STMicroelectronics N.V. (STMPA) Earnings Call Transcript & Summary

Good afternoon or good morning, everyone. Thank you for joining the second session of our Capital Market Day 2020, the presentation of our Automotive and Discrete Group, ADG. Today, Jean-Marc Chery, the Chief President and Chief Executive Officer, will start with a brief introduction of the session. Then Marco Monti, our President in charge of ADG, will go through the presentation of his group. It will be followed by a Q&A session where Jean-Marc and Marco will be joined by Lorenzo Grandi, President of Finance, Infrastructure and Services and Chief Financial Officer; and Marco Cassis, President of Sales, Marketing, Communication and Strategy Development. I would like to mention here that, due to the current sanitary situation, we have decided to adjust our setup to accommodate appropriate social distancing. As a consequence, we will host the Q&A using only audio. As usual, a few housekeeping items before we start: These live webcast and presentation materials can be accessed on ST's investor relation website. A replay will be available shortly after the conclusion of the event. This presentation will include forward-looking statements that involve risk factors that could cause ST's result to differ materially from management expectation and plans. We encourage you to review the safe harbor statements in ST's most regulatory -- recent regulatory filings for a full description of these risk factors. [Operator Instructions] Now before I turn the floor over to Jean-Marc, ST's President and CEO, let's go with a video as a reminder of our key strategic objectives. [Presentation]

Good afternoon or good morning to everyone. It's a pleasure for me to be here to introduce the second step of our Capital Markets Day, focused on the Automotive and Discrete Group. This comes after an extensive round of the communication for our third quarter earnings, where we have discussed already in detail what we are seeing in terms of current market dynamics, especially for automotive. During today's presentation, we will touch again on the short-term dynamics, but we will also focus on the mid- and long-term perspective and the related actions. So let me briefly recap on the short term. In automotive, as you know, global demand pick up faster than what we're expecting in July. This acceleration was driven by car production volumes, which continued to increase in China and South Korea and are restarting faster than expected in Europe and in the U.S. The legacy part of our automotive business is more closely tied to car volumes. This part was heavily impacted in Q2 by the lockdown. Then in Q3, we saw better-than-anticipated recovery in car productions. Today, we see a run rate of car production close to 2019 levels. You will see in Marco's presentation that also car sales, which are regularly tracked by industry analysts, are following the same trend. Clearly, we cannot say if this is a sustainable trend. Our current view for car production in 2020 is to be in the range of 75 million cars; and in 2021, in the range of 80 million to 85 million. So looking at the mid- and long term, we see clearly and still an acceleration of the transformation of the car industry towards safer, greener and more connected solutions. As a result, the mega trend of electrification and digitalization are also accelerated. They are the driver of increased semiconductor content in the car and remain long-term growth drivers for us. You will see soon from Marco's presentation that key asset ST has and the action we have taken to leverage these accelerated trends. In electrification, and this is valid both for the automotive and the industrial end markets, we now have the full set of power technologies to address the needs of our customers: wide bandgap materials with silicon carbide and gallium nitride and silicon with IGBT and high- and low-voltage MOSFETs, which we offer in a wide range of packaging options, including modules, both standard and custom solutions. In digitalization, for ADAS, we rely on our strong partnership, like the one with Intel Mobileye for vision-based processors and Autotalks for V2X connectivity, while leveraging our other product families, radar, in-cabin monitoring. We also address the needs of new vehicles' architectures with our Stellar microcontroller family in legacy automotive. We have taken action to adapt our business model, aiming at increasing our market share and distribution and moving the business model with some OEMs from COT ASIC to more ASIC ASSP. In Marco Monti's presentation, you will see how ADG contributes to ST's strategic objectives. You will also see, first, that over the last 3 years, ST has managed to grow faster than the market in automotive; second, how ADG has anticipated the market requirements, focusing on smart mobility applications driven by electrification and digitalization and on power technologies, supporting also our strategy in industrial; third, the actions that ADG has taken to reshape R&D investments in these strategic areas and to capture all benefits from the acceleration of the electrification and digitalization trends. And now let me hand over to Marco.

