Thank you for joining us at our annual Capital Market Day here in London, or via our live webcast, by the way. We do appreciate your interest in STMicroelectronics. Hosting our event today is Jean-Marc Chéry, ST's President and Chief Executive Officer. We have as well, several members of the ST Executive Committee joining in. Can we get the agenda of the day, please? Here is the agenda. You'll see that first, Marco Cassis, in charge of sales and marketing, communications, and strategy, will give a view of our end market and market strategy. You will hear as well our three product group heads, Marco Monti, in charge of Automotive and Discrete Group. We will have Benedetto Vigna, in charge of Analog, MEMS and Sensors Group, and Claude Dardanne, in charge of Microcontrollers and Digital ICs Group.
They will discuss their innovative product portfolios, their key IPs, their technologies, and by the way, some of the products that will help drive our growth are located in demo areas. Throughout the day, we encourage you to go and visit and discuss with our technology and management, and I'm sure you will get a better understanding on how the value they will bring to our customers. After, we'll get Orio Bellezza, in charge of technology, manufacturing, and quality. He will cover our manufacturing strategy. Last but not least, we will have Lorenzo Grandi, our Chief Financial Officer, then we go through the financial results and priorities. Then we will have a general Q&A session, and this will conclude our plenary session and the webcast. The event, we will have a lunch served immediately after the Q&A session.
Today, the plenary presentation are being broadcast live and over the web and can be accessed through the ST website. A replay will be available shortly after the conclusion of this event. As usual, our presentation include forward-looking information involving risk factors that may cause ST results to differ materially from management expectation and plans. We encourage you to review the safe harbor information and statement presented here, and also ST most recent regulatory filings for a full description of those risk factors. I'd like now to invite President and CEO, Jean-Marc Chéry, to take the stage and kick off our Capital Market Day.
Thank you, Céline. Ladies and gentlemen, good morning, and thank you everyone for being here or on the webcast to attend our 2019 Capital Market Day. I would also like to thank our chairman of ST Supervisory Board, Nicolas Dufourcq to be here with us today at this very important event for the management team. My presentation will be about ST value proposition, our ambition, the strategy, and the associated strategic objective I am convinced we will achieve to sustain this ambition. Of course, I will come back on the 2019 visibility and the financial target we have, and I will speak about our midterm financial target. Then I will let the stage on all of my colleagues of the Executive Committee. Let's start my speech sharing with you ST value proposition. Our value proposition.
About shareholder first, we are committed to deliver in line with our sustainable, profitable growth objective, the value to the shareholders. For our customer, our value proposition is to provide them with a key differentiating enabler they need to succeed in their own market. This enabler are technology, IP, product, application now and associated ecosystem. Our value proposition as well, is to provide an independent, reliable, and secure supply chain to support their growth, to support their strategic growth. For all stakeholders, people, communities, and security at large, our value proposition is a strong commitment to sustainability. Addressing challenges presented by a complex and a strong organization, very complex mobility system, continuous demand in energy, safety, security in data, smart manufacturing and infrastructure, and more. Our sustainability is encompassed in our shared value, integrity in all our business activities.
People, human to be always at the center of what we do, and the excellence in execution of whatever we are doing. I would like to say that ST fundamental are really solid. We serve a secular growing market, supporting really critical and key societal needs, and we have the strong determination to make ST a more solid, stronger company. We have simply the ambition to outperform the market we serve in order to build a sustainable, profitable, $12 billion company in the midterm. Our strategy focus on three high growing application. Smart mobility, power and energy management, Internet of Things. All of them are strategic electronic demand enablers. Our strategy aims to make stronger and accelerate our leadership in four end market we serve. Broadly, as a broad range of supplier for automotive and industrial, and selectively for personal electronics and communication infrastructure.
Our strategy leverage both proprietary technologies, but selectively external one, enabling our product portfolio and roadmap. Automotive specific ICs, both digital, advanced digital, and analog. Discrete products and power devices, either on silicon or wide bandgap materials. Analog, industrial, power conversion ICs. General purpose and secure MCUs, and industrial application MPUs. EEPROM, MEMS specialized imaging sensors, and ASICs develop a proprietary technology of our company. Our strategy takes advantage of our internal manufacturing, both our internal manufacturing with our 11 scale site around the world, and selected foundry and OSAT partners. To build, again, I would like to insist with an independent, a secure, and a reliable supply chain to support our customers.
I confirm our strategy is fully consistent with our business and operating model based on, first, organic growth with small strategic targeted acquisition, as we have done recently with Norstel or two years ago with some IPs of IMS. Internal design and manufacturing at the core of our operational activities, but complemented with, as a target, 30% of external activities with few selected partners. Definitively to keep a very solid capital structure. This is the business and operating model confirmed, our strategy is totally consistent with this model. Of course, my colleagues will shortly go into more detail on all this, but I would like to share with you in few minutes my conviction. My conviction about ST, my conviction about ST on making stronger our leadership in the electrification and digitalization of the car.
Clearly, we are seeing a game changer in the electrification of the car with silicon carbide power devices addressing this market, and mainly the electrical powertrain market. ST will make its leadership on this market stronger. I am totally convinced about that. Digitalization. Definitely, Marco will share with you what we are doing with Mobileye or Panasonic on advanced digital powerful processors. I would like to speak and to share my conviction about a major disruption in the architecture of the car, moving to domain where our MCUs, enabled by our 28 FD-SOI embedded PCM technology, will make the difference. I am convinced that we will become stronger and stronger in embedded processing solution. On top of the MCU portfolio we have developed with more connectivity and security embedded, either in system-in-package and system-on-chips.
We have developed, introduced recently, supported by our strong ecosystem, an industrial application microprocessor in order to increase our leadership in this field of product and activities. We will accelerate our growth. I am convinced we will accelerate our growth in analog and sensor for industrial application as well, thanks to the strong product portfolio we have. We will expand definitely our power and energy solutions to address the industrial market, because silicon carbide and soon gallium nitride will be a key differentiating factor to address this market. We will develop, we are developing innovation in IGBT as well, so we are a strong competitor in IGBT, and Marco will go more in detail about that. We are expanding our offer on power modules. All this leadership stronger. This acceleration to address industrial market, for sure, we accelerate our position with industrial OEM.
Definitely, we will continue to support the mass market through a very close and strong partnership with the distribution channel. I am also convinced about ST on winning high volume socket in smartphone, wearables, and accessories. Thanks to our focus, offering high-performance sensor, so both motion MEMS and environmental sensor, and specialized imaging sensor, and you will see the technology enabling our device. Secure solution, clearly, embedded SIM, near-field communication, and embedded secure element. Power management and analog. Now, in addition of these areas, we are also addressing important opportunities for now and the near future in 5G, RF, mixed-signal technology for front-end modules. Last but not least, in communication infrastructure as well, we will take advantage of cellular 5G and satellite communication. Before Lorenzo Grandi will go into detail into the financial, I would like to touch upon 2019 and our midterm targets.
I confirm that we expect 2019 SAM to be between -3% decline and +1.3% growth. I qualify it as a soft market conditions like in 2015, with backloaded, with H2 market condition stronger than H1. I confirm we plan a revenue, looking at risk and opportunities, a revenue range between $9.45 billion-$9.85 billion, with an operating margin in the first half of the year about 10%, and we expect to move slightly above 14% at the midpoint in the second half of the year. For the midterm, we believe our share of market will grow around 4%-5% average in the coming year.
Clearly, with all the enablers, the programs, the product technology differentiation, IPs, partnerships, this will set condition for ST in order to reach mid-term a $12 billion company revenues with an operating margin in the high teens. Definitively, I am pleased now to leave the stage to some of my colleagues of the ST Executive Committee to share with you our views, our plans in order to make our company stronger. As I told you a few minutes ago, with the ambition to outperform the market we serve, becoming a $12 billion company delivering high teens operating margin, while demonstrating in the first half of the year, resilience in current market, which I classify soft market conditions. I would like to thank you for your attention. I am now inviting Marco Cassis to come on the stage.
Waiting for Marco coming, I just would like to inform you that Marco will not stay with us at the Q&A session because very unfortunately, Marco faced a very sad personal event this Sunday, and I really want to thank my colleague to have the courage to make this presentation today. Thank you very much.
Thank you, Jean-Marc. Thank you for the nice wording. By the way, it's really a pleasure to be here this morning. Good morning, everybody. Jean-Marc just outlined our strategic objectives. My presentation is designed to give you more details on our markets and on our strategy, and also on the approach by end-market. Let me start by saying upfront that we are convinced that ST is well-positioned to grow faster than our served market. We have a very clear strategy and plans by end-market, and we are convinced that we can leverage through our technologies and product portfolio, the disruptions taking place in these markets. These disruptions are driven by, as Jean-Marc said before, by car electrification and digitalization, greater need for power and energy management solutions, and the IoT.
On top of this, 5G is a particular game changer here and a great opportunity for ST across all the markets we target. Let's begin. First, the market. We know very well that the semiconductor market is cyclical, and we are now in the soft part of the semiconductor market cycle. Our assumption for this year is the one of flattish or slightly decreasing SAM, between -3% to +1.3%. After two years of growth, if we compare ourselves with 2015, ST is in a much better position than it was in 2015. Lorenzo Grandi will elaborate more on that later. Looking beyond 2019 in our served markets, we see that automotive and industrial are forecasted to grow between 6% and 8%.
These are the markets where we are broad and that represent already today 60% of our revenues, more or less in a balanced way, 50/50. Personal electronics and communication equipment, computer peripheral are flat. Our selective approach is allowing us to target a faster than the market growth by focusing on specific opportunities. Overall, by the way, we are starting from good foundations in terms of positioning. You can see here that in product areas where we target to be broad, we have leading positions across our portfolio. As you will see in the presentation of my colleagues from the product groups, our roadmap will allow us to grow our business end position in all these areas. Just to mention a few, MPUs for industrial, strong roadmap on automotive microcontrollers, expansion of our analog offering, both in general purpose as well as application specific.
In those markets where we have a selective product approach, we are leaders or in the top 3. So strong position here. As a quick reminder, we have a very strong base, a very strong customer base. Today, we serve over 100,000 customers with a differentiated model depending on the customer type, with companies in our top 10 that are leaders in the market they serve. Our sales are balanced across the regions in which we are designing our products as well as by customer type and channel. So extremely well balanced and spread over a wide variety of customers. Let's look at our end markets. In automotive, we leverage our broad and deep knowledge of automotive systems to develop optimized solution. We go to the market collaborating with car manufacturers, with tier 1.
We offer solution with our distributors, and we have partnership with key players in China in order to grow there, too. What about our strategy? For automotive, as Jean-Marc said, our strategic objectives are leadership in car electrification and leadership in car digitalization. The trends that we see in this end market are the following. The semiconductor market growth, thanks to volume and to a larger extent, to the content increase. Traditional car content is increasing, aligned with market volumes. Car electrification and digitalization are driving much faster content growth than the car volume increase. The introduction of wide bandgap materials causing a disruption in car electrification and the digital architecture of vehicles is evolving from distributed electronic controls unit to powerful domain controllers. Our strategy is the following. First, continue to invest in technology, product, and manufacturing capabilities that directly support our leadership objectives.
Second, leveraging the key disruption with ST solutions, silicon carbide power devices for car electrification, and advanced automotive domain controllers for the transformation of car architectures. We also continue to partner with ADAS and V2X leaders as we have successfully done for a number of years already. On top of this, of course, we will maintain our leadership in traditional areas, thanks to our solid product and technology roadmaps. I will give you some more details. The overall automotive market growth over the next three years is forecasted at 8%. As you can see from the graph, some areas are growing more strongly than others, those associated with car electrification and digitalization. ST has a long time strong position in key areas of this market, like logic, power, and analog. You can see the positioning.
In areas where we have more recently entered, such as automotive microcontrollers and sensors, we have fast-growing business and significant room to grow further. The automotive market is a story of car volumes and silicon content growth. Over the next three years, car volumes are forecasted to grow modestly at just over 2%, while the semiconductor content per car will grow at over 5%. Key for ST is to be highly present in the areas where this content growth is coming from. If we look at the growth over a longer period, we can see that the majority of the growth is clearly coming from car electrification and digitalization. This is why we have the strategic objectives of leadership in these areas. These areas have an exponential growth, not a linear growth.
In the legacy, by the way, we keep growing linearly with car volumes, with a higher exposure to the market cycles impacting the numbers of cars sold. If we look in more detail all the car electrification opportunity, we can see that is driven by a very strong growth in electric car volumes, coupled with an increase in average content per electric car. This translates to an average growth to the ST served market of 20% per year over the next five years, more than doubling the market opportunity for us. Silicon Carbide power devices is, of course, one of the key disruptive technologies and products that enables us to lead in this area, but we also offer IGBT and other solutions that Marco Monti will talk about later. Moving now to car digitalization.
ADAS clearly is one of the key driver of content growth here and in the next five years. The forecast is in the next five years for the penetration of ADAS future in the cars to double, while the content per car will be multiplied by more than 2.5 times. This results in an ST served market growing on average over 16% and nearly doubling in the five years time frame. Here we are talking about ST solution for ADAS, like vision processing and radar, car connecting with V2X based on Wi-Fi and cellular, as well as sub-meter positioning that are needed, and with the most powerful 32-bit microcontrollers that is needed now to support the change of the architecture towards domain control. Moving to industrial. Here we use our deep industrial knowledge to develop solutions optimized for specific applications, leveraging also on our extremely broad portfolio.
We target industry leaders with leading-edge products adapted to their needs. After that, we spread these products to all the market, to the mass market. As already outlined, our objectives here are extremely clear. Leadership in embedded processing, acceleration of growth in analog and sensors, expansion in power and energy management, acceleration of growth with industrial OEMs. The trends that we see in this market are the following. The semiconductor content growth in industrial is driven by three key trends. One is automation, factories, buildings, and homes, the need for higher power efficiency, and data-centric industrial IoT application. Of course, this market is highly fragmented with many different applications, but if we look at a high level, there are healthy growth rates across the key application where ST is focused. Our approach is the following.
First, we are investing to maintain our leadership position on key technology processes for industrial. Second, we are expanding our product portfolio to address key trends and disruptions. In parallel, we have reorganized our sales and marketing teams to sharpen our industrial focus while expanding our field application expertise across all regions. Much more focus on the field to push this end market, which is fundamental for us, where we want to grow extremely fast. All these, by the way, is creating a more compelling offer and support for industrial OEMs as well as for mass market for our distribution partners. As I just mentioned, we see three key trends in industry that are driving semiconductor content. First is the provision of autonomous systems that are more intelligent and aware, integrating, of course, artificial intelligence.
That is valid both for existing systems as well as for new system that will incorporate higher levels of automation. This will result in a safer working environment and new man-machine interaction models. Second, the need for higher power efficiencies everywhere in industrial applications, for all kind of machines in factories, as well as for system and appliances in buildings and at homes. In addition to dedicated power technologies, there is a trend towards digital power control and optimization. The third trend we see for industrial machine and devices to become connected in a very secure manner so they can leverage the Internet of Things. Connection of machines inside the factory and to the cloud allows, for example, to capture and exploit the machine condition data for value-added services and predictive maintenance.
The industrial market is very broad, but if we look now at the application where ST is focused, we see that they are among the largest and highest growing ones in the market. Manufacturing and process automation and power and energy management, for examples, are both growing faster than their industrial market average and offer for us many opportunities in the areas where we are targeting to accelerate growth, embedded processing, sensor and analog, and power and energy solutions. Looking at the market by products, we see that like in the automotive market, where we are also broad, ST has long time strong position in these key areas. These includes power discretes, analog ICs, and micro components. Here, our position is based purely on our MCU portfolio up to now, and will strengthen it in the coming years with our new microprocessors portfolio.
