Welcome everyone, also from my side. I'm glad we have this face-to-face Capital Markets Day today. This will allow us to give you insight in what has all changed in the last two years. Therefore, we thought it's a good moment to have another Capital Markets Day to give you an update where we are, what we're going, and give also more detail on all the expansions we have put forward. In order to do that, I've brought today a team of X-FAB. Let me walk you over the presenters that will be here today. First, there will be Jörg Doblaski, CTO. He will talk about our technology innovations and how this will drive growth.
Next, we have Luigi Di Capua, VP Product Marketing. He will explain how all these innovations are used in applications and how that drives our revenue. Next to that, we have Agnes Janke, Product Marketing Manager for Wide Bandgap, who will tell you about how X-FAB is disrupting the silicon carbide market. To close in the morning, we have Volker Herwig, VP of the Business Unit MEMS, who will explain you how we create value and how we grow our value within the MEMS business. In the afternoon, after lunch, we'll talk about operations. I'm glad to introduce you Damien Macq, who is new to X-FAB.
He joined two months ago as COO. I brought him on board to particularly look at the whole of our operations, all the expansions we are doing, because we're really getting into a new phase. Together with Dirk Drescher, VP Operations, they will explain all the expansions projects. Finally, to close up, Alba Morganti will, our CFO, will explain you how that all translates into the financials. We also brought today for you Sophie Le Goadec, as our VP Procurement. Because she can answer whatever questions you have about logistics and supply chain, because our supply chains over the past year have been stressed a lot, and with all these expansions, how that turns out, what are the challenges there?
He, Henrik Schroeder, VP Human Resources, who will explain how we deal with talent, because in the end, at X-FAB, it's all about talent and making sure we have the talent on board to support our growth. Now, let me talk to you how we delivered on our commitments since the last Capital Markets Day. A lot, that was about two years ago. A lot of things have happened since then. Let me walk. As X-FAB, we kind of transformed, and the business changed a lot, and we also are at a point of huge expansions ahead of us, and how that all will turn out. What did we explain to you in the 2021 Capital Markets Day?
We told you that we would strengthen our position as the number one foundry for automotive, and I will tell you what we did there. What we stay committed to the industrial market, that we will make medical the next automotive, and that we stay the number one silicon carbide foundry. Finally, from a financial perspective, we told you that we would grow through the semiconductor cycles at about 10% year-on-year, reaching $1 billion in 2026, with an EBITDA more than 27%. Let me talk about how we did on the delivery of that. First, in the automotive segment, since 2019, pre-COVID, we grew 17% year-on-year. In fact, strongly limited also by our capacities, and therefore, we're also expanding on all the sides. We can tick the box there.
The medical. We said we will make medical the next automotive. In fact, what do we mean with that? When we look back 30 years, automotive, there were not so many chips and semiconductors in cars, and we were really there as a foundry to participate and supply semiconductors for the automotive market. Over the years, more and more semiconductors came into cars, and we followed that whole journey. Well, we believe in medical electronics we're at the same stage. We are also at a point where semiconductors will strongly grow into medical health applications, and X-FAB is committed to be part of that. In the industrial sector, we booked 18%, compound annual growth rate, driven by a very broad range of customers.
Many of those are using our technologies that are particularly designed for automotive, but fits very well for industrial applications. This is driven by the overall electrification of our society. Then last, on the silicon carbide, there we have shown a 63% growth in quantity of wafers out year on year over the last four years, resulting in a 30% revenue growth. The revenue growth is not as spectacular as the number of wafers, because in the last couple of quarters, we have been using more and more consigned wafers, so we only charge for the value add. From a value creation perspective, this is fully intact. From a revenue perspective, it represented a growth of about 30%.
We are with all the investments we are doing, we believe we are very well positioned to continue on the same growth over the next two years. When we look at that from a financial perspective, looking at the first quarter, 2023, we showed 16% year-on-year growth for the whole of our business, and 28% on the automotive, industrial, medical, so our core markets, so largely exceeding our 10% goal. We reached $208 million for the first quarter, so well on track to reach the $1 billion soon. We achieved 28% EBITDA margin, so it's above the goal that we fixed out in the last capital markets day. Now let me explain what has changed in the last two years.
Everybody has heard, it's in the news everywhere, that there is a chip shortage, and this is persisting and continues. We, as X-FAB, we are in the midst of this. You hear governments in the US, Europe, also Japan, putting forward CHIPS Act plans to try to stimulate semiconductors and reshore the industry into Europe and US. Everybody noticed that semiconductors are strategic, semiconductors are difficult to make, so it's not an easy fix to reshore. There everyone puts forward long-term plans. As X-FAB, we're looking at those CHIPS Act plans, but it's too early to report on anything there. Let me talk about the structural shortage.
As X-FAB, we are in the 200 millimeter technology space. Those factories have been built typically in 1995, 2000 timeframe. They were built to make memories and logic. Since 2000, gradually, those technologies evolved into 300 millimeter and other new factories, leaving a lot of capacity on 200 millimeter empty. Some of those factories were closed, some were continued to operate, but at a relatively low level. In all those factories, people started putting in new technologies, analog mixed-signal technologies, just like we did. We acquired those factories and put automotive, industrial, and mixed-signal technologies in place. There was enough capacity over the last 20 years. Nobody invested in factories, in those 200 millimeter factories.
The underlying business in automotive, industrial, and so forth, more and more applications started using those technologies. All those applications started growing in volume. We had an underlying exponential growth. Suddenly in, around 2020, everybody noticed that there is, in fact, not enough capacity in the world to serve all this. This resulted in a significant chip shortage that is still continuing because this growth continues. The investments have started. Everyone is catching up with their investments, but it will take time, as Derek and Damien will explain what the timelines are to put in place additional capacities. Let me explain the transformation of X-FAB. What has changed in our business?
We have transformed our business since 2016, where the majority of our business was still consumer, computing, communications, we call it CCC. Towards today, our business is 89%, our core markets, automotive, industrial, medical. This gives us a much better visibility, much more longer-term projects, and more value add. Next to that, in the last year, we started a win-win discussion with our customers. Over the last 30 years, the way of doing business was that our customers gave us a forecast and they delivered purchase orders just in time. Those forecasts, they were not binding, so if business changes, then we have empty factories. When they show us growing forecasts and we invest, we might have an empty factories.
Now we change that model in a win-win, where we have binding agreements with our customers to allocate certain capacities, also growing capacities, because we are investing. As X-FAB, on our end, we have a guarantee to have minimal loading in our factories. This lowers the risk and gives us also more confidence to do the investments. Let me talk about a few parameters of those long-term agreements. They're typically structured over three years for the CMOS side of our business. They are typically five years for the silicon carbide business. They are all long-term agreements with a take-or-pay component, it's a commitment both on X-FAB's side and on the customer side.
There is a participation and a pre-financing of the investment. On the CMOS side, it's typically 15% of the overall contract that they pay as a prepayment. On the Silicon Carbide, it is roughly 75% of the associated expansions that we put in place for those capacities that are subscribed in the LTAs. We have about 70% of our capacity on 200 millimeter CMOS allocated under LTA agreements, with about ten largest customers. On the Silicon Carbide, we have three contracts signed up and more into negotiation. Let me talk about the way forward. Semiconductors are key in resolving challenges of our time.
One very important item is the decarbonization and the deal with the global warming, reducing CO2 emissions, reaching the zero emission in 2050, and there the components that we produce contribute in so supporting the, this change. The other challenge is the growing population, where that also requires more healthcare and better healthcare, and healthcare at a different scale. There also, electronics and semiconductors come in, and we're, as X-FAB, also committed to support that. One theme that you will hear today from several of our speakers is that X-FAB is very well positioned. On the left side, you see a study from McKinsey showing semiconductor revenue in 2021, and the outlook for 2030. You'll see there that automotive is the strongest growing segment within the semiconductors.
Second most important is industrial. As X-FAB, these are our two most important segments, we are very well positioned to enjoy this growing demand in those markets. Another important factor is another important tool for us, is the visibility we get through our prototyping. We have a very strong prototyping, it continues to grow over time. We do about 250 prototypings per quarter. These are each times, new contracts, new engagements with a customer on a specific product, giving us visibility on future production to grow. Also giving visibility on new application that will come in the next years. This is also a very good tool to predict future growth.
With all these elements, we also need to make sure that we put the right capacity in place to support this business, and we are expanding throughout the on all our sites around the world. Overall, $1 billion for expansion CapEx, and Damien and Dirk will tell you more about that this afternoon. With that, we have all the building blocks in place for our success. First, we have unique technologies to support in addressing important challenges for the end markets, automotive, industrial, medical. We have a very strong structural growth in those markets. We have a very diversified customer base. 90% of what we produce, we are the sole source. We constantly optimize and further improve our product mix towards higher value creation.