Good afternoon and good morning to everyone. As Jean-Marc just mentioned, in my presentation you will see how ADG contributes to the ST strategic goals to be leader in automotive and industrial market. These, thanks to our specific focus on electrification and digitalization and integrated solution to serve power and energy conversion. Let me start with a quick overview about ADG and how we have performed versus the market. ADG represents more than 1/3 of the ST business. And in particular, 68% of ADG revenues are in automotive, where we cover substantially all application with a wide technology portfolio. Industrial is the other major focus area, and here we serve the market with our power and discrete products. In the first 9 months of 2020, ADG had $2.33 billion in revenues. After a difficult market condition in Q2, ADG came back to 17% sequential growth in Q3 but still down 4.9% year-over-year. This is very much in line with the automotive market dynamics. If you consider car sales: Global demand picked up faster than expected, starting from July, and electrified vehicles has been a strong driver for this growth. In September, car sales grew to 7 million units, coming back to the level of 2019. As the result, the revenue of ST [ Automotive and Discrete Group ] increased 30% sequentially, with microcontroller and smart power growing sequentially about 35% and 40% and ADAS maintaining a solid sequential growth of about 15%. I would like to describe now our performance versus our competitors. ST is the only company among the top automotive players that grew in average in the last 3 years by more than 10% per year. The industry in the same period grew by 5%, and the top 3 players by 2%. Also, our analysis suggests that ST gained again market share in automotive in the first half of 2020, driven by our effort on electrification and ADAS where we clearly outperformed the market. So now I would like to take you back to what we presented at our 2019 Capital Markets Day. A few things in our market have changed since then and other remain unchanged, but the dynamics of the market confirm the fundamental of our strategy. The choices we have made have been confirmed to be the right ones. And these put us in the position to plan for incremental steps to better adapt to the changing environment. In more details. What remain the same in the market? Clearly, the overall trends toward smart mobility driven by electrification and digitalization of vehicles has remained strong. This requires, on one hand, power product, a trend we see not only in automotive but also in the industrial market. On the other hand, it requires digital product adapted for car which are more connected, safer and with an architecture more suitable to support these features. These trend are pushing the car semiconductor contents, partially compensating the reduced car volumes, which were as strongly impacted by the macroeconomic dynamics, but at the same time, the market had some changes. The pandemic-related lockdown reduced the ability of the carmaker to produce in 2020, impacting their sales and their capability to invest. The reduced budgets of the automotive industry had a visible effect in delaying many programs for full autonomous driving, but at the same time, this pushed the pervasion of mid-level L2++ ADAS system to entry-level car models. We also noticed that the trend toward electrification has continued at full speed. The dynamics of the market substantially confirm our strategy but require some additional action that I will describe in more detail shortly. So what remain unchanged in our strategy? First, our strong commitment to automotive and industrial markets. Second, our investment in technology and innovation, with a special focus on power technologies. And third, our ambition to lead in smart mobility; and in contributing to make industries smarter, greener and more efficient. Importantly, the pandemic do not compromise our customer engagement, innovation speed and R&D execution. But what we decide to change. The reduction of car sales generated pressure on traditional automotive products that are still an important part of our portfolio. To anticipate this, we decided to strategically accelerate the investments to support the automotive macro trends with action that I will soon describe. In particular, we strongly accelerate our R&D programs in car electrification, extending our product portfolio, including silicon-based and new material solutions. As an example, we added in our product portfolio power GaN products with a new specific focus on this disruptive technology. In thermal market, Asia became even more important on our strategy, and we established partnerships with Chinese leaders in electrified mobility. Let me now move to our market expectations in the near and midterm. The market recovery after the lockdown is visible now, and we are back to the pre-pandemic rate in term of car sales. Based on what we see in term of customer dynamics, we are slightly more optimistic than the current analyst views. We believe that, in 2021, car sales will be closer to the 2018 level. One important factor substantially confirming our strategy is that, despite the overall negative trend for car sales in 2020, electrification and ADAS are consistently growing. In fact, while internal combustion engine car volumes declined by 20%, mild hybrid grew by 85% and full electrical by 35%. At the same time, while [ low ended ] has declined by 20%, in line with car sales, Level 2++ (sic) [ Level 2+ ] grew by 12%, confirming the pervasion of this application in a larger range of car models. As a consequence, the good news is that the low level of cars sold did not impact the importance of the automotive trends, confirming them as the main driver for semiconductor market growth. To confirm this, the electrification-related semiconductor TAM will grow to more than $5 billion and for ADAS to about $9 billion by 2023. So what about the impact of these market dynamics in our strategy? During our last Capital Market Day, I mentioned our strategy to concentrate our efforts on the new automotive trends. We made this choice in order to reach within 5 years at least 50-50 share on our revenues, with 70-30 being our starting point. Today, thanks to our action, we are ahead of our plan. At the end of 2019, we were already at 65-35, and our goal is now to be 50-50 