competitive landscape. And then of course, as we move toward trench architecture, how you see market share moving around that, how do you anticipate that.

Yes. I think the leader, according also to the independent research, the leading competitor regarding silicon carbide-based revenue is especially leading with 1 customer in the area of automotive. And we all know that this is an xEV customer which manufactures -- or which is headquartered in Silicon Valley and being the leader in xEV as we speak. That is very remarkable success. As said, on the automotive side, the big design wins and the revenue increase, still to come. So the increase of revenue now predicted in '21 will come from our first automotive customers ramping. On the other side, as I mentioned, the major chunk of the revenue we already have shown or I have shown for the ending fiscal year is from the industrial space. And I think that's a good message because it's a very diverse customer base, so via distribution, but also via direct customers, I think on the slide was depicted more than 60 direct customers with revenue of more than EUR 10,000. That is also translated into that these customers are, to largest extent, in industrial customers. And if I do the math, with the recent numbers from the market researchers, I come to the conclusion that, in the industry space, we are already leading the silicon carbide camp. Our own calculation lead me to a market share of around 20%. There is, of course, more room to grow, but there will play simply the game of portfolio, right? Because nobody like us has the ability to bring product variants in the area of discretes and the area of modules. And then, of course, in the area of modules, we have the ability to combine silicon and silicon carbide diodes. Remember, we have products, IGBT and silicon carbide switch, meaning MOSFET inside one module. And that's simply the portfolio game that we play which make us very resilient and gives us also a competitive edge because there's not another player that I can see who has the ability of combining and generating this amount of products. But of course, looking forward, the automotive growth, that will be key also for us to execute according to this growth. And related to your second question with regard to the trench concept, I think that is one of the reasons why we are now quite successful being the latecomer in the areas of the MOSFET. Because our customers realize that the best-performing MOSFET is the Infineon device, and that simply provides additional value to them in the system because they're able to integrate more power, or at the same power range, being in the situation to dissipate less energy which comes from losses because our device is more efficient. And therefore, I'm somehow also convinced, since we see on the competitors' road maps these trench devices appearing, right? Everybody, every competitor has the device on their road map. And today, only Rome is the competitor who's already manufacturing in the successful manufacturing process for trench devices. That would be my answers.

The next question is coming from Aleksander Peterc, Societe Generale.

I'd like to stick to the silicon carbide area for now. Can you give us an outlook for 2024, '25 kind of time frame, where you see the overall silicon carbide device market to be? In a couple of billion dollars, I suppose. And what proportion of this market, in your view, will be automotive and what will be in nonautomotive parts? And then secondly, on Siltectra. Can you update us on the time line of when you see the qualification of a solution and when you intend to put in practice in high-volume manufacturing?

Yes. So I start with the second question. So when we acquired Siltectra 2 years ago, we gave ourselves a 4 -- 5 years' time frame, where I said that is a start-up. And we knew what we acquired, not at all manufacturing-ready, but very valuable IP. So I think we made great progress regarding professionalizing the equipment environment, having meanwhile a the pilot line setup, which is manufacturing -- is able to manufacture in a pilot line setup, which is also able to fulfill standards being required for qualification. These tools are available, that process flow is already available. And we are already in an R&D mode regarding regular manufacturing of wafers, providing of course then this feedback loop during front-end manufacturing process. So I would say we definitely will keep this 5-year time frame that we communicated. And as we speak, we are on a faster track. And I now don't want to commit any time lines, but I promise I will let you know in 1 year from now because, of course, we have an accelerated track where we would be even faster than those 5 years as mentioned. So regarding the market growth, I think the recent numbers, which were published by Yole are quite close to what we and how we see the market. I do not have it right away all in my mind, but I will look it up and maybe provide an answer in a couple of minutes. So the market growth from today's perspective is in the -- anticipated to be in the range of around 30%. I think it's rather 25% to 30%. And you can debate forever. It's not an exact math. And in that range, the market is by far outgrown by the xEV, by the automotive application. And if I remember correctly, the automotive CAGR in this market is well above 30%. And I think it's in the high 30s. So a little bit above -- below 40%, but I will look it up in a minute. And the rest is then distributed through the whole variety of applications. But even down the road, we see also other areas of traditional silicon applications being penetrated by silicon carbide. There's a big but, the overall market still being today small, right? And talking about a multibillion power market, talking today about a market which is in the range of still below EUR 1 billion market overall, significantly below. Even with the high CAGR, silicon carbide will be only a certain portion of the overall power market. And from our understanding today, there are of course a variety of applications which are very keen on using the technological advantages of wide-bandgap silicon carbide, but maybe also gallium nitride, but there's an area of very, very cost-driven, medium, and I don't want to say, low performance, right? But very, very cost-sensitive applications. For example, general-purpose drives, that is in the industrial space, the biggest single market still and will remain the biggest single market. And from today's perspective, there is no way that from the cost -- from the pure cost position and also from a value proposition, that silicon carbide or another wide-bandgap material will be able to enter this cost-driven market. So that would be my answer.

