We would like to draw your attention to the screen for the safe harbor statement and today's agenda.
What does it take to be at the heart of a technology that makes the world work? It takes perseverance, commitment, and steadfast leadership. What does it take to know the problem before anyone thinks it's a problem? Unrivaled experience, a unique perspective.
Those are a lot of the drivers of what needs to happen.
Meticulous planning. What does it take to solve our customers' challenges before they become obstacles? Trust built on established relationships.
Reported revenue of $8.3 billion, grew 24% year-over-year, and we're very confident where we are given the uncertainty in the market.
A foundational knowledge of customer needs and the ability to execute a strategy. What does it take to respond to dramatic change without breaking a sweat?
Shares of onsemi all-time high this morning.
Calculated flexibility, hands-on expertise, and a level of unyielding confidence forged in proven performance.
Top priority for 2023 is all about execution, serving the customers, getting the parts in their hands, making sure that they can deliver what they need to their end customers as well.
What does it take to drive resolute business growth against all odds? Concentrated predictive analysis, thoughtfully crafted plans, and the skillful guidance to remain ahead of the curve. Navigating the challenges of today requires focus, determination, clarity of purpose, and relentless innovation. Ruthlessly challenging the status quo and boldly refusing to be satisfied with just being good enough. Sustaining success means intentionally and strategically investing in what we must do today so we can create a better tomorrow.
Ladies and gentlemen, please welcome on stage President and CEO of onsemi, Mr. Hassane El-Khoury.
Welcome. This is not a mistake. The slide is from the last Analyst Day we had where I was on stage here telling a lot of you who are here with us today, what we're gonna do, what is the formula, how are we going to transform the company into the leadership and technology that is going to drive the future for mobility, the future for the industrial market, and the future in all of our strategic markets using Intelligent Power and Intelligent Sensing as our core competencies and core technologies. Very simple formula. Playing to win. Implement structural changes, so we can sustain what we don't know yet. If you've read a lot of the news, we are in unknown, uncharted territories. you know, some can argue we've been in those uncharted territories for the last two years.
Structural changes are the only way to be able to sustain through them and plow through them. We have to create value and capture that value and reflect that value, so we're able to reinvest it in a better future, but none of those will be possible without excellent and focused execution. All I can say is, if you look at everything on this slide, I'm very, very proud to stand here and say, "We have done exactly what we said we are going to do." Of course, I am very, very subjective because I said those things.
Saying, claiming that we've done all of them, obviously, it's a pretty easy thing to say, but I'm not the only one who agrees that we have executed, and we have delivered against a backdrop of positive, you can call it, 12 months ago or uncertain in the last two quarters. In an uncertain environment where the next one's going to be. You have acknowledged the work, and I wanna thank you for that, for going on this journey with us two years ago to where we are here to highlight the next phase in this evolution of this great company with the technology that we are able to provide for our customers and the value we're able to create for our shareholders, our employees, and our customers.
When we look at the markets we play in, I am sure all this data, all these charts, stack bar, line chart, %, CAGR, all of this is what everybody's been used to. This is what we are used to seeing. This is what historically even our strategies have always been based on. Get the data, spin it in Excel, it spits out. No matter how you look at it's going to spit out a market that is growing about 4%. We can argue half a point here, half a point there. It's irrelevant. That's the way the world has historically looked at it, we don't look at it that way.
The way we look at it is from a very different lens, a lens from a sustainable ecosystem, 'cause we are projecting a future where value is created by providing value forwardly in order to create that better future with technology. That future is about mobility. I'm not only talking about the electrification of mobility, but I'm talking about the added intelligence, the safety of mobility. Something that we all here in the room are users of and really benefit from as that technology transforms that industry and creates that value for our customers in order for them to transport it to their customers. It is about the industrial and capturing the energy, generating, storing that energy in order to enable the automotive transformation. We today, if those two do not happen together, one is going to impede the other.
If we don't have a solid infrastructure of renewable energy and the ability to store it and the ability to really deploy it across a network, the mobility or e-mobility and electrification of vehicles is not going to deliver what we all think it's going to deliver. That harmony and that tidiness between the two is what's going to drive that growth. All around the data that is enabled by such a transformation, where technology is going to be the primary value driver. Again, it is the value driver that our customers want in order to deliver it to their customer, who are all of us in this room. That is the sustainable ecosystem, and it all has to work together. One cannot work without the other.
That's creating this opportunity that is about intelligent power, intelligent sensing, and that is the new world we live in. That is the new world that we're going to deliver to because this new world is not about a 4%. This new world is about tremendous growth and tremendous opportunity for companies like onsemi with a portfolio of intelligent power, intelligent sensing to deliver on that growth. Because you heard it in the video, just good enough is not good enough for us. Four percent is not good enough for us. We have the content to deliver all of that. Our breadth of portfolio, our innovation, and really the attention to the execution gives us the ability to deliver the content that without which this future will not happen. We're at the core of it, and we enable it with content across the board.
Again, of course, I can be very, very subjective sitting here talking about all the stuff we have, but I am not alone. Our customers gave us the testament of the ability to help them change their world in their markets that they want through the breadth of portfolio and the innovation that we have been delivering in those markets that form the sustainable ecosystem. Relationships that historically would not have been possible between us, between OEMs directly, between customers that three, four years ago were not even customers, but now are, given this opportunity that we all have ahead of us.
As you know, Mercedes-Benz is picking up speed to become a technology leader in electrification and software. The last years showed us how supply chains can be affected by external events. We therefore want to strengthen our supply chains, and on this journey, we need strong partners.
onsemi and Mercedes-Benz have had an indirect partnership on the procurement side through tier ones for many years. In the last two years, a direct relationship and engagement with onsemi was formed following the height of the chip shortage and semiconductor crisis. The result of this relationship has led to a high degree of credibility and trust between companies. In 2022, the onsemi team has supported us greatly to ensure Mercedes-Benz production. On the technical side, there has been already a long-term technical collaboration between the companies, and we have now chosen onsemi as one of our key partners for future silicon carbide devices, thus securing preferred long-term supply, technology and quality of this decisive semiconductor component for our electrification. One example of this collaboration is onsemi's VE-Trac SiC module, which increases the efficiency and lowers the weight of the all-electric Mercedes-Benz Vision EQXX traction inverter.
This has a positive effect on the electric vehicle's range. The EV completed a 1,200 km distance. It holds the record for longest distance traveled on one single charge.
We are looking forward into direct sourcing and growing partnership together with onsemi.
Good enough is just not good enough for us, and it's definitely not good enough for our customers. Therefore, we need to keep pushing the boundary of technology, and you'll hear about it throughout this event. We have to look at markets that matter. The 4% adding to a focus of technology that we enable gives you a much better growth in the market, and it extends even beyond that as we focus on the sustainable ecosystem, which is the future of this industry. For us to grow beyond, we have to anchor on technology in markets that matter. For that, you see our growth being more indexed by capturing that value, by capturing these markets and targeting those mega trends that are enabling these markets with the electric vehicles, the EV trend.
That is not just, again, about the EV, it's about the EV ecosystem. The ecosystem, it does create from cars all the way to chargers, all the way to energy generation. The machine vision that's going to improve throughput of our factory automation, and it's going to improve the safety of automotive. That is the core of the sustainable ecosystem that's going to drive outsized growth, driving our overall growth at 3x what the market is. Why? Because good enough is not just good enough for us. The formula does not change. You know, why change something that's working very well? I think we all can agree it worked pretty damn well. Why change it? What we do in every one of these is different. We still will play to win. That's who we are. That's how we operate.
We always focus on everything we do being structural, so it's sustainable, it's consistent. It's all about value, but none of that matter unless we can execute to all of it. One thing, hopefully, after this, we're not gonna have an argument about is, can we execute? I think that's a nonevent. We can argue about everything else, but our ability to execute, I will give credit to a great team we have worldwide that has done nothing but execute against all odds and all uncertainties. What is it? It's about playing to win in the markets that matter, but doing it much better than we've done it yesterday, than everybody else is doing it. Improving upon what we do, setting our own benchmark, and doing it well through vertical integration to deliver that innovation across the bill of material and capture that value.
Also never forget to build a sustainably future-driven workforce. Some of you may not have caught it, our announcement with Penn State about a very strategic investment we've done with Penn State, the leader in material science in the country, for a silicon carbide center for onsemi, for both of us to drive the future of this industry. It's not about what we do today, it's not about what we do next year and the year after and the next five years. It's about what we do in a roadmap long beyond that. It's about implementing structural changes and going to market differently than historically all companies have done it. It's about going broad. It's about leading with digital, giving us that efficiency to scale fast, which we have done.
That ability to scale fast in a very sustainable way and outgrowing revenue and outgrowing the bottom line through those efficiencies of using technology to our advantage. Investing in opportunities to expand our TAM, Sudhir will talk about that, and really optimize our manufacturing footprint in order to maximize that output and give us that value and that structural improvement and that structural stability that we have achieved in the last two years and we will maintain and continue to achieve. Then capture that value. It's not just about creating, it's about capturing that value. Capturing that value through a faster go-to-market with the best tools in the industry, whether it's simulation for power or an SDK or a software toolkit in our sensing.
It's the ability to get every customer that values the product in that market to get to market with a low-touch model and extract that value from that market, a value that is driven by the innovation technology that we create every day, and subsequently expanding that gross margin and delivering the best-in-class ROIC. Most importantly, again, it's about executing. It's about building this tight and powerful organization that is accountable for what they do, but the autonomy to make it happen. That's the ownership we have. That's who we are. That's the culture we drive. Everybody in the company believes that and drives to that. They all will and have and will always get rewarded for doing this, doing the right thing, and creating that value.
Because we have a tremendous opportunity ahead of us that is expanding, forever expanding with over $17 billion of LTSAs while maintaining that predictability of finance and performance that you have gotten used to us delivering quarter after quarter. That is the formula that has not changed, but components of it will, in order for us to maintain that trajectory and accelerate the trajectory that we have been on. That formula will deliver the undisputed leader of the new world by executing day in and day out.
Hello, and good afternoon. By way of introduction, my name is Simon Keeton, and I manage the Power Solutions Group for onsemi. Today, I'm going to talk about high power products. Specifically, I'll touch a lot on silicon carbide. To do so, we'll look at three primary topics during our journey together over the next half an hour or so. The first topic will be a refresher course on the EliteSiC Advantage, otherwise known as the four S's for silicon carbide success. That may ring a bell for those of you that joined me in Hudson last year. The second topic, we will look at innovation, both at the die level, but also at the package level, which is incredibly important.