In sensors and actuators, where we have only recently started building our industrial portfolio, we are growing fast, and we see extremely good potential for further growth. Looking around across all our portfolio and our market position, you can see we are in a unique positions to offer complete solution. We do master sensing, we do master the brain, and we are mastering brain becomes more and more intelligent, and we do master also power. A perfect unique position to offer solution to the industrial market. Allow me to emphasize particularly the opportunity in power semiconductor that we see in this end market. If we look at the market overall, we see that about 25% of the semiconductor content in industrial applications is power related. There are a number of application where we are particularly focused, such as motor control and energy storage, where this content exceeds 40%.
You will hear more about how we address this application, both from Marco Monti and Benedetto Vigna in their presentation. Moving to personal electronics. Here, our main focus is on smartphone application specific that also are suited for other personal devices. We are also taking advantage of our general purpose portfolio for the broader personal electronic market. We build dedicated products for some top smartphone players and have a number of market leading products for other players. Here we go with our strategic objective, and as Jean-Marc said before, we have two objectives here. First, leadership in selected high volume smartphone applications, addressing them with our sensors, secure solution, power management, and analog solutions. Second, we want to capture opportunities in 5G that our RF mixed-signal technologies can address in front-end modules.
Looking at the market, we see that smartphone volumes will be flat or not growing that much, but there will be content growth in specific peripheral areas, such as user interface, power and energy management, security, and 5G RF front-end modules. 5G and SET will grow fast within flat overall volumes so that it is clear an opportunity. Other personal devices also are using same architecture and peripheral components, same as the smartphones. This will allow us to sell these devices also to wearables and this kind of devices. The ST approach here is, again, to invest to keep technology and product leadership in selected areas, focusing on five areas for the smartphones. First, specialized imaging sensors, MEMS sensors, secure solution, energy management, 5G RF. We can also leverage our portfolio to address other personal devices, as I mentioned before.
Our focus in smartphones include the four areas we have been focusing on for already many years, and where we have a market-leading position, as I said to you before. Market-leading position driven by leading IPs and leading technology. Wireless charging, secure solution, including secure element, NFC and eSIM, sensor and actuators like our motion and environmental sensors, as well as our specialized imaging solution based on our FlightSense technology. A new area here for us is 5G RF, where we have opportunities to use our unique mixed-signal RF processes and IP to address the RF front-end evolution in 5G smartphones. If you compare the 5G analog front-end with the 4G analog front-end, you have much higher silicon content there, and the requirement in terms of specification are much higher. Claude will talk more about this later.
On the last of our area of focus, last of our end market, communication equipment, computers and peripheral. Here again, the approach is to use our deep knowledge of specific application to develop dedicated products leveraging differentiated technology, capabilities, and IP portfolio. We are leveraging, of course, our long-term relationship with key players in this area and target specific customer and product opportunities for networking and satellite communication. We also engage in partnership to address specific technologies and markets. Our strategic objective here is twofold. First, transform our approach to communication infrastructure, moving from digital wireline application to cellular and satellite application requiring mixed-signal RF. Secondly, we want to capture specific opportunities in 5G RF front-end modules.
The trend we see here that are relevant for us, of course, are 5G is driving a disruption in the communication market, and it's creating a need for new advanced technology to meet the more challenging performances required by 5G. Low-earth orbit satellite communication needs similar technology to 5G. Our approach, of course, ST, we do not have the portfolio to be leader overall, but thanks to our differentiated technology, we can be successful in selected areas. We are leveraging specific ST mixed-signal technology to take advantage of opportunities in 5G and satellite communications. At the same time, we are continuing our traditional business where we are a leader, such as power management for data storage. The 5G disruption, I will not go in detail, but it's clear the 5G, in particular, is bringing a disruption that will impact many end devices, applications, and services.
The well-publicized improvement in performance in the network enable new use cases and services, and will also be a catalyst for the connection of many more devices to the IoT. For ST, we see opportunities coming with 5G in all our end markets. In automotive, it is a driver for digitalization with ADAS and vehicle connectivity and for the architectural change in the car that I mentioned earlier. In the industrial end market, we see the potential for many more smart connected devices in factories, buildings, and homes that require embedded processing with integrated connectivity and security, plus sensors and analog and power. In personal electronics, in addition to the opportunities common to other smartphones, we can address the need for high-performance RF front-end module components with 5G.
In the communication infrastructure market, we have opportunities in millimeter wave communication links, power management for base station, and with our RF GaN and RF SOI technologies overall. To conclude, we are well-positioned to grow faster because we can leverage technology disruption, and we are focused, selective, and we are able to execute. With that, I would like to hand over to Marco Monti, who will talk about ADG. Thank you.
Thank you, Marco, good morning also from my side. I will guide you on the commitment of my organization to grow, outperforming the market. This is thanks to our strategy in automotive, of course, but also in industrial and personal electronics. Let's just start with a few numbers related to, let's say, our result by 2018. I think you are very familiar in how we are organized. In my perimeter, there is one specific group that is dealing with a specific automotive product, APG, that last year grew 14% with EUR 2.2 billion. Then we have one group that is dealing with the power and discrete for the full market application, and of course including also what we do in automotive.
For example, the activity we do on silicon carbide for automotive traction is inside this subgroup that last year grew 19% with $1.3 billion revenue. For the nature of our business, as you know, we are very much exposed to industrial and automotive domains that represent for me, let's say 95% of the total turnover with the limited exposure to personal electronics. With a strategy that is very clear, in automotive and industrial, we would like to be broad range, and of course, as Marco mentioned, we would like to be focused on the highest growing application. This is thanks to the technology offer that we have covering traditional silicon technology as well, a new material as I will try to describe in my presentation.
Marco mentioned at the starting with automotive, the dynamics of the automotive business that is going through a part of the business that is growing substantially linearly with the vehicles, and the part that is totally disruptive, and it is going through the silicon pervasion. This is very much evident in my perimeter because if it is true, as you see from the chart, that the vehicle will grow in the next five years in the range of 2.6%, you clearly see the electronic legacy is substantially going linearly with that at 3.5%. You see the electrification is growing nine times more than the legislation of the vehicles, and the digitalization is five times more. Of course, we would like to be broader range, but we want to concentrate on highest growing parts of the market.
Clearly the development of our, let's say, strategy is going across the year and is visible the effort that we will put in the highest growing segment of the automotive market because these new trends of the digitalization and electrification will reach, let's say, in five years more or less from now, the 50% of our total turnover being today in the range of 30%. This is for us extremely important, not just to, let's say, boost our revenue and of course also our margin, but also to be independent from the fluctuation of the vehicles in the market as you saw is happening this first part of 2019. The implementation of our strategy is clearly visible on our 2018 numbers because overall in my organization, in the automotive application, we grew in the range of 18%. That is double than the sum we are serving.
You see that among the top 10 players in semiconductor for automotive, we are the one that is growing more. You see by application also the implementation of our strategy because it's very visible that we are outperforming the market in all the different application domain, but clear in electrification and others, we are growing much more than the market. Just few number, powertrain, I think probably you have the feeling that we are growing in electrification just because of silicon carbide power MOSFET. Of course, we are a strong believer of silicon carbide power MOSFET, but it's not just that. We are overall growing in electrification with all our technology portfolio, including clearly also IGBT. In the chart I just put the incremental revenue that we had excluding the billing of silicon carbide.
Excluding what we did in silicon carbide in 2018, we still grew 150% versus the previous year. Another point that is for me peculiar is the growth of microcontroller. You know, microcontroller we are keep growing. Last year we grew 50%, and we reach a total revenue of EUR 230 million for specific 32-bit for automotive. We are in the path, as I'm saying also last year, to be among the market leader in this application domain. Now start with electrification. The first message is that the electrification is substantially boosting the power content in the vehicle.
If you take the powertrain that is probably the one that is more representative for this trend, you clearly see that in few years from now, in five years from now, the power electronics will represent more or less 50% of the total, when today is in the range of 38%. Electrification is substantially going with the main inverter of the car, and we perfectly know the main inverter could be made or by traditional IGBT solution or by silicon carbide power MOSFET. As you know, we are a strong believer in silicon carbide for the reason that you see in the right part of this chart. At the end, what is important is that the silicon carbide will contribute in cost cut for the car maker.
Independent analysts are estimating that using silicon carbide power MOSFET for the electrification of the car, they are going to save EUR 2,000 because of the better space utilization, battery utilization, better cooling system. This will boost the market that we believe that by 2020 will exceed EUR 10 billion. We are a strong believer of silicon carbide, we perfectly know that not all our customers are ready to use a silicon carbide solution. They will still use IGBT solution. As you see in the bottom of the chart, in the pie chart, IGBT by 2023, we still represent a significant part of the powertrain market. This for us is perfectly okay because we have the right flexibility to support our customer, of course, through silicon carbide power MOSFET solution, as well through IGBT, thanks to the innovation on technology that we are bringing.
I will describe in a second. Of course, silicon carbide, which is the message on silicon carbide? That we are perfectly in the path already described to have a 30% market share by 2025, when we believe that this market will be in the range of EUR 3 billion, will exceed even EUR 3 billion. Our ultimate goal is to exceed EUR 1 billion revenue in silicon carbide power MOSFET by 2025. We are in line with this strategy. Last year, as I committed, we did more than EUR 100 million. This year, we'll do EUR 200 million. Also in this case, sometimes I have the impression that silicon carbide for us is just a very limited number of customers. This is not the case. Today, we are serving in production more than 10 OEMs, and we see a big traction in design award.
Today, we are engaged with more than 20 car makers worldwide. Of course, many European that are probably leading this trend. I want to mention specifically the alliance with Renault in Europe for this activity, with Hyundai in Asia for this. Big traction for, let's say, award that is boosting our sales in the next few years. In terms of execution of our strategy. Also, in this case, we are perfectly in line with our wish. We are in production with the third generation. We are perfectly in line with the qualification of the next generation that you know will be a trench technology. Through the acquisition of Norstel, we are in the path to be vertically integrated, and this is, of course, to boost our manufacturing, to protect our supply chain, but also to push the R&D activity.
Ultimately, the goal that we have is to be, as soon as possible, in 8-inch manufacturing with the silicon carbide, the power MOSFET. Also to protect our supply chain, you saw we did a multi-year agreement with Cree to support our volumes. Definitely, I would say Cree is a strong partner for us in the developing of the market of silicon carbide. Outside silicon carbide, what we do in electrification? In the chart, you have many examples referred to different technology, but I just want to concentrate on IGBT for the benefit of time. IGBT for us is a reality in the electrification of the vehicle. We are in production with more than 15 OEMs just for electrical traction, or let's say hybrid electrical traction. This is thanks to the innovation we are bringing.
As you know, last year, we brought in production the IGBT with the trench version that is having a lot of market success, and this is substantially boosting our commitment to grow more than 30% compound for the next three years just for IGBT in the electrification domain. This is, of course, also supported by our manufacturing strategy with the new 300-millimeter plant that we are building in Agrate that Orio Bellezza will describe in next presentation. Module. Clearly, module is a big subject for automotive. You must complement module with the silicon strategy. In this case, big flexibility. We have a big variety of module qualified for automotive application, in companion with a big variety of technologies to be inside the module.
In such a way that we are capable to offer really flexible solution, even integrating in the same module, for example, silicon carbide BCD or IGBT BCD. To offer to our customer or fully standard plug-and-play solution or fully customized solution. This is a very important item because we perfectly know that our traditional customer will keep asking fully customized solution. We know there are a big number of player, particularly in China, and I would say not necessarily linked to the automotive domain, that have a specific IP in the electrical motor control that start to offer plug-and-play solution, integrating electronics and mechanics to the car makers. Of course, we must be ready. We are ready to support this new player with our plug-and-play solutions. ADAS. ADAS, as you know, is a very crowded market.
There are many companies claiming to have outstanding market share in the perimeter of ADAS. We want to take a very consistent approach. We want to define what we consider the ADAS domain. For us, ADAS is the product that we specifically design for this application domain. Radar, let's say sensor and let's say MCU, vision processor specifically made for ADAS domain, specifically smart power product for ADAS domain. We are intentionally excluding the standard product, the typical standard voltage regulator, or the typical microcontroller that are used also in other applications and by chance could be used also in the ADAS domain. Having this perimeter defined, you know that we are growing a lot. Last year, we substantially exceeded the $300 million. We are in the range of $300 million with a market share, again, in this perimeter of 30%.
We are representing about one-third of the total market in the ADAS domain. Just few examples, in the vision-based processor, of course, our partnership with Mobileye is continuing very successfully. Last year, we put in production jointly 78 new car models. Overall, end of 2019, we shipped in the market 32 million EyeQ, very good success. Now EyeQ5, as you know, EyeQ5 is the first product of the market in 7 nanometer for automotive applications. This is perfectly in line with our plan, the product has been delivered to our customer, and today is in the field test. Another important activity, just an example that I would like to show you, is the joint activity with Panasonic.
Again, this is another good example of cooperation between ST, that is a, let's say, market leader in the integration of complex digital domain for automotive applications, and Panasonic, that is having strong IP portfolio in the area of, let's say, vision-based processor, visual-based imaging processor, and deep learning. Very good cooperation, and the aim that we have together with our partner is to support vehicles to be launched during Tokyo 2020 Olympic Games. Now, ADAS beyond the vision processor. Also, in this case, we have multiple examples. I just want to mention what we did or what we are doing for the connectivity of the car to our partnership with Autotalks. You know that in this application domain, we were originally supporting the Wi-Fi. Now we understood that we need to support also the 5G because we need to be multi-standard.
Our product today is capable to support the two, let's say the two standard with the two modem integrated, the 5G automotive modem and the Wi-Fi modem. This is becoming really the node of the connectivity for the car, supporting multiple standard applications. You see a big traction. We have already 4 major car makers supporting our solution through more than 10 Tier 1s. In this case, we'll enter in production in 2019 with major volumes materialized in 2020 and in 2021. Now, the subject of the domain, the change of the architecture of the car. This is a big subject in automotive. Today, the architecture of the car is very complex. A typical car is having up to 130,000 ECU with a traditional 32-bit microcontroller. This is very expensive. Also, the management of the software is very complex.
Software is the first mechanism of failure in the car today. It is impossible to reflash the car by air, every time you have a bug, you need to go to the dealer with a big cost associated. Also, the communication speed inside the car is in some way limited. The architecture is evolving through the concept of the domain. There will be a limited number of domain, four or five domain. Each of those managed by a microcontroller with a power in the range of 10 times more than the current microcontroller. Let's say a super domain, let me call in this way, that is managing the ADAS functionality of the car. In this case, you need to have a processor capable to more than 100 trillion operation per second. You clearly see that this transition is perfectly matching with our strategy.
Our strategy through the EyeQ6, that is perfectly fitting the domain controller of the ADAS. Our new strategy on 32-bit micro, the Stellar, based on ARM architecture, ARM core, and based on our 28 FDSOI with the PCM memory embedded. In order to, let's say, describe to you how this transition can be enabled by our technology, I just want to give you one example, probably the most easy, the example of the powertrain. If you take the left part of the chart, today the traditional powertrain is made with the proliferation of ECU managing the engine and the power, let's say, the normal combustion engine, the electrical engine, the transmission, the vehicle control. All of this will be moved under one single domain, managed by, let's say, a super microcontroller, capable of, let's say, computational power about 10 times more than a traditional one.