in less than 3 years. This will boost our revenues, improve our profitability; and it will make our sales less dependent on car volumes while benefit more from silicon pervasion. To achieve this goal, we took 3 major action that I will now describe. First, we increased our innovation effort in traditional automotive technologies. Here we are constantly moving the focus from traditional application to innovative system, like for example, solution for EV battery management, drivers for our power components or power management for ADAS products. Second, we have expanded our product offer to better serve the new automotive trends, leveraging a disruptive technology portfolio. A few example of our diversified silicon carbide portfolio: the new product offer on power GaN, the extension of the microcontroller Stellar family, addressing now multiple applications. Third, we redeploy resources from automotive legacy product to expand our product coverage on new technology like GaN, SiC or low-voltage power MOS. As an examples, thanks to the experience gained on planar technology, like our BCD, we gave a strong impetus to our GaN product road map. Also, we extended our market coverage in Asia with a new dedicated local R&D and application team to leverage customer intimacy. These 3 action are progressively moving our R&D investments to support the new growing trends. R&D to serve radical innovation now represent about 60% of [ ADG ] global spending, while just few quarters ago, it was about 30% of the total. Let me better detail the action I just described, giving you some color on 3 technology cluster: smart power, digitalization technologies and power. Starting with smart power. Here, as I mentioned, we are accelerating our focus on new and fast-growing applications. A first example is intelligent battery management products ranging from 48-volt hybrid vehicles to full electrical ones. A second example is our smart power driver for IGBT and silicon carbide MOSFET specifically designed for functional, safe automotive application. The third one is power management products for ADAS processor, like the companion chip for our EyeQ family. Moving now to car digitalization. So what the market is asking for in this field, I can simplify in 4 cluster. First, Level 2++ ADAS system [ to equip ] from high-end to entry-level cars. We estimate that, by 2025, more than 40% of vehicles will be equipped with these systems. Second, ADAS supporting full autonomous car, for example, for robo-taxi, where the semiconductor content could exceed $5,000. Third, microcontroller to service software-friendly vehicles. This is an important breakthrough in the automotive industry because it will drastically reduce recall by enabling software upgrade over the air, but as a consequence, it will require a drastic architectural change. And finally, car connectivity to use the huge amount of data available to improve safety and driving experience, also enabling mobility as a service. Our strategy in digital product is perfectly matching these market requirements. In Level 2++ ADAS system, [ we are a 30-year partner ] with Intel Mobileye for EyeQ products. We already sold over 54 million units, and in this area we are the market leader. Now we complement our offer with radar solution including FD-SOI RF receiver and transmitter and with dedicated microcontroller from the Stellar family. For full autonomous driving, we are developing the next-generation EyeQ family based on 5-nanometer FinFET technology. Also, our Teseo solution for precise satellite positioning is instrumental in supporting full autonomous driving. In microcontroller, I'll give you more detail later, but here I just want to mention that our ARM-based Stellar family is now adopted by multiple partners like, for example, BMW. In infotainment, I want to mention our Telemaco family extremely successful in supporting multi-standard connectivity in the car. As I described before, the change of the car architecture is a key breakthrough for the automotive industry. It brings several advantages like the simplification of the software, the over-the-air reconfiguration. It will contribute to increased safety and the optimization of fuel consumption, reducing the car weight. The change of the architecture had 2 major impacts in the semiconductor market, one related to microcontroller, and the second to smart power products. In fact, the new architecture needs much more complex microcontroller with very high computational power and with an embedded memory capable to implement the software [ refresh ] without interrupting the car functionality. This high-performance microcontroller will generate an higher TAM versus today, with an estimated 30% increase in microcontroller value. Second, the smart power portion will significantly increase in value. In fact, the new architecture requires dedicated power management lines for each domain. And these lines must be protected and robust against possible malfunction to assure the proper safety of the system. This totally new product will significantly increase the smart power TAM by more than $1 billion, in our estimation. Also in this case our strategy is perfectly matching and anticipating the need of the new car architecture. For microcontroller, our Stellar family is built to satisfy the new system needs. 28-nanometer FD-SOI permits high-frequency operation. This grants the computational power need for concurrent software execution in the [ multiple-core structure ]. The PCM embedded memory enable fast access time and robustness; at the same time, permits over-the-air software update without interrupting software execution. As you may have seen in a recent joint press release, our Stellar family has been endorsed by Bosch, an automotive market leader and a solid ST partner. The same in smart power. Our new [ M-11 ] VIPower family perfectly fits the in-vehicle power grid requirement. It is capable to protect the different domain of the car, controlling the current and at the same time minimizing the cable size. This will significantly contribute in reducing car weight and then fuel consumption. We are already engaged in several customer programs in this area. Before opening the chapter of our offer in power technologies, let me just share a high-level view of the wide product offer to serve electrified mobility beyond silicon carbide. I'll just give you a few