The third and the final topic, we'll look at how onsemi is the undisputed leader in delivering our customers solutions that are optimized to that end-use application, particularly in the two secular hypergrowth areas of electric vehicles and energy infrastructure. My goals for the day, let's share, let's learn, let's have some fun, and I'll be available afterwards as well. We'll start out with the EliteSiC Advantage or the four S's for silicon carbide success. S number one, supply. onsemi has a history of vertical integration. We are, in fact, the only integrated device manufacturer that grows their own silicon boules and silicon carbide boules. Looking at silicon carbide, not only do we have the boule growth, we have the wafering, the substrate, the epitaxy, the device fabrication, and then we marry it with best-in-class package technology, including modules, discrete, and Known Good Die.
That's S number 1. S number two is scale. We have a very broad manufacturing scale and a multi-decade experience in automotive and industrial. We know how to make high-quality products in high volume. Case in point, over the last decade, we've made over 500 million automotive and industrial modules. We know how to do high volume and high quality. That allows us to ramp in line with customer expectations into these two secular hypergrowth areas. The third S is scope. We have a very broad portfolio that we bring to bear for our customers. A suite of sustainable solutions all the way from energy generation, think solar, to energy consumption, think electric vehicle. A depth and breadth of power technologies. Low voltage FETs, medium voltage FETs, high voltage FETs, IGBTs, silicon carbide, discrete, module, known good die. I can continue on and on.
At the end of the day, what's happening is customers want to reduce that supply chain footprint. It's complex. With onsemi, we have LTSAs with literally hundreds of devices on a single LTSA for a single customer. They want to reduce that complexity in the supply chain. We have a very broad supporting portfolio for optimized solutions. Sudhir will talk about that too in just a minute. The fourth S, and perhaps the linchpin of all the S's, if there was a linchpin, it boils down to superior technology. For onsemi, we lead with die performance, but we differentiate with that package technology. So you take die plus package, combine that capability together, and onsemi has the most optimal solution on the planet for our end customers. That makes all the difference in the world.
Let's dive a little bit deeper into superior technology. When I'm discussing superior technology, you need a baseline capability to be able to deliver. Let's jump in our time machine and go backwards 18 months. Where was onsemi at? Well, we had a very good capability when it came to fab and package capability through brownfield development that Hassane touched on. We also had a lot of new product releases in die, in discrete, in modules, and we were almost complete. I say almost. There was a missing piece to that puzzle. That missing piece was substrates. We solved that missing piece with the acquisition of GTAT 18 months ago, now known as onsemi Hudson. We have that base from which to develop the superior technology. The question we often get, how's it going? It's going really, really well.
Since the acquisition, if you look at substrate output, up by a factor of 10. Same time frame, die output, up by a factor of 12. Package output, up by a factor of 4. Yields, up by a factor of 1.7, covering boule growth, wafering, and wafer fab. Exceptional performance. 3x the number of new products in that time frame as well. This is why we're so confident on the path to the first $1 billion in silicon carbide. The onsemi manufacturing machine is firing on all cylinders. Let's go into topic number two, looking at innovation, die, and package. We lead with the die performance, but I'd be remiss without how we actually approach it with a customer. The first thing we do, the first conversation that we have with our customers, what's your need? What's your problem?
What are you trying to solve? We have a very broad toolkit that we can bring to bear, and we look at it from a customer perspective. Customers want to solve a particular problem. We want to, let's say, extend range in a vehicle. Okay, let's translate that back to die performance, higher efficiency. They want to reduce size or save weight, as you heard on the testimonial from Mercedes. Well, let's increase the switching speed. Let's reduce the size of the magnetics and give you a smaller, lighter solution for your electric vehicle, and do so with the highest reliability, with the ruggedness. It's reliability on two fronts. Yes, it's the quality aspect of long-term reliability, yes, indeed, but it's the reliability of supply as well. With the vertical integration, we can do that all the way from powder to power.
Here is what we're working on, which is five generations of silicon carbide. With each generation, an increase of performance, and we won't stop there. How do we deliver it? Through cell structures and channel engineering. With cell structures, we're actually shrinking the unit cell dimensions and getting more current through a particular area, basically pushing more into a smaller area and using geometry to help us as well. We talked about this in Hudson a little bit. M1 was a square shape cell design. M2, it was hex. M3, stripe. M4, trench, and so on and so on. Here's a hint. Why are we able to do this so quickly and so effectively? Well, we've been doing it for 30-plus years. The same concepts that were applied to silicon are now applied to silicon carbide.
It's a different material, absolutely, different properties, but there's a lot of carryover from silicon to silicon carbide that allows us to execute to this roadmap particularly quickly. Look at channel engineering, for example. It's a very deliberate process that we have in manufacturing silicon carbide. This allows us to shorten those channel lengths and increase the channel carrier mobility by reducing the density of what's called trapped charges. And we do it effectively well. And what does it mean at the end of the day? It means a figure of merit, and a figure of merit is a quantitative performance metric that says, "Okay, how does it work?" Well, onsemi compared to what's on the market today is 30% better figure of merit. And that translates directly to the performance that the customer is looking for. You build a better mousetrap, but we're not done yet.
Lead with die performance, differentiate with packaging innovation. Think of it this way. Let's say this clicker is a piece of silicon carbide. It's a great piece of silicon carbide. Let's say I didn't have the history in packaging that onsemi has, and I put it in some package. You won't get the heat out of the device like onsemi. You won't reduce the parasitics like onsemi. You won't have the performance like onsemi. Again, thinking from a customer perspective, they want range, for example. How do you get the range? Well, you need the performance of silicon carbide. How do you get the performance? You have to extract the heat out of the die and through the package and out to the fluid. We've been doing this for decades. We know how to do this extremely well. Customers want flexibility.
We can give a solution to a customer, one footprint that starts out with an IGBT-based solution, can migrate to a silicon carbide-based solution in the same footprint, an upgrade path built in for our customers, allowing them to scale in power levels as well and do so with that inherent reliability and ruggedness. If we actually look at the different packages we offer, and this isn't an exhaustive list, but just a few of the things that we offer. You'll see this transition. On the very bottom, it's a gel-filled case module, and we'll talk about this in a bit more in just a second. That's a great way to start, and it's been an industry standard for years. We supply these. It's a great solution. Up to about 150 degrees Celsius for that die, maybe 170 if you push things.
You want to get more out of silicon carbide, run it at 200 degrees Celsius. How do you do that? Transfer-molded modules, either discrete or in module format. We have the dual side cool, the half bridge going to the full bridge with the cooling channel, successive generations of improvement, reducing the thermal resistance significantly, allowing you to drive more power, more efficiency, better range, lower cost. I can go on and on. Let me grab a couple of show and tell examples just to drive a point home, and you can play with these in the demo room after I'm done. My left hand here is that case module in white, and what you see on the bottom is a cooling channel needed to cool the device. My right hand over here, this is a dual-sided solution. They're functionally about the same.
This is a full bridge. This is also a full bridge. Just so happens this one uses three half bridges. It's just an architectural concept, but they're functionally equivalent. Now, if you see me shaking it all up here, I'm not nervous. This thing weighs about 5 pounds in my left hand. It's pretty heavy, and you'll get to pick it up for yourself. When we develop a dual-side cool solution with a cooling channel, it's a 50% reduction in weight. It also has higher performance, higher temperature, higher power, significantly reducing the weight in a vehicle. I'm gonna put this one down now before I drop it. Actually, I'm joking. I'm picking up another one, and I'm coming right back. Do we stop with that? Nope. This is about innovation. Back again. My arm may start shaking. Let's see. This is a polymer cooling jacket.
This is something that we created. It uses a capability called jet impingement. You'll see a inlet and an outlet. When you take a look at it downstairs, you see these little holes, and those little holes will spray the fluid on the backside of the module. Pretty cool. We just wrote a paper on it. We patented it. What's really cool, it's a 90% weight reduction. Nine zero. 90% weight reduction as compared to a metal cooling channel. Pretty cool. We're working with our OEM friends on this as well. You'll get to pick these up, and it's radically different. If you think about it, what we're looking for is an optimized solution. We say optimized solutions are in our DNA, and here's our helix here. The blue strand is a die technology.
The red strand that's about to pop up is a package technology. The way that we develop these technologies is a little bit different. We develop two concurrent next generations in these technologies at any one period of time. silicon carbide, today we have M3. We already have M4 and M5 in development concurrently. IGBT, people ask, "Is IGBT going away?" I will tell you no. We are investing full throttle in insulated-gate bipolar transistors at the same time as silicon carbide. Today, we have Field Stop 4. We already have Field Stop 7 and Field Stop 9 in development. Couple that with the package capability, discrete, gel-filled, transfer molded. You can mix and match between these technologies and the packages for the most optimized solution, the key word being optimized. It makes all the difference in the world.
Let's take a look at these optimized solutions applied to two of the largest, fastest-growing secular growth markets with a tailwind behind them. The first one we'll look at is electrification. Again, coming at it from a customer perspective, let's say I'm an OEM for a day. Most OEMs have multiple platforms. They're not trying to solve for one thing, they're trying to solve for many different platforms. This is a genericized look at what those platforms could be. Entry-level, mid-range, high performance, ultra-high performance. What you see as you go up the triangle is an increasing level of content. Makes intuitive sense. As you go down the triangle, what you see is an increasing volume, unit volume or SAR for these vehicles. This is what our OEMs are trying to solve for. How does onsemi apply the optimized solutions for the OEMs?
We power every architecture. Let's review some of these architectures together. Ultra-high performance. Case in point, this is a four-inverter architecture, one inverter per wheel, so one motor per wheel. The highest performing cars or trucks or vehicles, the fastest zero to 60, the longest range, the top tier. onsemi has the design wins and LTSAs using our quad silicon carbide architecture. That's 4 silicon carbide traction inverters, one per wheel. Let's take it a step down from there. High performance. Now we're moving from a quad solution, so a quad motor, quad inverter. Now we're going to a dual motor solution, front and rear. In this case, we're looking for high performance with good range and acceleration. Silicon carbide used on the rear axle motor and on the front axle motor. Great. onsemi has the design win and the LTSA for this architecture today.
Next step, back again with a dual motor solution, this one's slightly different. You notice silicon carbide on the rear axle motor, IGBTs on the front axle motor. Why would you do that? It's a slight drop in performance. It's a good mid-range performance with on-demand acceleration. The rear traction motor is silicon carbide. That stays on all the time. The IGBT kicks on when you need it. Power on demand. Think of merging onto a freeway. You hit that accelerator, you need to come up to speed, it turns on, and now you're gliding along, it turns off. It's a hybrid solution using both IGBT and silicon carbide, one on each axle. We have the customer design wins and the LTSAs today. We're looking at entry-level vehicles with extended range. Single motor, many cases, rear axle motor, and this is with silicon carbide. Why silicon carbide?
Extended range. Onsemi provides the single silicon carbide traction inverter with the customer wins and the LTSAs. Here is another version of a single motor inverter solution, rear wheel. This is the entry point vehicle. This time, IGBT. Cost-effective IGBT-based solution with an upgrade path that I mentioned earlier. Start out with a module, IGBT-based. You want the upgrade path or that next up vehicle? Okay, same footprint now with silicon carbide. Takes it up a notch. We have the design wins. We have the LTSAs. Next up, something interesting that people are talking about right now and is coming out a lot in the market. It is an evolving need for increased performance at the entry level. This is a blended solution using IGBT and silicon carbide in a single traction inverter.