Up to eight kilo MIPS per second. Again, what do you need to implement such a big computational microcontroller? Substantially, you need to have a very dense technology, silicon technology, because of course the product is very large. You need to have a real-time capability because this is required by the engine control. You need to work with a frequency higher than 400 megahertz. You need to have a low power consumption because you don't want to have a very expensive liquid-cooled solution. Then you want to have a flash cell very large in size and with a fast access time. You see that it's perfectly matching with our strategy in technology, because we will use an FDSOI technology that intrinsically is very robust and is consuming a very limited amount of power.
It's the first embedded memory technology in the market, 28 nanometers, extremely compact and capable to reach more than 600 megahertz working frequency. Definitely integrating the PCM flash that is totally disruptive, that is extremely dense, extremely reliable for automotive application, and capable to very fast access time. This technology is the base of our strategy in microcontroller through the Stellar architecture. Again, today we are offering a large number of product, very good market success. We have more than 10 project with the leader in car makers and in Tier 1 to support our, let's say, mission to become market leader in this application domain. Now moving to the industrial domain. As Marco said, the power technology are becoming a real enabler in term of industrial application because the power technology represent today more than 25% of the total sum.
You see that we are starting from a good position. We are at the second in the market for power and discrete product for the industrial application. As you see, we have many leadership position in the area of automation, energy management, and motor control. This chart is just to tell you that we have a very compelling and broad offer based on our technologies, again, standard silicon technology as well as new material. You see that we have a tailored product for every single application in the industrial domain. A very wide offer based on our technology and our product offer. I just want to give you one example, of course, mimic what we do in automotive. The side effect of the electrification of the vehicle is substantially boosting the need in the networking to recharge this vehicle.
What we see is that by 2030, there will be the need to have an install capacity of 280 billion kilowatt-hour. It's a huge amount of power. You see strong bias on China, clearly in terms of number installation. This will require a total investment in the range of EUR 50 billion. For sure it's a big segment that is showing up. You see that we have a very flexible solution based on our module approach that is stackable to go from the entry level charging system in the range of 60 kilowatt till the upper level of 350 kilowatts charging station.
Also in this case, the main message for us is the flexibility, because we can support our customer with traditional IGBT silicon solution when the cost is the main driver or with silicon carbide MOSFET when the high power, high efficiency is clearly the game changer. Also in this case, modules is a big subject, because of course, you need to support the technology with a module approach to facilitate the life of our customer. Also in this case, you see the big variety of offer that we have from the entry-level application of a washing machine to the super energy consumption of the solar inverter and wind turbine. You see a big product offer in order to support our customer in all the different application domain.
Also in this case, with the right flexibility to support all silicon carbide power MOSFET solution or let's say traditional IGBT silicon solution. I'm going to finish the time dedicated to my presentation, but I cannot leave the stage without reinforcing that we are also supporting, in personal electronics, the strategy of the company to be market leader in the 5G domain. In our case is the power RF with our GaN on silicon solution. You know the partnership that we already disclosed with MACOM, this is going to, let's say, be in the market of the base station and the networking, where our aim is to exceed the 50% of the market share in the 5G domain.
Thanks to a technology shrink and 8-inch manufacturing strategy to support the market of the, let's say, mobile handset, thanks to our, again, solution based on GaN, on silicon that is extremely competitive in term of performances and price. This is concluding my presentation. I think I gave you hopefully the sense of DG as an organization that is committed to grow, to exceed the market and the expectation of the company. Of course, in automotive, that is our reference application domain, of course, supporting the transition and transformation of the mobility, but also in the industrial market with our big variety of product and technology in power application. Again, supporting the strategy of the company to be market leader in the 5G domain, thanks to our GaN on silicon technology.
Of course, as you clearly understood from my presentation, for us, technology innovation, and as you see from Orio, complemented by really a world-class manufacturing organization, is really a key enabler for us to grow more than the market and to support our customer. Thank you. I would like to invite Benedetto Vigna on stage.
Thank you. Good morning, everyone, and welcome to AMS World, the world of analog, MEMS, and sensors. One year is gone since the last time we met. Many things happened. I am proud to share with you a little bit of what we did, but most importantly, what we believe it makes sense for us to do in this field to support the growth of ST. Also, I think, the presentation here will go through three main messages. The first part of this presentation will remind you a few financial figures about this group and remember what is the strategy. The second part, that is the core of this presentation, will address what we intend to do with the three legs of this group: analog, MEMS, and sensors.
The last one, I'll show you how and what we believe it must be done to support, to make it happen, what is our strategy. Let's start with this chart. This chart is the only one with numbers. Our CFO, Lorenzo, later will show you all the detail. This chart is telling you that this group, that represents more or less one-third of the company in 2018, billed around EUR 3.15 billion, out of which EUR 1.8 billion in sensors, both MEMS and specialized imaging sensors, and EUR 1.35 billion in analog. You see also a lot of percentage numbers. These represent the growth year-over-year of the group overall, as well as the analog and MEMS and Sensors. We can count on the trust of around 70,000 customers, thanks to which these numbers are materializing. This is the chart about the strategy.
You have seen that, as Jean-Marc said before, we are addressing four market segment. The strategy of this group is very simple. There is a very strong focus on analog and industrial. We are pushing large opportunities like the power management mentioned before by the colleague in personal electronics, and we are looking at a selective approach for communication equipment and computer peripheral. The product in this group, the tools that this group will use to make it happen are analog, general purpose, and application-specific, or the sensor, either MEMS or specialized sensors, imaging sensors. Let's start from the analog. I was making a consideration while listening to the colleague before. We are in Great Britain, we are in a place where 250 years ago, two guys, first James Watt and then Maxwell, put the base of what was the first industrial revolution.
Here, I will show you what we are doing in this group in terms of application-specific product, either like wireless charging or automation or motor control or connectivity to make it happen, what is our strategy. A quick chart on general purpose analog. They are not so famous like the application-specific, but the end of the story, they are the glue, the basis of what is the analog. We are having in our group, in our company, we are having three kind of product families, usually classified like signal conditioning, power management, or interfaces, that you can find a little bit everywhere. Operational amplifier, USB interfaces, DC-DC converter. These are analog products that are spread all around the four market segment, and where we had a lot of growth, especially in the last years, thanks to the strong support of the sales organization of Marco.
Let's start from the application-specific product. I would like to mention one example that is starting from personal electronics, and it is moving to other market segment, where there is a lot of hype, a lot of discussion in these days ongoing. Then I will talk about the ASSP that we are doing for industrial market. Wireless charging. We want to address. We are addressing this market. We have several solution on the field already, more than 10. The roadmap here of this product is relatively simple. You need to make sure that you transfer more and more high power from a few watt, mostly for wearable and hearable, to more than 15 watt, that is what is needed for the smartphone. Today, we have clearly engaged programs with key OEM and with many small customers.
Today on the market, we reach 15 watt, we are heading to the next challenge that is 30 watt. This is very important because we want to get rid of the burden of the wires. Let's go now to the industrial world, to the industrial world where, let me say, our solution can make easier, in a certain sense, the Industry 4.0 transition. Here, when we talk about industrial world, there are three areas of focus, you have seen also in the previous presentation of Marco before, when he was talking about power discrete. We have power and energy management, we have motion control, and we have automation. Here we have three product for which we have a leadership. We have power line communication, where we have a market share that is around 38%, and for which we shipped more than 120 million unit.
We have motor drives, for which we shipped more than 1 billion unit, we have a market share that is more than 11%. We have also intelligent power switches that are meant to address the switching of the power, for which we have a market share more than 10%. Last year was very important for us because more than 10% of the sales were coming from new product. Passing 10% in industrial market is, we believe, a really important threshold. Let's start from the first examples of application-specific power and energy management. Here we have chip that work or standalone or in conjunction with the power discrete to manage the power, analog or digital. Examples are the drivers for silicon carbide devices or gallium nitride devices, or even AC-DC controller that take the power directly from the mains. This is example number 1.
Example number 2, this is not the steam engine of James Watt. This is electric motors. Here we are talking about chips done in BCD technology, smart power, that are able to move any kind of motor, either stepper or with brushes or without brushes. We have shipped more than 1 billion unit. We have more than 130 product. Why do we believe they are well-placed for the future? It's very simple. We are in a company where there are product that can combine like a puzzle, like a Lego, okay, to offer miniaturized, efficient solution. We have the microcontroller of Claude. We have the power MOSFET silicon or silicon carbide of Marco. We can combine this with the galvanic isolation and motor control of AMS. We can do a unique solution. We can do micro module.
We can do solution that go from, again, a few watt motors to several kilowatt. This is really one important area where we saw a strong growth last year. We are pretty confident for the future. We talk about factory automation. When you talk about factory automation, there are four important blocks. There is, for sure, the microcontrollers or the MPU. Claude will detail about that. There are sensors. There are two important blocks that are about analog chip. One is called connectivity, and one is called actuation. When we develop this product, there are four objectives that we need to keep in mind when we develop the product. The reason why when we ask the customer, they are selecting us.
Number 1 is the flexibility of the solution we offer, the robustness, the fact that this has distributed intelligence, last but not least, it fits the safety requirement that are mandatory in industrial world. This is factory automation, where safety is all about, let me say, the proper connection of the chips of the devices in the factory. When we talk about factory automation, there are well-consolidated trends. There are also new trends. Again, they fall under the umbrella of this Industry 4.0 or what I would say, IIoT, Industrial Internet of Things. There are, you know very well, there are many different applications. It goes from metering, from real-time monitoring, predictive maintenance. If we go to see now, which are the trees, the mushroom that are becoming trees, the new vertical to asset tracking and predictive maintenance.
What can we do to address these two needs? We can address this need or offering wired or wireless connectivity solution. Here, we have to cope with different standards or IO-Link. That is a standard well-known in industry where you reach the last mile of the factory. In the last mile of the factory, you have sensor and actuators. How do you reach them? With IO-Link. Or in some cases where the wire you cannot reach with the wire, where the time is not so critical, you can reach with wireless communication. This chart is very important because it brings to an important point. Last week, we had an important announcement with one of our partners, that is Microsoft. When you talk about industrial market, you talk about key OEM, but also you talk about a lot of small companies. The market is fragmented.
How do we want to address this? We want to address this with two important tools that are, as announced last week, connected also to the cloud. We have two examples of two wireless sensor nodes. I invite you to see also in the demo area, the first one over there, the strange fastener, because this is industrial, that has a lot of sensor, the microcontroller, the wireless link, as well as the IO-Link. While on the other side of the chart, you see a disk that is called a SensorTile, is a conglomerate of all the products that we have in ST that is meant to make easier the way people can develop their application. We have also a nice shell around that is, we see simple things that is making easier the life for the developers, either in the startups or in the big companies.
This chart brings me to the other important product of this group, the sensors. Here you see there is a microcontroller, there are connectivity, analog solutions, but then there are the sensors. Here again, when we talk about AMS, we're talking about a group that is making sensor and microactuators. We talk about a group that in the last 20 years, put on the market 5 billion microactuators and 17 billion sensors, either MEMS or specialized imaging sensors. You heard me several times now in this event telling that we start from personal electronics and we move, we pervade, we push these technologies also in the other field. Last year, 2018, has been very important for us. This has been a little bit anticipated by Marco Cassis before. We passed the threshold of the three digits with the sales of our MEMS in automotive and industrial.
We are very proud of this because a few years back, we were depending only on personal electronics. This is a journey that we are doing. There are many other opportunities. As you see in the ranking, we are still below. We have opportunity to do better. The same strategy is what we want to apply, and we are applying actually to the specialized imaging sensors. Also in the booth in the demo area, you can see some interesting demos of what we are doing, not only alone, with partners. Partners are key to make these things happen. Let's start also, since you see a lot of times, the sensors, they start from personal electronics for a simple reason. You need to lower the cost of this solution. You need to master the production process to enhance the yield. Which better playground than personal electronics?
Here I want to show you what has been done last year. In the center, you see the traditional personal electronics product that all of us have, and then there are the wearable, the hearable, and the glasses. A different maturity level. Well, in this kind of product, you can find analog solution, voltage regulator as well as, let me say, motion control. There are some phones on the market that are pushing out the camera, but you see also our FlightSense product. You see our motion MEMS, you see our pressure. Let's say, we continue to make sure that our products are on the leading edge because in this market, as well as in the others, it's all about product leadership.
If we keep 50% market share, as IHS Markit supports, 50% market share in motion sensors and for personal electronics and for automotive telematics, we can do this thanks to unique technologies Jean-Marc was referring to this morning. We have some unique technologies that enable us to address the guidelines of the product development, increase accuracy, lower the power consumption, and offer some, let me say, smarter solution. These are the products that you see over there, are some examples of products we put on the market last year. I would like to do one focus, in particular, on the high accuracy sensor we put on the market for industrial application. Now there are, in our product portfolio, a lot of products that start with an I.
I stands for industrial, because here we have to guarantee 10 years longevity to the customers, and we need to make sure that accuracy is such that we can address challenging applications like, for example, measurement of inclination. A few years ago, we were out of this market. We are entering this market, and we see that more and more inclination and vibration are important. Accuracy is important. You see also what we are doing for applications in automotive. Accuracy, we are addressing it, and the second driving force of this product line is what? Is power consumption. We announced it in February, a product that has embedded a decision tree machine learning core that can work also with our microcontrollers in a way to address application either in personal electronics or industrial robots. What does it mean?
It means that a lot of times, the motion that you have to recognize are not so complicated like the sound or the image, for which you need a lot of processing power. The motion that we do, that the machine does, a lot of times they're not so complicated. A decision tree in the chip can make the job. The result, we can lower the power consumption, depends on the use case, from 10 to 1,000 times. This is the MEMS. Let's switch now to the photons, to the light. Specialized imaging sensors. Here, we are talking about different technologies. We are talking about Time-of-Flight, for which we've been the pioneer a few years ago with our SPAD. We're talking about structured light, and we're talking about ambient light sensors.
There is a very nice demo outside for the ambient light sensor that is very interesting also because last year I told you we were entering this field. We made it. We have several products running in the market with our ambient light sensors. What we want to do in this field? Here is a chart. We were, as I said, the pioneer of time of flight. At that time, we started to offer solution in time of flight with a few pixels. Now, we want, and you have seen, let me say, there is a progress in increasing demand on the market, where we want to increase the number of pixels. We want to go from a few hundreds of pixels to megapixel.
We want to increase the field of view, the possibility to detect what is around, from around 30 degrees to more than 70 degrees. We want to do this either with SPAD or with fast photodiode. This is what we want to do in the time of flight. When it comes to structured light, the other approach that I think many of you are having in their pocket or on the table, well, for this kind of technology, we want to increase the quantum efficiency for the sake to lower, let me say, the power consumption of the system. Then we want also, last but not least, on the global shutter, we want to make sure that we are able to detect properly the people when they are not in the car, for example, there is a demo outside, where, let me say, in low light condition.
These are the things that we want to do in terms of technology. In terms of market, well, with all the digitalization that Marco has been explaining you before, for sure, we have to leverage our unique position to complement the offer of Marco before in automotive. Here you see three blocks. We are having infrared and visible cameras. The visible cameras are not the cameras for the parking. The visible cameras are the camera for, let me say, e-mirror. The infrared one are meant or to do the lidar, because when you want an autonomous vehicle, you need a lidar. You need a lidar as well as you need a radar. Instead, you need to see also what are doing the people in cabin.
