Good morning, everyone. Welcome to our 2017 Analyst Day in Phoenix. I also welcome those who are joining us on webcast. As many of you know, my name is Parag Agarwal, and I'm Head of Corporate Development and Investor Relations at ON Semiconductor. We have a very impressive agenda today. We will start the presentations with Keith Jackson, our President and CEO, who will provide a strategic overview of the company. His presentations will be followed by presentations from three business unit heads, who will talk about their product line. Finally, Bernard Gutmann, our CFO, will close out the presentations with financial overview. We have allocated sufficient time for Q&A, feel free to ask away. Now let me turn your attention to the obligatory safe harbor statement. I'm not going to read it, but feel free to read it at your leisure.
With that, it is my pleasure to welcome Keith Jackson, our President and CEO.
Good morning, everyone.
Good morning.
I can tell you we've got a day which, as we were putting together these materials, excited us. We think the company is in a great position for sustained growth and profitability and cash generation, we're going to tell you a bit of that story today. As Parag said, we're glad to take questions as we go along. Key focus for us, as we look at our strategic directions, is making sure we continue to be an execution machine, doing the things we know how to do well, and taking advantage of all the investments we've been making to generate additional cash flow and bring that to our shareholders, through increased value in rapidly paying back our debt right now. What's happened with the company, we have really been transforming ourselves. We're along a journey. It's not through yet, but we're well into it.
To go from being a supplier of many multi-sourced products to one that is really a key enabler in analog and power, we believe we've made a great distance and journey on that already. Most of the strategic markets we're focused on today are really reliant upon ON Semiconductor and our specific technology to drive their markets. Secondly, we have built a large company. It's definitely diversified, both from a market, customer, technology, and product perspective. We have a lot of opportunities, and we're able to optimize our investments to the highest growth and highest margin areas, which gives us the opportunity to, again, without growth in the markets, improve our margin profiles. The Fairchild acquisition is one that we are extremely excited about, that is going very well for us.
We've had nothing but positive surprises, we'll give you an update on that as we go through the day. Lastly, just making sure we reinforce, that really the opportunity this year and next year and the year after is for margin expansion and accelerated free cash generation. I talked about the evolution. This just gives you a quick snapshot. If you go back to 2006, the company was less than half analog power and sensors. If you come fast-forward to fourth quarter of this year, over two-thirds of the company was in that category. We've been definitely focusing both our organic and inorganic efforts, in those areas where we think have the highest growth rate for the industry, and also making sure we have the technologies that are most relevant for those industries so we can drive margins.
From a market mix perspective, you can see we were just over half of our focus markets of auto, industrial, and communications in 2006. Now, again, three-quarters of our business are in these marketplaces. It's not just something that's on our wish list. It's not just a strategy on paper. It's things we've been executing to. Now we think we have most of our assets in the sweet part of the market that will get the fastest growth in semiconductors over the next few years. I mentioned diversified from a customer perspective. This is also pretty key. There are cycles in the business, what we've observed even beyond cycles is that there are winners and losers, and they change places from a customer perspective from time to time. What we've shown you here is the size of our top customers.
You'll see that none of them are bigger than 5%. You'll see a mixture from communications, automotive, and industrial on this chart. We have exposure and good content at everyone in the markets that matter, and certainly even those that matter the most to us do not have substantial or overarching impact if they fail or stutter in their marketplace. I talked about innovation. This is something that is really key if you're going to be driving margins. At the end of the presentation today, Bernard will be talking about how we're driving the margins up. We are in competitive markets. We'll always be in competitive markets. What differentiates you? There are several levels, I'll talk about those today, but one of them clearly is our ability to enable innovation. We've been investing in advanced technologies that are really enabling our customers to reach new heights.
All of you are very aware of ADAS in automotive. You're also very aware of the image sensing portions of that. We're going to talk to you about how we provide the content and enablement for all of those systems. Vehicle electrification. Yes, it's something that is growing quite rapidly. You may not know just how significant that is from a semiconductor perspective. An electric vehicle can have $200 plus more content for us than a straight hybrid or combustion engine. From other disruptive technologies, there's other things in cars. Of course, there's the push for increased gas mileage at all times. We've got some new parts now that will enable our customers to remove tens of meters of cable from their car, removing hundreds of pounds, and this is something, again, that we find disruptive, something that we're uniquely positioned for, and gives us confidence in our growth there.