Depending on the mission profile, where you are in the city, on the highway, you're using either the IGBT or silicon carbide in the same module. New concept, new capability, yeah? Think you might have heard about it. We've been doing it for three years in the industrial market. We're just gonna take and then apply it to automotive, because we have that toolkit. Next, I just wanna take just one or two minutes, and we're gonna segue a little bit into motors. onsemi isn't getting into motors. I'll tell you that right now. We're semis. We can help with some issues with motors, allowing choice, allowing optimization, 'cause that's what we do. I'll tell you briefly what it is, and then we'll get into the details a little bit. It's called an exciter module.
Think of it as a companion chip to our traction inverter solutions. Those are traction inverter solutions over there that I had in my hand. On that same cooling channel, you can add this additional device called an exciter module, and it can allow an OEM to change their EV motor. Why would you wanna do that? Let's take a look. Most EV motors today are called permanent magnet motors. These permanent magnet motors are very efficient, and they're very good. It's made of two primary components. You have the outside component is the stator, which is static. The internal component is the rotor, and that rotates. The way that that thing works together is from that traction inverter, you're electrifying that stator. There you have the traction inverter, and it creates this rotating electromagnetic field. That's what you see rotating around.
With the permanent magnets, that also creates a magnetic field from the magnets. You'll see that kind of rotating around itself. Here's the good news and the bad news. You want to use a very, very strong magnet when you do this. In many cases, magnets that are used are made out of a substance, a rare earth metal called neodymium. You might have heard of it, maybe. Here's the issue with neodymium. There's a lot of industrial waste. According to a Harvard study, 1 pound of neodymium to make 1 pound of it can create 2,000 pounds of industrial waste. That's a problem. Don't get me wrong. Many people are trying to solve this issue in many different ways, and this is just one of them. It also has a regional dependency to it to source that neodymium.
As a result, governments don't like it. What's one option? Well, one option, you remove the magnets altogether. Hey, let's get rid of those magnets. We don't want to use neodymium. Okay. Problem there is you don't get that magnetic field on the rotator. What you can do is replace that permanent magnet rotor and put it in, instead of magnets, with windings. It's called a separately excited synchronous motor. When we do that with the rotor, now there's no permanent magnets. There's no neodymium. There's just windings, and that goes back into the stator. Now you have to excite that rotor to create a magnetic field because you need the two magnetic fields working together, and that's what drives the wheels.
That little blue guy at the end in the animation, that's the exciter module sitting right next to the onsemi traction inverter. What that little blue guy does, it completes a solution because it excites the rotor and creates that second magnetic field. You're back in business because you have the two magnetic fields working together, driving the propulsion for the vehicle without the use of rare earth metals. A couple things that are cool about this. Number one, being repetitive. Design wins and LTSAs on this product. Number two, it allows a greener choice, which is pretty cool. Solutions for a sustainable future. Optimized choices, different solutions, different approaches. We give the customer the choice. Let's talk about the power of the portfolio. I've talked a lot today about the traction inverter itself. There's a lot of content in that traction inverter.
Unfortunately, I don't have the time to talk about everything else in that portfolio that onsemi has to offer. Solutions for onboard charger. Solutions for the DC-to-DC converter. 48-volt e-compressor. EPS, the power steering. The PTC heater. We have every solution. Again, what does that lead to? It leads to LTSAs with hundreds of devices because we can solve the entire problem. I even go beyond this. Sudhir will take it even further. At the end of the day, what it means, 160% annual growth rate for the past two years and this year. Extremely high growth driven by customer wins and LTSAs with optimized solutions. Next up, we talked about electrification and now energy infrastructure, whether it be generation, whether it be storage, or whether it be charging.
The first example, and before I go into it, everything that we have talked about before, those technology innovations, both at the die level and the package level, are applicable to energy infrastructure. Same thing. We lead with scalability in energy infrastructure. In this case, let's look at a solar example. Residential, got it covered. Commercial, covered. Utility, covered. What you see is a broad solution set from 6 kilowatts all the way to 320 kilowatts and everything in between. Look at the technologies. silicon carbide, IGBT, hybrid. Familiar story? Absolutely. We lead with the scalability, but we differentiate with optimization. Let's dive into one particular example. This is with a specific customer. It's a case study. Let's walk through this case study together. A 320 kilowatt, 1,500 volt utility scale solar inverter. That's the starting point. Why is that a big deal?
Nobody makes a 320 kilowatt solution or they didn't until we did. What do the customer need? Performance, low weight, reduced size, a lower system cost, go-to-market speed, and supply assurance. Kind of the typical stuff. Well, what do we come up with? You see this hybrid solution, IGBT, silicon carbide and silicon. What's the big deal about that? Silicon carbide was used for the boost portion of the architecture. IGBT was used for the inverter portion of the architecture. What was our approach? What did onsemi deliver? Time to market. Just to go into this for just a minute. Imagine I'm a solar company, and I'm looking for this 320 kilowatt solution or whatever solution it is. The usual approach is I have a concept. Okay, I could solve this with IGBT.
Okay, I'm gonna go build a solution using IGBTs. I have to physically build it. I gotta prototype it. I test it. I put it in the field. I see if it works. Oh, it didn't work, back to square one. Try it with silicon carbide, do the same thing. I build it. I prototype it, put it in the field, see if it works. All right, not quite what I want, back to square one. We're talking six months to a year to go through that development stage. onsemi authors the Elite Power Simulator, and this simulator doesn't just work off of datasheet specifications. Doesn't work just off the device specifications. It works off device, manufacturing parasitics, device parasitics, system parasitics, all combined into one simulator that the customer can then use in their system.
Instead of taking six months, nine months, 12 months to figure out the most optimal solution, you can get a good direction on this, two hours. It's that fast. Just another value proposition that we add to our customer set. Module design. This optimized technology in this example with a combination, the best combination with silicon and silicon carbide. Again, I mentioned silicon carbide for the boost circuitry, IGBT for the inverter. Best of both worlds. When you have the right toolkit, you can provide an optimized solution. Technology. We talked about the concurrent development of two generations of next-gen silicon carbide and IGBT. We have the right technology for the end customer optimized with that leading performance. Then again, don't forget about the power of the right module and the packaging. We developed a specific solution for solar with the best thermals.
With packaging, it's about extracting that heat, reducing the thermal resistance as compared to die, which is reducing the resistance of the on channel in the circuit. A 15% reduction when it came to thermals. The net-net, an unoptimized solution will be at least 30% worse than an optimized solution, and onsemi was able to deliver it. In fact, this is the only 320 kilowatt solution available. Industry first. We apply the same mindset, whether it be any energy generation, energy storage, ultra-fast charging, with what does a customer need and bringing that very broad toolkit and that manufacturing capability to bear.
At the end of the day, we have LTSAs with eight of the top 10 energy infrastructure customers driving a 70% annual growth rate in each of the last two years and this year. With that, we talked about electrification. We talked about energy infrastructure. We talked about these hyper-growth markets. Lot of talk, but actions speak louder than words, and our customer actions prove that we have value with $9 billion of silicon carbide LTSAs in electrification and energy infrastructure. It's a testament to what we can bring to our customers. With that, I'll say thank you.
Good afternoon, everyone. It's super exciting for me to be here. I will say that I kind of feel like the third guy in a four by 100 relay, you know, where the first two guys just break out in an outstanding lead, and really all I have to do is not face plant. Here goes. I'd like to talk to you a little bit more about the sustainable ecosystem. Hassane described the sustainable ecosystem as the new world. You can also think of that sustainable ecosystem as being the onsemi flywheel. With each push, new product that we introduce, we generate more momentum, and in doing that, create value for our customers and for our investors. What I wanna talk to you about is what we are doing beyond silicon carbide.
What are we doing to accelerate that flywheel? What we're doing is we're starting with our core two value drivers, silicon carbide and silicon power in the intelligent power realm, and that's what Simon talked about, and image sensors in the intelligent sensing realm, and I'll talk more about that later. Now, you guys understand that these are the core value drivers for onsemi. It's what you think about, it's what you write about, it's what you value us. The question I'd like to answer is what beyond? What are we doing beyond that? We'll start with the power realm. As we talked about, silicon carbide and silicon power is obviously a source of strength. We're number two in the marketplace today, but number two with a bullet. We're planning for 38% CAGR in revenue over the next five years through 2027.
Now we think about what can we do to create more value beyond that silicon carbide and silicon power device. It really comes down to delivering what customers want. In the power realm, the most significant parameter, what customers value the most is efficiency. Every little bit of efficiency improvement matters tremendously. In fact, 1% can be the basis of winning and losing designs. It has a profound impact. You may say to yourself, "1%? Is that really that big a deal?" Well, think about it in the bigger picture context, first starting with electric vehicles. Now, we all understand that electric vehicles are going to grow in terms of deployment over the next several years.
If you think about that in actual miles driven between 2022 and 2030, we're gonna see roughly a 5x increase in the miles driven. Now you heard from our customers and from Simon that what efficiency translates into for electric vehicles is additional range. The impact in 2022, if you were to get a 1% efficiency improvement, would be an additional 3.2 billion miles per year driven in electric vehicles. Pretty substantial at 1% efficiency improvement. If you look further at the data center and think about that same 1% efficiency, the backdrop is this. Data centers consume about 205 terawatt-hours of electricity per year. That's the equivalent of 19 million homes, an astounding amount of electricity.
Here, the benefit of 1% efficiency is measured in reduced power consumption, and as a consequence of that, reduced electricity costs. That 1%, again in 2022, would translate into just under $400 million less in electricity costs in the data center. These numbers are huge. They're impressive, but they pale in comparison to the kind of benefits you can get by improving efficiency by 1% in industrial automation. In that area, industrial motors consume 50% of the world's electricity, of their energy. 1% improvement, again in 2022, would translate into a $25 billion reduction in electricity costs. We're having this conversation in terms of costs, but we could also talk about it in terms of, again, in terms of sustainability.
Whichever way you look at it's obvious that 1% is profound in its impact. The question becomes, what are we doing? How can we impact that efficiency? Simon talked about it already in the context of silicon and silicon carbide, but what we're doing is essentially augmenting, starting with those. In any power delivery system, silicon carbide or the silicon power device is really the core of value. That's the place where you're delivering value to the load. Simon talked a lot about optimization, what we do in die, what we do in package, what we do in modules. It's all about optimization. What we're doing is taking that optimization to the next level by adding more devices on the BOM. The first is the gate driver.