This is coming to the regulation that is mandating from January 1, 2022, the monitoring of the people, the activity of the people in the car. The last part of this presentation, I would like to condense it in one chart. We believe that AMS is well-positioned to grow, both with the help of the key OEM and the distribution. We believe this for three main reasons. Number one. We have a strategic focus in automotive, in industrial, with our products, analog and sensors, but also thanks to the support of the products you have seen from Marco and you will see from Claude. Number two, we have the broadest portfolio of sensor and microcontrollers. Here, really, we can sense all the different dimensions through the sensor we have. Number three, innovation is our DNA. You have seen different kind of products.
This innovation is because of two helix. One is the IP, and one is the technology. I would like to stop here and to leave the stage to Claude. Thank you so much.
Good morning. I will give you an overview of the MDG Group. MDG Group is a microcontroller plus digital activities. Just information related to what we were doing last year. Last year, we, on the microcontroller and also what we call MMS, all product based on a non-volatile memory technology, we achieved EUR 2.3 billion last year. On digital, we achieved EUR 650 million. In fact, there are two different game here. First of all, the first subgroup, MMS, we are growing since years focusing on microcontroller, secure micro, and memories. We want to pursue our growth the next few years, and I will explain how we are going to do that. For what concern digital, remember, this group was set up three years ago when we decided to close the activity on set-top box.
Of course, the past few years, we are dropping the set-top box activity, but you have seen that we have been able to maintain our revenue stable despite this closure. I will explain, now it's time to redeploy this activity, and I will explain how we are going to do that. First, let's check a few numbers and worldwide position on the market. Main activity for MDG, of course, is a general purpose microcontroller. STM32, which is a main driver for their growth. We are very solid. We are enjoys the number 2 worldwide position last year. Of course, the goal for us is to become the number 1 player in that domain. Second point, on the secure micro, we are number 3 worldwide suppliers.
You will see this market is very challenging, but we have everything in place right now in order to become also number 1. For EEPROM, of course, memories. We enjoys the leadership position for years. We have more than 30% market share. The goal for us is to pursue our effort and to consolidate our leadership position. Last but not least, the wireless communication. What I call, before it was called digital, now we are focusing on wireless communication, and this is a brand-new activity where we are moving people to pure digital to this type of activity from a pure digital process to RF mixed-signal type process. What do you want to do with those activities? First of all, we want to become the leadership on embedded processing. Embedded processing means microcontroller plus industrial microprocessor. Until now, we are not addressing this market.
I will tell you what we are going to do. Second point, we are right now migrating a lot of people who used to be involved in a set-top box, digital activity, et cetera, in a RF mixed signal type activity. To make a long story short, to redeploy a lot of resources in order to focus more on faster-growing market. Of course, we are going to keep our leadership on EEPROM. When you check all the market, the target market we have. First of all, if we want to become the worldwide leader on embedded processing, well, we must be strong where the market is growing faster, which means we must be extremely strong in industrial with our capability to be leader in embedded processing. Of course, if we succeed, solution will be deployed on personal electronic, communication, computer, and other consumer type application.
Main driver has to be successful for us. Second point, for what concerns secure micro, the market is moving. The main market driver is a personal electronic smartphone. When you have to pay with your smartphone, you have to have a very high secure solution, secure element. You need to have eSIM, and you need to have NFC controller. This will be the main driver. Of course, all this advanced secure solution will be deployed all over those activities, automotive, industrial, and communication. RF EEPROM future will be on RF EEPROM, embedding near-field communication interface. We are moving on in that direction. We have this new activity where we are talking about advanced RF mixed signal process in our end. We want to become a key player in a 5G in certain domain.
We want to succeed because we feel that we have the foundation in order to achieve this target. This, of course, will be deployed in a communication infrastructure as well as on set. Last but not least, when you are leading the game on a high frequency type, RF interface, you try to bring some innovation. We'll do that with a very advanced product for short range, high speed connectivity capabilities, which are going to address first the industrial market. Let's go to embedded processing first. Embedded processing, it's a very large market, highly fragmented, where microprocessor and microcontroller are the brain of each application. That's a key point. If you have the brain of application, if you are able to supply customer with the brain of their application, you are opening the door for many, many other product. That's our goal.
Of course, evolution of this market will be driven by IoT perversion, and this IoT perversion will be accelerated by the 5G deployment. We have also the capability of AI, which will be embedded in those product. This is running today. Is not the future, is running today. We have a lot of microcontroller embedding AI capabilities. Of course, when you will have to have those connection to all those product embedding AI, you will have to bring the highest level of security. There is no connectivity possible without any security. Here you see that we were addressing until now the microcontroller business, which is in blue here. We are growing a lot, but the sum was not large enough for us. We are addressing now the microprocessor to complete our job, and our sum, of course, is increasing a lot.
To make a long story short, we will address the sum, which will be in the range of $12 billion in 2021. Where do we stand? We enjoy the past three years, nice growth on STM32 as well as STM8. Today, our general purpose micro, our market share is above 20%, according to WSTS. Remember in 2007 when we started this activity, our share was 2%. We have been able to multiply in 12 years the share by 10. Critical point for us is to keep this share, and even to become number one, we must grow and to achieve at least 24%-25% market share. We enjoy 25% average growth the past few years. We have to confirm that, and we have to keep going with a wave of new product. We have to serve today 60,000 customer worldwide.
As of today, STM32, we shipped more than 4 billion unit. The goal for us is to bring 10 new product in the STM32 this year, and we are able to deploy three new product in Q1 this year. Of course, to move our product to more advanced process, meaning that we will achieve, we will design our product in the 18 nanometer FDSOI PCM type application, software, technology within the next few years. Second point, we want to enter in the MPU, microprocessor business. The first product has been launched this year in Q1, and I will say the market acceptance is at the right level. We were expecting very nice feedback from customers, and we see now a lot of activities. When I said microprocessor industrial, it doesn't mean ST is going to reenter in the application processor. Let's be clear.
When you have your smartphone, you have one application processor. In general, today, you will have 4 billion transistor embedded in this application processor. Die size will be in the range of 90 square millimeter and based on the 10 nanometer process. We are talking here about industrial processor, meaning instead of 4 billion transistor, we have a 60 million transistor. Instead of 80-90 square millimeter, we have 25 square millimeter. Instead of 10 nanometer process, we are in 40 nanometer process. There is a gap of 60-70 between both product. Don't mix up everything. We are targeting industrial market. The way those design are done are suitable for industrial market. We do capitalize on the knowledge we have on microcontroller to pursue our effort in MPU. Here is just the fact I just explained. We have two product line, which are key.
We have to be the best in each product line. Of course, STM32 MCU will remain the main driver for us. We have a standard architecture, and right now we are focusing on, first, more wireless connectivity, more advanced security, and more capabilities in term of AI. When we are talking about AI embedded in our STM32 platform, in that case, we are talking about operation up to 300 giga operation per second. Now we are moving to the MPU. MPU, we are going to reuse a lot of IPs, which were used in MCU, but we bring a lot of computing power. Here we are talking about dual core. In that type of device for industrial, you have two core, two Cortex-A core. If you compare with your application processor in the smartphone, in general, you have six to eight CPU core, major gap.
We have in this application, still some real-time operating system capabilities via a Cortex-M utilization. You have two core for computing power, one core for real-time operation. We have the two core are running on Linux. You see totally different architecture. Of course, we brought the first product in Q1 this year, but we are moving with more security. We are moving with more local autonomy. In other words, we are going to bring more computing power dedicated to AI, and we are moving with far more advanced graphic capabilities. Those two product line today are critical for us to become embedded processing. We are going to generate a wave of new product around MCU, around MPU, and this will have to be associated with the best possible ecosystem.
Designer, embedded designer, in general, are selecting their product, if we are able to give them the best possible tool to develop. Let's go now in more detail for what concerns wireless connectivity. Wireless connectivity, we can differentiate in two different activities. First of all, for communication which are not cellular type of communication, where there is no need for embedded SIM, no need for certification, where we are going to have embedded in one STM32 platform, a standalone block, which will perform all the function to drive the wireless interface. Doing that, we will address Bluetooth, Thread, Zigbee, Wi-Fi, all those standard. For us, it's not to select one standard, but we want to address all the standards since there is no leader in those standard. Also we want to address LoRa, Sigfox and 802.15.4.
We are using the same concept of dual-core architecture, one for the application, one to drive the wireless connectivity. Of course, we go ultra-low power. We have to support AI capabilities in those devices. Right now, this year, in Q1, we introduce the first SOC STM32 embedding Bluetooth low energy capabilities. 2nd quarter, we are going to introduce a LoRa chip. Of course, we have a lot of people right now developing application on module using an STM32 associated with some connectivity block. For what concerns cellular communication, here this morning, we spoke about a 5G deployment. We spoke about major IoT pervasion, generated by the capability provided by 5G.
There will be a connected world using a cellular approach, meaning in each application, you will need to have one STM32 equivalent, you will need to have one modem, and you will have to have embedded SIM, secure embedded SIM. Today, we have the STM32, we have the secure embedded SIM, and we have also a partnership in order to generate the interface, the modem compatible with Narrowband-IoT or LTE Cat-M. Right now, we are able to provide our market with a module. This is a phase 1. Assuming that we have this module in production, we can serve the market which are listed here, city light, logistic, et cetera. Also Industry 4.0. You see the application targeted with a cellular IoT approach are totally different. We do believe that this is going to be a great growth the next few years.
Let's switch now to AI. AI, we have a lot of people talking about AI everywhere. We are right now provider of AI solution at the edge. First of all, we launched in Q1 this year, a very advanced software called STM32Cube.AI. The goal of this software is to transform your pre-trained neural network in executable on STM32 platform. It's a very sophisticated software tool. Doing that, you can use your knowledge, you have an STM32. You have to be capable to develop your application, this is going to open a lot of option around your STM32. In other words, AI is accessible, is simple, accessible, innovative solution to many platform based on STM32. Of course, this is opening a lot of capabilities in term of innovation.
When you have AI software available, assuming you can train your neural network, you will have a lot of new application accessible with your platform, which is a brand-new approach. Right now, with our STM32 and the most advanced product, can replace very complex application running AI on microprocessor platform. The goal for us is to try to move up the STM32 we have, embedding a lot of AI capabilities. Last but not least, of course, we are working right now to add on the existing MCU and MPU platform, specific hardware accelerator, which are going to be key to go to very sophisticated application. Here is the tool provided to the mass market. The goal for us is to give our customer the capability to develop AI application in a very easy way. Here you have STM32 Cube AI, of course.
If you are not familiar with pre-trained neural network, we can give you some example. You will see that on the demo, after this presentation. We have some demonstration of example of application running, which are usable by customers. In addition, we have access to community and also to partnership. In other word, we provide all information for customers who are not knowledgeable on microcontroller to develop their own application embedding AI capabilities. Just here is the chart I was showing you a year ago, where I was explaining where we were able to use MCU and where we are going to be able to use NPU embedding hardware accelerator. On the bottom of this slide, you see a lot company and application. In fact, all those application right now are in production using STM32 with software developed with STM32 Cube AI. We have Sagemcom, [NNF, et cetera.
I don't have time to go in detail to all those application here, but those application are in the demo room. You can check, we'll explain how it works. Now, I said the third element of evolution of STM32 is security. I will not go in detail on security. When you have a security knowledge, when you have communication, when we have Industry 4.0, for example, we have to be sure that you guarantee the right level of system integrity. In order to support that, we have microcontroller embedding secure boot, crypto device, in order to give the right support to our customers. You have to be sure that you are going to address authentication and verification of what you are doing. For that, you have to provide secure key to generate, as a crypto.
We have this available within Secure Micro and within STM32. Of course, we have to secure all the communication channel. For that, we have also the right security level. Just to summarize what we have done, for basic security, we have a lot of features embedded in STM32, crypto, trust zone, protection of memory array, et cetera, protection of secure key. When you have the need to go with a very high level of security, we have a dedicated specific secure micro, which is called STSAFE, able to bring the highest level of security in your application. To complete my presentation on data processing, I told you that we must provide access to the best possible ecosystem. On this chart, you see it's a very complex chart. I will not go in detail.
I just want to share with you that for each product we bring on the market, we are giving specific hardware development tools dedicated to the product. We are giving flexible prototyping capabilities for this product. We are able to provide software development tools dedicated to the new product. We are giving, same time, a lot of user application, vertical application, AI, graphics, communication to the cloud, communication to the stack to drive wireless communication. When we launch mass market a new device, we have all those software, all those tools available for customers. If you want to win in the mass market, you must bring that, or you will never be there. Ecosystem remain a key point. You can have the best possible portfolio. If you don't have the right ecosystem, you go nowhere. Everything has to be synchronized and coherent.
Let's move to secure solution, standalone secure solution. This market, so-called a smart card market, was sluggish for years. In fact, the market is dropping. There is a new market emerging. We see a less basic SIM. We see a less activity in that domain. We see that the single wire protocol SIM was not successful. We see some specific growth in this market. The growth is mainly driven by personal electronic, mobile security transaction driven, NFC, near-field communication, embedded SIM, embedded secure element, which is going to address a smartphone first and also wearable application, watches, et cetera. If you want to be solid, strong in the secure market, you must be solid, strong in personal equipment. We have where to address with a similar device, automotive, and industrial market.
I will not go in detail because last year we gave an overview of what we are doing. We are provider right now, the main provider of embedded SIM in the smartphone industry. In other words, we are able to provide customers with our secure element embedding the right software to drive all the SIM functionalities. That's the main point. That's the chart in the middle. When you go to the smartphone on the right, I show you that we are going to ramp up production second half. This has one combo chip performing all the basic function, near-field communication controller, embedded secure element to drive the secure transaction when you are going to pay with your phone. Third, embedded SIM capabilities. One single chip provided by ST using 40 nanometer type technology is able to drive all those function in one smartphone.
In other words, you will have only one chip performing all those function, where today you used to have three chip, then two chip, now only one. We will ramp up production with one smartphone maker in the second half this year. Of course, doing that, we have established ourself as a strong leader in secure element. As of today, more than 1 billion secure element has been shipped overall in the world. Of course, focusing a lot on personal electronic, but today, we see the same concept deployed in automotive with Marco Monti. Industrial, where we perform machine-to-machine type application. In this situation, we are provider of the full solution, software plus embedded SIM, hardware plus embedded SIM software, and also on a trust platform module for computer market. The goal for us is to pursue and moving those device from 40 nanometer down to 28 and 18.
Of course, we have the driver, which is personal electronic, and today we have a lot of product which are well adapted to the market we target. We have a automotive version grade of the secure element to address automotive, and we have also various option to address industrial and authentication. It was the end of my presentation dedicated to product based on non-volatile memory technology. Now, I will switch to totally different story, which is the future of digital. Future of digital, I told you that we have some very good strengths roadmap available, and the goal for us is to become key player in the 5G domain. 5G is a disruption. All people are telling that everywhere. In fact, when you go to 5G from 4G, what you need to do, first of all, better coverage, internet for all.
There are other solutions than the 5G, by the way. Better coverage, very high data rate, above 1 gigabit per second, reduced latency, less than 1 millisecond, and high density connected node. Right now, this type of targeted features is generating a 5G front-mounted market. There are 3 level of 5G according to us. First one, when you are below 1 gigahertz. In that case, we are talking about application which address massive IoT market, where you will have Narrowband-IoT type evolution, low power, low data rate, low cost. Such a market we target with STM32 cellular, embedding STM32 modem plus embedded SIM. I spoke to you about this issue. Right now, let's go to the this will be the largest business, and this will generate a lot of IoT application. In other word, million of users by square kilometer.