The next question is coming from Johannes Schaller from Deutsche Bank.

I think, Andreas, you alluded to the outlook into next year being relatively positive, and you also called it more of a seller's market maybe than a buyer's market, mentioning the lead times are going up. I mean, we're coming out of a year with relatively high idle costs, but I guess it's not out of question that, in 2021, we could even run into an allocation situation if things progress reasonably well. So one question, maybe more on the near-term capacity. Can you help us understand a little bit, with the additional ramp of Dresden and what is going on as well? Both for PSS and for IPC, kind of where we're actually heading in terms of capacity for these 2 divisions over the course of 2021, and how much demand do you think you can actually serve? And then on capacity, more a kind of longer-term question. When we look at the ramp of Dresden, that was more driven by IPC and then, back then, PMM. I guess, for filler, it's going to be more automotive initially. How much of the capacity for fillers, do you actually expect to get for your 2 business lines more in the medium term? Is that going to be significant? Or do you currently expect that to be really mostly on automotive fab?

Yes. If I may answer part one of the question. So how much capacity is being prepared for the growth? So in a nutshell, depending a bit on the mix, PSS is prepared to grow 20 percentage points, give or take, throughout the next or the running fiscal year relative to the last fiscal year. That's what we are prepared for, and that's why I also feel very much comfortable with to give the winners with whom we intend to grow, also the required capacity flexibility to allow also over-proportional growth rates amongst those partners. And the other question, I give to Peter.

Yes. Capacity-wise, it's important to understand how we handle the increase in capacity. So while the company is continuously investing to fulfill our growth objectives. We have the following principle internally. Whenever capacity is being added, especially on the silicon side, 300-millimeter, then basically, we have the fair share principle, which means, according to the share of revenue we today run, and of course, ATV being the biggest part of it, followed by PSS and then IPC, we are asked, do you want this capacity? So everyone has the right to get his, so to say, fair share. And if somebody refuses, we either do not invest or the other party has the ability to get more. So meaning, if the demand stays healthy on PSS and also on IPC side, we will also participate in the increase in capacity. And that is a good news. And of course, if the demand is not as high, then we adapt accordingly. Maybe -- question answered?

Yes. I think that makes it much clearer, yes.

Okay. Maybe you give me briefly the opportunity because I looked up the numbers, right? And the recent update from the Yole report exactly states for the growth rate for silicon carbide, total market, 30% in average CAGR from 2019 to 2025. That would mean EUR 2.5 billion market in 2025, whereas xEV is EUR 1.5 million (sic) [ EUR 1.5 billion ] out of these EUR 2.5 billion in 2025. And the CAGR, I memorized correctly. So xEV growth itself would be 38% CAGR then. But interestingly, highest growth is then -- or even higher growth rate is predicted for the charging infrastructure, which is expected to grow with a 90% CAGR. So very big growth numbers, all attractive for us.

Maybe a quick follow-up just on your last comment on the capacity situation. So that would mean -- what you said would mean that, for filler, you could very well end up in a situation where you have IPC and PSS combined have a similar share of the filler capacity as automotive has, if I understood your comments correctly, if you find the right applications and the right demand for that.

Correct. Yes. Exactly.

Last question is coming from Sandeep Deshpande from JPMorgan.

My first question is in IPC on the appliances business. This conversion to IGBTs or MOSFETs in the appliances market has been ongoing for a long time. So does this market in the near future transition to being a less-growth market and grows with the growth in appliances, which is typically GDP-like as such? Or is there still a lot more penetration towards using these power-saving devices in the appliance market? And my second question is, I think there was an earlier question on this, but a follow-up on that. On the RF business. Clearly, you have some RF devices in base stations, et cetera. What are you supplying today in the smartphone market on RF? Because you did sell your power amplifier business, so I would guess that these are LNAs, et cetera. So maybe what you sell today and what the intentions of this business are going forward.