Over the last years, we have changed the contracts with our customers so that we have now a natural hedge, there is no risk on exchange rates anymore. Through our prototyping pipeline and the long-term agreements, we also have better visibility on the future. This also gives us confidence to do the investments, that will drive us acceleration of growth in the next years. With this, on the short term, we have set a goal that next year we will reach a $1 billion revenue, with an EBITDA in the range of 30%. Beyond that, with all the expansions that we're putting in place, we have the capacity to produce for about $1.5 billion in 2026, with an EBITDA that is in the range of 30%-35%.
Damien will explain about all the expansions, and, now I would like to pass the word to Jörg Doblaski, who will talk about all our innovations. Thank you.
Also from my side, welcome to all of you to our Investor Day here in Paris. I'm Jörg Doblaski, I'm the CTO in X-FAB. I have to say, I'm really proud to see our technologies so much in the focus of that, let's say, transition to a carbon-free society. This, let's say, a transition to electric vehicles, to renewable energies, and also to the digital health, as Rudi just mentioned in his presentation. During the next 20 minutes, I will give you some more detailed insight into our technology offering, and also explain how these technologies really drive the growth and the profitability for us today, and also in the future. Over the last 30 years, we have really focused on specialty technologies for specifically the markets, automotive, industrial, and medical.
This has been our strategy from the start, and it is still really the basis of our success. This timeline here shows just some of our main achievements in the time, and I selected those because they for sure demonstrate technology leadership, but also they are all still very relevant to our business today and in the future. Already, when we started as a pure-play foundry, we started to invest a lot in customer support and design support, and I think we can say we even invented that business model of a pure-play foundry with a strong support, enabling our customers to compete with IDMs, with integrated device manufacturers, which are often the main competitors in those markets.
We have continuously invested in that design support over, let's say, all this time, and we are still, I think, leading edge here. On our technology offering, we were an early mover into MEMS, with our MEMS foundry, and later on with MEMS integration on CMOS, and also 8-inch MEMS manufacturing. Volker will later on tell much more about that, how this has further evolved. Also, on wide bandgap, we were the first foundry to offer 6-inch silicon carbide manufacturing for our customers, and here Agnes will explain how this has further evolved and will evolve in future. On the CMOS, it was always our strategy to develop feature-rich platform technologies, really made for our core markets. We did that in 350 nanometers with XH035, a high-voltage CMOS platform.
We repeated it in 180 nanometer with XH018, both processes that are under high demand for our customers and that really drive volume in our fabs today. We also see a lot of demand in the future, backed up by long-term agreements, as Rudi mentioned. We were an early mover into SOI, so BCD on SOI on 180 nanometer with XT018, is now, I would say, our most popular process, where we see most of the NRE revenue by our customers who are adopting that node. We have continuously enhanced that platform over the years with more non-volatile memory features, more high-voltage features, as one is also shown here, the 375 volt module in this BCD on SOI platform, which is especially attractive for some of the medical applications like ultrasound.
This roadmap, or these technologies really drive our growth today and in the future, and, let's say, fill our fabs today and in the future. This success with this strategy continues now with XT011, with our new 110 nanometer BCD-on-SOI platform that has just been announced a couple of days ago, and which will see volume production later in this year. This is again, a platform that fits extremely well to the needs in our core markets, and it also fits very well to our manufacturing capabilities in X-FAB France. I believe this will be, let's say, the basis of our growth for the next 20 years. Before I share more details on this XT011 platform, I would like.
explain a bit more about the main factors of our success in terms of this technology enablement. What makes our offering here really unique for automotive, industrial, and medical? What is our value proposition here? For sure, it starts with the technologies, they need to fit to the product requirements and to the general constraints in these markets. I will spend a bit more time on that. It's also our ability to customize, which is very important, and which really can create very attractive business here. I mentioned customer support already, a very important pillar, and also very important to be able to offer a long lifetime here, a long process lifetime, since products in those markets usually have a long lifetime as well.
We have the ability, and we focus on innovating together with our customers to really create attractive solutions and stay ahead of the competition. Lastly, first time right. On top of this, we aim to do the things first time right, and enable customers to do first time right design, also a very important topic. I will dive a bit deeper in all of those pillars, starting with the technologies. I mentioned already, we focus on automotive, industrial, and medical, and for us, this means we do not just look at one feature, at one topic, like the digital node scaling, which drives other foundries to go quickly from one digital node to the next. You can see for us, we still have a volume production in one micron technologies, and we've just announced 110 nanometers, our latest node.
Instead of looking just at this one feature, we enable really a wide range of features that can be combined. Analog, high voltage, embedded flash memories, just to name a few that are relevant in the automotive space, and there are many more. For sure, you can see some of those or even many of those features also in the presentations of other foundries. What does really make our offering here unique? I just take the example of an automotive product. Let's say someone wants to design a smart motor driver product, they need digital, they need analog for sure, and they need high voltage to drive the motor, maybe for the window lift or whatever.
They need high temperature and the robust, let's say, reliability to fulfill the target specification of their customer. Already you see the competition is getting thinner. This product should be designed on a BCD-on-SOI process to really be able to fulfill the latch up requirements and noise immunity and things like that. You see, competition is getting much thinner. For sure, it should have automotive flash, because I want to program that and use it for many different car OEMs, really to penetrate the market well. To have all of this on one wafer, this is really what makes our offer unique in the CMOS and SOI space. Analog, high voltage, NVM on one wafer, supporting automotive requirements.
Besides that offering of CMOS and SOI, we have for sure our MEMS and heterogeneous integration solutions and also our wide bandgap solutions. All of this together creates the most feature-rich set of technologies, ready for the requirements of automotive, medical, and industrial, and ready for the application needs, as Luigi later on will highlight in his presentation. On top of these platform technologies, we enable a lot of customization, as I mentioned, and this really unlocks highly attractive business for us. I have one example here with me, which demonstrates this quite well, I think. On that picture, you see a cross-section through a wafer, showing or part of it, showing some of the metallization layers, so the top metallization and some of it below. Here, the very top layer is the passivation.
Each chip needs this, it seals the chip from the environment and it, let's say, is required to manage the stress during packaging. Usually, such a passivation would not be flat, but have really some landscape, some. It would be an uneven surface, following a bit of topology of the top metal layer, which is totally okay if you want to use a standard packaging, standard, let's say, bonding, but it is a no-go if there should be some post-processing. Being able to have this super flat passivation really opens the door to applications like heterogeneous integration, or, also very important if there needs to be like an interface layer to some biocompatible product.
For us, as a foundry, and for sure, there is a lot of process IP here included to come to this very flat layer, but it's for us, a relatively low effort to do this based on a platform technology. For our customers, however, this really changes the perspective, and it is a huge enabler for them, and really only makes some of their products possible. By the way, you will see some examples later on in Focus presentation, showing really nice examples of that. Most importantly, it changes really the perspective. Customers come to us and ask us: "Can you do this? Can you help me with that?" Such discussion really shifts the perspective towards the technical feature, and then price is maybe not the first topic one talks about.
Customer support, I already mentioned. What we frequently hear from customers via our design hotline, we get quite a good rating regarding our PDK. They say our design kits have a very good quality. We have maybe the largest offering of ready-to-use design IP for a foundry. This, together with our with the robust design flow that we support, I would say, enables really already a large number of products just to run to a tape-out and then to volume without much handholding from our side. We can, for sure, do much more than that and really extend our design support up to a level where we run concurrent product and process development together with our customers.
All of those pillars, all of those building blocks of the design support, enable us to engage with more than 400 active customers. 400 customers that are in production with us and that are also designing new products. More than 1,000 new tape-outs each year, new designs are created by those customers, which eventually translates then into 1,000 new products in production. For me, as an R&D guy, also quite remarkable, most of those products really implement our proven design IP, so with more than 2,000 of our memory blocks instantiated and used in those designs. This strong support layer is already a very important, let's say, factor to enable first-time right design, which is one of the other pillars. Even on top of all this, we do even more.
We have a unique toolbox, I would say, to tackle some of the challenges, especially in automotive. We can say we are the only foundry to have a really a holistic solution to deal with that. This creates such a huge benefit and a win-win for our customers, that it's really key for us here to invest here. For our customers, it is for sure, it reduces the risk for their development, it reduces effort. Even more important, if they avoid a design iteration, they can be fast at the market. They reduce their time to market and take advantage of being the first mover and win a bigger market share.
For us, for sure, it is as well a competitive advantage, I mean, I as the CTO, I like to have my guys working on new stuff and not fixing things, so it frees up resources. The most important thing is we are able to move business, move new technologies faster into volume. Our fabs will see the volume much earlier and benefit from, let's say, more higher attractive business earlier. As another item here to mention, the long lifetime support. It is typical for automotive, for medical, and industrial to have long qualification times. It takes a long time until the project is really ready for the volume production in the market, then the product is also running over a very long time.