Second point, we have below 6 gigahertz type approach. This is in a continuity with the 4G. Of course, there will be a much better data rate. Average will be 10 times what we see on 4G. Latency will go down to 1 millisecond. Of course, we are going to improve a lot mobility and security. In other word, we'll be able to phone from the train up to 300 kilometers without any problem. Here is a continuity of the 4G. We will address this market with RF front end, the same way we are addressing the RF front end of 4G.
The third level of activity in the 5G is linked to 24 gigahertz type technology, where we can talk here about a real disruptive innovation, because this will require millimeter wave frequency, new architecture, beamforming, and MIMO type techniques, and where you will target data rate above 10 gigabit per second. In other word, 4G is at the level 1. Sub-6 gigahertz will be 10 times faster. 24 gigahertz will be 100 times faster. Such a general features. What do we do, or what do we plan to do? First of all, we have added the technology, the silicon technology, in order to perform RF front end module. RF front end module are the module which are driving all the antenna of your smartphone. We have the right technology, and we have been in production with this device.
We started the production on 4G, with our process 130 nanometer RFSOI. This is in volume production and will last until the next of years. We are in a production also with the first generation of 5G OEM product using 65 nanometer RFSOI technology. This will focus on sub-6 gigahertz type application. We are in volume production now, huge volume production planned next year. Also we are in the prototyping phase with a 50-nanometer, silicon germanium technology, where we are prototyping a lot of OEM as well as people for infrastructure. What do we do on RF front end solution? It is going to be used in the OEM as well as in the base station. To make a long story short, we have the right technology roadmap to serve 4G, 5G sub-6, 5G 24.
For us, we want to pursue this effort, and we have the capability in term of manufacturing to be very solid in that domain. Here we are talking about silicon for RF module. The next one is more complex. In fact, we started years ago with companies targeting LEO satellite constellation in order to give internet access to all the world. We have done a lot of work with those companies. Doing that, we have been obliged to work on the beamforming technology, MIMO technology. We are right now ready to start volume production of those devices. This will be done in second half this year. That means second half this year, we are going to deploy this type of product dedicated to a satellite constellation, LEO satellite, and we are going to ship in volume starting second half.
The key point here, we have acquired the beamforming technology techniques. We have acquired the MIMO techniques, which are the solution to develop 5G 24 gigahertz. Right now, we are trying to capitalize on this know-how in order to become one of the key players in the 5G 24 gigahertz domain. Critical point here is, in term of antenna, you will have a lot of chip per antenna. I mean, the market size will be much larger, and you can see that there will be from 128 to 1,000 RF chip per antenna. The market will not be small. That's the reason we believe that we want to go there.
Still to capitalize on what we have done the past 3 years on RF mixed-signal design, we discover that we have a lot of people in the industrial market trying to have a very complex communication, fast communication, transfer of data in a very mobile environment. Robot, for example. And we were trying to find a way to replace cable and connectors in those application. We have it right now. I spoke to you about this a year ago. We are sampling the first product. We are sampling right now the second product, which has very optimized billing of material. In other word, more functionality embedded. And those product ST60, which used to be called SparkLink years ago, are bringing the capability to transfer in a very, very easy way. There is no protocol of communication, is like NFC.
Very close, few centimeters, you will transfer up to six gigabit per second. Very fast, with a very low power and with very optimized billing of material. Those product are in the prototyping phase right now, and the mass market will be done in early next year. Just to conclude on my speech. First of all, we want to reinforce our embedded processing leadership position. We want to be a key leader in industrial microprocessor. We have to pursue our effort on the STM32, which means we have to keep running to remain number 1 on the Cortex-M architecture. 10 product will be released this year, and we have to be sure that people will be able to deploy AI activities on our platform. For what concern the secure solution, I told you that we have right now product in our production embedding NFC controller, secure element, embedded SIM.
We want to deploy that in volume and to succeed, and also, of course, to capitalize on this position on NFC to sell more RF component product. Last but not least, we will contribute for sure to the 5G deployment. First of all, we have the key foundation. We have the silicon architecture in order to succeed promoting RF solution in front-end application. Second point, we are going to ramp up production for the LEO satellite second half this year, and we want to capitalize on this strength in order to rise the 5G market. Thank you.
Thank you very much, Hervé. Now it's time for a break. I hope you have to digest all those presentations. We will reconvene at 11:30 A.M. About half an hour to go to the demos, hopefully have some coffee, and do whatever you want. Now we will begin the second part of the session with Orio Bellezza for the manufacturing strategy.
Okay. Morning, everyone. Welcome back to work. After the business group presentations, now I'm going to share with you ST manufacturing strategy. Our strategy is based on a balanced utilization of internal capacities and external sources, both for front-end and back-end. The internal production relies on six front-end and five back-end sites that are mostly focused on the manufacturing of specialized and proprietary technologies, supporting our business group's product differentiation strategy, as you have appreciated in the presentation of my colleagues. At our major sites in Europe and in Asia, we have established integrated competence center of manufacturing and R&D to timely develop and introduce new products and accelerate volume ramp-up. The partnership we have established with silicon foundries and assembly and testing subcontractors are also instrumental for us to serve our customers with multiple and flexible sourcing, both on standard technologies and leading-edge FinFET, CMOS technologies, and others.
Following a very short and general introduction about ST manufacturing footprint that most likely you know already, the rest of my presentation will focus on four selected key programs supporting our strategic evolution, and these are: the wide bandgap technologies based on the new material like silicon carbide and gallium nitride, the evolution of our 300-millimeter manufacturing strategy. The manufacturing on power modules and our packaging outsourcing. I'm sure you recognize in this track what my colleagues of the groups have been already sharing with you in terms of product and technology strategy. In silicon carbide, we have already accumulated impressive experience in 150 millimeter volume manufacturing. We are expanding our capacity, and we're moving ahead for the vertical integration of the silicon carbide substrate supply. Gallium nitride, where we are progressing in our technology developments.
We plan to enter in volume production next year in a 150 millimeter for radio frequency application. We prepare for the evolution to 200 millimeter by 2021 for both radio frequency and power applications. I will cover our 300 millimeter strategy. As you know, we have a well-established high volume manufacturing fab in Crolles, and we have recently announced that we have started the construction of a new fab in Agrate devoted to smart power MOS, and HBT. This fab is planned to deliver the first production wafers in 2021 and to ramp up volume from 2022 according to the market demand. Our internal capacity in 300 millimeter is complemented by outsourcing in foundries. I will talk about power modules manufacturing that is instrumental to our product group strategy, both for silicon and silicon carbide device in terms of content.
Finally, I will report about our plans to qualify additional technologies in foundries, including analog and power products, to increase our level of front-end outsourcing to 30% and above. In back end, concerning the utilization of OSAT, this objective is already achieved, and we plan to stay at this level. Let's have a deeper look to the silicon carbide device manufacturing. Having started volume ramp-up in 150 millimeter two years ago, we will reach this year accumulated production output of more than 100,000 wafers, multiplying by three our capacity in 2019 compared to 2017. The Catania fab is performing at the best levels of our industry standards with very good yields comparable to traditional silicon-based power families. So far, we have delivered to our customers more than 20 million parts of automotive grade power MOSFET and diodes.
We continue to capitalize on the strength of our manufacturing and R&D competence center in Catania to introduce and ramp up new products and new technology nodes. At the same time, to support our growth ambitions, we are planning to establish soon an additional production line in Asia. Another key ingredient of our strategy is the access and control of the silicon carbide substrate sourcing. I will tell you a little bit more on that in the next chart, where we are summarizing our overall silicon carbide strategy. The strategy is twofold. First, device manufacturing. Our technology roadmap will drive the production mix evolution in power MOSFETs from the current planar transistor to the introduction of trench-based generations in the next couple of years.
In terms of fabs, as said, we are expanding our capacity in Catania, and we will timely invest in a second 150 millimeter line in Asia. We also want to play as a leader in the industry innovation internally and in collaboration with our material and equipment suppliers in order to drive the transition to 200 millimeter production, delivering a breakthrough in terms of cost. In back end as well, we are expanding our capacity at our Shenzhen plant, and we are activating a second source always in Asia. Concerning the substrate supply, the second leg of this strategy, we assume that the market will remain constrained in the foreseeable future. In order to support our growth and our ambitions, we have adopted a clear and robust supply strategy. First, today, we have already four qualified supply in production, so we are safe for the short term.
Second, we have signed a multi-year supply agreement with Cree, and we want to enlarge our base of committed supply. Third, we decided to go for vertical integration with the acquisition of a majority stake in Norstel, as already reported by Marco Monti, my colleague. The purpose of Norstel acquisition is for us to establish internal manufacturing of 150 millimeter wafers and to spearhead the evolution to 200 millimeter wafer diameter, consistently with our device production strategy as reported above. Moving now to gallium nitride. In the radio frequency application, we are completing this year the qualification of the technology, and we are co-investing with our partner, MACOM, to prepare 150 millimeter volume production in 2020. We have also defined a plan to move in 2021 to 200 millimeter wafer support as respect to grow in volumes.
We are also accelerating our development in power conversion GaN application by investing in a new 200 millimeter pilot line at our Tours plant, where we also plan to initiate production in 2021. Those technology developments, both in diodes and transistors, are conducted in partnership with CEA-Leti based in Grenoble. Power modules. Power modules is another important area of focus for us in manufacturing, for our backend manufacturing to support our power technology and product strategy, as described by Marco. We are in place since many years a competence center in Catania, with the capability of design, development, and prototyping of power modules. We are today capable to master production at our plant in Shenzhen of power modules of various complexity, and we have full control of the related supply chain and material and subcomponent.
We can produce both standard and customized solutions based on the embedding of both silicon or silicon carbide devices, or the combination of the two. Production capacity in Shenzhen is under expansion, and we also have activated a second source in subcontracting. Here we have, let me say, good level of robustness. Going back to front-end manufacturing, I will discuss our 300 millimeter strategy evolution. Today, our 300 millimeter footprint is based on our Crolles 300 fab at a number of collaboration with Silicon Foundries. As already announced, we have started the construction of the new 300 millimeter fab at our Agrate site in Italy, having the mission to develop and produce smart power BCD, power MOS, and IGBT.
Going forward, we will extend our 300 millimeter footprint within the frame of a multiple sourcing cluster, where we'll mix and balance internal and external capacities as described here. According to this model, Agrate and Crolles will have specific technology missions. Crolles will continue to drive the advanced digital technologies such as specialized imaging, FD-SOI, and advanced embedded non-volatile memories, radio frequencies, and other CMOS derivatives, while Agrate will focus mainly on analog and power. On top of that, at the same time, both fabs will also manufacturing a set of technologies that will be common in order to allow a certain degree of loading flexibility between the two sites. The ST 300 millimeter cluster will include the Crolles and Agrate fabs, as well as the network of Silicon Foundries.
In this model, the foundries support technologies in common to the ST fabs, as well as certain advanced CMOS node that we are now producing, such as FinFET, for example, but also others. In this chart, we report the current status and the expected evolution of our 300 millimeter internal capacity. In Crolles, we have doubled our production since 2015, mainly in specialized imaging sensors and microcontrollers. Going forward, the fab can be further expanded in a quite modular way through a step-by-step extension of the current building and facilities. Agrate 300 is forecasted to start volume production by 2022, then grow according to the market demand as already said.
As you can see, considering both fab contributions, we have the capability to set internal infrastructures to potentially multiply by 2.5 times the total internal 300 millimeter capacity at support of our differentiated product and technology portfolio in both digital and power. Including also the expected growth in Silicon Foundry, clearly our 300 millimeter production base is set to materially increase in the future. Here we have some more details on the Agrate 300 millimeter fab. As said, the main mission of this fab is the development and manufacturing of smart power BCD, power MOS, and IGBT. In the first phase on the investment, we will install a development and first industrial deployment line to prepare for process and product qualification.
The construction started last year within our current Agrate campus perimeter, is progressing fast, and the building will be ready to receive equipment in the second half of 2020. We plan to be ready to process first silicon the first half of 2021, and to complete the first product qualification by the end of that year. After that, we will modulate the production investment and ramp up according to the demand of the market. Today, in order to accelerate the learning curve in the scale-up of the analog and power technologies to higher diameter wafer, we are leveraging on our experience and capability in Crolles to anticipate the development of specific and critical process modules. Concerning our strategy in foundries, the partnership with silicon foundries are key ingredients to our strategy, mainly on two aspects.
First, to provide our customer with multiple sourcing of both industry standard technologies and a number of ST proprietary technologies that we transfer to them. Second, to access to advanced CMOS FinFET nodes below 18 nanometer, where the huge cost of development and of the investment for the capacity are not justified by the size of our business. Our current level of foundry outsourcing is about 20%, and we have the objective to increase it to 30% in the midterm of the total front-end production value. To support this plan, we are currently executing the transfer and the qualification of additional technology families, such as embedded non-volatile memory in 40 nanometer, IGBT, and we are planning for others. On the investment on CapEx.
Our 2019 CapEx, as being communicated, is set in the range of between $1.1 billion and $1.2 billion. These investments are finalized to support both the execution of our short-term business plan as well as our strategic initiative as described so far. About one-third of the 2019 CapEx are in support of our major key programs that I described today. I mean silicon carbide and gallium nitride as well as power modules. The 300 millimeter strategy, including the fab construction in Agrate and the expansion of Crolles for specialized imaging. About 40% of the CapEx is dedicated to the capacity and support of our sales growth in second half of this year and the first half of 2020.
This includes the expansion of power MOSFET and IGBT capacity in Singapore, the 28 nanometer FD-SOI in Crolles for new business ramp up, as well as the completion of the acquisition of the 200 millimeter fab from Micron in Singapore that we announced back in 2017. The remaining slice of the pie is devoted to support process mix change of our BCD technologies, technology development, as well as several programs of automation, improvement, and modernization of our operations. In conclusion of my short note, I confirm that manufacturing is a key enabler to support ST strategy and customer satisfaction. I think my colleagues has been already very vocal on this from all of the groups of the company. Our production sourcing is based on a balanced make or buy model, so between internal capacity and outsourcing.
We have identified and described today four areas of focus supporting our strategy. I repeat again, investment in production, expansion, and technology development for silicon carbide, gallium nitride power modules, and the vertical integration of the silicon carbide substrates. The expansion of our 300 millimeter manufacturing footprint with the construction of the new fab in Agrate for analog and power technologies, and the modular growth of Crolles in a cluster model that also includes silicon foundries. We're targeting to achieve 30% of foundry utilization, and for this purpose, we are executing programs to transfer more technologies to the foundries. Coherently, of course, our 2019 investments are supporting these strategic programs as well as our short-term capacity and technology mix evolution. I think I can conclude here, and I give the word to my friend, CFO Lorenzo Grandi. Thank you for your attention.
Thank you, Orio. Thank you very much. Still on time. Good morning to everybody. It's, of course, a great pleasure to be here again in London with you. The good news, this is the last presentation of the day. Let's start with this presentation. I think this morning my colleagues have covered in great detail many things. Our strategy to address our served market, our product portfolio, broad and diversified in automotive and industrial, and selective in personal electronics and communication equipment and computer peripheral, our manufacturing strategy and roadmap. Overall, I am sure that they have convinced you that ST is a unique company that is playing a leadership role in the semiconductor industry. In particular, that we are addressing secular growing markets with key intellectual properties, core competencies in technologies, and innovative products supported by strategic manufacturing assets.