A gate driver's job in life is simply to drive that power transistor, whether it's silicon carbide or silicon power. If you do a really good job of co-designing, co-delivering these parts, you can actually deliver meaningful customer benefits. If you are the provider of the silicon carbide or silicon power device, then you can provide information back to that gate driver, electrical information like voltage, current, and temperature, and a whole bunch of information. With that gate driver, you can drive that device to maximize performance, which translates into efficiency. You do that just at the cusp of losing reliability. If you're pushing it to the limit there, you're going to deliver an optimized solution. We're going to do that with gate drivers. We won't stop there. Beyond gate drivers, we're going to do controllers.
Controllers are essentially the brains of the system. They manage the power flow in any subsystem. There are a number of different control techniques that are used to manage this power flow, ranging from analog to digital. Regardless, if you do a good job of co-architecting, co-designing, and co-delivering this solution to your customers, you deliver meaningful value. First and foremost, in terms of efficiency, where we can deliver anywhere from a 1% to 2% efficiency improvement by doing an exquisite job with these components. Also reliability and ease of design in that we're solving problems that customers would otherwise have to solve for themselves. In doing this ourselves, we also can reduce some of the passive components in the system, and that has the consequence of reducing system cost. This is exactly what we're planning to do.
As we do, we're building from a leadership position today. We've talked about where we are with respect to silicon carbide and silicon power. It turns out that in gate drivers, we've been doing this for a while as well, but with respect to silicon. We've been doing this for silicon power devices, and we've already, in that realm, delivered some of the advantages I talked about on the prior slide. We also have controllers. Today, customers value us for our controllers. The vast majority of what we do is control in the analog domain, and I'll talk a little bit more about that in a moment. We have this leadership position across the board, and what we're doing then is to build or expand on this leader position in each one of these categories of devices.
You know what we're doing in silicon carbide. In that semiconductor power device domain, we're inevitably gonna get to gallium nitride as well. I have no news to report here. It's really the same thing that we've been talking about in that we see GaN is something that's going to be required in the future, but at the right time and at the right investment level. With respect to gate drivers, we're going to extend the portfolio and do an exquisite job of gate drivers that are tightly coupled with our silicon carbide transistors and products, and then as well extend those to gallium nitride. In the controller domain, we're going to transition. We're gonna migrate from analog to control to digital control. This is hugely important.
To give you a sense for why that's important, I have to kind of give you some historical context. Historically, the control technique, digital versus analog, was a decision about what you want to optimize for. Digital offered all of the benefits that you know from digital. It's typically more configurable and therefore more flexible. It tends to be lower cost if you do it right. It's easier to design. For us, easier to design. If you wanted to extract every bit of performance, every bit of efficiency out of a system design, then doing that in the analog domain in the past was the right answer. What we're doing is to introduce digital control techniques in a no compromises way. We're gonna deliver all the efficiency. You say, "Well, how are you going to do that?
How are you going to deliver digital with all the efficiency that you have in analog?" The answer is, we're gonna leverage that analog expertise that we have and couple it with techniques in the digital domain. You can kind of think about onsemi and the engineers here as being kind of in two classes. You've got the wise analog, kind of, you know, I won't ascribe age to it, but the wise analog guys working with the young bucks in the digital domain, and the analog guys kind of think of the digital guys as being kind of cute and, you know, driving innovation. It's those guys working together that really deliver this no compromises solution.
When we do, we're gonna deliver an array of power ICs, ranging from AC/DC converters to point of load and PMICs, multiphase controllers, switches, and eFuses that take advantage of these digital controllers. When we do it, we're gonna do it in a way that we extend across all of our target market segments, from automotive to industrial to cloud. Ultimately, the impact of doing all of these things is to dramatically expand our TAM opportunity. When we think of the power ICs that I'm just talking about, the TAM opportunity that gets exposed and becomes available to us is $14 billion. That's just talking about the power ICs, not the silicon carbide and the silicon power devices.
Controllers, not all of them, but the higher end of controllers take us into very high gross margin territory, where the industry has already established gross margins of being 70%. By proliferating this power portfolio, we're creating value for our customers, we're creating value for onsemi, and we're creating value for our investors and accelerating that flywheel. How do the customers think of this? Ultimately, the customers are interested in the efficiency benefits, but they're also looking for a vendor that can do more for them. The implication of proliferating this product portfolio is that we can provide solutions that solve all of the customer problems. The customers think of the power subsystem in the systems they design as a tree. What we're doing essentially is delivering a full power tree solution. On the left side, this is an automotive example.
On the left side, you see AC/DC. That's in an automotive use case, the on-board charging. On the right side, you see what happens in the DC domain, all of the boxes represent the specific power functions that get fulfilled. The colors here refer to the colors that I've shown on the prior slides and highlight what we can do when we proliferate our product portfolio. It's not just automotive. That kind of capability, that kind of coverage of the power tree extends into industrial and cloud, you can see that front end, that AC/DC component looks remarkably similar across each of these, the backside of it in the DC domain looks different. When we proliferate these products, we're able to solve all of these problems.
Hopefully that's clear for how we will ultimately deliver value to customers and grow our business. Shifting from the power realm to the sensing realm, as I said before, we start with a position of strength. That position of strength is our image sensor technology. With our image sensor technology, we have a number one global position in the markets that we care about, automotive, industrial. Those are the markets that are part of the sustainable ecosystem that we've been talking about all afternoon. We have about 46% market share in automotive overall and an even higher 68% market share when you look specifically at ADAS. That's a very, very strong position. I'd like to stop and talk a little bit about how we earned that leadership position.
Simon talked about the competitive advantage in silicon carbide, and I don't have something that is as effective or pithy as the four S. There are also four specific levers or advantage elements to our solution. The first one is superior technology. Here it's about doing a good job with three specific capabilities. global shutter, doing an entire image at one time, which you have to do when there's a lot of motion. rolling shutter, exposing rows at a time, which you have to do when you are optimizing for high dynamic range. image signal processing for readying that image prior to human consumption. You have to have all of these, and we do. The second dimension of competitive advantage is just the focus. Unlike some of the other competitors out there, we are exclusively focused on automotive and industrial.
That means that every innovation that we do has a root in solving a customer problem in automotive and industrial. Here we have a very large installed base. For example, in automotive, we have over $450 million automotive sensors deployed. Obviously have scale. The next lever is solution enablement. I'll talk about this a little bit more in the next few slides, but the basic idea is this. We give customers everything that they need to design their image sensing solution, and we do a better job than the rest of the field. The last dimension is really simple. It's the school of hard knocks. We've been in this, you know, when you think about our predecessor companies, for over 45 years. We've leveraged those years of experience and focused innovation to generate trust in our customer base.
Trust that we'll be able to solve problems going forward as well as we have done so far. Let's talk a little bit more about the superior technology. I wanna give you a glimpse into the specific aspects of that technology that we do exceedingly well. I'll start with dynamic range. Remember, I talked about this. The basic idea of dynamic range is you wanna capture the brightest of the brights and the darkest of the darks in any image. The better job you do in terms of that range, the better performance that you'll have. If you look here, you have a visual indication of that. You can see the image on the left is kind of washed out. The image on the right, which is one that we produce with our very high dynamic range, actually, you can see all the clarities.
You can see the differences between the lights and the darks, and everything has absolute clarity there. That's the benefit of high dynamic range. That's what customers come to us for. Additionally, we do an exquisite job in low light performance. Low light performance is exactly what it sounds like. It's the ability to deliver high-quality images in an environment where there is relatively low light. Obviously hugely important in automotive, but in a wide variety of use cases. To be excellent at low light performance, you not only have to use world-class design techniques, but you have to be world-class in terms of manufacturing cleanliness. This is an area where we distinguish ourselves as well. Being good at the manufacturing side as well as at the design side. The next area within tech-technology is low power performance.
Here, particularly with our latest generation device, we use techniques to intelligently manage power in the sensor and in the system overall. What that means is, based on the situation and the needs that the customer has, you can reduce the power consumption throughout the chip and leave only portions of it on, if you think are operating, and process images at low resolution. Then when, for example, there's a motion event, you wake up the rest of the chip in the system and get to the full level of performance. As you do that, depending on the amount of time that you spend in that low power state versus the high power state, you can save as much as 30%-70% of your power, or at least become that extent, to power efficient to that extent.
That's the kind of capability that we're driving. Beyond all of these things, you have to introduce system functions that are relevant, specific to automotive and industrial. Those system functions are cybersecurity and functional safety. There are ISO standards that govern this. We are the only supplier that meets all five cybersecurity threat mitigations. We have functional safety. We meet ISO 26262. These are essential in the automotive and industrial environments where functional safety and security matter a great deal. That's just a glimpse into what we do at the actual sensor level. The value that we create extends beyond that sensor. Let's talk a little bit about that solution enabling.
Think about it as kind of a value stack above the sensor. It's this value stack that creates not just stickiness, but presents a bit of a competitive barrier. We'll start with the lowest level of that stack above silicon, and that's DevWare. This is a software development kit. As with any software development kit, the idea is to place the power in the hands of the customers to extract everything that they need in terms of performance and functionality out of your device without having to understand all of the details of how that device operates. We put our learnings into this software development kit and then take that software development kit and put it in the hands of our customers, and then our customers can use a very familiar interface and very simply configure systems that meet their full range of performance requirements.
When we do that, we have the ability to scale the business. We're not talking now about something that's just in the domain of very experienced system-level engineers, but we are basically lowering the bar in terms of technical sophistication required to implement an image sensing solution, and in doing that, creating the ability to scale to thousands of customers. We don't stop there. Beyond these, we do reference designs, both in the automotive world and in the industrial world. In the automotive world, our customers use our reference designs essentially as a development platform. Ultimately, they'll do their own designs and take those into manufacturing. In the industrial case, we do full module designs that they themselves can implement in their systems. Customers themselves can implement in their systems, and some do.
Again, not just the sensor, but as well the software and the industrial design, and on top of all of these things, our customers can take advantage of a complete ecosystem that ranges from lens vendors to discrete image signal processing vendors, providers of SERDES interfaces, SoCs and FPGAs, and independent software vendors that write software for this platform, all who have worked with onsemi, all who can be used to extract all of the system performance that our customers need. It's really all of these things that we do that create that advantage. As I mentioned before, all of this comes based on 45 years of image sensor leadership, 45 years of innovations in this area. As I said earlier, it's the school of hard knocks. We've graduated that school, so our customers don't have to.
We translate all of those learnings into product features and capabilities, whether at the sensor level or in the software that we create or the reference designs. Okay, that's the basis for our advantage. That's sort of the core value driver in intelligent sensing. As with the power realm, we're building on that and extending and creating more value in the form of new products. What we're doing first is to add depth sensing capability. With depth sensing, we basically are using time of flight to gauge depth. This is used, for example, in lidar applications. We use different techniques to solve for long-range lidar and short-range lidar. In each case, we improve the perception capability that our customers are embedding into their solutions.