It is absolutely key to be able to have the process stable over that long time frame. This chart here shows just, let's say, the case on X-FAB, the quantity of wafers per product over the product lifetime. What you can see here, there is really a large number of products that runs much longer than ten years. The average lifetime is more than 12 years, and there are many products that even run in our fabs in volume more than 20 years. I want to highlight this one dot here. This is a still a 1-micron product, 1-micron technology, running still in high volume now for more than 18 years in our fab in Texas, which is really remarkable for me.
The processes behind those, the wafer technologies, we need to keep stable over a time, more than 25 years, or a quarter of a century. We have to keep them stable, we have to keep them in good yield, it's much more than just developing only a process. We have to keep it stable, we have to know how to do that. Now, after this, coming back to XT011, let's see how, let's say, all these pillars, which create our success, how they link to the new platform. XT011 is a feature-rich 110 nanometer platform with embedded non-volatile memories and with high voltage. Really now the flagship in our, let's say, technology offering. It will be the backbone of a lot of customization.
Already now, while the process was still in development, we are engaging with one of our lead customers for an industrial application, where we enable a customer-specific solution on top of that platform. We have already our full feature design support available already during that time and also including a very rich set of design IP already at that early time frame. Certainly, the process will get all our attention in terms of full or long lifetime support, since this is key in those markets. It is really the first 110 nanometer BCD-on-SOI platform with high voltage, with NVM offered by a foundry. For us, it's really a key innovation as well.
Finally, our first time right toolbox for sure, we took care to have that available, even for this concurrent product process development that should take place, while the process was still in development. The best thing to prove all this is with a customer quote, so I'm really happy to show here the feedback from Melexis, one of our lead customers in XT011, provided by the program manager of the lead product, and you can read it in your printout. What I would like to highlight here is really XT011 is seen as the perfect match for the next generation of IC products, and also quite remarkable, a first-time right IC design was achieved with that platform.
For me, it's really amazing, and I'm really proud about also my team, who helped to enable this together with our customer in a really good teamwork. XT011 will be our growth driver for the future, and it will be the process of choice, especially for products which need more digital scaling, a larger digital part. To quantify this in terms of area reduction, I took here an example of a product family. Obviously, I cannot show really floor plans or something like that. This would disclose too much, we have here a product family, which we support over multiple generations and for one of our customer.
The move from XH018, our high-voltage CMOS platform, to XT018, created an area saving of 45%, thanks to the advantages of SOI. With XT011, with the next generation, the main focus for our customer was to keep those products as compatible as possible, so they had to be pin compatible for a fast adoption in the market. Even with those constraints, they were able to reduce the size by another 34%.
What we can say, depending on the, on the product, let's say, requirements and the product set up to 50% smaller area is possible compared to XT011 compared to XT018, which also translates into twice the number of dice per wafer, which is, I think, also an important factor in times where there is still a shortage in especially automotive. This process will really enable our customers to develop more attractive products with better performance and a smaller die size. XT011 is also a good example of how we innovate together with our customers, and for sure, we will not stop our journey, our innovation roadmap here. There will be enhancements of that platform.
We are already working on, so next generations of high voltage and more non-volatile memory offering to come. Also in other areas, we are constantly innovating to create new solutions for our customers in opto technologies, in silicon photonics, heterogeneous integration, and silicon carbide, to name just a few. We are continuously innovating to keep really our edge here, our technology leadership, and to enable future growth and profitability in our core markets. With this, I will now hand over to Luigi, who will explain how, let's say, these technologies link to the application needs in automotive, in medical, and in industrial. Thank you for your attention.
Good morning, and welcome to our review. I'm delighted to be here today. My name is Luigi Di Capua, VP, Product Marketing. Today, I will explain how we are going to meet our growth targets by serving the fastest-growing markets, namely automotive, medical, and industrial. For each of these market segments, I will explain our value creation, our growth perspective, and showcase some applications enabled by X-FAB. Let's start with our technology portfolio offering. Automotive has always been the foundation of our business, and the technologies that we have developed for automotive are well suited for the industrial and medical market as well. We can sell those technologies into industrial, and we are able to offer a high level of customization in medical. York just highlighted a unique technology portfolio covering CMOS, MEMS, and SiC.
As you can see on this slide, the CMOS and MEMS technologies address the free market segments, while SiC is used in automotive and industrial that requires solution for high-power applications. Let me now talk about our three core markets, let's start with the automotive. Our DNA is and has always been automotive. According to McKinsey, from 2021 to 2030, the automotive market is forecast to show a 14% CAGR, the automotive market is the fastest-growing market within the semiconductor industry. In Q1 2023, our automotive revenue grew by 35% year-on-year, accounted for 58% of X-FAB total revenue. Our revenue growth in automotive is driven by three main pillars: electrification, safety, and comfort and convenience. Let me now look at the semiconductors technology nodes used in automotive. These charts show the semiconductor technology nodes used in automotive.
The mature nodes from 110 nanometer and above are highlighted in dark blue colors, while the smaller geometries are in light blue colors. Smaller nodes are mainly used for digital applications, like microcontrollers and memories, while the mature nodes are mostly used for analog sensors or power applications. Our rich technology portfolio, ranging from 110 nanometer to one micrometer, can address more than 75% of the automotive wafer channel. Let's now look at automotive applications. These charts show the wide range of automotive applications enabled by X-FAB. It's all about sensing, actuating, and powering. Our CMOS and MEMS technologies can support a large range of sensors. The TPMS, so Tire Pressure Monitoring System, based on our CMOS and MEMS processes, is one of the most successful application from X-FAB, and we have more than 50% worldwide market share, a great value for X-FAB.
Our solution for actuators are mainly based on CMOS. You can see those application under the comfort and convenience. For instance, here we are talking about intelligent key system, climate control, seat control, adaptive cruise control. For powering solutions, our BCD-on-SOI and SI technologies are well suited to support the shift to electrification. Let's focus on this next. These days are all about EVs. According to HSBC, by 2035, the penetration of xEVs is expected to globally rise to more than 75%. The semiconductor content of an EV is much higher than the one in a traditional car. You can find up to 180 chips made by X-FAB in some premium EV cars. EV empowers. We have the right technology mixed with our BCD-on-SOI and SiC to support EV.
Several customers have selected EV for their inverter solutions and our BCD-on-SOI for the battery management systems. The battery is a key component in an EV and needs to be well managed. Here is a short video from X-FAB to explain what we mean by well managed. This video has been recorded by Ulrich, one of my team member.
The battery of an electrical vehicle is actually built up out of several thousand of individual small battery cell, and each of these battery cell has to be monitored by a so-called battery management system. Lithium-ion batteries, you probably know that from your smartphone, are very sensitive. Temperature, voltage, charge, all these parameters need to be controlled very tightly, and the better this is done, the longer the battery lasts and the further you can drive.
Our BCD on SOI is the best choice for battery management systems. It allows our customers to design smaller chips with better performance compared to BCD bulk. Let me now show you our average revenue per car. These charts show the evolution of our average revenue per car. From 2016 to 2022, the average revenue per car has more than doubled, from $2.31 to $4.78. We expect these figures to strongly increase, thanks to the conversion of X-FAB France to automotive and the stronger adoption of SiC in EV. Now that I have set out our value creation in automotive, let me talk about another very exciting market, the medical. As we say many times before, our ambition is to make medical the next automotive.
A strong growth in medical relies on three main pillars: personal medical devices, medical equipment, and lab-on-a-chip. Personal medical devices and medical equipments mainly rely on our CMOS technologies, while lab-on-a-chip requires the combinations of MEMS and CMOS. This chart illustrates the wide range of medical applications enabled by X-FAB. Let me illustrate some of them. Pacemakers and cochlear implants both stimulate nerves or muscles, and both uses our high-voltage CMOS technology with low-power memory. One of the customer has designed the world's smallest solution for medical-grade thermometer sensors, and this solution can be integrated into a smartwatch. Volker will show it soon. For medical equipment, our BCD-on-SOI technologies enable state-of-the-art solutions for ultrasound imaging by providing excellent noise and high isolation.... Let me then zoom into one of the very exciting application, the lab-on-a-chip.
At X-FAB, we have developed a unique expertise in lab-on-a-chip devices. What is a lab-on-a-chip devices? The lab-on-a-chip devices consist of integration of electronics, biosensors, and microfluidics. Those devices can be used, for instance, for early detection of sepsis, tumor cells, for drug development, or for doing DNA sequencing. These tests are fast, accurate, and non-invasive. Volker will showcase some of these applications soon. In 2022, the revenue for lab-on-a-chips at X-FAB accounted for 32% of X-FAB medical revenue, and it has grown by 33% per year on average since 2018. Let's move to the industrial market. The industrial market is a very fragmented market and filled with a large variety of applications. These markets need solutions for going green and going smart.