example, for the benefit of time. Our IGBT family perfectly complements our silicon carbide products, serving application like cost-effective inverters. In this area, our strategy on power modules is very well defined with an offer covering both silicon carbide and silicon-based solution. To properly serve the market, we have standard automotive modules like SPAC and SPAC driver made in our own plans. This strategy is complemented by partnerships with leading module maker for high-specialized power solution. In this area, I am pleased to disclose our partnership with a power module market leader, SEMIKRON, to both IGBT and silicon carbide solution. I also want to mention our specific offer in microcontrollers tailored to support electrified car. Let me now open the chapter of power technologies that are strategic for the market we want to serve. In fact, in automotive, power technology will represent a TAM of $8 billion by 2024, with silicon carbide and IGBT growing more than the average. Also, in industrial, power products are key. In fact, there are several application where they represent more than 1/3 of the total semiconductor BOM. So I'm sure it's clear that power products are strategic for a market leader in automotive and industrial. Thanks to the investment we made in the past, complemented by our most recent actions, our offer in power product is very wide, covering silicon-based and new materials to match multiple application requirements. We have also a well-defined and very competitive manufacturing strategy. Silicon MOSFET are made in multiple 200-millimeter fabs. IGBT is in 200 millimeter today, and it will move to 300 millimeter in Agrate. Silicon carbide is today in 2 fabs at 150 millimeter that are already compatible with the 200-millimeter process flow. About GaN, I will describe our manufacturing strategy in a while. Let me start with silicon carbide, giving you a quick update on our activity. Today, we are the market leader in automotive and in industrial applications. We are accelerating our award rate with more than 68 active programs equally split between automotive and industrial. For this aspect, we are today partner of many market leaders in automotive and industrial with high volumes production programs. In this chart, you have just some of those that are supporting our ambition to reach $1 billion by 2025. Our market success is driven by more than 20 years of innovation with the best technology of the market today in term of performance and cost. Because we already reach manufacturing yield in line with standard power technologies, we are very confident in our ability to keep our competitiveness also in the future. I want also to update you on our activity on silicon carbide supporting our ambition to be vertically integrated, including substrates and MOSFETs. Our plan is to have at least 40% of our production vertically integrated by 2024, with a program in advanced phase of execution that I will disclose in detail in the next few months. Let's now move to gallium nitride, another disruptive power technology, where we recently increased our focus. Thanks to our most recent action, our GaN product coverage is now very wide. We already disclosed our partnership with TSMC on 650- and 100-volt applications that complements our internal development, ensuring the right time to market. More recently, to increase our product portfolio, ensure the right IP setup and shortening the pathway volumes, we acquired a GaN leader, Exagan. This acquisition boosts our application coverage and provide access to already engaged customers. I want to assure you that our ambition on GaN-based product is very strong; and we plan to cover many applications from automotive to industrial, to personal electronics. To fulfill our leadership ambitions, we have set up a manufacturing strategy based on 2 different path. A 200 millimeter with our proprietary technology in the manufacturing line we are bringing in our existing plant in Tours, France, leveraging already available competencies. This is complemented by a 150 millimeter manufacturing activity in TSMC using the technology where we are cooperating. I cannot leave the subject of high-power technologies without briefly mentioning our activity on IGBT that is complementing our silicon carbide strategy. I need to stress that, despite the strong success of silicon carbide, IGBT is and will remain an important technology, first, because it perfectly fits some cost-effective consumer and industrial application like, for example, induction heating or home appliance; second, because also in automotive there is still room for entry-level solution based on traditional silicon. For this reason, we push innovation on trench IGBT technology in order to have multiple competitive process options to fit different application needs. Without going too much in detail: We specialize our IGBT technology to cover application from the cost-effective induction heating to home appliance, to automotive. In our IGBT portfolio, we have now more than 600 products. And we are constantly enlarging our offer to support our goal to grow 3x faster than the market in the next 3 years, keeping the pace we have had since 2019. Before concluding, I just wanted to give you one last example of our strategy in power technologies, low-voltage power MOS. This technology is strategic for the simple reason that in industrial market we see the same trend toward electrification that we have seen in automotive. Most professional industrial tools are now converted to be battery operated, where in the past they were mains operated. To support this conversion, low-voltage power MOS are key. In fact, the TAM of this product is forecast to represent about 60% of the total semiconductor BOM in professional tools. In this product category, thanks to the already secured business, we plan to keep growing at the same pace we had in the last few years. To conclude. Despite the pandemic that strongly penalized our automotive market in the first half 2020, ADG managed to grow faster than the market across all applications. We have anticipated the market requirements. We focus on smart mobility application, driven by electrification and digitalization; and on power technology, supporting our strategy in the industrial market. And we have -- take action to reshape our R&D investments in these strategic areas. Thank you.