I would like now to share with you how all these ingredients come together to deliver sustainable, profitable growth. Today, I will cover three main areas. First, ST financial results. Second, our outlook for 2019. Third, our midterm financial model. Let's start from the results. The past four years have marked an important transformation for the company. Over that timeframe, we deliver significant revenue growth, improved the profitability, and strengthen our financial position. To briefly summarize, revenues grew over 40%. We gained market share as our revenues grew at a compounded annual growth rate of 12% compared to the 9% growth of our served market. Gross margin improved 620 basis points, thanks to negligible unused capacity charges, improved manufacturing efficiency, and better product mix. Our strong leverage on operating expenses translate into a 1,290 basis points increase in operating margin. Net income increased by EUR 1.2 billion.
We significantly improved our financial position by generating more than EUR 1.4 billion of free cash flow in the period after investing EUR 3.6 billion of CapEx. We exit last year, 2018, with a strong net financial position. In particular, in the last two years, ST has shown a strong acceleration of revenue growth. In fact, we have delivered many quarters of double-digit growth, consistently outperform our served market, and register constant year-over-year improvement of our operational results. On a full year 2018 versus full year 2017 basis, our end market focus translate into a balanced revenue growth. Among our product groups, AMS led the growth with 20%, thanks to imaging and analog. ADG grew 16%, with automotive strongly benefiting from auto digital, the ADAS and microcontrollers, and semiconductor pervasiveness in the car.
Power and Discrete strongly grew both in industrial and auto, also driven by silicon carbide power device in the car electrification. Finally, MDG posted 11% revenue increase, mainly leveraging its leadership in microcontrollers. Among our region of origin, EMEIA grew 15%, driven by automotive and industrial. Americas by 24%, thanks to personal electronics, hard disk drive, and communication infrastructure. Asia posted 9% growth leveraging on mass market personal electronics RF application. Our distribution channel increased to 35% of our total revenues, allowing us to reach a large variety of customer worldwide. Last year, when we met here in London, we outlined some expected result for 2018. Today, I'm proud to report that overall, we met our expectation. First, revenue would grow between 14%-17%. Our result was 15.8%, so slightly above midpoint. Second, gross margin would be stable at around 40%.
Indeed, our second half result brought gross margin for the year to 40%. Third, our operating margin would improve by about 360 basis points from H1 2018 to H2 2018. We increased 370 basis points, bringing operating margin for 2018 to 14.5%. Fourth, CapEx between $1.2 billion-$1.3 billion. It came in at $1.26 billion. Finally, our free cash flow increased 73% to $533 million. Exiting 2018, we had around $2.6 billion of liquidity. This amount is after distributing $216 million of cash dividend and after starting the fourth quarter of 2018 with a $62 million purchase of our common shares. In regard to our $750 million share buyback program, cumulatively, we have repurchased 10.4 million shares for EUR 141.4 million. Our net financial position at the end of last year was $686 million, well above the amount entering in the year as we have anticipated.
Supported by strong revenue growth and improved profitability, our common shares have outperformed our peers in the SOX index. From December 31st, 2015 to April 30, 2019, our total return to shareholders, including dividends, increased 196% versus 149% of the SOX. Let's stop now to talk about the past. Let's move about the future. Since the second half of last year, we have witnessed some weakening in the market. In particular, we started to see a slowdown in China for our general purpose microcontroller, and if you remember, we were already communicating this dynamic in September last year. As a result, inventory began to build up in our distribution channel. Meanwhile, macroeconomic conditions has been volatile. As we enter in 2019, some economies and regions began to deteriorate and several indicators turned to negative. The slowing of economic growth and the contraction of certain end market drove estimates downwards.
Even today, there are still well-known situations negatively impacting business and consumer confidence. For example, the U.S.-China trade war. Some industries, such as automotive, are seeing dislocation and imbalances as part of the industry transformation and changing dynamics. The electrification of the car, for example. After two strong years of revenue growth, we now see 2019 as a year of lattice revenues. However, we see this year a temporary pause in our path of revenue growth. The significant different market conditions compared to the first half of 2018 are impacting our 2019 first half billing. Q1 revenues at $2.08 billion, declined 6.7% year-over-year, substantially in line with our expectation entering the first quarter. For Q2 revenues, we anticipate a sequential growth of 2.4%, but still a 6.3% year-over-year decline. However, we are showing more resilience than in the past in regards to financial performance.
Gross margin in Q1 was 39.4%, better than expected. Our Q2 gross margin, which will be significantly impacted by unloading charges of about 80 basis points, is expected to be around 38.5% at the midpoint of our guidance. This means our gross margin will average for the first half of 2019, about 39%. About one percentage point lower compared to the first half of last year, but with $290 million or 6.5% lower revenues. We plan for a strong sequential revenue growth in the second half of the year across industrial, automotive, and personal electronics. This expected level of growth is taking into account already engaged customer programs and new product introductions. Our business plan also assumes improved market condition in the second half. This improvement will fuel revenue growth, especially microcontrollers, analog and power and discrete.
As we mentioned at our Q1 earning announcement, we began to see, and we continue to see, some positive signs in improved order entry and an increase of sales, the POS, at our distributors. Regarding our gross margin, the level of inventory will materially increase in the first half of 2019, production will smooth in the second half, negatively impacting our second half gross margin. Additionally, our exposure to external foundry is still suboptimal to mitigate the demand swings and to help to lower the impact of unloading charges. This is the chart. No, it's okay. It was like that. For the full year 2019, our plan is for revenues to be in the range of about $9.45 billion-$9.85 billion.
This means a strong H2 over H1 increase, up around 30% in revenue to $5.45 billion at the midpoint from the $4.2 billion of H1 at the Q2 midpoint of our 2019 expected revenues plan. To account for risk and opportunity, we see growth in the range of 25%-35% in the second half compared to the first half. About 65% of the $1.25 billion of expected growth in the second half will be driven by already engaged programs and new products from our customers. The remaining 35%, or about $440 million, is expected to come from improved market conditions. Based on the evolution of the market in the second half, we may experience a possible range of variability between $70 million-$90 million of opportunities and risk. We do expect to have the following main drivers fueling the growth in the second part of the year.
Personal electronics with our specialized image sensor, analog, secure element, microcontrollers, and RF front module. Increased revenue in satellite constellation deployment with engaged customer programs. Silicon carbide power device will benefit from additional capacity available in the second half, allowing for higher revenues with existing and new customers. The already engaged and new programs with our customer can also be impacted by general market condition or by their product traction at their final customers. Regardless, we remain confident in our ability to strongly perform in the second part of the year. At the midpoint of our guidance for Q2, our gross margin for the first half will be about 39%, impacted by around 40 basis points of unloading charges.
We already started in Q2, we will continue in the next quarters to scale down our production in order to free up resources from our working capital, reducing inventory. During the second half of 2019, we will run our manufacturing operation at suboptimal level of saturation. As a result, we will continue to register significant unloading charges. Also, at these expected saturation levels, our front-end manufacturing efficiency is negatively impacted. Overall, unsaturation and related manufacturing inefficiencies will negatively impact our second half gross margin by about 150 basis points, only partially mitigated by some productivity improvement, particularly in our back-end operations. Additionally, in the second half, we envisage an increased price pressure negatively impacting our gross margin, only partially offset by improved product mix.
We do expect gross margin for the year to remain at the midpoint of our revenue expectation of $9.65 billion, slightly above 38%, implying a second half running slightly below 38%. At the high end of our 2019 expected revenue range, we plan for a gross margin in the year close to 39%. At the low end, with higher impact of unsaturation and substantially no further reduction of our exposure to foundry, our gross margin will move down, will still remain above 37%. It is clear that revenue growth and operating leverage will be the drivers of our operating margin improvement in the second half of 2019. Even accounting for potential risk to our business plan, we do expect a significantly higher operating margin in H2 versus H1. We do plan to be careful and attentive in controlling expenses.
As a result, we confirm that we will maintain our average quarterly net operating expenses this year between $620 and $630 million. Operating margin in 2019 is expected to be in the low teens, at the midpoint of our 2019 revenues range, and slightly above 10% in the low end of the range. Due to a slower than expected start to 2019, we adjusted our capital spending from our initial estimate. We moderate our 2019 CapEx plan by $100 million to a range of $1.1 billion-$1.2 billion. The decrease is essentially impacting short-term capacity addition. At the same time, we are protecting our strategic products. New 300 millimeter fab in Agrate, expansion of installed capacity in Silicon Carbide, and new image sensor generation. In 2019, after three years of revenue growth, the two last years being strong growth, ST will experience a year with revenues that are substantially flat.
Nonetheless, we expect to still deliver good results. Revenue at the midpoint of $9.65 billion, gross margin between 37%-39%, disciplined control of operating expenses, operating margin at the midpoint of revenues for the year in the low teens, protecting our strategic initiatives, and delivering a significant positive free cash flow in the year, while securing a solid balance sheet. Today, ST is a more resilient company, delivering sustainable profitable growth. We are ambitious. We are focused on making further significant progress as we move forward. My colleagues, in their previous presentation, have shown how the company has all the ingredients to be successful: market strategies, innovative product portfolio, key technologies, and world-class manufacturing infrastructure. I mentioned earlier, our end market focus has translated into a balanced revenue growth. It has also positioned ST for a significant growth in the future.
As my colleagues shared today, we have many catalysts and opportunities to further improve our revenues and profitability. Assuming our servered market will restart its growth at a pace between 4%-5% in the next few years, ST has all the ingredients to outperform the market to reach, in a midterm period, a revenue run rate of around EUR 12 billion. This revenue level can be supported by our current manufacturing infrastructure, exploiting the full capacity of our fabs and moving incremental production to foundry to around 30% of our front-end production value. At this revenue level, we will enjoy significant improvement in our economic and financial performance. Our operating margin was 14.5% in 2018. We can, and we will do better. As we look on a midterm horizon, we see the potential for improvement.
We will face, as usual, detractors such as price pressure and inflation of material and energy cost. Our path to improve and include the significant benefits from manufacturing efficiencies driven by higher volumes and strong operating leverage on expenses. The path to our midterm financial model is based on a solid business plan. As we have detailed today, we expect to rapidly restart our growth. Our ambition over a midterm horizon is to reach EUR 12 billion of revenue, substantially thanks to organic growth. At this level of revenues, our gross margin is expected to be between 40%-41%. This translates to an operating margin between 17%-19%, thanks to the significant leverage on expenses that over the midterm will increase at a normal rate of inflation compared to the current level.
Net income with an effective tax rate between 15%-17%, we reach EUR 1.7 billion-EUR 1.9 billion, and free cash flow well above EUR 1 billion. After a level of investment supporting the expected revenue growth, the financing of our launch strategic initiative and small acquisition to complement our technology and product IP portfolio. To support our midterm financial model, all our product groups are expected to contribute to the growth and improvement in profitability. Leveraging revenue growth and specific margin expansion drivers, ADG will move from 12.1% in 2018 into the mid-teens. AMS from 15.5% into the high teens, and MDG will move from 18.6% to about 20%. Our capital allocation plan is simple: sustain our growth and with the profits generated by our business, provide our shareholders with a return on investment.
In the near to midterm, our CapEx will maintain our world-class manufacturing machine and will support the key initiatives to drive further growth. As already said, we do look to increase our outsourcing to foundry from the high-teen level of today to over 30% of our production value. On M&A, as mentioned, our plan of record is organic growth. However, we target a smaller strategic acquisition like the recent Norstel deal to complement our portfolio. Having said that, we will remain vigilant in evaluating larger potential opportunities that make strategic and financial sense. On May 23rd, our shareholders will vote on the resolution for a cash dividend of EUR 0.24 per share, payable in equal quarterly installments. We do intend to continue to use, in the next years, cash dividend as our primary means to distribute wealth to our shareholder, consistent with our expected cash generation.
Additionally, we have also launched in Q4 2018, a EUR 750 million share buyback program over the next three years. In conclusion, I hope my colleagues and I have today offered to you a compelling value proposition. Indeed, ST has a bright future ahead, and we are aiming for a stronger and more resilient company. We see significant opportunities to grow our revenues, improve our profitability, and make ST financially stronger. Ultimately, we are focused on the creation of long-term value for our shareholders through the sustainable profitable growth of ST. Thank you for your attention.
This is achieved the plenary presentations, I will now call our executives to come on stage to go through the Q&A session. If you don't mind. Any question in the room? Janardan?
Hi, good morning. It's Janardan from Liberum. Thanks for taking it. I think everyone's question will be, what is midterm, when you say that you'll reach this EUR 12 billion target in the midterm? If you could give us some guidance on what you see that. Also, about 60% of your revenue is from industrial and automotive. In that midterm horizon, when you reach that EUR 12 billion, would you see that mix remaining roughly the same, or would you see industrial and automotive continue to grow as part of the mix, perhaps to 65% or 70% of your revenue? Thank you.
No. For the next two years, we do believe that the overall mix between product groups, so automotive, industrial, personal electronics, and communication infrastructure, basically will remain the same. However, it's more important that inside each product groups, as Marco has shared with you, we will see the acceleration of electrification and digitalization, clearly. In industrial market, you will see by product, the acceleration of our power solution proposal, and analog and sensor as well. The investment we are doing in industrial OEM is, let's say, more long-term investment than the short-term investment. As Benedetto said, when you have already 10% of new product in industrial market buyer, is already a great achievement. This investment on OEM industrial will take a longer time than a short-term period than two year.
It is definitively a very strong determination we have, to work with industrial OEM, and we do believe the value proposition we share with you will pay back in the next five years, of course. No major change overall between OEM, top 20 other OEM, and distribution. No major change by product group, but you will see dynamics, very strong dynamics inside the product group in order to sustain our long-term profitable growth.
Midterm?
The description of what is midterm?
Of course, when we talk about our model for midterm for the company, as you can easily understand, it depends also how the market will move in the next few years. If we assume that the market will restart to grow in the range, as I was saying during my speech, between 4%-5% in the next years, we do believe that the company has all the ingredients to reach this midterm model, either in the second part of 2021, as a run rate, or in 2022. Of course, I repeat, it depends on the macroeconomics condition. It depends if the market will restart to grow at that speed, at that pace that we believe is the one for our secular growth in the market in which the company is at, the market that the company is addressing.
Barclays, Andrew. Sorry, Andrew. Forgot you're
Thank you. Andrew Gardiner from Barclays. I was just wondering if you could give us a little bit of greater update in terms of the visibility you're seeing in the end markets at the moment. You obviously highlighted the macro uncertainty that we can all see. Just how have things changed over the last month since you reported? Also, Lorenzo, I caught that you said something about a EUR 70 million-EUR 90 million range of risk, or something to that effect. Can you just give us a bit more detail as to what you mean there in terms of the near-term dynamics? Thank you.
I will take the first one, and then he will come back to the range. Compared to what we said three weeks ago at our Q1 result announcement, clearly, we confirm the positive sign, making us confident for H2. We have spoken about the POS, the sales of our distributor. We confirm to you that, in Asia, China and America, the POS is showing a positive sign. Means the end demand is there. Where we are not still seeing positive sign is in Europe. We know that generally speaking, Europe is less reactive than the Asian one and American market. We confirm to you that our booking in Q2 is showing a much better trend and sequential growth compared to our Q1 booking.
These two signs, the order bookings plus the POS trend positive in China and America, make us confident for the second half of the year. For let's say, custom design programs, on which we forecast approximately two-thirds of our growth for H2 versus H1. Now, we have a very good visibility for Q3, in our backlog. The visibility for Q4 will come, as Lorenzo disclosed to you, it will depend as well, from the end demand of our customers and the successful start of their device. That's the reason why this range of revenue, certainly we will confirm at our July result announcement. As last year, we confirm late in the year a range. Why? Because we know that our customer can adjust up to the last moment in the year.