As for automotive, our CMOS and MEMS technologies can support a large range of sensors, and our high voltage technologies are mainly used for power conversions. Let me now highlight one of the major challenges for the industrial market, the renewable energy. The energy crisis is a catalyst for the energy transformation. The EU Commission has set a target to increase the share of renewable energies from 22% in 2019 to 42% in 2030. The transition to renewables is driving the adoption of SiC, which is the best suited technology. According to Yole, from 2021 to 2026, SiC in renewables is forecast to show a 19% CAGR. Agnes will talk soon about our SiC offering. Let me summarize: we confirm our strategy to focus on the three fastest markets, automotive, medical, and industrial.
We stay in the sweet spot for automotive semiconductors, we continue to drive innovation in medical, and we contribute going green and smart in industrial. With our recognized expertise in analog, high voltage, CMOS, MEMS, and wide bandgap, we will play a major role to support the future of mobility, healthcare, and energy. With this, we are perfectly positioned to meet our growth targets. I would like now to hand over to Agnes, product marketing manager for SiC. Thank you very much for your attention.
Hello, everybody. Thank you, Luigi. My name is Agnes Janke. I'm the product marketing manager for silicon carbide and gallium nitride, and I want to give you some insights about X-FAB silicon carbide market. I want to explain you why SiC is such a hot topic today, and which role X-FAB plays in this market, and how we add value with silicon carbide for our customers and for us. You heard a lot about silicon carbide, but what is it? Silicon carbide is a semiconductor material that is used in power electronic devices, such as diodes, MOSFETs, or power modules. Those are used in systems that convert power, so either from the direct current to the alternating current, like in solar inverters, or from different voltages, like for example, from the battery of an electric vehicle down to the lower voltages of the car systems.
The domain for Silicon Carbide are high voltages, and we see the shift already today to devices with 1,200 V and above. The shift is driven by the huge demand from automotive OEMs and industrial customers, changing, for example, to 800 V applications. Today, the market is dominated by the integrated device manufacturers like STM, Infineon, or onsemi. They are serving over 90% of the market, but with its growing market and also the growing technical requirements, more and more of the fabless companies are gaining market share, and drive the Silicon Carbide adoption with their innovative products, and those are our customers. As the demand for higher performance products increases, this will make the market for Silicon Carbide to quadruple in 4 years, in both in wafer revenues and also in wafer shipments, producing a growth rate of around 25%.
Today, silicon carbide is present in only 8% of all power electronic devices. By 2026, it will increase to 20% of all devices and increase further to a shipping of around two million wafers in 2026. You might ask, "Only one-fifth of all power electronics is based on silicon carbide? Why is the transition so long?" Let's have a look at some details. As I already mentioned, the domain of silicon carbide is higher voltages. In the past, silicon carbide came from a legacy of 650 volt devices, trying to replace the silicon-based IGBTs that were struggling to achieve that voltage range.
They managed, and today, the silicon at the 650 volt range is particularly a battlefield for power electronics, as it can be served by the silicon-based IGBTs, superjunction MOSFETs, and also the wide bandgap devices, gallium nitride and silicon carbide. Here, price pressure is the highest. Silicon carbide is moving towards higher voltages of 1,200 volt and above. Those devices show a higher performance gain in the system, thanks to its lower losses. The cost added for those, compared to the silicon-based counterparts, is better. The devices, and although they also will remain more expensive, the silicon carbide will remain and is more expensive than silicon-based devices, the benefit is increasing with higher voltages. What changes significantly and where really the gain is in the system cost.
With higher rated devices, you can use a different topology, you can take advantage of smaller or passive components, cables, other battery sizes. You can change the passive cooling systems and therefore reduce the cost of the overall system. For the application, that means that the products, that all the products using silicon carbide, will improve. For example, electric vehicles can either save battery cost or, if you remain with the battery, extend the range of the car. Charging stations will charge faster at higher power. Solar or wind renewables energies will create more energy out of the same system. With regards to railways, silicon carbide enables electric trains with lower weight and less noise. Why does it take so long for silicon carbide to gain momentum? Manufacturing of the wafers of the substrate is hard and expensive.
Processing of the wafers is complex, and it needs specific equipment. The system implementation is a challenge. You cannot simply exchange the silicon carbide transistor within the system. The whole system needs to be redesigned to gain the system benefits. The benefits of silicon carbide start to overcome its growing pains, and silicon carbide is starting to move from niche to mainstream. What role does X-FAB play? How do we contribute to this growing market? Let me show you in the next slide. We are the number one pure-play foundry for silicon carbide. We are offering access to silicon carbide markets to players that would not have had the possibility to participate with their own products. In a world of customized products, we enable them to create their unique product, which is superior to the more generalized white label products of the IDM.
Our customers' products exceed IDM products and technical characteristics like RDS(on) or voltage classes. They achieve really outstanding device performance with best-in-class specifications, and this with industry-leading high yields, both on wafer and on device level. The foundation for this lies in our technology expertise. Since over ten years, we developed silicon carbide technologies with and for our customers. We were early adapters of silicon carbide. In 2013, X-FAB has been the first pure-play foundry to industrialize the silicon carbide processing. Successively, we added more and more silicon carbide-specific equipment into our 6-inch line in wafer in Lubbock, Texas. We are converting more and more of its capacity to silicon carbide. In 2020, we added our first epitaxy tool, as we saw the need to also optimize that part of the supply chain.
The epitaxial layer and the silicon carbide process are highly dependent on each other, a lot of synergies can arise when you optimize both together, and which improves on the end performance of the product. Although we gain and also a lot of scientific knowledge, our customers really come to us with their IP. They can always be sure that their IP is protected. X-FAB is not a product company, we will never sell the products, this is why we are not a competition to our customers, and this is very important to them, as they might, for example, experience in other non-pure-play foundries. Our technology and experience make, in the end, the better product. Better efficiency and planning, prototyping and production, Jörg explained it, make a faster time to market.
We guarantee their supply by securing long-term agreements, not only with our customers, but also with the substrate suppliers. All this considered together, gives our customers a very good cost efficiency, and us, stability and predictability for the demand. A win-win situation for our clients and X-FAB. We are powering innovation. We create a real value add, and this is why customers are coming to us. X-FAB is in the sweet spot of the supply chain, and we choose to be there, to focus on where we can create the most value with our expertise and leveraging of our existing CMOS fab. That's why we focus on three pillars within the supply chain: the process technology and IP development, the silicon wafer epitaxy, and the most critical part, the wafer processing.
We believe that the Silicon Carbide wafer manufacturing, the substrate manufacturing, including the boule growing, ingot slicing and substrate grinding, this will become a commodity dominated by the Asian players for the non-captive market. Although, of course, knowledge is required to grow a wafer, a good wafer quality, all wafers are similar and quality is highly scalable. It depends heavily on energy cost and available capacity, and the prices for the 6-inch wafers are expected to drop soon. Upstream of the supply chain are the device and system manufacturers, our customers. They are responsible for the device design, the qualification, and the packaging before the transistors or diodes can become an electric system. In contrast to the separated supply chain, there are the fully vertically integrated IDMs, integrated device manufacturers, that cover the whole supply chain from substrate manufacturing to system.
This requires tremendous investments, as read in the press, maybe like the billion-dollar investments announced on Infineon, Bosch or Wolfspeed. X-FAB invested constantly over time in our silicon carbide capacity expansions and technology-specific equipment. With our 6-inch wafer fab in the U.S., we do not need to construct a new building, we do not need to construct a new line. We can leverage from our existing 6-inch fab, and we convert that simply to silicon carbide capacity. By adding a specific equipment, like an implant or a furnace, we can increase the silicon carbide capacity significantly. By arranging long-term agreements, we have a clear visibility of our future roadmap and capacity needs. Where does this bring us? Our success is based on our customers. We grow with our customers, and they grow with us.
We are now serving around 30 customers worldwide and are running over 50 silicon carbide process families or products. We secured around 75% of our capacity expansion plans with LTA customers, around 25% is reserved for potential new customers and also the development of our current customers, increasing the number of customers to around 45 in 2026. Last year, our silicon carbide revenues reached $55 million. We nearly doubled our silicon carbide wafer shipments year-over-year, and with a significant share of those wafers also based on our in-house epitaxy. In 2026, we target to increase the revenue from silicon carbide to around $300 million-$350 million, depending on the share of the products and consigned wafers. This will be roughly 20% of X-FAB's overall revenue.
We are investing $200 million into our silicon carbide capacity expansions to support a capacity build-up of around 210K wafer starts per year in 2026. In comparison to the overall wafer shipments and 6-inch equivalents as reported or forecasted by Yole, this brings us to a market share of around 7% today, and we target to rise this share within the highly growing market to around 11% in 2026. All this makes us the number one pure-play foundry for silicon carbide. Let me summarize the main messages to you. The silicon carbide market is growing. SiC will be in around 20% of all power electronic devices in 2026. Silicon carbide is going from niche to mainstream application, gaining market share and higher power applications.