We're taking this depth solution and extending across not just automotive applications used in Lidar, but also industrial applications as well, where depth sensing is important. Beyond this, we have a suite of sensor interfaces. Think about them as analog front ends that sit in front of those sensors and translate those sensors into meaningful information. We're number one today in ultrasonic sensors. We're number one in inductive position sensors, rotary position sensors in automotive. Each of those sensor interfaces have application in the industrial realm. What we're doing is extending into the industrial realm with those solutions. Just as you saw in the power realm, we have the ability in the sensing realm to dramatically expand our TAM coverage.
Here, again, specifically within the sustainable ecosystem, we open up a TAM opportunity of $5.3 billion by 2027. Very significant growth opportunity for us and one that we can get by building on the leadership position that we have today. We've talked about intelligent power, specifically the silicon carbide and silicon power devices, the gate drivers and controllers, the power ICs that add value in the system in the form of efficiency. We've talked about the image sensor being kind of the core of the intelligent sensing realm, and value proposition and expanding that with depth sensors and a variety of sensor interfaces. Beyond these things, onsemi has a number of parts that we have provided for many, many years that solve very specific problems in our customers' intelligent power and intelligent sensing solutions. Think about this as accessories.
We're accessorizing, you know, customers can take advantage of these components to build out their solutions. The effect of all of these devices together is to make us indispensable to our customers. They really rely on us to solve a variety of problems that they face at the system level. Silicon power and silicon carbide devices now power ICs, image sensors, including depth sensors and sensor interfaces. Let's look at what this means in terms of customer opportunity. I'm gonna start with the data center. Here's a picture of a data center server, you can see the different subsystems where we provide components. The artificial intelligence accelerator, part of most data centers today, obviously, that's gonna be a rapid growth trajectory. The AC/DC, high-power AC/DC component, the server board itself, the motherboard, the networking unit.
If you look across all of those boards within a server, the opportunity that we have today, if you were to sum the boxes on the left, is about $37 per server. Where we're going in terms of the product portfolio is to enable over $150 of content per server, and that translates into an opportunity of $5 million per data center. A very significant growth opportunity for us. As excited as I am about cloud, I'm much more excited about automotive. Here, Simon talked about the power of the portfolio and talked about all of the solutions that we can deliver within these fields, within these specific subsystems. If you translate what Simon talked about to very specific devices that are designed in, you get a picture that looks like this.
Here we're talking specifically about silicon power and silicon carbide devices. Obviously, quite a few products designed in, devices designed in. What we're showing you now from this point is an actual vehicle, a BMW i7. All of the products that I'm going to talk about represented by these dots are designed into a BMW i7. Remember, this was just part of the picture. As we build out the car, you see the number of devices that we are actually designing into and have been designed into. Here, we range from power ICs, sensor interfaces, image sensors and the other, what I termed as accessory parts. This is the kind of coverage you see. There are almost 500 devices designed into this BMW i7, almost 500.
I don't think anything else encapsulates my message to you better than that. In fact, I just often have to pause here 'cause I love this slide so much and look at the dots and admire the dots and that number of 500. Just kinda have to let that soak in a bit. That's the opportunity in front of us. It's actually an opportunity that we're realizing today. To sum up, we talked first about what we're doing to expand our value creation in the intelligent power realm that's getting into silicon power devices that silicon ICs that allow us to build complete power trees, leveraging our strength in silicon carbide and silicon power devices, and in doing that, expose ourselves to even more of the TAM, $14 billion of TAM, in product categories that scale up to 70% gross margin.
We talked about building on leadership in intelligent sensing through focused innovation, whether that be in image sensing or depth sensing or sensor interfaces for automotive and industrial. We talked about the fact that we continue to have a variety of other parts that allow our customers to accessorize their solutions in a way that make us indispensable to them.
All right. Thanks everyone for joining us today. For those of you that I haven't met, I'm Thad Trent, CFO. I think Sudhir has hit stride. I've just got to get it across the line by taking everything that you've heard and stitching it all together to tell you what it means for us financially. We'll jump right into it, but before I do, it's really important to acknowledge what we've done over the last two years. The last two years have actually set up the growth for the next phase of our transformation, and the structural changes actually make this possible. If you look at the performance since we started this journey, we've outperformed the SOX by 147%. Hassane talked about the winning formula, he didn't tell you what it meant for us financially.
He went through the components of what we've done and what we're going to do as we continue to march down that path and continue that transformation journey. Let me take you through the points of what it's done for us financially, because I think it is compelling, and again, it sets the foundation for what we're gonna do next. First, let's talk about the revenue growth. What you have here is four years of revenue. It would've been easy for me to go back to 2020 and anchor on that, but that was the COVID year, so I put 2019 on here. I'm telling you today, we are a radically different company than we were historically. I'm gonna prove it to you as we go through this. You can see the revenue growth that we've gone through here.
It's compelling. We doubled down on automotive and industrial. For 2022, we were almost 75% automotive industrial. In Q1, we were 79%, so we achieved our target there. Over the last two years, as we've executed our transformation, we've grown at a 26% CAGR over the last two years. At the same time, as you've seen, we've been doubling down on silicon carbide, and that's working out pretty well for us. Also, in these numbers, we've walked away from approximately $300 million of non-core business that was low margin business, highly competitive business. We've walked away from about $300 that's not reflected in these charts. If you take that forward to gross margin and operating margin, our gross margins since 2019 to 2022 have improved over 1,300 basis points.
This is again, we exited that low, low margin business. We executed our FAB-lite strategy, our FAB Li ter strategy. We also closed the gap on the price to value discrepancies. We are underselling the value of our products. We're below market. We've closed that gap, and we're now delivering the value out of the products that we deliver for our customers. You can see the trajectory. We ended 2022 at 49.2% gross margin for the year. If you look at the operating margin, you can now see that the operating margin actually is up in the realm of where historically the gross margins of the company used to be. That's why I'm telling you this is a radically different company, and we've structurally changed this to go forward and grow on.
Our operating income actually grew 5.3 times faster than revenue. We also reduced our OpEx through structural changes. We reallocated resources into high growth areas. You've seen us make some structural changes in there. We're running about 15% now, and that's through us being very intentional, taking out inefficiencies in the company, structurally changing it and doubling down in areas of growth. From a free cash flow standpoint, we've actually grown our free cash flow from 2019 to 2022, 10x. The same time we invested in silicon carbide, we've been bringing on our 300mm fab capabilities in East Fishkill. In 2022, we started buying back our shares again, and earlier this year we announced a $3 billion share repurchase program that was authorized by our board that will go out through 2025.
Clearly, cash flow is the outcome of everything that you've seen on this slide. Again, this sets the structure for what we're gonna do next. At the same time, if you look at what that means from an industry perspective, we're now generating industry-leading performance. The free cash flow during this time was number one among all the global semiconductor players. Our ROIC for 2022, which approached 39%, is the top four of global semis. Our operating income growth, which grew 3.7% or 3.7 times during this time, is in the top three. We're getting elite performance out of our structural changes as well. Taking the market, the sustainable ecosystem.
Hassan did a great job of describing this, focused on auto and industrial, intelligent power, intelligent sensing, the structural changes that we've made inside of our businesses. I'm gonna tell you now, we've tuned up these businesses to the point that they're premium businesses. They're high-value businesses. I'm gonna walk through each of them briefly here for you. You've seen each of Hassan, Simon, and Sudhir talk about these individually, I'm gonna sum it all up for you on why we think these are premium businesses. For silicon carbide, Simon didn't say it, we're gonna grow 2x the market at a 70% CAGR from 2022 to 2027. Staggering growth. We're vertically integrated. He talked a lot about that, we're taking advantage of our long history and our DNA of packaging and module capabilities.
We've been able to leverage that and accelerate our ramp very quickly. It's been in our DNA. We're ramping quickly. We've been very successful. I think Simon did a great job of talking about that. In silicon power, another great opportunity here for us. We're number two position in market share, and we're growing. We're expanding our BOM coverage. This is the area that we exited some of the price-sensitive business, so we will not be playing in that market. We've got about $350 million more to exit this year. By that time, we think we're done with the exits and that the business is totally tuned up at that point. For Power ICs, Sudhir talked about it. We're gonna double down on the $14 billion TAM. This is a TAM that has 70% gross margin.
It's a big opportunity for us to get a margin uplift out of this. We can optimize our system cost with gate drivers and controllers. We'll expand our leadership through the power tree. On the intelligent sensing side of things, I've talked a lot about this over the last several quarters. This is a totally different business. It's focused on auto and industrial. Sudhir walked through it. The team's done an amazing job of actually turning this business around. We've got number one position in auto and industrial. We're growing in machine vision, big opportunity for us. We're playing where we can win and where we can differentiate. We've got 68% market share in ADAS, 27% market share in industrial, so clearly the market leader. We've got 8-megapixel ramping, and the ASPs are two and a half times the 1 megapixel.
Big opportunity for an ASP uplift and a margin uplift there as well. We're number one in ultrasonic sensing and in inductive sensing. When you roll all this together, the team's delivering high-value products for our customers that are highly valued, high margin. We said on our last call that gross margins are now exceeding 50% in this business and continuing to go up, and it's the most profitable sensor business in the industry. We'll continue to expand on that. Taking the premium businesses, bringing it back into the market. This is going to drive our shareholder value. Our premium businesses will drive premium results. You've already seen it. I pointed out industry-leading results over the last several years. We'll continue to do that. You've seen our ability to execute.
When you think about it from a shareholder value, there's really 4 vectors we're gonna focus on. Revenue growth at 3x the market, expanding gross margin, accelerating free cash flow, and shareholder returns. I'll walk through each of these and give you a little bit more color. Starting with the revenue growth. I showed this slide, talked about the sustainable ecosystem. Our growth there is 21%. Overall, company is gonna grow 10%-12%. The market is forecasted during this timeframe of 2022-2027 to grow around 4%. That's the 3x the industry here. You can see where we're focused. You've seen us where we focus, we execute, where we exit, we win. Or execute, we win. We'll continue to do that. That gives you kind of the indication of what we're looking at.
Let's break this down a little bit further. We report in auto, industrial, and other. We've been doing that for a while here. Let's take the 10%-12%, and let's break that out on where we see that growth coming from. On the other market, there's a really key component in this bucket. It's our 5G and cloud power. That's forecasted to grow 22%, so that's significantly up from what we told you two years ago. That is a strategic part of our market, even though it is in other. We report it in other. On the other piece of that business, it's actually going to decline because we're not investing in the broader market. We're going to engage with strategic partners where we can provide differentiated solutions and win, but we're not going to go broad.
We'll be very opportunistic in this bucket. Over time, because we're not investing, it will slowly decline, but it's offset by the growth in cloud and 5G. Industrial, it'll grow 10% CAGR, 2022-2027. Industrial is energy infrastructure, factory automation, EV charging. Automotive, growing 19%, that gets you to the 10%-12%. In automotive, it's electrification obviously, silicon carbide, IGBT. It's increasing ADAS adoption for our sensors, and it's power management, LED, and advanced safety as well will drive growth in that market as well. That gets you to the 10%-12% over this timeframe. Now if we actually break it out by the product categories we talk about, we talk about intelligent power, intelligent sensing, and other. Let's build this one out as well. Same timeframe.