Yes. Thanks for the question. And thanks for bringing up the relevant slide. And you're totally correct, right? So the appliance market is, growth-wise, not super attractive. But thanks to the fact that, due to the energy regulations, CO2 savings, et cetera, as mentioned, we see the need for this inverterization in appliances. And that is already ongoing, and that still will continue for the years to come until we come on a global level basically on kind of saturated, to a saturated level regarding the inverterization. But as the graph nicely shows, this drives now for the foreseeable future, for the next couple of years, an overproportional growth of semi content in appliances. And that is exactly why we think, why I think that's an attractive market to be in, and where we're simply being not a very strong player in the product of IPMs so far. But nevertheless, we grew our market share and being now holding now #3 position behind Mitsubishi and behind ON. Why we think that there is very nice growth potential looking forward and where we can make reuse of the existing technology-based on IGBTs, but also on silicon carbide. And I would like to highlight also the second aspect which is interesting, that is now the appliance market. And there, we sell typically the IPMs, the integrated power modules. But also for low-power drives, we see a request for this kind of molded module products, where we now also recently qualified our first silicon carbide-based IPM. That is, from a cost perspective, not a sweet spot for home appliance because it's a very high-performing product, but that can be also used for the industrial space. So our strategy is here also to reuse this kind of product type, which we have then for other applications. And with that, Andreas, you take the second part.

So talking about radio frequency and smartphones, in smartphones, we are mostly present with LNAs, RF switches and antenna tuners. As far as our business in telecom base stations is concerned, so RF business in base stations. So here. We are more talking about antenna tuners. We are selling there millimeter wave chips that go there for backhaul communication. And last, not least, that's not yet in big volume, but under our R&D, gallium nitride-based power amplifiers. You know that Infineon is holding a lot of IP and know-how in the area of gallium nitride, and we're also about to proliferate that one into the area on gallium nitride on silicon, in 8-inch, leverage for radio frequency applications and here in power amplifiers.

Okay. Thanks, everyone. I think that's pretty much everyone done in terms of questions from the line. Just a couple of questions just to finish up for me. One thing we haven't touched on, on the call so far is around Cypress. We've touched on a few of the product areas, but I just wanted to ask more generically. And I guess, Andreas, given you've consumed some of the division or some of the business, how do you feel the integration has gone? And particularly in the context of trying to integrate a business during disrupted travel, unable to send your own management teams across, how have the teams been working together? And how do you feel team's progressing partly on a synergy side, but also in terms of being able to drive some of the cost synergies out of the business as well?

Yes. Cypress integration is, at the moment, that's a little bit surprising, you may say, running according to time plans as they were constructed pre-COVID, if you will. So how comes? And I want to mention 2 elements here. Point number one, Cypress culture and identity, if you will, is pretty close to the Infineon identity, we started realizing. And that makes it then easy to walk into integration discussions, if you will. And here, we were very positively surprised. So that was, even though no travel is permitted, accelerating integration on WebEx and Zoom meetings, if you will. And the other topic, for sure, goes in direction of harvesting synergies. So mostly cost synergies, while we speak, but also the first revenue synergies are here. So these 2 areas are pretty much within what has been planned before. So we are executing, and I would consider slightly positive news relative to what somebody may expect.

And one final just for you, Peter, in terms of the -- coming back on some of the broader silicon carbide discussion. You've talked at length about the differentiation you have on trench and you showed the chart of RDS(ON). One thing you haven't mentioned I just wanted to understand a little bit about is, is there any intrinsic difference between planar and trench just when it comes to the end cost of the device? As in, how expensive is it to make trench versus planar compared with potential cost savings you get from a slightly smaller die?

I think that is very much similar to silicon. From the electrical properties, there's certain applications where still our customers like and who want to have planar MOSFETs, still. That is legacy business which Andreas entertains on the silicon side, which we inherited from International Rectifier, that is okay. But not growing, nobody invests into it. In the end, always cost matters. And for the trench technology, you pay a little bit higher effort on the device integration and device manufacturing. But that is by far justified and outperformed due to the performance, plus cost advantage. So overall, regarding amps per square millimeter of silicon carbide, the business case is totally clear, and that is also the reason why everybody in the industry is heading towards the trench-based device. And also, if you -- if we look into our road maps and into the ideas, we have ideas for the next 3 to 5 generations for trench technology, which, again, is very similar to what we learned and what we executed on the silicon side, right? Andreas, I think -- I don't know if you're already manufacturing SFET 8, but I think so, right?

Yes.

So many, many generations to come also for the trench-based MOSFETs on silicon carbide side.

.That's great. Thank you very much to everyone for joining us. It's been great to hear so much from both Andreas and Peter. I don't know if either of you had any closing remarks you wanted to give before we end the call, but thank you, everyone, from my side for joining. And if any one of you would like to close, please feel free.

Yes. Thanks a lot. It was a lot of fun to talking to you. And a special thanks to your moderation, David. And of course, to the very interesting Q&A. Looking forward to reconnect...

Yes. Likewise also from my side. So I hope to see you soon, and I hope that we will be coming together in a context of looking back to what been having evidence on hand that it turned out to be a seller's market, and that Infineon also had the ability and did translate that into the bottom line growth. Thank you.

Thanks, everyone, for joining us.

Looking forward to the next face-to-face, right? So stay healthy.