X-FAB enables fabless companies to compete in this market with highly competitive products and at very good yields. We are in the sweet spot of the Silicon Carbide supply chain. Our focus on silicon wafer, epitaxy and chip processing allows for very pinpointed investments. We were early adopters. This is why we are successful now. We are growing constantly, and we will grow further. We are targeting around 20% of our revenue contribution in 2026 to come from Silicon Carbide. As you see, we are well on track to maintain our number one position as number one pure-play foundry for Silicon Carbide. With this overview, I hope I could give you a short introduction to Silicon Carbide, X-FAB's business, and I hand over now to Volker, who will present you our MEMS business. Thank you.
I have a wafer here. I hope you can see it. There are about 5,000+ dies on it, and this wafer is coming from a special place, or a place we care about. It's coming from a fab. In this fab, there is a large team, highly educated, highly motivated, well-trained, and they work on a super expensive tool set in a super expensive clean room, and they try not to waste time and energy to capture all the economic benefits of high-volume semiconductor manufacturing. You may have noticed the wording, "not to waste time and energy." This is the definition of efficiency. In the semiconductor space, everything is about efficiency. Now it comes.
Very soon after this wafer is leaving a fab or X-FAB, this wafer will be diced, will be separated into 5,000 plus individual pieces, and those individual pieces will be processed one by one on the way to a final system. Now you may start to wonder or you ask yourself: Is this efficient? In order to tackle this issue, or if you could tackle this issue, you perhaps could capture value. The definition of value here would be for a customer, either a cost down, performance envelope increase, system size reduction, or time to market improvement. For X-FAB, one of the metrics is for sure, what type of multiple on top of a CMOS wafer we can achieve, whether it's 1.2, 1.5, 2 or 3x.
In other words, by using wafer-level system integration, adding massive value add on top of an already quite valuable X-FAB wafer. What is required to do so? First of all, you need a wafer fab. Sure. It turns out to be that the very same process steps you need to have to do a pressure sensor, a thermopile, a gas sensor, a MEMS microphone, are required to wafer-level system integration, and that's where MEMS comes in. Welcome. My name is Volker Herbig. I'm running since 2016, the business unit, MEMS. Having outlined my agenda, it's obvious this puts us on a journey. This journey started in 1997, when X-FAB acquired the first MEMS toolset to do pressure sensor. Actually, one of those pressure sensors released in the late 1990s is still in production as of today.
MEMS business is good for us. It's really good for us. We have captured a 25% worldwide market share in automotive pressure sensors. We do a ton of inertial sensors, thermopiles, gas sensors, you name it. About eight years ago, we spotted this opportunity, I mentioned already, to reuse our MEMS technologies to add value on top of a CMOS wafer. In doing so, we had to expand on our capabilities, and we added some more, like we heard about it already, ultra-flat passivation, noble metal, through passivation via, through-silicon via, just to name a few. Based on those capabilities, we are able to acquire a number of projects, and these projects today are responsible for our massive growth we do experience right now. That's not where our journey is going to end, actually.
Since a while, technologies are merging, and customers are asking us: "Can you bring things together which couldn't be integrated in the past?" What am I talking about? Are so-called chiplet technologies, where you print, transfer print, either a sensor, a CMOS sensor, or a III-V material like indium phosphide, gallium nitride, gallium arsenide, or an LED on top of a CMOS wafer, X-FAB wafer, on top of a panel and top of a flexible substrate. Applications demanding this type of technologies, this stuff, quantum computing in the photonics space. Next-generation quantum computing will rely on this, exactly on this type of technology. Displays, advanced automotive and medical sensors rely on this technology and will hit... First products from X-FAB will hit this market space in the 2024, 2025 timeframe. We heard already from Luigi, X-FAB is working in a semiconductor market sweet spot.
What does it mean for MEMS? Frequently, we are getting approach from customers working on those cutting-edge application you can reach here, sensors, microfluidic, lab-on-chip, wafer-level advanced packaging. Quite often they work with the RTO community. These are the Fraunhofer, LETI, IMEC, and they mature products in this application space with them. They call us and ask us: "Can you bring those applications, those products for those applications into volume production?" Volume production means for us, we take over responsibility, we take ownership to deliver wafers in quantity, in quality, at a agreed moment in time, at a agreed price, almost forever. Yeah, that's what we heard. And we feel exactly this capability for this application space puts us in a quite unique position. Why is that? What you need is, you need a 200 millimeter centric fab ecosystem.
You need a CMOS fab, tightly integrated with a non-CMOS fab. From a business perspective, it's quite diverse. You have medium-sized products or projects. What I try to tell you is, you need a foundry approach to really make a good business out of that. Now, let's take a look at the solution perspective here. I mean, the key to enable this, wafer-level system integration is customization. Unlike on the CMOS side, where customers design in a frozen and fixed process, here we allow, we do entertain either process modifications or we even develop a new process to tailor this process to the customer design needs. In doing so, the customer has a significant advantage. He can push the performance envelope. He really is getting an edge by doing that. However, for us, this is a big deal. Why is that?
Doing process development takes up a lot of resources and is adding complexity in the fab later on when you do volume manufacturing. Also, from a scalability perspective, we can run just a limited amount of R&D projects at any given time because it's so resource intense, therefore, we are extremely selective. Every project has to fit, and we do expect a larger-than-average return on those projects and massive commitments already in the R&D and the production phase from our customers. This slide is answering a question, I raised in my introduction. I was asking, what is the multiple, what is the value add we are able to capture here?
What it tells is that if you take a $1,000 CMOS wafer, by applying a few but high-value MEMS CMOS layers, you can turn this wafer into $4,000 heterogeneous integration wafer. Let's take a look, what is this going to do to our top line? In the time frame from 2019 to 2025, on the back of a booming MEMS business, this is really good for us. We have turned on just 4 system integration products since 2019. We expect wafer volume more than double. The same is true for the ASP, for the average selling price. This is happening on the back of the value-add business I just described.
All in all, do the math, we are ending up with a 4.4 ratio or ex-revenue grow. To put a number to it for you, it's about $160 million-$170 million revenue, what we are target in this time frame. We have covered the business model. We looked at the value creation. We looked at the top-line numbers. The exciting part for me is starting of this presentation. I give you some insights into the applications. You can determine for yourself what is the impact of those applications for X-FAB and really also for yourself. I mean, they perhaps make a difference in your life. These are all applications done by my team.
They are either in volume production or in start of production, or one or two examples are still in a, in a R&D phase, but I, I think it's worthwhile to show those. Let's get started here. On your sensor suite of your personal medical device, you wonder what this is going to be. What I'm referring to, it's a phone, it's a ring, it's a, it's a smartwatch, or it's a earbud. There is a missing capability. What is missing? It's body core temperature measurement or medical-grade temperature measurement, and this is what this is all about. We have developed this with our customers, and this wouldn't be impossible, or it wouldn't be possible without wafer-level systems integration to reach the performance, the required performance and the form factor.
This is going to come on more products, whether it's phone, earbuds, rings, whatever. Those sensors will emerge. This is about putting 26,000 LEDs on top instead of 64, 128 discrete LEDs, putting 26,000 LEDs on top of an X-FAB CMOS wafer. Actually, I wouldn't call for an automotive headlight application, and I wouldn't call this any more a headlight. This is a projector, and this projector is used as an automotive headlight. What does it do? You can project signs on the street, but also you can really bright illuminate the scene without blinding the incoming traffic. Expect cars featuring this application, this technology, later this year on the road in Europe. This is about developing for the industrial infrastructure domain, really, or the next generation ultra-high brightness in and outdoor displays.
What do I mean with ultra-brightness? Those will be a magnitude of order brighter than existing displays. You can look at those in bright sunshine, and you have a really vivid color impression. This is a very personal one, actually. I mean, for people in my age group. This is about a sensor. This sensor, it's a CMOS sensor, is printed on a flexible substrate. This flexible substrate is wrapped around a guide wire together with a clot. If you have a stroke, you have a clot, with a clot removal toolset, the emphasis is on toolset. This is introduced in your artery system, in your blood vessel system, and this may go, depending where the clot is, it may go up to your brain.
That means you're ending up with this sensor in the brain. What is the sensor going to do? I tell you. The doctor who runs the surgery needs to make a decision, and because there are two different clot types, one is red, one is white, they have a different composition, and they require a different removal method. In the past, there was a 50% chance that the doctor was right and a 50% chance he was wrong, or based on experience, perhaps the ratio was a little bit different. Now, he can make a reading and can really determine what type of clot he sees, he has, and can do the right removal method. I mean, time is of essence if you have a stroke. This is helping a lot.
This is an application where the wafer level system integration is really helping. Lab-on-a-chip is really showing how powerful can this be. This is for high-speed DNA sequencing. It's being used. Prime application space is population genomics, where you have to sequence a genome from 1,000, 100,000s, even millions of people. This is being done with an X-FAB flow cell, which is a disposable. This is great business for us. Large chip, disposable. Yeah? Also, it's. Think about how this has been done 20 years ago. I mean, this has been. There was this Human Genome Project, perhaps some of you remember. It took 13 years to do the first genome, and it was costing just $300 million. The overall program was $3 billion. Now, this is that fast. It can be done, and this is on record.