Intelligent power is going to grow 16%. Got SiC growing 2x the market, 70% CAGR. We expect by the time we get out to 2027, we've got 35%-40% market share of the SiC market. Silicon power, we've got share gains in auto and industrial, and we'll accelerate our power IC growth as well through controllers and gate drivers, which will drive growth as well for us. Intelligent sensing, I think we've covered this quite a bit here, but we're going to grow in machine vision, focused on auto and industrial. That's the majority of that business today. Advanced safety for ADAS and in-cabin vision, factory automation, and we'll be opportunistic in human vision. We're only going to play where we can win and differentiate. We're not going to get into the dog fights down in the low end.
Sudhir did a really nice job of showing the competitive advantages that we have at the high end, and that's where we're gonna continue to play. That's what will continue to drive high profits out of that business going long term and grow at 8%. On the other, I look at this as very attractive cash flow business. Sudhir called it accessorizing, but I think about it as kind of bedazzling, but it's the same thing. How do we have more BOM coverage? It's complementary to everything that we do as well. You've seen some videos from some of our customers. We have a highly diverse customer base, geographically diverse as well. You can see some of the names here, the marquee names in the industry. We're engaged with everybody that we should be engaged with. We've got a broad channel.
We've got a broad direct sales staff as well that covers these customers. The guys have talked about what customers need, but kind of stepping it up at a little bit more of a higher level, they need best-in-class performance. I think you saw that between all the slides. They need suppliers that can scale, reliable with them. They want fewer suppliers, not more. They want a tighter supply base. Everything's about system cost. How do we put it all together at a cost-effective solution for those customers? What we deliver is obviously the industry best products, most efficient products. We are vertically integrated. As Simon talked about, we acquired GTAT about 18 months ago, gave us vertical integration on the silicon carbide. That's very important to our customers.
We've got a broad portfolio of products, as you saw from the dots on the BMW i7. We optimize the solution cost across the BOM as well. Our top 20 customers represent about 35% of our revenue today. We have no 10% customer. Those top 20 customers, on average, purchase 800 products. That demonstrates the breadth of the portfolio. As we continue to expand, we'll continue to expand that number as well. You can think about us signing LTSAs that span multiple years with hundreds of part numbers, which gives us really good visibility and tight integration into our customers on how do we align with them. It's a more strategic relationship and partnership. Moving on to expanding the gross margin.
Almost two years ago, I stood on this stage, and I introduced the concept of FAB liter. I'm here to tell you that we're done with FAB liter. We've executed very well. We executed faster than we expected during this timeframe. If you ask, "What is FAB liter?" It was really getting more flexibility, getting a low cost, fixed cost structure for us to be able to leverage and grow off of. In 2022, we divested four fabs in one year. We thought it would take us much longer. We executed very quickly, and were able to divest those. We'll exit those four fabs over the course of a few years, transition that product into our existing network. During this timeframe, we actually took our capacity down by about 17% as we exited these fabs and tuned up our FAB liter strategy.
We're using our captive capacity, our internal capacity for differentiated products, and we're flexing to the outside for anything that is a common technology. What that allows us to do is make brownfield investments that drive a high ROIC and a high return off of those investments. The other thing that this does is by eliminating this fixed cost structure and optimizing, it allows us to reduce the gross margin variability and the volatility that we've seen in the company in the past. I think as you've seen with the market uncertainty and the softness, we've actually been holding margins. I think that's an example of what we can continue to do. The question is, FAB liter is done, what's next? Where are you gonna go next? While we were doing the FAB liter, we're actually growing revenue.
Although our internal capacity was going down, we were actually growing revenue, and that's a part of the mix. Now what's next? The next is fab right. We're gonna move from FAB liter to fab right. What's fab right? Fab right is we're gonna actually optimize now within our footprint. We're gonna drive efficiencies, and we'll continue to derive best-in-class ROIC on the investments that we make and on the assets that we deploy today. We've got to get the mix correct within the manufacturing footprint for the optimal cost structure. We've been talking about we've been moving IGBTs out of Korea into East Fishkill. It creates the capacity for silicon carbide. Very low investment at that point because it's brownfield investment again.
As we think about this, the red is where we're gonna grow strategically, silicon carbide, East Fishkill, primarily. We'll make those investments, but we've got great visibility as we actually bring capacity on because of the LTSAs. Again, we're very tied to the hip with our customers and their volumes, and that gives us the ability to know what's coming at us and build capacity as we need it, not build capacity and hope to fill it later. With this, the fab filler strategy of the past is dead. fab right is not about filling a fab, it's about optimizing the footprint. Also, our EFK, our 300mm fab is the only U.S.-based 300mm power semiconductor fab. We're also bringing sensing in there, that actually expands it as well.
At the same time as we do this, we'll extract that $160 million of fixed cost from the divested fabs during this time as well. That gives us a nice little margin tailwind. As we optimize here in Fab Right, we'll drive our cost down, reduce our fixed cost, optimize within the four walls, and drive investments that drive a higher return. We'll use external flex to optimize the internal and then flex to the outside when we need to. We'll use outside for common technologies. The goal here is to maximize our utilization first and then flex to the outside. We'll do that, and that allows us to go up and down with market conditions. Historically, we've manufactured about 65% of our product in-house. The last couple of quarters, we've been running somewhere between 55% and 60%.
When you get out a couple more years, we'll be back up to about 65% in-house again as we do this. Again, we've got that outside capacity up and down to modulate. Low investment, high return on that. That's Fab Right. That's the next strategy here when it comes to the manufacturing footprint. Fab Right is also our path to net zero. We've said we aspire to be net zero by 2040. We've been doing a lot of work on it. I think the team's done a great job. If you look at what we've done over just the last year, we've reduced our Scope 2 emissions by 21%. Our waste reduction has gone down 23% in 2022. Our water usage has gone down 19% in 2022.
As we continue to optimize our fab footprint in Fab Right, we expect to get an 18% savings on energy between now and 2027 when you think about it in terms of a kWh per wafer. We start to get more efficiency, which drives down energy consumption. This helps us start to achieve the net zero by 2040. There's obviously a lot more that we have to do, but this is definitely putting us on the path to achieve that. As we make investments, we think about this in terms of the equipment we bring in, the energy usage, and how we continue to drive that down as well. What does that do for gross margin? Our previous gross margin target was 48%-50%. You guys have already read the slide.
The next target, the next milestone here is 53%. How do we get there from where we are today? Anchoring back on 2022, 49.2%. We've talked about 2023 being a transition year. I'm not gonna touch on 2023. I'm gonna talk from 2022 to 2027 and talk through the different components. One of the drivers is just mix and new products. As we continue to focus on auto and industrial, that drives higher gross margin. You've heard us talking about expanding on new products, high margin products. That will continue to give us a margin uplift there as well. Next piece is silicon carbide. As we ramp silicon carbide, these margins are at or above the corporate average. We absorb that ramp cost that we do report in our non-GAAP results.
The last piece is Fab Right. Again, we recognize the $160 million of savings from the four divested fabs. We optimize, so we drive our cost down there as well. We also have a boost from utilization as utilization goes up over this time frame as well. We maximize that, and again, we flex to the outside where we need to, and it gives us that flexibility. That gets us to 53% gross margin by 2027. Going from gross margin to free cash flow, it all flows through to cash flow. We'll accelerate our free cash flow as well. I showed you this chart earlier on one of the previous pages. During this time frame from 2019 to 2022, 10x growth in free cash flow.
We're looking at free cash flow margins of 25%-30%, we're upping that target as well. How do we get there? Outgrowing the industry by 3x, expanding the gross margin as I just covered. When we think about our operating expenses is very targeted, focused on investments. We'll continue to make investments in R&D. High ROIC investments. I'll show you more on that. Working capital management, which includes inventory as well. By doing that, we're looking at 2x growth in our free cash flow. You guys are all sitting there probably going through your models trying to figure out what that means, I'll just give you the punchline. $3.5 billion-$4 billion of free cash flow in 2027. The market success drives our investments.
We will continue to invest in Fab Right, in our 300mm fab. We'll expand for strategic growth, as we've talked about. We're gonna ramp 300mm, expand there, and we're gonna augment by investments in packaging as well, in advancing that as well. The other piece that we're gonna be looking at doing for the next two to three years is a $2 billion investment in end-to-end silicon carbide. We're currently evaluating the U.S., Europe, specifically the Czech Republic, and Korea. This is where we've got a footprint already. We can do additional brownfield investments there, which reduce the capital intensity. What does this all mean for capital intensity? Sorry. Capital intensity will be 15%-20%, consistent with what we've been saying, through 2024. Then it'll decline over time to about 11% through 2027.
That's higher than what we had previously, but Simon talked about $9 billion of silicon carbide LTSAs. As we continue to have success in the market, we'll continue to make those investments, and those are high ROIC investments. The LTSAs that we have with our customers minimize the risk of building capacity and not being able to fill it. Any government subsidies will be a benefit. It'll offset the CapEx and enhance our ROIC. As we think about our investments, we'll continue to invest for growth while focusing on best-in-class ROIC. Speaking of ROIC, how have we been doing? I mentioned it earlier, nearly 39% in 2022. Here's similar chart to what I showed you earlier, 2019-2022. If you went back further, the company, you know, ROIC was pretty close to the cost of capital.
We're now best in class. You say, "Okay, well, how did we do this? We've been expanding rapidly. How were we able to do this?" If you look at our capital intensity in 2022, we were about 12%. That's about half of the median of our peers. Again, how do we do that? It's through these brownfield investments. Brownfield investment is about 40% less capital compared to a greenfield, where we're actually bringing infrastructure in, shovels in the ground. What we're able to do is actually leverage that footprint, that infrastructure, and continue to expand. As we think about that $2 billion of expansion, it's that type of expansion, leveraging the infrastructure. Oh, by the way, while we do that, we accelerate time to market by two years.
We get questions on how were you able to ramp so quickly on silicon carbide? This is it. It's that we had less capital required to ramp, and we are faster time to market because we are leveraging our footprint. That's the value of the network we have that we'll continue to use. Shareholder returns. Gonna create a lot of cash flow. What are we gonna do with it? If you think about it from an investment and a capital allocation, we'll continue to invest in our business. We're gonna invest for growth, expand our competitive advantage. I've talked a lot about ROIC. M&A, we are in a consolidating industry. We'll look at strategic M&A. You've seen us do it already with GTAT and seen that we can actually execute very well. We'll look at technology tuck-ins where it makes sense.