Overall, I have to say yes, very positive signs, which make us confident for the overall market in H2. In China, in America, not yet in Europe, but we are not especially worried because we know that Europe is less active than the Asian market. On custom design specific programs, we have a good visibility on Q3. Visibility in Q4 is coming normally, okay, at the adequate speed. Maybe Lorenzo, you can be a bit more accurate.
Just to complement the answer of Jean-Marc. Yes. What I was saying during my presentation is that on the portion that is depending on the market, I was indicating a range that was moving from EUR 70 million-EUR 90 million risk and opportunity. Why? We do not believe that in the second half, let's say all what is related to the market should disappear, because we see today that we have some demand, some, let's say, improvement in this respect. It's true that on this area we may have some risk, and these risk are embedded overall in this EUR 200 million ± that we have envisaged. For sure, if you take in term of percentage, the EUR 90 million, EUR 70 million, EUR 90 million on the EUR 400 million, EUR 440 million coming from the market is the highest, let's say, range of volatility that we have in our plan, in this respect.
What I may complement, okay. For automotive, I will not go more on what Marco has said. Okay. We see that overall vehicles produced are basically flat, but in H1, okay, slightly decreased, but there is a boost from digitalization and electrification. On personal electronics, I don't want to comment. It will depend the introduction and the success of the various device. I think it's important to share with you, from the field, the feedback from our customers, either big OEM or distributors. Clearly, the feedback is the end demand is there. Since mid of Q3 last year and accelerating in Q4, in Q1 as well, they are in a inventory supply chain adjustment. All customer we visited confirmed to us and confirmed to me that the end demand is there. The end demand, there is no strong downside, deceleration, decline of the end demand.
It's more a mood where people are cautious on the overall inventory level and they decrease their inventory supply chain. We are more in this mood, which again, make us confident as far as the end demand is still there. Again, the POS is a good sign that in H2 we will have a recovery of the market and all these opportunities and risk are embedded in the range of revenue we have disclosed to you.
Stéphane?
Yes, hello. Stéphane Rioudaux. I have two questions, if I may. The first one is on the OPEX. You've guided for 40%-41%, gross margin, on the back of the EUR 12 billion of sales plan. Quick calculation, it makes EUR 2.7 billion of OPEX, EUR 675 million per quarter, barely the inflation. The question is, do you need more to achieve all your growth opportunities, or do you need to redeploy some resources? The second question is about the inventory correction in microcontrollers to see if we can have an update specifically on this point. Thank you.
I take the question about expenses. About expenses, we think that the structure that we have today in the company is adequate to sustain the $12 billion. It means that at the end, we do not envisage to have any mass significant addition in the company in order to push our programs. For sure, we need to complement here and there. We need, let's say, as also Claude explained, maybe to readdress some areas in order to put our resources in area where we believe to have a faster return on investment, but without very significant changes in the structure of expenses that we have today. For sure, when we look at the midterm, we need to consider that there will be some, let's say, rate of inflation, maybe a little bit higher than the inflation. You made that computation. I'm pretty sure that this is correct.
Let's say, that means that we need somehow, let's say, to some extent, let's say, increase our expenses. I repeat, without any significant increase in term of massive increase of our expenses. In term of dynamic of the inventory, maybe Claude, if you want to give a comment.
Yes, I can answer. Versus the business trend we see today with the worldwide distribution. The disti stock should be at the right level, I think that at the end of third quarter this year.
I would like to make stronger the message about OpEx. This is clear that Claude, during his presentation, has shared with you the fact that smoothly, the resources we have allocated to pure digital ASICs are going to support first industrial application NPUs, clearly, where we want to go. Clearly, also, RF mixed signal, either for front-end module or infrastructure. 5G communication or satellite communication. As well, Marco needs some resources for advanced digital for ADAS. This is point number one. We still continue to have a movement of design resources within the company, which is, let's say, very consistent with what we have done last year as a strategy revisit, to say we want to really focus on this end market, end product, and technology. Well, as a normal, let's say, business as usual company, we also want to act consistently.
We want to be, let's say, very aligned with our strategy. Definitely to focus more on OEM industrial, we'll need more field service engineer to support this ambition. On the other side, okay, we will make a lot of progress on our efficiency and productivity in some business, let's say, processes. We do believe that our OpEx will be remain this model at the inflation, because the company is really entering with a strong dynamic to be very coherent in term of resources with the strategy we support.
Jerome, you are the second Jerome.
Jerome Amer , BNP Paribas. Two quick question, one for Claude. The choice for phase-change memory for the microcontrollers at 28. Just want to understand if your client are agnostic to that choice because we see NXP, Samsung offering embedded MRAM, we see GlobalFoundries embedded MRAM as well, Renesas seem to move to embedded MRAM. I just want to understand if your clients are agnostic or if it will imply some choices going forward. The follow-up is for Benedetto. When should we expect a Time-of-Flight QVGA or VGA, and will it be a direct or indirect? Thank you.
First, PCM technology, we are engaged in this type of techniques for years now. We have a long experience, we know where we go exactly. Today we have ADG is already promoting their first product based on a 28 nano FD-SOI PCM type technology. For us, based on the result we have, we enjoy right now, decision is taken to move on to go down to 18 nanometer FD-SOI PCM technology. This is our choice, a strategic choice taken, we are moving on with that direction.
I take the second one. The second one is, as you know, we are already on the market with a small number of pixel. We are heading to, as I said before, to a megapixel. That means for us, the pixel size less than five micron or even less than four micron, and we are heading for end 2020, first half of 2021.
Probably. Let me just complement for automotive. You made a question on PCM. Automotive is the strong pusher of PCM. I think I need to spend a couple of words. I have to say that our customer are a strongly pusher of the new flash cell, new embedded memory. The reason is very simple, because this flash, as you know, is going to the back end of the process, is not, let's say, going through the full treatment, temperature treatment when you develop the wafers. It's intrinsically very robust. Is, for sure, the best memory to implement the concept of the functional safety in automotive. All our customer working in automotive are a really strong believer and pusher of this new flash concept.
Johannes, right in the middle. Hello? Hello. Johannes.
Thanks. Johannes Schaller from Deutsche Bank. In terms of silicon carbide, Cree obviously made a massive capacity announcement just a few days ago. Can you talk a little bit about how this impacts your silicon carbide roadmap in terms of material availability and potentially also pricing? How does it impact your plans on the Norstel side? Do you think actually you can be price competitive with Cree given they're stepping up scale so massively now? Second question, just in terms of your IGBT momentum that you showed on the automotive side. If you could talk a little bit about that, what has enabled you to get there now against some of your competitors, and then how are you catching up in this market? Thank you.
On silicon carbide, as I mentioned in my presentation, Cree, we consider them a partner in our strategy. We are very much aware on the investment they are doing. We are very pleased by this investment. They are simply following our demand of substrates. We really welcome their, let's say, moving strategy to keep investing in this technology. On Norstel, the acquisition of Norstel for us is important in three different area. One is, of course, to support our growth, so our internal volumes. One is to be competitive and aggressive on price. The third one is clearly to push on R&D, in order that we will be as soon as possible to 8-inch manufacturing, so to provide to Orio 8-inch substrate to manufacturing our silicon carbide MOSFET and diode at 8-inch. We don't want to tell you more.
Clearly, as Orio presented, the final goal is to be vertically integrated. We will not, say of course, do 100% of the volumes internally, because you know in automotive you need to have the concept of the multiple sources. We'll be one of the sources that we have, and clearly we would like to be cost competitive. This is one of the goal, as I explained. We have an internal plan to, let's say, integrate Norstel in our manufacturing. We are not yet ready to disclose you the plan that we have internally. I think in the next few months, our plan will be much more clear.
The momentum on IGBT.
Sorry, IGBT. Sorry, I was forgetting IGBT. IGBT clearly for us is a big subject. I think since more than two years, we introduced a new trench technology. It's a very innovative technology developed in our, let's say, manufacturing in Catania, in our R&D in Catania. We are extremely successful, I have to say. I was presenting the activity we do in automotive, but clearly we have exactly the same kind of activity for the industrial application. We know, of course, we strongly believe in silicon carbide. As well, we know that not all our customer are ready to move on silicon carbide, and we know that there are application which the pressure on price is so huge that will not move to silicon carbide in the next weeks, let's say. We are boosting our activity.
We'll continue the development of new technology in trench, in IGBT, and complementing this with the power module.
I'm sorry. You said go to the back because I don't see who is. Maybe right in the middle here. I'm sorry, there's light here. I don't see you, but I know you're there.
Good afternoon. Amit Harchandani from Citi. I just wanted to go back on the midterm financial model, and in particular, the gross margin evolution. You had about 40% in FY 2018. You're talking about doing 40%-41% in the midterm. Clearly a lot of puts and takes. Obviously top line leverage, depreciation, 300 millimeter outsourcing. I was just wondering if you could help us get some more granularity in terms of what the puts and takes are, because from 40, you are just going to 40 to 41 despite EUR 2.4 billion on extra revenue. That would be my first question.
Secondly, as I look at the segment-wise margin evolution across the three segments and where you would like the operating margins to go to in the midterm, is it primarily OpEx driven across all of them, or is there actually gross margin movement in any of these segments? Thank you.
I take this question. About gross margin, what we think in our mid-term model is that our gross margin will be in the range, as you rightly said, between 40%-41%. The gross margin for the company is a blend of many ingredients. One is related to the products, the mix of the product, the good mix of the product groups, is related to the fact that for sure, moving ahead, we will face some headwinds. One is related to the fact that for sure, cost of material and energy, as well as price pressure will be there. There will be also the impact of our investment to fuel the future growth of the company that will somehow impact our ability to improve in term of our gross margin.
I would say that we think that this level of gross margin midterm is the level that we may position ourself as a company. We don't see, let's say, strong opportunities to do better. On the other side, there will be different dynamic in term inside the various groups. I would say that in term of operating margin, for sure, the big leverage will come from the leverage on expenses. This will be in term of model, as you have seen, by the way, my chart, it will be definitely one of the most important drivers in term of improvement of our operating profit and operating margin.
Adi, you.
Adithya Metuku from Bank of America. I had a quick question on 5G. You talked quite a bit about that. Can you give us some color on how much dollar content you expect to gain when you go from 4G to 5G in the front end of the phone? Maybe talk a bit about the strategy you have to gain share from competitors who are currently in this space. Thank you.
Okay. For the 5G deployment, the first priority for us is to address RF front end module, like I said. Today, we are in the range of $100 million revenue, 4G plus 5G ramp up. We do expect to multiply by 2 to 3 this revenue within the next 2 years. This is for the RF front end module, 5G. Also we are going to see a drop in 4G, of course. For what concerns the. We believe that we can make it because, in fact, we have the technology roadmap we have right now. We have been able to promote those technology to the people involved in the 4G RF front end module, it was a very good start for us because they were just designing their product on our platform.
Right now, we are moving on with a similar approach for the 5G, where we are moving from 130 nanometer SOI to 65 nanometer SOI. We do that with 2 different approach in terms of business model. We are giving capabilities to our customer to design their own product on our process roadmap. COT business, COT plus business, or we do our own design. I think that today, the differentiation for us is mainly coming from the advanced technology roadmap we have.
Perhaps Matt in front.
Good afternoon. Matthew Ramsay from Cowen. Where are you guys?
He's here.
Had to wave a little. Hate to come back to gross margin again, just a couple clarifications. I think that the target model, you talked about it being back half of 2021 or perhaps 2022, it occurs to me that that's right about the time when the new 300 mil fab will be coming online and presumably at tiny volume. Maybe you could talk a little bit about the puts and takes as that capacity comes online, what that might mean for margins going forward from there. Secondly, I thought it was useful that you broke out electrification and digitization in the auto business as growing much, much faster and the majority of the growth. Are those products as a portfolio margin accretive to the business? Thanks.
If I well understood, your question is about, let's say, if in our midterm model, there is any impact on the new fab. No? Okay. No. In reality, let's say what we do expect is that the investment that we are doing in 300 millimeter is not significantly impacting in terms of contribution to the revenues, both, let's say, dynamic of the gross margin on this midterm model. We do expect that this will be the further step in which we will definitely start to enjoy in terms of capacity and in terms of benefit, in terms of cost from the 300 millimeter. Will be after this step, in the sense that at the end, it's still not embedded.
If you think to the time horizon that we are saying here, that is 2021, 2022, still at what we call our 300 millimeter in aggregate is not significantly contributing to this.
Probably for the electrification and.
Oh, if you want.
No, please. For the electrification and for the digitalization, yeah, of course, this product are contributing. As I mentioned into my presentation, this product is contributing to our top line, but also to our margin and even to be independent from the fluctuation of the car registration that normally you have. Yeah, for sure, let's say. Of course, particularly on the electrification, there are also associated commodity business, the diode, the protection and so on, that is not the mainstream that what we are presenting today. This is clearly not the super highest level of margin. Let's say for everything involved by ADAS, all the power stages, all the modules and so on, they are for sure accretive in term of gross margin.
I would like to highlight on this electrification and digitalization, the way ST is addressing these two high-growing market and application. We address it in a very good balanced way. First, look at electrification. Electrification, we address it in a good balanced way between strong innovation on silicon carbide power device, but innovation in IGBT on silicon.
We are addressing in a very well-balanced way between internal manufacturing and internal supply, but external one as well. We mentioned that we are qualifying a foundry for IGBT. For sure, for silicon carbide, we have a very great partnership with Cree, with Greg. I know Greg very well, but we are setting an independent supply chain for silicon carbide. Very well-balanced way. We will offer custom design solution for a car maker or tier 1, but we will provide a standard solution for the high volume player we have seen in Asia offering turnkey solution of the full, let's say, electrical powertrain. The engine, the asynchronous engine, the inverter, and the gearbox. Here we will provide standard solution. Custom solution, standard solution, internal manufacturing, external manufacturing, well-balanced IGBT silicon carbide. Well balanced business model to go. Digitalization is exactly the same.
On advanced digital, we use our partnership with Mobileye and we are, let's say, fabless model. We use TSMC mainly on this very advanced, let's say, FinFET technology. We start to diversify our, let's say, middle-end and low-end ADAS with, let's say, partner like Panasonic, and certainly some other will come. We leverage our internal capacity and internal technology like 28 FD-SOI, [ePCM], and we will have a next generation for high performance MCU to address, let's say, the domain, the change of architecture. As well here, we will qualify a second source with a foundry. The last but not the least, we will leverage a 5G disruption with our partnership with Autotalks on V2X between Wi-Fi and 5G. You see, we really address this high-growing application in a mixed mode, well-balanced business model, internal, external, well-balanced product family.
This is really something I would like to add light, and this will contribute to the midterm gross margin improvement and operating margin improvement of ST, thanks to this approach.
You see that my boss is very well prepared on automotive, it's a strong point. I think I still need to keep my job as a marketing guy in automotive.
Barlow? The gentleman here.
Very good afternoon. This is Dinesh Kathani from IHS Markit. Thank you very much for inviting us. I have two questions related to the SiC MOSFET production and the Tesla contract. The first question is, we heard that the SiC MOSFET production capacity is full because of the Tesla contract. Any comments on that? That's the first question. The second one is ST able to meet up with the demands from Tesla with regards to the SiC MOSFET? Or will it be one of the two suppliers, ST being the primary and Infineon being the secondary supplier for that company? Thank you.