I mean, this is, you can look this up. You can Google this gentleman here. They had to sequence his genome, and they did it in five hours and two minutes, his full genome. It's on record with the Guinness Book of Records. It's on record with the Guinness Book. This is an example where purpose and business really go hand in hand. Last example on the system integration side of things, internal nickname for this application is electronic mice. What is it? We put clusters of cells, or we put organoids on it. Organoids, these are clusters, three-dimensional clusters of stem cells, developed either into brain cells, heart cells, or retina cells. Then they do mass drug screening on this stuff or some research. Very interesting stuff.
Now, I started out my presentation with MEMS. I will close out my presentation with MEMS. This is an application, one of our customers, Melexis, released it, the most accurate pressure sensor ever made. This is the last application I wanna show. It relates also to you. I will explain in a second. It's about a spray head, and you now starting to wonder how exciting a spray head can be. There are two type of applications. One is basically for those inhalers used for treatments, drug treatments. If you use an inhaler like that, you have to solve two problems. You need to adjust the droplet size to make sure that your drug is ending up at your desired landing spot.
This is either the throat, upper lung, middle lung, and that's what this technology is able to do. You can really dial in, not on the fly, we have to do it per inhaler, the droplet size and the droplet distribution. But then you have to solve a other problem as well. The energy you nebulize the drug, the energy intake to do this nebulization is also quite critical. If you nebulize lipid-based nanoparticles, this is a, this is a critical element, and this technology is reducing the energy intake for nebulization by a factor of 10, by a magnitude of order. It's called mild nebulization technique. What is the spillover in the beauty phase?
Now I need to talk about a fragrance, a perfume, I have to admit, I'm perhaps not the perfect guy for that one, but I now need to talk about small droplets, not generating a wet surface and it's a different experience. The spray is supposed to be longer than with a normal perfume. To be honest, I really don't feel very comfortable. The purpose of this is, you have a chance to check it out by yourself. Over there is a goodie bag when you leave this event here. There is a pre-production sample of this perfume. It's from our company, it's the head is from Medspray and X-FAB. The flask and the perfume is from a company called Aptar.
I hope you have a great experience with that, and you remember this event, my presentation. That's all what I have. Thank you very much for your attention.
Thank you, Uta. Welcome to Kuching. Today, I'm very glad to have you here in X-FAB Sarawak. Well, with the continuous strong demand for our product, our fab is full, and we are running at full speed. That is why we are investing heavily to expand our capacity and to improve our productivity. Today, my team and I are happy to show you how we improve our fab. Let's start with my colleague, Cynthia.
The products we manufacture are highly complex. To complete a single wafer, it takes about four months and more than 600 process step. As we speak, there are more than 80,000 wafers running through our fab. These wafers are being produced for many different customer, in different quantities, for many different applications, and using all kinds of technology. The whole fab has more than 400 equipments. It is a challenging task to make this work every single day, 24/7, and yet we are always looking for ways to do even better. That's why we introduced Scheduler, a system that helps our operator to set the right priorities by analyzing the queue. We are proud that our site has been chosen as a pilot fab. With Scheduler in place, we make our wafer processing much smarter, which will help to increase our capacity.
We can't wait for it to go online at all sites. Pushing our capacity limits, our AI team made another big step. Let's take a look.
Even the best machine is not 100% accurate. In our line of work, our operators have to analyze and review thousands of microscopic images every day, including defect detection on our wafers. What takes a trained employee a day of tiring work, can be done by advanced computing within a split second. That's why we developed a groundbreaking technology we call AIR. AIR is an AI-based software, which improves the accuracy of in-line defect classification by 20%. This changes the way we work tremendously, and we are gaining insights that help improve our machines and workflows.
Our machines are running 24 hours every day, all year long, and they are consuming a lot of electricity. To ensure a stable power supply, we installed four new power units with a total power capacity of 4.8 MW. We are now able to protect our most critical manufacturing facilities, even when there are external power dips. The key system of our fab can continue running without the risk of losing wafers. With this, our production lines are more resilient than ever before.
What we do in X-FAB Sarawak is a result of about 1,400 committed people working together as a team. For them, we are dedicated to creating a positive and innovative working environment. Our result pays off. Our annual attrition rate as of last year, was less than 5% as compared to 15%, which is the national average. The employee satisfaction rate, according to the latest survey, is about 80%, and for the last eight years, we received ten HR award. Six of them are for Best Employer or Best Place to Work.
To serve the continuous strong demand for our product, we are also expanding in new manufacturing capability. We invest heavily in building a new clean room to increase our capacity to 40,000 wafer start per month by year 2025. With this additional clean room space, we can further expand our capacity for another 10,000 wafer start per month in the future. With all these projects in place, we are more productive than ever before. Now, back to Paris. Have a nice day.
In the past, now we are moving with a new X-FAB process to a setup where you need in the range of 35 layers to produce a typical process, as described by Jörg Doblaski early this morning, to be able to implement the non-volatile elements, to be able to involve, to develop the flash on the chip. There is a lot of additional mask that are coming on top of this more complex process. Together with this, we have also some process, as described by Volker Herbig this morning, that are even more complex. One of the example that you see on the slide, for one of our DNA chip, you need up to 54 mask layer to produce one wafer of this device.
This increased manufacturing complexity implies more process steps for each silicon wafer we produce. From what you saw earlier on the video, what we define as the process steps is every time you see a wafer in the video of Sarawak moving to an equipment, it is a process steps. You see the number of process steps that is growing quite significantly from 500 to 1,000, depending on the flavor that we are producing. This manufacturing complexity create additional value we provide to our customer and attractive price multiplier for every wafer we produce. It's not only about counting the wafer, it's also understanding what is the flavor of the wafer that we are selling. I think this morning, Volker presented this factor 4 multiplier.
Here, we say between two and five, depending on the process that we are considering. Being close to the center of Paris, the center of France, I think it's very relevant to start our virtual fab tour by the site nearby in Corbeil. I have to say, the CEO of the site in Corbeil is in the room, Justine Wasner. Justine, I'm talking under your control. If I miss something, don't hesitate to interrupt me. The transformation plan on the site is already in full motion. In fact, as a reminder, the activity on this site used to be concentrated on what we used to call CCC products. Now we are migrating more and more steadily and faster towards X-FAB process.
Having secured the transfer of our X-FAB 180nm technology on the site, we have now new process enablement. The 110nm is produced out of our site in Corbeil, and the first wafers are coming out of the lines, not yet in production, but already enough to release the process at this point. It is important to notice that the copper metallization of our 110nm leverage there, the asset utilization that we have on the site. Basically, this will create more capacity and more value for X-FAB as well. The availability of the technology was announced earlier. On June the first, there was a press release. There were more details this morning on what this process can bring and can mean for our customer.
I would say the main element relies to the number of chips that we can get per wafer. In terms of capacity, this is a significant change and a game changer for X-FAB. Associated with the site transformation, we execute on automation. We have to upgrade the pack of equipment. We have to standardize the whole equipment that we have there. All in all, we are still $170 million that we will continue to invest between now and 2026 to implement the phase of our expansion plan. We will have, at the end of this plan, an addition of 14,000 wafer start per month coming out of the site.
I would like to zoom in a bit more on the transformation we are talking about. Very often a figure says much more than any word. You see on the graph the illustration of the revenue that are generated out of Corbeil. In light blue, you have the revenue generated from CCC, the old legacy process that we had in Corbeil. In dark blue, you have the revenue generated from the new ITSfab technology, so the umbrella of all new ITSfab technologies that we transfer there. This transformation plan was already presented by Jörg Doblaski in 2021 in a similar meeting here, or it was virtual, I think. It was virtual. Since then, and driven by the mega trends already presented by Rudi and Luigi Di Capua, in fact, we had to accelerate the transformation.
If you compare the plan that we had in 2021 with the plan that we have now, we have much more reliance on the ITSfab technologies. Today, you see the arrow pointing at today, more than 80% of our revenue out of Corbeil are already made out of material that is built on ITSfab processes. This huge transformation was a significant challenge for our teams in Corbeil, even more so in a supply and resource-constrained environment. Now that the ramp-up of our first technology is confirmed, as you can see on the picture, we can see a very positive future for the fab and for ITSfab overall. I think Rudi mentioned this morning the contribution of Corbeil in the automotive and in global revenue of the company.
To translate that in more numbers, you see that on the picture on the left, the number of wafers produced every month in Q1 2023, and you see the direction that we are taking for Q4 2026. You see that we add capacity on our 180 nanometer node, but in even more relevant and even more important, we really multiply substantially the capacity that we have in our BCD-on-SOI processes. Here, we are summing up the contribution from the 180 nanometer and the 110 nanometer together. We produce already some of the 180 nanometer HT BCD-on-SOI in Corbeil today, but you see that by Q4 2026, we will produce 13 times more wafers out of the fab. This is a significant boost for that technology.