It will also be accretive to our financial model, it's gotta be strategically important and financially accretive as well. Because we have low leverage, we've got firepower if and when we wanna make a move. Third piece is just balance sheet flexibility. We'll maintain the flexibility to continue to invest back in our businesses. We're going to keep our credit ratings the same, BB+, Baa1, no change there. We've got shareholder returns. Our policy remains unchanged here. 50% of our free cash flow will return to our shareholders. As I mentioned earlier, we have $3 billion authorization through 2025. You can see even as of last quarter, we are putting this to work. What does that mean from a financial model?
anchoring on 2022, $8.3 billion, 49.2% gross margin. You can read it. We look forward, you've already seen the 10%-12% in terms of revenue growth, growing 3x the industry. 53% gross margin. This is through mix, fab right, silicon carbide ramping. Our OpEx will be 13%. This demonstrates the leverage in our model. As we drive this, our OpEx dollars are actually going to continue to increase, but our percentage will decline because of the leverage that we have there. We'll continue to invest in our people. We're also going digital first, where we're investing in new capabilities in data analytics. Our operating margin, if you do the math, actually will outgrow the revenue growth. CapEx, 11%.
Previously, that was at nine, we did take it up slightly, but we also took up our free cash flow margin from 25% to 30%. Strong returns to shareholders. As I said earlier, about $3.5 billion-$4 billion of free cash flow by the time we get to 2027. Premium businesses driving premium results. Take the markets, intelligent power, intelligent sensing, automotive, industrial, cloud, 5G. Take our premium businesses, silicon carbide, silicon power ICs, image sensors, put that all together, you got premium results and shareholder value. All right, we'll open it up to some questions here. We've got a couple mics, so raise your hand, and we'll get the mic runners to run them out.
Thanks. Vivek Arya from BofA Securities. Thanks for a very informative analyst day. I actually had two questions. One, you know, very strong gross margin performance over the last few years, but that was also a time when your customers were dealing with a lot of shortages, right? Your, you know, yourself, your peers benefited from pricing as one of the key drivers of growth. What is your assumption about industry pricing as part of that 10%-12% growth? Do you think it stays supportive? Then I have a quick follow-up.
Yeah. Look, we look at pricing as, stable, because just look at it from the LTSA perspective. Our LTSAs are multi-year, in duration, where both volume and pricing is locked in, so that anchors on the value. Number two is the new products. You know, if you look in that time horizon, as new products become a more meaningful part of revenue than they are today, we know those products already are accretive
Last but not least, a big driver of new products is on silicon carbide, and as they scale, we've always said it's at or ahead of the corporate margin. All of these together give us that confidence that pricing is not a detractor in it. It's very important to note when you talk about the environment in the last few years, we talk about pricing, we talk about pricing very differently than some of our peers. When we talk about pricing, it's really that price-to-value discrepancy.
Proof in point, you know, the stuff that we wanted to lose, you know, we, quote-unquote, "price ourselves out of the market," yet here we are not at the same rate, that gives you where the value is coming from, and even customers are willing to pay that higher price because of the value. It's bringing those value up, and that's sustainable over that long period of time.
All right. For my follow-up, Hassane, if I did the math right on silicon carbide and your market share assumptions, you're doing roughly, you know, $1 billion, right? That's your target for this year. If I follow it out to 2027, it seems like it'll be about $2.7 billion-$2.8 billion roughly in that year as kind of that run rate. When I look at market share, it stays in that 35% ± throughout that timeframe. The question is that, you know, is your assumption that the market dynamics of silicon carbide, the competitive landscape essentially stays as it is over the next several years?
Do you think as more players enter into this market, bring on more capacity, you know, are you really able to hold to that 35% market share? Is that kind of a bottoms up or a top-down, number?
Most of it is, I would say, bottoms up, because it's, again, it's supported by the LTSAs which extend through that period of time. Let me just cover a little bit on the market dynamic. You know, if you look at IGBT today, there are two players and then everybody else. When I say everybody else, you're talking about low single-digit % for everybody else except the top two players. Why? Because of everything that Simon talked about. Silicon carbide is now a different level. It's new. Talking about in the next, even let's talk just the next five years. Talking the next five years about the competitive landscape is going to be any different than it is today, let first the current competitive landscape scale properly. I know what we're doing.
Others have some challenges on their own. We don't see that changing. We don't see new entrants. We're not gonna see new entrants that are credible at the high quality, reliability, and the scalability, you know, the S's that the four S's that Simon talked about. That if I extend beyond, you know, that window that we talked about through 2027, I mean, we're projecting from an EV market perspective, 50% penetration by 2030, you know, give or take a few years depending on which report you go to. The runway of that growth is beyond even 2027, and that's what's gonna keep propelling us through that. Again, it's not a, it's not a destination. It's a milestone. We do see our strength.
We have established a very strong infrastructure and a very strong foundation today, with the $1 billion in 2023 and the LTSAs moving forward, and we're gonna sustain that.
Thank you.
Hi, guys. Over here on your right. Ross Seymore from Deutsche Bank. A couple gross margin questions. First one, I think for Thad. Two years ago, when we all sat here, you talked about a 45% peak target, you said you hope to hold a 40% trough cycle target. Now it's 53% on the high end. I think recently on calls, you've talked about a 45% trough. How does that trough change going forward? How much variability cyclically do you expect in the gross margin between now and that 53 mark in 2027?
Well, look, there's a lot of uncertainty in the market right now, and we're holding our gross margins, right? We've said that our trough should be at the mid-40%. We continue to execute that. Now as over time, as we execute towards that 53%, my assumption would be that trough continues to come up in a relative basis. Right now, I think we're testing it, and we're proving that we're holding it.
Great. I guess a follow-up for Simon. You had a nice, I guess, pyramid chart talking about the value going up and the volumes rising, different ways you guys attack that, which are IGBTs and silicon carbide, et cetera. Can you just talk a little bit about the value contribution at the high end per vehicle versus the low end, how much variability that is between that? Maybe whether it's you or Thad, is there a huge gross margin difference between that? 'Cause investors are worried that the mix of vehicles have been so high-end over the last few years. As we go to lower priced cars, you can play in that, but is there a negative financial implication? Thank you.
Sure. If you, if you look at that ultra-high performance, for example, that's using the quad inverter solution, that's got a lot of content in it, and in cases it's north of, you know, $1,500. That ranges from $1,500 or so down to several hundred dollars for a single inverter-based solution. That'll also change between silicon carbide and IGBT. The, the value proposition remains the same throughout that stack, right? Whether it's the ultra-high performance or the entry level, there's a value that the customer is trying to solve for, the margin is quite stable across that.
Thanks. Thanks for taking my question. This is Gary Mobley at Wells Fargo Securities. I wanted to maybe read between the lines from what you wrapped up in your prepared remarks about the $2 billion in silicon carbide investment. That sounds like it might be a new greenfield wafer processing facility. My question is, do you see the brownfield investments that you've made so far to support the revenue you have today and maybe a few years into the future, do you see that having some limitation on the revenue? What is the risk of not investing in greenfield quick enough?
First of all, the $2 billion is brownfield. The three locations that we're looking at is where we have infrastructure already, and we have a footprint. It's end-to-end is the way you should think about it for silicon carbide. We believe there's a lot of brownfield investments we can do before we have to move to a greenfield investment, right? We have that large footprint, we can continue to leverage it, but that is clearly brownfield, which will reduce the capital and again, high return on that investment.
That's a pure end-to-end type solution, wafer growth or boule growth, wafering, epitaxial growth and fab. The whole thing. The full suite.
Thanks, gang. over here, love to see the 40% operating margins. I just had 2 brief questions. The first is just to drill down on the pricing comment you made earlier. Are you guys assuming that the semiconductor industries you operate in go back to normal historic pricing and given the LTSAs, you blow right through it or something else? Just want a little more clarity there and a quick follow-up.
Yeah. I can't speak generally on the semiconductor market, I can speak on the businesses that we're operating in, specifically. One, obviously the LTSAs gives us the visibility, the stability, and really the outlook on where the pricing is. Our new products give us the layer on top of that. Beyond that, we are operating where we bring value, which means this expectation of, you know, on January 1st, you know, set your clock, January 1st, different price for us on our approach and the mindset that we have with the customer, those days are over for us. There are efficiencies that will be gained.
We are, you know, you heard, us talking about efficiencies, whether it's fab output or cost or energy costs, those are efficiencies that we will capture, and those will give us more competitive new products. An expectation of pricing just for the sake of, pricing, that's the, I would say, the old world.
Great. Then just briefly on the $9 billion in LTSAs in silicon carbide, I assume that's not the next 12 months. Can you just tell us how far out that goes?
It's a range. LTSA is a range of years. Typically, they're three to five years. The number for the $9 billion, it's lifetime. Vast majority of that is over the next five years.
Thanks.
Yeah, thank you. Roger Gill from Needham & Company. Great presentation. Just on the margin commentary. You mentioned that mix is a big part of it. As of the most recent quarter, less than 80% of your revenue is coming from auto and industrial already today. What are the expectations of auto industrial as a percentage of total revenue over the next two, three years? Is that gonna go up to 90% or maybe you could elaborate that.
Yeah. You're right. Last quarter, 79% of our business was auto and industrial. I would tell you part of the reason for that was the other markets were soft right now, right? I think when you return to normal, you're going to see auto and industrial growing at a faster rate, obviously. If you model that out over the time horizon, you're looking at, you know, auto and industrial being north of 85% of the total company.
Also the one thing, you know, when we talk about mix, it's not necessarily a market mix or a market shift mix. You also have the new product mix, 'cause that is again, you know, we've been doing a lot of investments in the last few years. We've, you know, Thad and I, you've heard us talk about how we focus on new product from a margin perspective tied to the value, you know, that Sudhir and Simon talked about. There is a mix within those markets that's also going to drive the margin.
Hi, Will Stein from Truist Securities. Thanks so much for a very informative analyst day and a great message. First, I'd like to ask about new products. You talked about Time-of-Flight sensors. Is that Lidar? Is that the same thing, or is this a different sort of sensor that maybe some of us are less aware of?
Time-of-Flight sensors is the basis for Lidar, yeah. It is.
Okay, great. The follow-up is, maybe a different way to look at, I think the question that some people have been focused on around margins, but less around margins, more of this concept that the whole industry may be perceived to be overearning a little bit right now. You know, the other way you get there is through units. I wonder where you perceive ON to be in that cycle, when you think about inventory levels on your own balance sheet and at customers. Should we expect a sort of meaningful correction on the path to 2027? Do you think the whole supply chain is actually much tighter than people are concerned about?
Yeah. I mean, I'll cover a little bit on the dynamic and then Thad, if you have any more info. Overall, we've been taking a very cautious approach on inventory. You know, you've heard us talk quarter after quarter about inventory, whether it's in the channel or with the customer, including as late as last quarter. We're gonna continue to do that. We're not doing that, you know, it's, you know, you call it good years, soft years. We do that as a matter of structure. Structurally in the company, our approach with the customer for managing inventory is to solve a very specific problem. What you'll see us do is be very dynamic with it.