Thank you, Marco. We will now start the Q&A session that, as a reminder, will be audio only.

The first question comes from the line of David Mulholland from UBS.

Just 2, if I can squeeze them in. Just coming back on the comment you made on the IGBT business in autos and potential to grow 3x faster than the market growth. Can you just help us understand, how big has that business grown for you today? And just in dollar terms just so I can put that in some context of the market. And where do you think you're making the biggest inroads in terms of gaining penetration in the market? And then secondly, on silicon carbide, obviously you've done very well getting established with Tesla. And there's a lot of opportunity for the market to grow going forward, but obviously as this market matures and we start seeing more customers coming along, traditionally there's a lot of dual sourcing in the automotive market, so how do you see the kind of risk? Obviously there's opportunity with all the other customers that you're working with, but is there any risk that, that customer may look to dual source in the future even on silicon carbide?

Yes. Thank you for your question. It's Marco here. On IGBT, today, we are starting from a base of, let's say, [ $150 million ], if you consider what we do in automotive and industrial. And let's say, as I mentioned in the presentation, our ambition is to grow, let's say, by 2025, exceeding, let's say, $300 million. So you see that the grow is particularly wide. And this is including pure silicon-based or module-based solution. I will say that the grow is equally split between automotive and industrial application. In automotive, clearly what we are serving with IGBT is entry-level, let's say, solution for a main inverter; or even 2-axis engine -- electrical engine car, for example, that need in the front axis the silicon carbide and in the back axis the IGBT solution. In industrial, there are plenty of solutions for which we are covering with our IGBT technology, moving from 650 to 1,200 to 1,700 volt. In term of silicon carbide, yes. Let's say, normally, silicon carbide solution are highly specialized because, as you know, to obtain the maximum from a, say, silicon carbide solution, you need to work very tightly with the system maker or, even better, with the carmaker in order to have an optimized solution versus the system of the carmaker. So this is an area where, honestly, a double source is very much questionable because, again, the solution is very much customized. And we have a very well visibility. Of course, we are playing in automotive, so there is a very well visibility on what will happen in the next few years. So let's say the path that we described by the $1 billion by 2025 is very well sustained by solid business awards that we have today across our customer portfolio.