Let's now move from France to Malaysia. We were in Malaysia in a movie with a presentation from BC, the local CEO of the fab. You have seen from the movie the expansion plan that is ongoing. We built a new construction over there, this is combined also with a completely reshaping of the clean room space that we have right now. It's not only about new construction, it's also making sure that the existing space is used in a more optimal way, in a very efficient way. All in all, we will have more than 6,000 square meter of clean room space available over there, this will allow us to increase the capacity from 30,000 wafer a month today to 40,000 wafer a month by Q4 2026.
It's also important to notice that the building that we have and the floor space that we have there will allow us in the future, not yet part of this investment plan today, but it will allow us in the future to move up to 50,000 wafer start a month on the site. It's substantial in terms of space that we gain on the site and in terms of number of wafers that we are adding compared to today. Like in all the other place, we continue to secure the automation of the site. That is a constant effort on this side. That is a constant effort also to make sure that what we do on one site is reused and in synergy with other sites.
I think Dirk will also repeat that term, and it was a strong focus for the team of operation in the past and also in the future, is to make sure that we really can leverage and exploit the material that we have, the assets that we have there in the most efficient way. In total, we will invest close to $550 million on the site. It's quite important over the period 2023 to 2025. If you combine the extra wafers that we produce there with the additional manufacturing complexity I was referring to before, we will have an increase of 50% of our mask layer process per day in Sarawak. The level of activity of the site will increase quite substantially.
It doesn't mean we'll have people moving wafers, because automation allows us to do this in a much more efficient way. Translated again in more numbers, this slide show the growth in terms of wafer delivery with a zoom of the two flavors of the 180 nanometer nodes. Again, notice here the significant expansion that we have in our BCD-on-SOI process that is really successful, gaining traction with all our with the full customer base that we have. Very important to mention about this expansion, during this expansion, we will make sure that, one, we maintain the overall execution excellence that we achieve in Sarawak. Right now, Sarawak is really on the benchmark side and considered as an excellent fab with really good productivity.
Also very important, we will make sure we leverage the best practice that are developed there and deploy them to the other fabs. These two things have to happen on top of the capacity expansion.
X-FAB Dresden is our smallest 200 millimeter site. Nevertheless, it is extremely important to deliver our specific very successful 350 nanometer technology to our customers for automotive, but also to support the growth in MEMS, as Volker explained. In Dresden, we have been really busy in the recent two to three years to add additional 2,000 wafer starts per month. We are right now performing at 10,000 wafers per month. Right now, we continue this growth plan to put another 1,000 wafers on top. In order to do so, we optimize our existing cleanroom. We build new cleanroom to make room for the new equipment. We invest in new equipment, overall $30 million. We also focus on automation to increase the efficiency of the installed base capacity.
It's just an important note, imagine we take these 1,000 additional wafers, explained by Volker, valued on the CMOS side for $1,000, and we can turn them into a MEMS solution for our customers and create more value by a factor of 4, yeah. This is exactly what this means here. Next slide. X-FAB Texas, another extremely exciting location where we massively grow. X-FAB Texas is our growth engine in silicon carbide, as explained by Agnes in the morning. Specifically, this fab, which has been operating CMOS for many, many years before on 6-inch wafers, is the best place for us for a conversion towards silicon carbide. Why? Because the installed base equipment running on 6-inch is able to convert towards silicon carbide at low costs. This is the ideal place to run silicon carbide.
What we do right now, we run massively, we are massively busy to convert fab space into new cleanroom space. We are talking about more than 3,000 square meters there. We order equipment that is very specific tailored for the silicon carbide technology, install it. Overall, we invest $200 million in order to grow our capacity by 340% in silicon carbide. Same thing here, automation is key in order to get maximum efficiency out of the equipment. This is also where we benefit from any other fab's development in automation, in new solutions, that we do not invent the wheel twice or three times. We copy from other fabs who are leading in the specific automation space. Next topic, MEMS. Also here, in operations, we are driving massive investments in our capacity.
The two sites here where we have this skill and the technology is X-FAB Erfurt and X-FAB Itzehoe. X-FAB Erfurt is the location where we do the R&D, where we develop the technology. In Itzehoe, it's a relatively small site. Nevertheless, this is our key site where we have our integrated magnetic concentrator technology running and also stepping up. In Erfurt, we recently created additional cleanroom space, which is ready for equipment right now, and we are installing equipment for MEMS. We are talking about $50 million investment here, and we will also install new equipment in our site in Itzehoe. In both sites, we plan to ramp up our capacity by 84% in Erfurt and by 63% in Itzehoe. Massive, massive growth here in our MEMS technologies. There is one important.
Going through the whole fab tour, one important topic I want to mention here, to some extent, we are also a new X-FAB on the skill and engineering side. At this moment, the level of collaboration, the level of technical exchange, and also the support, the secondments we enable between the different sites to help equipment installations, challenges in specific problem-solving, this is ongoing currently on a completely different level if I compare this with three or four years before, for the pre-COVID times, though COVID was really a big struggle for us, where travel was not possible.
We are performing on a different level, and with the challenges I had in the operational execution, this collaboration and this exchange of engineering resources and know-how, helps us to face these challenges and to master the timelines.
... Thank you, Dirk. So now that we have scanned all the sides of X-FAB, the six different sides, the different investment on all the sides, we want also to give you a bit of explanation on where we stand in terms of expansion plan from some additional engagement we are taking with partners. What we talk about here is about creating quickly capacity for specific products, specific modules of our mature technology in 350 nanometer, 180 nanometer, and make sure we can create fast capacity and offer it to our customer. This is a true win-win-win for the different parties involved.
If we look at it from our supplier, partner perspective, they have empty capacity that is left over from more volatile business, and are ready to fill in the capacity with additional material and get additional revenue or utilization out of it. If you look at it from a customer perspective, it allows them to get more wafers, more capacity, but also to secure some business continuity plan because they will get material from different fabs, which is sometimes more and more required by customers of our customers. Last but not least, the win for X-FAB is that we can offer more capacity, we can protect some business, and we can do it with no additional CapEx investment. It is something that we are entertaining. Right now, we have already a first project that is running and in good shape.
We will have risk production by the end of the year in 350 nanometer. Only this risk production and this specific project allowed us to win or to save somehow, $120 million CapEx that we would have to invest if we would have to build this capacity internally. It's a way to very efficiently use our cash and make sure that we can serve our customers efficiently. Allow me to conclude this presentation. We described the different element of our capacity ramp expansion. All in all, we will add more than half a million wafer every year for each fab by 2026. Half a million wafer, extra capacity. If you sum up all the numbers that are there, you multiply them by 12, more than half a million. It's a significant number.
This expansion cover all the sides of X-FAB. I think we have give you some we have given you some details about how much per side, how much capacity do we get per side, with all their specificity. We do it in CMOS, we do it in SiC, we do it in MEMS. This expansion involves outsourcing, which is a new way to make sure we can secure business, as I presented earlier. We will do it when it makes sense. This expansion is definitely the most tangible commitment from X-FAB, supported with long-term agreement that we have with our customers, to serve our core markets with the technology differentiation they expect in quantity and in quality. Last but not least, this dimension, this expansion has two dimension. I think you can see on the numbers there, we grow in capacity, we grow in process complexity.
Growing in process complexity means also growing in value. This has been explained already many times. All of this will allow us to double our revenue by 2026. Let's now take a deeper dive into the financial associated with this growth together with Alba. We want to thank you for your attention, before, I just want to repeat, "...", we will do what we have to, "...", remove it. I hope we were able to convince you that we are convinced, we are personally acting in all this with a lot of trust, we want to thank you for your attention.
Ladies and gentlemen, I'm Alba Morganti, the Group CFO of X-FAB. A lot of you know me already, and I see a lot of well-known faces in the room. My role is to make sure that we will remain financially agile to support our strong growth. You have seen several presentations already today. I hope you enjoyed your journey with us so far. Now, let me walk you through our road to growth. The major theme of today has been our unprecedented CapEx expansion. I will show you today, now, in my last presentation, where this expansions plan will take us in three years from now, but also how we will finance them, and what this will mean for us from a financial perspective.
I will show you how we will ensure we will deliver our promises and materialize all you have seen from these very ambitious presentations just before. Last month, we published our very satisfactory Q1 results, showing EUR 208 million unprecedented sales on a quarterly basis. This is the result of a deep transformation of our group. We can now see the fruits of our hard work, and the business is delivering the numbers. These numbers are far different compared to six years ago when I joined the group. Our financial model has changed since our last Investor Day. We are now a brand new X-FAB, and as we will go through my presentation, I will set out how we will deliver our guidance.