When we have a very strong ramp coming up, we're going to stage inventory. silicon carbide is a perfect example where a lot of, you know, you mentioned balance sheet.
A lot of the inventory on the balance sheet is the growth in days and dollars is silicon carbide getting ready not just for the $1 billion, but for next year's growth on top of the $1 billion that we're projecting in 2023. You're gonna see us use that as an asset in order to stage and support our growth. You're not gonna see us, for example, doing it blindly from a manufacturing perspective. You know, the fab filler days are over, and you've seen us do exactly that even last quarter when we talked about utilization being at 71%. We're going to manage very, very tightly agnostic of where we are in, you know, up or down years from an industry perspective.
We're gonna use this as a tool for our own business as we wanna run it and as we see it with the LTSAs, which is a very, very clear view of where demand is.
Yeah, I think you're absolutely right. I think if you look at what we started doing Q2 of last year is we slowed starts down, right? We slowed it down before the market got soft. If you look at the inventory on the balance sheet, we're building strategic inventory, as Yazan said, for silicon carbide and also for our fab transitions. If you look at the rest of the inventory, you know, it's running lean. The channel's running lean as well. I think you're gonna see us be much more proactive on that side of things than reactionary.
Hi, Josh Buchalter from TD Cowen. Thanks for taking my question. I wanted to ask about silicon carbide on the substrate side. First, can you confirm that we're still on track for 50% on internal substrates exiting this year? Then longer term, you know, what's the right ratio of how we should be thinking about your substrate mix internal versus merchant to hit your longer-term targets? Thank you.
Sure. Certainly I'll confirm that we'll exit this year north of 50%, so we're still on target for that. That percentage will grow from internal substrates, while we'll also keep a variety of external substrate capability for flex capacity. I think that's important to do. But with internal substrates, we don't have the margin stacking, right? We're gonna keep driving that number up while still keeping the flex capacity available.
Chris Rollins, Susquehanna. Thanks for the day. Regarding silicon carbide, I guess I'll just throw out a couple hot topics, some that you mentioned, just to get your thoughts on. The first is, I think you mentioned trench. Would love to know how you feel about that, the reduction in die size. Second is wafer splitting on the silicon carbide side, feasibility, economics behind it. Last is Chinese wafers, Infineon now using 2, that they think are pretty good. How do these things affect the industry and your outlook? Thanks.
All right. I'll start off with the trench. We looked at the different generations of technology, and M4 technology for us will be a trench-based technology. I'll also remind you to think about it a little bit differently in that not all trench is the same and not all FETs are the same. That figure of merit that I showed, that performance capability, that was comparison of M3, which is planar, against competitors and some of those competitors with trench. We still outperformed with planar, our planar versus competitor trench. Trench isn't necessarily a magic bullet. However, it is a step forward, and we've seen it in silicon. We will see it in silicon carbide as well, but not all trench will be the same.
One of the concerns that you have to have in widespread adoption is planar is very well known. It's very rugged. With trench, there's other issues that you have to deal with, especially things like current crowding. We have a very unique approach to taking care of things like that. At the right time, we'll be introducing trench to the market, very reliability, very highly reliable. We can scale it, and the performance levels go to the next level. Not everything is the same in the market. Your second question, please?
Wafer splitting.
Splitting. Yeah. Splitting, that's another technology development that we are aware of, that we're working with various partners with. We've also got internal developments around those areas. At the right time and the right place, it could make sense, but we have to look at what can it give? Is it reliable? Can we do it at a large scale, et cetera, et cetera. We're aware of all the developments going on in the outside world. We're also conducting our own developments internally. Right place, right time, we'll be there if it makes sense.
Chinese?
Chinese.
Meant for, yeah.
I'll start off with, as I said before, we have multiple external suppliers. Some of those are Chinese. We understand where things are at. You have to take a look at the bigger picture is can you reliably get high volume or potentially are you getting the cherry-picked stuff? Just something you have to think about. At the end of the day, I know that when I sign an LTSA with a customer, right? That's an obligation that I have with that customer. That's a trust factor, and I know where I can place my bets and feel very confident that we can supply. I think having that internal capability is paramount. Could you get it elsewhere? Yes. Would I bet the company on it? No, I would not.
Hi. Thanks. Jed Dorsheimer from William Blair. Thanks for the presentation and also taking my question. First question, just as a follow-up to the previous gentleman's question on the substrate capacity. It looks like you've had a remarkable progress in the yield increase, the 1.7x. If I look at the $1 billion of target, that's about 250,000 wafer starts. When we were up at Hudson, it's roughly, you know, had space of about 400 furnaces, plus or minus. I guess my question is, most of the yield increase of that 1.7x on the substrate that you've seen, 'cause it seems like your yields downstream are pretty good, you know, all along. I'm just curious, where has the progress been?
Then I have one follow-up question.
Sure. I'll try and address that. It, it's not one area, right? It, it's this incredible, very deliberate focus that we have across boule growth, across wafering and across fabrication. That 1.7x represents across that entire manufacturing spectrum. It's important that we do that because focusing only on one area means you're not getting that full capability. We have to complete the full product. We attack on all avenues at the same time. It's incredibly important.
Thank you.
Got one more.
Thank you. Tore Svanberg from Stifel. My first question is for Simon. Simon, as we think about that additional $14 billion TAM in power, and specifically the controller part of the market, there's some pretty established players there already. I was just wondering if you could, you know, add a little bit on how you intend to capture that additional TAM with some share gains.
Sure. I think maybe you intended to ask that question of me. The point that I think I was trying to make there is, if you have the silicon carbide or silicon power switch and you tightly couple the gate driver, and then as well tightly couple the controller such that you have three ingredients that are optimized to work together to solve those customer problems and deliver the efficiency, it's that optimization that becomes the source of competitive advantage. The optimization is only possible if you have the fundamental power switch, be it silicon carbide or silicon.
Thank you. My follow-up is for Hassan. Hassan, I saw your list of customers. There's a few Chinese on there, and I was just hoping you could just talk to us about either strategically or philosophically, how you feel about, you know, going after more Chinese customers over the next few years given the current geopolitical turmoil. Thanks.
Yeah, look, from our side, we've always said, and we've been consistent, that whether it's LTSAs or even our revenue, we're geographically distributed, customer distributed within those geographies and even within a customer, platform distributed. Because it's not just about the geopolitical or kinda geo-geography. Even within a customer, you have to be distributed over a platform because what if this model doesn't really hit the market, but the other one does? You have to win with the customer across the board. That is very important from a strategy perspective, which goes back to the diversification of our go-to-market. I talked a little bit on the formula, the go-to-market being deep, broad, and very fast with the tools that, you know, Sudhir mentioned and Simon mentioned, whether it's simulation or SDK for image sensing.
That's from a strategy perspective, it's consistent across the board. From a, from a geopolitical aspect, obviously, we look at it just like everybody else, a lot of our peers, where it's two ways. From a manufacturing, we are in a very, very good spot. We, you know, manufacturing, we have it across the board. Our fab right, which is optimizing, where we have the same device in multiple fabs. Because again, for us and business continuity and really for the customer, it's not just about geopolitical. What if you have a flood? You have to have that continuity. We look at when you look at it from a business continuity perspective, it really wipes out any geopolitical 'cause you have to do the right thing for the company regardless of geopolitics, and that's how we have been approaching it across the board.
All right. We probably have time for two more questions. Yeah, let's go.
I could tell Parag was reluctant to give me the microphone. Thanks for allowing me to ask a follow question. So compliments to establishing a great automotive business, but there are some holes that you could fill, no doubt. I'd love to get your updated thoughts on the radar sensor market and then maybe extending the conversation further into the compute domain as it relates to fusion processing for all the sensor modalities.
Yeah. I'll start with that. We don't have any intention to do radar. It's something that we looked at in the past, and we're making the very conscious decision not to pursue radar as a sensing modality. We did talk about depth sensing and of course, depth sensing, using Time-of-Flight and just to elaborate on the prior answer. Basically, what we're talking about is measuring the Time-of-Flight of optical signals. In that realm, when we're doing that, we actually do some of the processing. We do the image signal processing for traditional image sensing. We do processing as well to measure the Time-of-Flight for specifically optical images that relates to the depth sensing capability.
To the extent that we combine those together at the system level, there's sensor fusion that happens there. What we're not doing is, talking about moving to that next wave of really complex processing yet.
You know, from a strategy perspective, obviously, you know, I anchor always, you know, personally and for the corporate, for the company, we anchor on focus. Do what you do it well, double down and focus on it. That's part of, you know, what Sudhir mentioned. We go deep, we lead, and we innovate.
All right, last question. Parag's got it.
Hey guys. First of all, thank you for three-plus tremendous years of execution. We appreciate that, so do investors. Last time, we heard a lot about silicon carbide. I felt like this time we heard a lot more about silicon and packaging. I'm assuming that you're pretty happy with where silicon carbide is in the state of affairs. The question is, when you talk to your customers, how important is the need for vertical stack of silicon carbide? I think you've taken share from non-vertical silicon carbide producers. Do you expect to continue to take more share as time goes on? Secondly, as a house having silicon carbide, and packaging under one roof, are there other competitors, U.S. or foreign, that you can think of that can challenge you in the future with or without government subsidies?
I've got a follow-up.
I'll start the answer there. If I think about silicon carbide, people tend to hyper-focus on the device itself, and it's incredibly important, and that's where I showed we have the differentiation on the die technology, giving the performance advantage. You need to think of it like a tripod to be successful, right? You need boule growth, you need the device, and you need the packaging. Those three things have to be there to really give the value proposition to the end customer. Today we wanted to talk about silicon carbide, but there's silicon carbide out there. Okay, that's fine. Can you get it in the right package, right? That's the next step. It doesn't just float in midair as a piece of silicon carbide.
You have to get the right package in it, that's really bringing the value up and why the vast majority of our revenue today is coming out of silicon carbide modules. If we look at the competitive landscape, I think we stand alone, and we've looked worldwide, right? It's not hubris, it's just a fact. That's where we are.
All right. I think that wraps things up. We invite you to join us in the demo area, take a look at everything we have there, and thanks again for coming out today.
My pleasure. Just before. Okay. Before I lose everybody, I just wanna take a few minutes. We've had two hell of a year. It's a two positive years, against a very challenging environment. I know from a P&L and from results and so on, it's been great. I can tell you it's been a hell of a lift for our employees who are worldwide, that are going through a transformation, through a journey while managing this backdrop. I do wanna thank everybody in the company for supporting the transformation, delivering the results with their unwavering commitment. Specifically, of course, my direct team, some of which are here.
We have a solid executive team that is again focused and committed to achieving those results and the ones we just highlighted. More importantly, the team that's been pumping day and night to deliver this event here that you see here, but also downstairs. I just wanna take that time to say thank you, and enjoy the rest of the evening.