Of course, as you can understand, we cannot comment on our customer. The only message that I have is that we are a very well-diversified customer base. As I mentioned in my presentation, we have more than 10 OEMs today in production. Let's say our capacity is going through the needs of our customer, overall, let's say all the customers that we have in this segment for both automotive and industrial domains. As mentioned during the presentation, we are progressively solving the problem of the bottleneck of the supply chain, so substrate and other materials. Of course, you can imagine we cannot comment on the specific programs that we have on silicon carbide.
On the other side of the room, there's François-Xavier Bouvignies here.
Hi, thank you. It's François-Xavier Bouvignies from UBS. Just a quick question on silicon carbide, I'm afraid, again. When we move to trench from planar in your roadmap, do you expect any change of competition? Your design wins have been very strong at the moment. As you move to trench, do you see any change of market dynamics?
No. I have to say, today we are very happy with the second or the third generation that we have in production that are planar. I have to say, the movement from, let's say, the planar technology to the trench technology, I would say, is more to standardize our production flow because all our technology power domains are based on trench, the low voltage, the high voltage, the IGBT. The silicon carbide will be in trench, we'll have a better utilization and reutilization of the tooling. There will not be a substantial change for our customer. There will not be a substantial change in the market that we are supporting. It's exactly in continuity what we have today. It's a normal development of our strategy through the roadmap.
Specifically, you don't see Infineon maybe, which is more focused on trench, getting more aggressive on the future deals?
The trench is a different way to develop the transistor, it's not changing the performances. Again, no, we don't see, of course, we have a lot of respect from our competitor, we think that with the design that we are building will boost our future for the goal that I expressed by 2025.
Just in front, we can keep the same side with Julia.
Hello. Just maybe to come back on 5G. You gave us these very helpful numbers for silicon carbide, EUR 100 million going to EUR 200 million, and we mentioned the opportunity on the front end. Is there an opportunity in RF power as well? The question would be, could you size the 5G opportunity for you guys, the same way maybe you sized the silicon carbide opportunity? Because obviously, that's expected to be a big driver, not just the front end, the overall.
Yeah. If you are referring to the part of RF with our GaN, no, we gave a target for, let's say, for the base station and for networking. The target that we have, let's say, in partnership again, with our American partner, is to gain 50% of this market. I think you can easily calculate how much will be the revenue in the midterm. We want to, let's say, to enter in the market that is new for us because the power RF is new for us, for the mobile handsets, still with the same technology, to a technology shrink and going to eight inch. Also in this case, of course, it's an opportunity for us, but we prefer not to give you a size. We will prefer to be a little bit more consistent when we will be in production.
We start production at the end of this year with volume in 2021. At the moment we will be in production, we develop our production, we'll give you the target of the revenue.
Okay. For the 5G, I said in my presentation that right now we are planning to have a significant ramp up on the RF front end module for the 5G sub-6 gigahertz, which is the first phase. We see in that domain for us, EUR 200 million-EUR 300 million within two years from now. This is the first phase. In fact, the critical point is to know when we are going to see the deployment of the 5G millimeter wave, 24 gigahertz, where we have a lot of prototyping activity at the level of infrastructure as well as onset. Where we don't see significant growth within the next two, three years. It's difficult to anticipate today. We are not able to share with you what will happen.
We have a lot of activities, a lot of prototyping. The deployment in volume is questionable because the infrastructure will be far more complex. Here we are talking about a disruption versus what we have seen in 4G and 5G sub-6. This is a very complex situation. Right now it's premature, I guess, to give you some numbers. The key point is, of course, we want to capitalize on the know-how we have.
Well, clearly, of course, I am fully aligned with my colleagues and the team for the short, medium terms. What is important for the company is to invest R&D and to invest resources in business and technology, which can bring EUR 1 billion additional revenues. That's the reason why we invested in Silicon Carbide, because again, we do believe that with the market reaching EUR 3 billion in 2025, 30% market share clearly will give to ST this EUR 1 billion revenue opportunity. Revisiting last year our strategic plan and confirm that we want to accelerate, we want to allocate resources from pure digital where we were, let's say, only incumbent and absolutely not a leader on RF, let's say, mix signal to address front-end module.
Infrastructure, the expectation the team have and I have is to, in the near term, to bring EUR 1 billion additional revenue to the company. It is not okay within the two years planning horizon, of course. This will bring to ST certainly the opportunity to continue to grow because as I said during the introduction, the value proposal we have to our shareholder is to leverage, deliver value based on sustainable and profitable growth. This will participate for sure.
Gentleman in the back.
Thank you. Just a question on your involvement in 5G handsets and 5G infrastructure. When I look at the history of ST going back over the last 20 years, you've dabbled obviously with infrastructure, with no success. You obviously had significant involvement in handsets that did not end up very well. I just wondered, given the growth opportunities you have in automotive and industrial, in general purpose microcontrollers, why going after shiny growth markets where returns are highly uncertain? Is it free cash flow or am I missing something? Thank you.
In fact, we have a very nice technology roadmap dedicated to this type of application. I will say that when a decision was taken to reallocate digital resources to other task, we are checking what we have in our hand. In fact, the technology we have, which has been developed internally, is giving us some strength in order to address this market. For me, we are not going to address the application processor and everything in the smartphone. We are just focusing on the front-end module, which is a very specific application.
Because we have today the process to do a switching low-noise amplifier, and also in complement with Marco power amplifier, we said, okay, there is a business opportunity where we can capitalize on what we have done in R&D, generating short-term revenue, which is the case today, and without any major improvement or major investment in terms of additional R&D. Is what I said, the business model is a COT, so we are just giving access to our process to our customers. We are giving them all the PDK, and they can start their design. Of course, we are planning to do some design. As we say, this is more, I will say, an opportunistic way and also a way to reuse, capitalize on what we have in our hand. This is good for 4G because we have been able to start 4G.
This is good for sub-6 5G, I think that it will be good for the also other version. We are capitalizing on advanced technology we have, I think it's key point for us.
Again, it's a question of leadership. Clearly, we totally share with you that on automotive and industrial, we will address this market. We are addressing this market with a leadership position, we will accelerate our leadership position with a broad range approach. Again, the broad range is either leveraging our internal technology and supply chain or selectively external one, like a seven nanometer FinFET with the TSMC. On personal electronic and communication infrastructure, we want to be leader, so it's a question of leadership, but being very selective. We know that when ST offers the best technology, offers the best product, we are winning high volume socket, which generate a strong cash from operation in order to support our value for our shareholder to leverage the sustainable and profitable growth.
Each time we offer this differentiation and this best technology, we are a winning company. That's the reason why on personal electronics and communication infrastructure, with the strong know we have on RF mixing technology, where ST offers the best technology, RF SOI, 65 nanometer SOI, BiCMOS 55 with SiGe. Sooner or later you will have RF embedded in a 28 FD-SOI or 18 FD-SOI. We have the best technology to offer differentiation to our customer. It's just a question of consistency with the strategy I introduce or Marco has described to you.
Any question left in the room?
Thank you. It's Achal from Credit Suisse. Just a question on your radar business. Can you give us some color as to what the size of that business is, in terms of revenues or in terms of market share? I guess you had a very strong position in 24 GHz a few years back. You were slow off the blocks for 77 GHz. What kind of progress you have made in the last 12, 18 months on 77 GHz specifically? Some anecdotes in terms of what the design momentum is in that business.
We describe, let's say, the total ADAS business for us is in the range of EUR 300 million 2018. I would say this is equally split between what we do in vision and what we do in radar. In radar, we are let's say market leader on 24 GHz. Unexpectedly, I have to say, this business is not declining. I would say we have multiple customer that are entering in production with new application at 24 GHz. This is, again, unexpectedly not declining. We still have a design win in 24 GHz. On 77 GHz, we have a very, let's say, compelling approach in the sense that we have two different way, let's say, to measure the need of the 77 GHz market.
One is with the pure bipolar, for just receiver and transmitter, and one is with the 20nm FD-SOI, when the receiver and transmitter is included in a baseband processor. We have the two. Of course, our market share in 77 GHz is not today comparable with what we have in 24 GHz, but you will rapidly see in the next few years our volumes growing because of the design award that we received.
Thank you. Maybe another question on the time of flight technology. I guess, we've seen a number of new smartphone launches, which are now adopting time of flight for front-facing autofocus. So far we've seen piecemeal approach from different suppliers, and now I think the market is moving more towards integrated approach. Can you again help us understand what your technology offering specifically is and how you are changing your offering from last year into this year? Thank you.
Thank you for the question. I think that we've been pretty consistent in the last 3 years to say that when we talk about 3D sensing, we talk about our offers of two technologies: time of flight and structural light. Clearly, we have to keep in mind which are, number one, the market application. Today, the most spread application is Face ID. Number two, which are also, let me say, the direction of the customers.
The good point that I can reply to your question is that we are able to master both solution, the structural light with global shutter, but also with our ability to integrate in a single module different component, while in time-of-flight, we have been working on our unique technologies, SPAD, since many years, and now we are also working on a fast photodiode to address both direct and indirect. As I said before to the question of Jerome, we don't want to stop at 38,000 pixel or 77,000 pixel. We want to go up to 1 megapixel for a simple reason, to detect the face or to enable killer app, you need the right resolution.
Question to Alex.
Yes, many thanks. Alex Duval from Goldman Sachs. A couple more on silicon carbide. I wondered if you could talk a bit about the bottlenecks to scaling it up further, and in particular, any challenges you might face ramping silicon carbide on eight-inch, and how you'll deal with those. Secondly, you obviously have ambitious targets in terms of getting to billion of revenues, but could you talk about any verticals other than automotive that are most exciting for silicon carbide for STMicro? Many thanks.
Concerning the migration to 200 millimeter wafers, today, bottleneck are not in the device fabrication, but is in the substrate fabrication. The migration from 4 inch to 6 inch took time, and we can foresee a similar effort based on what we know about our suppliers and what we are learning by working with Norstel, but this will be the major challenge. Concerning the device fabrication, we don't foresee nothing special. Of course, there will be a specific effort for what concern the epitaxy kind of technology. For the rest, I would say that most of the tooling is quite similar to what we use for the non-silicon carbide kind of technology. Concerning market outside automotive.
Yeah. Automotive clearly is the first market that we are approaching, we were approaching at the beginning with silicon carbide, mainly because of the mandates. There is the mandate, particularly 2021 in Europe, for which at the end, carmaker are obliged to enter in production with electrified vehicles. You saw from my presentation that, again, one of the most efficient way to implement electrified powertrain is because of silicon carbide. I have to say that in the industrial domain, the opportunities are even bigger than what we have in automotive, because every time you have a power conversion, a high energy, high voltage power conversion, silicon carbide is huge. I can tell you the strong push from our Chinese customer to convert their industrial application to silicon carbide is really huge.
That's why for us it's absolutely a priority to be vertically integrated to have all the substrate that we need and all the, let's say, manufacturing capability that we have to support this kind of market. I have to say again, today, from the evolution of the size, industrial is even higher than what we can see in automotive.
Another question right in the middle.
Thank you. Just a follow-up question with regards to the CapEx over the midterm. If I look at the second half of this year, you're planning to do about $5.5 billion in revenues on an annualized run rate of, say, $11 billion. On an outsourcing perspective, you plan to raise it by $1 billion. More or less, if I take your internal capacity plus outsourcing, you're close to the $12 billion, and yet you envision spending $1.1 billion-$1.5 billion over the next, say, three to four years. Just trying to understand in terms of the CapEx spending over the next three to four years, to what extent is that for adding more capacity? To what extent is it strategic initiatives? Shouldn't you be ending your midterm with a revenue run rate capacity significantly more than $12 billion? Thank you.
In term of CapEx. Yes. Of course, what we said is that in term of infrastructure, the company is in the position to sustain the $12 million. It's even true that for sure, we have a portion of our CapEx that is devoted to have what we call maintenance of our infrastructure, manufacturing infrastructure. This portion usually represent about something in the range of the 7% of the total level of revenues. This is needed also to fuel the R&D. There is a portion of our CapEx that is devoted to the fact that we need to improve our mix, to change our mix inside the fabs. This CapEx is another portion that is important because, of course, there is an evolution in term of technology, there is an evolution in term of needs that we need to follow. We need also to complement our capacity somewhere.
Let's say we have not fully balanced. You have not to forget that once we will reach the midterm for our model and midterm for our company, the game is not over. We will invest in these strategic initiatives. We need, let's say, in the future, in the next few years to fuel the company in order, let's say, to be ready for the future and to be ready for the next step of challenges. That's why we have indicated that this range in term of CapEx, this range of term of CapEx is encompassing substantially all these ingredients. We said that it will be between $1.1 billion to till $1.5 billion because, of course, if we need, let's say, start to
Facilitate and, let's say, put equipment inside our 300 millimeter. This will be a significant portion of our CapEx.
Maybe a last question, if any, before we break. This lady here.
Hi. Thanks for taking my question. Karina Satito from ARCT Research . Just thinking about your 5G RF, you've answered quite a few questions on that. You haven't really specified exactly the products that you're thinking for the front end. I know FD-SOI and legacy has been used for switch and LNAs and antennas, from my understanding, one of the big opportunities for 5G is 5G transceivers on SOI. Is that something you're looking into, or is it more of growing your legacy product portfolio on LNAs and switches? Thanks.
We are a provider of the product for the transceiver, RF front-end module. In fact, today, when we are talking about FD-SOI, we have far more opportunities in capitalizing on the strengths of this process, is what I show you this morning. In fact, right now, on the 28 FD-SOI, I think that the goal for us is to deploy further as many product as possible capitalizing on this process. For example, today on 28 FD-SOI, we have a lot of activities on, we say, satellite, and this type of application. This is where we can capitalize a lot at FD-SOI. We won't remain only with RF front-end module. We will have a lot of other activities.
To make it very simple, it is well known from a technical community that 28 FD-SOI is the best technology with RF native device. Much better than the bulk. Certainly, the further shrink and node, including on the FD-SOI, will present challenge to offer very good RF features. Clearly, the team of Claude, focusing on this RF mixing approach. Again, today, on 130 nanometer with RF SOI 65 and BiCMOS 55 is leveraging this specific feature of 28 FD-SOI, addressing the market, okay, we want to focus on. It is well known that FD-SOI 28 is really one of the best technology for RF devices.
With this, if there is no more question in the room, or at least you can go now, exit the room and ask your question to the management. Jean-Marc, if you would like to just conclude.
First, I would like to thank you for your attendance, for your attention, and all the question you ask to the management. Clearly, the main message today is that the team is really focused and engaged to deliver this year, this range of revenue between $9.45 and $9.85. In soft market condition, clearly, H1 is a pause in the growth of ST because this market condition. We strongly believe H2 will grow compare H1 at midpoint $1.25 billion. Two third on engaged program, one third because of market condition. Again, discussing with the stakeholder, clearly, it is a year of supply chain inventory adjustment. The end demand is not collapsing. We are addressing a secular growing market. We described to you where we believe that the market will grow. We have in our hand 10 macro programs which will contribute to our growth.
Key programs in automotive, in industrial, in personal electronic, in the near future, will be boosted by the 5G infrastructure deployment. It was important to share with you where we do believe in vehicle to X communication, in industrial IoT, in personal electronics, in communication infrastructure. ST will leverage the 5G deployment. This 5G deployment will arrive soon and will accelerate strongly. The last but not the least, we are strongly convinced that leveraging a market, the market we serve, growing average 4%-5%, we will deliver $12 billion revenue second half of 2021 or 2022, improving our operating margin high teens, keeping our operating expenses under control because the company has a strong dynamic to reallocate resources to improve our own efficiency. We are transforming ourself.
Last but not the least, we rely on 40,000 skilled and engaged people, which is really the core of our competitive advantage. Thank you very much.