Can I just follow up on that then? Just in terms of the engagements that you have elsewhere beyond your largest customer. How do those break down between who you're working with? How much connection you have all the way through to the OEM. And in terms of the design of those solutions, are they following a similarly customized approach? Or are we starting to see some more standardization and, I guess, more -- is it still discrete products? Or is it more module solutions that are starting to be looked at in obviously production 2 or 3 years time from now?

No. To be honest, with see the opposite. We see more and more fully customized solution. We have, of course, the programs that are running with a usual Tier 1 that we have in our customer base, but more and more, we see carmaker directly involved. And we have even cases for which we are supplying just, let's say, silicon or dies to our customer. And they will use their internal capability to build fully customized modules. So as you see, the offer is very well diversified. Carmaker, for sure, are very much involved in this approach even if sometime, again as I mentioned, there are Tier 1 involved. And again there is really a big variety of solution that, again, are more and more pushed on the full customization.

The next question comes from the line of Aleksander Peterc from Societe Generale.

I just have a few. First one, could you give us a kind of a broad outlook for automotive semis, where you see this market growing, say, in terms of CAGR over the 2020 to 2025 period or something similar; and by what percentage points you think you can outperform overall automotive semi market growth? And then the second question, maybe it's a bit early to see that, but do you see any increased competition from the Infineon-plus-Cypress combination? Or is that not a factor in the markets where -- [ that we're discussing here ]?

Let me say, in term of market dynamics, as you can clearly understand, it's quite difficult to comment in these days because, unfortunately, we suffered in H1 from a major reduction of cars production. So this is clearly, let's say, changing the -- completely the level of statistics. As mentioned in the presentation, we think we will go back, in 2021, more or less at the same level of 2019. On this perspective, we think that the market will evolve from, let's say, a compound average growth from 4% to 6% in the next few years. And what is clearly visible is that more and more the market -- and you clearly see it reflected in our strategy. The market will be driven by the pervasion of silicon more than car volumes. We substantially model our evolution in term of, let's say, revenues, more or less having almost a flat level of car sales for the next few years, again plus, minus. And then what will drive dramatically the market dynamics will be the silicon pervasion. Then clearly electrification and digitalization, as we tried to represent in the presentation, will be the major driver because this is the content that is boosting semiconductor pervasion in the car.

On competition of...

Yes, on competition. Clearly, we respect all our competitor, and particularly we respect a lot of the competitors that you mentioned. And I think we have the elements to keep growing at our pace and, hopefully, to keep overperforming the average of the market. So I think we have clearly the elements to keep growing, again, even if, of course, we respect the competitors. And I will say that a recent add-on in our strategy in power, including new materials like silicon carbide, clearly, but also again product, will clearly boost the sales in the next few years. Hopefully, we're performing -- keep overperforming, I guess, our competitors.

[indiscernible] -- yes. Jean-Marc speaking, okay. And for the other, let's say, product line and overall under the processing solution addressing mass market or other industrial and so on, we don't see yet, okay, any effect, okay? It's too early to measure.

The next question comes from the line of Achal Sultania from Crédit Suisse.

First one is on this, again, silicon carbide opportunity. I guess, when we think about some of these car companies that you have partnerships with, and there have been a few in public domain now like Hyundai and Renault, Nissan and BYD, can you just help us understand how should we think about the timing of the ramp of some of these projects? And what kind of solutions are you providing? Is it more around onboard charger? Is it more inverter? Any color around that will be helpful. And then secondly, on the margins. When we look at ADG, it's still the lowest-margin group within the whole company, so how should we think about the margin profile given that there are clearly structural growth drivers of -- what's the key driver for growth, yes, or improvement in margins going forward? And does -- some of these new products like GaN or silicon carbide, are they going to be margin dilutive over the next few years or similar to group average -- or not group average but auto average? Yes.