This plan is just the beginning of the journey to bring X-FAB to the next stage of its evolution and become a $1.5 billion sales group. This is the new X-FAB. We have changed significantly in the last three years, in three ways since the last IPO, and our business is now centered around high-value-added products and markets. We have progressively refocused our core business into our three key end markets, that today represent 89% of our total revenues, and they are all growing markets. It reflects the shift to high-value-added products. We also phased out the legacy business, especially in Corbeil, as you already saw, and freed up the top line growth to grow strongly.
With replacing the legacy business in Corbeil gives us the chance to go for more X-FAB technology business, which is, of course, bringing a higher value. We also neutralized our exposure to foreign exchange volatility by increasing our euro sales share in sales, and creating a natural hedge. Thus, volatility is no longer an issue. Ever since the IPO, volatility in foreign exchange has been an issue and a source of uncertainty. This is no longer the case. We have focused on creating a natural hedge by growing our sales in euros.
This has been very successful, and we have been growing from 11% of EUR sales in Q4 2016, up to 42% today, which covers all our EUR costs, as you can see from the pie chart, the currency pie chart in the right side. Now let's have a look at how profitability has improved. Our profitability improved thanks to strong top line growth, an increased share of high-value-added business as a contributor to profitability, economies of scale coming on stream, higher prices as a result of the paradigm shift with our customers, who are now prepared to pay higher prices in order to secure their suppliers. With this, we protected our margins from the various cost increases. We also have been hard at work on productivity improvements.
We have several optimization projects to increase efficiency, and as you can imagine, automation is a big part of this. You have just seen how our profitability has improved since our last Investor Day, and profitability is set to continue to increase going forward. Let's have a look at the levers we still have to work with. Continued sustained top line growth, thanks to the continuous growth of our core markets, which commands a higher average sales price. We expect our core business to grow to more than 90% in the future. In this first quarter, we recorded an all-time high prototyping revenues of $26.4 million. The productivity improvement at all the sites of X-FAB will lead to an increase of the wafer output by implementing optimization and automation projects. This will lead us to an improved profitability.
The way we have managed Corbeil is a key part of our story. We have always been clear that Corbeil needed to be transformed, and we acquired it for its potential. When COVID hit and the supply chain was impacted, we were forced to put a lot of the transformation projects on hold due to CapEx, but also to hiring freezes. We have now emerged from this period. As you have heard from Damien and Dirk, demand has exploded, and we are now accelerating the process to meet the demand. Today, we have successfully converted the legacy business at X-FAB France into high value-added X-FAB technologies, which reached a share of 89% in Q1 this year. We estimate to reach an EBITDA breakeven at our French site at around $200 million sales by 2024.
Corbeil has still a big potential for further growth, as the site is forecasted to deliver about $340 million sales by 2026, which is, by the way, tripling the 2022 sales. Finally, economies of scale are possible, thanks to cost advantages and all the production output in all existing sites. On this slide, we wanted to show the impact of increasing our share in high value-added business. Here, you can see how the average revenue per wafer has increased 68% since 2018. This is mainly driven by two factors: the change in technology, of the technology mix, with an increasing share of higher value-added business, and the price increases we have introduced to protect our margins. Now let's talk about how our cost composition has evolved.
The overall share of fixed costs fell to 60% in the past from, in, right now, from 66% in the past, due to an increase of the variable cost, driven by indirect material expenses. On the one hand, we have to buy the expensive silicon wafers, silicon carbide wafers, for a proportion that is of our SiC business, only for a portion. As explained before, we have customers which are consigning part of the wafers. We have to source more and more expensive SOI wafers for the automotive business we are ramping up in X-FAB France. As a comparison, the substrates for the legacy business used to be consigned by the customer. With this split of variable and fixed costs, an increase of $1 in sales leads to a bottom-line improvement of about $0.60.
Despite the challenging environment, we have been able to keep our cost composition rather stable, thanks to a good cost monitoring, as you will see in the next slide. We have been able to cover all cost increases thanks to price increases. As you can see here, comparing the historical with the actual cost breakdown, bless you, we have kept our costs well under control and demonstrated that we managed to keep it stable, and there were no significant changes. What can we expect going forward? Which trends do we see? The share of material expenses is trending upwards, mainly driven by the increasing quantities of expensive silicon carbide wafers that we have to buy in line with the strong growth of that business.
Depreciation will go up in line with the CapEx spending, finally, financing costs are increasing in line with the financing needs and the current market condition for raising funds. On the other hand, we see the share of the cost of fixed assets, facility cost, and employee-related expenses decrease as economies of scale come into effect. Now let's let me walk you a closer look at the CapEx, which is needed to support the major capacity expansion programs. As you have already heard from several of my colleagues, we must expand our capacity in order to meet and support customers' and X-FAB's demand. We have initiated capital expenditure on an unprecedented scale for X-FAB. In 2023, we expect CapEx to come in at approximately $350 million.
which is already a significant increase compared to the previous years, as you can see in the bar chart. Over the three years spanning 2023 to 2025, we will invest $1 billion to expand our various capacity requirements. The maintenance CapEx for that period, which is an additional investment of about $200 million, will come on top. It's important to note that with the CapEx we already spent, invest until the end of last year, we are already able to grow our revenues to $1 billion level, which is only organic, so without any further acquisition. The additional CapEx plan for 2023 to 2025 will allow us to grow by additional $500 million sales per year. This being said, let's have a look at the CapEx ratio of X-FAB.
Historically, under normal circumstances, we spent 15% of annual revenues equally spent between maintenance and expansion CapEx. During 2022-2023 to 2025, the CapEx ratio goes up to 30%-35% of the sales, which, out of which only 5% is maintenance, and 25%-30% will be expansion CapEx. Once this CapEx program is completed, we will be back to a more normalized range of 15%-20% of annual sales. This is to sustain a 10% annual revenue growth. Every new US dollar CapEx spend generates an additional $1 to a $1 extra revenue, which implies $0.30 of additional EBITDA. The payback period is rather good. The main question is, where all these investments will take us?
At our last Investor Day, we said we wanted to reach $1 billion revenues and an EBITDA margin of 27% by 2026. In fact, as already said by Rudi, we have outperformed our previous forecasts, and we should be very close to this target already by the end of this year. As you will see in our yearly guidance at the end of the presentation, this is well on track. By this will not be the end of our growth. Our ambition is to reach $1.5 billion sales organically, and generate an EBITDA margin of 30%-35% by 2026. Basically, half a billion of EBITDA.
As you will see in the comparison of the 2023 perspective to the 2019 numbers, we have made the right decisions, and the key elements of our strategy are bearing fruits. Next question is: How will we finance all these CapEx? The financing of the CapEx is built on four pillars. The main proportion will still be covered by the operation cash flow. We estimate to use about $400 million-$600 million out of the total cash that we will generate out of the operations. We have also the LTA prepayments amounting to $350 million, approximately.
We will raise additional funds via new credit lines, amounting to we think about $300 million-$600 million or $500 million, still to be finally decided, but that's the range. On top, of course, of our already existing EUR 200 million credit line that we successfully closed on December 1, 2021. We are currently looking at the best financial solution, taking into consideration current market conditions and currently raising interest rates. By this, I thank also all the bankers in the audience that are supporting our growth by putting a lot of efforts to guide us into this direction. Finally, we have a strong cash position amounting to $350 million as of the end of Q1 this year.
While we aim to keep our cash position at a healthy level, we might occasionally use part of our own cash to cover some of the gaps. The pie chart demonstrates that X-FAB can cover more than the half of the total capital expenditures with our own cash. Once the CapEx program is completed, the picture will change significantly, as you will see on the next slide. Looking ahead, our ambition is to more than double our EBITDA. Once the CapEx program is completed, this will completely change, and in 2026, we forecast capital expenditures of about $280 million. Compared to an EBITDA of half a billion, as you can see from this slide, the EBITDA that we will generate in 2026 will be more than the double compared to this year.
In other words, from 2026 onwards, we will generate a very significant amount of cash as the CapEx spend will start to show a positive impact from a cash flow perspective. Now we are at the end of the financial section, and I would like to summarize our outlook. I hope that these slides have shown that the efforts of all these years are now paying off, and we have planted the seeds for even a stronger future growth. As you will see, the guidance for the full year of 2023 remained unchanged, with a revenue between $880 million and $960 million and an EBITDA margin of 23%-27%.
The free cash flow is expected in the range of 10% of annual revenue, positively impacted by the long-term agreement prepayments. Let me reiterate that our goal is to reach revenues of $1.5 billion by 2026, which is more than doubling last year's sales with an EBITDA margin in the range of 30%-35%. In 2026, we expect our free cash flow to come in at approximately 15% of annual sales. I am very happy to be, together with my ixfabulous team, be part of this journey and to support the bright future of X-FAB from a financial point of view. I thank you for your attention. With this, I hand back to Rudi.