Well, let's say, about your first question on silicon carbide, clearly we show a portfolio of customers that we have. Some of those are, let's say, already in production. Some of those are ramping up in volume in Q4 this year, and some of those will ramp up production across 2021. I will say that in automotive, the major ramp-up of new customer, you will see in Q4 this year. Application are the ones that you mentioned substantially. So of course, the major one also in term of, let's say, quantity of silicon is the main inverter of the car for both, let's say, electrical car and for a few example of hybrid cars. The other big area where silicon carbide is clearly giving a strong benefit is in the area of the DC-DC converter; and then of course, onboard charger, again if we remain in the automotive application, if you want to expand a bit on the industrial application also the infrastructure. So the charging station in the infrastructure is also a potential good, let's say, example of application where silicon carbide is clearly boosting the solution. In term of profitability, before probably to leave to Lorenzo, just one reminder. ADG is a combination of multiple product segments. We have rich application like ADAS, again electrification, for example; and also we have a consistent business that is serving pure commodities like we have in our discrete family, okay? So of course, the result of the profitability of the group is also the combination of this products family. And at the end, across the market dynamics, this is the performance.

Yes. Maybe I can add something here, Marco, if you allow me. For sure, it's true that, in this moment, the ADG is the segment is the group in our portfolio that is suffering more. As it was, let's say, underlined and explained during other occasion that we had to discuss together, for sure, today, one of the impacts, say, that we are suffering in ADG is the impact that we have on our legacy [ fab or on our ] legacy, let's say, production manufacturing, in which what we have seen today is definitely a significantly decline in the legacy products, creating [ unloading ] but also creating [ inefficiency ]. The combination of growth, let's say, on new products driven by as what's explained by Marco and, on the other side, let's say, improvement in this manufacturing portion of our footprint, manufacturing footprint, that will -- a combination, of course, of increased demand Now we are really in a low-end part of the demand. And change of some mix inside this manufacturing will improve, let's say, the efficiency. This combination will definitely be the driver for our expected growth in term of profitability. The main portion that will drive this improvement in profitability, in a way, is expected to come in term of operating margin from the power, let's say, portion of our portfolio. This will definitely be the one that will increase more during the time frame of our plan.

The next question comes from the line of Sandeep Deshpande from JPMorgan.

I have 2 questions, if I may. A couple of questions, I mean, firstly, regarding the ASIC business in automotive that you do, for instance, with Mobileye and with others. I mean, how do you see development occurring in that business today? Clearly you have major wins there already, but are those customers continuing to design future generations of products with you such that this business remains continuing, that they don't move this business to somebody else in the future? And then secondly, regarding the silicon carbide business, my question will be that when do you see the next big customers starting to ramp up in silicon carbide. You've already signed up with a bunch of other customers beyond your lead customer in the automotive market, but when do you expect substantial revenue to start to come from those other customers?

Well, let's say, Sandeep, your first question, on ASIC. No. We will -- let's say there is no, of course, plan to reduce our exposure to in particular digital ASIC for automotive. We see more and more instead a full customized solution in the area of ADAS for automotive. And so again our relationship with Mobileye is extremely good. As I was trying to describe in my presentation, we are -- keep working on the next-generation L2++ ADAS, so the EyeQ6 [ and ] 2 different configuration. And we are already planning to move on the next generation that will be a 5-nanometer FinFET technology product to, let's say, drive the full autonomous mobility. So again, we will continue to work. Mobileye is -- as you know, is one example of the portfolio of customers we have in this area. We are diversifying also our offer. Of course, we count on -- specifically on Mobileye that is today market leaders. With them, we have more or less 70% in the vision-based solution for ADAS. We are also working on specific ASIC for radar application in our FD-SOI technology. So again we don't see reduction of the momentum in this specific area. About silicon carbide. Again as I mentioned, the momentum is extremely strong, so today, we do not see more the need to, if you want, promote the concept or the advantages of silicon carbide against the traditional technology. So we have already a large portfolio of customer on which we are working on something that will enter in production across 2021; and customers that are already in production that are ramping up volumes, as I mentioned, in Q4 this year and then consistently in 2021. So again I think we, you should expect to have, if you want, a kind of linear growth from now to what we disclosed, 2025, [ $300 million ]. So more and more, for example, European OEMs will enter in production between '21 and '22 with the silicon carbide solution for [ main inverters ].

The next question comes from the line of Andrew Gardiner from Barclays.