Industrial market. This is going to be changing rapidly. If you look at the big motors that are involved in the industrial market, that's where most of the energy losses occur in industry. We've got not only all the technologies they need for today, we're also leaders in the new wide bandgap technologies in GaN and silicon carbide, enabling for them solutions for the future. We're also automating our industrial constantly. One of the big things that has been happening is reaching some standards on wireless connectivity in the industrial space. We've got all the presence we need there to enable our customers to do monitoring with the sensors, communicate with their CPUs, and then actually manage power in all forms. We've got end-to-end connectivity in industrial. Lastly, in communications, certainly there is a slowing of the handset growth.
We understand that, but we still have some significant opportunities to increase our dollar content in each of those phones. We'll be talking about several of those today. They're not limited to things like wireless charging or fast charging, but we think there's some ways we can continue to drive content there. The USB Type-C, which is not just a phone phenomenon, it's going to be on all mobile devices. There's some great content for us that we think also drives that much above the market rates. We think we have opportunities with our focus markets to outgrow semiconductors. What are the priorities? Growing profitably. We get a tremendous amount of leverage with our model with growth, and by investing and growing in the profitable areas, you can accelerate the cash generation portion of that.
We will continue to invest in those areas from a market perspective and in the leading technologies from a product perspective. We've now got some great size and scale, that really helps us from a sales reach and distribution partnerships. Our network of global Field Applications Engineers has been growing at double-digit paces for many, many years now. We can not only take our great products to our customers, but we can help them design their systems, getting to market faster, ensuring that we get a good share. From a cost structure perspective, we'll talk a little bit more about that later, we've got a tremendous amount of leverage there. We are one of the largest producers of products in the marketplace, if not the largest.
An operation there that focuses on just getting points of improvement on the margin side has a great deal to do with the drop-through we get to the bottom line. Fairchild has given us new opportunities to accelerate those reductions, we'll talk about those. The manufacturing operation is focused on net margin improvements, all of this should be accelerating our free cash flow. Lots of questions on Fairchild. One of the ones I always get is, "Well, what's been different than you expected?" My answer is pretty consistent. There's really only two things that were different than we expected. One of them is they had some great technologies that we're going to be able to leverage with ON's sales force and manufacturing base. We're pretty excited about that. The second was the sales opportunities that we see out there.
That was a bit of a surprise. We knew they had been losing share. One of our key concerns was how do we turn that around? As we went out and talked to their customers and the distributors, we found very willing partners in that. In our very first full quarter of Fairchild, the bookings we did for their products were up 49% year-on-year. This has been a very pleasant surprise for us. We're very excited about that. That is helping us raise our estimates, which Bernard will share with you later, on the synergies and cash flow we can drive from that. From a gross margin perspective, really no surprises. We knew there were big opportunities with ON's manufacturing network to start squeezing costs. We've already been doing that.
We got a very rapid start on it because of the delay in closing the deal. We're well ahead of plan, we're quite excited about the Fairchild integration. What did it really bring to us? Really, on the power side, you're going to hear today the difference the portfolio made on the technology side. ON was largely a low voltage and some mid-voltage types of power solutions. Fairchild had been focusing on the high voltage side of that equation. What it makes for us now is a provider of power products that can do everything our customers need them to do. Scale and cost structure, we were already quite large, shipping over 1 billion units a week last year. We're now going to be upping that as we go through this year to more than 1.5 billion as we exit the year.
These are rates not seen in the rest of the industry. It gives us the economies of scale and the leverage with suppliers that we need to continue to drive margins. Key technologies, you'll hear us talk a lot about power, of course, but also when you look at the silicon carbide, which that adds to our portfolio in the USB Type-C products they have that complete the full portfolio in USB Type-C, making ON one of the few folks that can do end-to-end solutions. Lastly, with all of this goodness, the free cash flow, we should be seeing an extra $260 million a year just by 2019. I also get a lot of questions on industry consolidation. We certainly believe the industry will continue to consolidate. There are way too many inefficient players out there.
The industry certainly has much opportunity to return more to shareholders with consolidation. Scale is relative. I think right now there's a lot of digestion going on by big players, but these big players, at some point in time, are wanting to get bigger. The market dynamics have not really changed in the last year, so at some point in time, that'll reengage. We're feeling very comfortable at this stage that we have the size and scale we need to grow faster than our competitors. Nonetheless, we'll always be looking at when we should participate and how we should participate. What I can tell you is we certainly will be only looking at opportunities that give us significant returns for our shareholders, and the investment yielding way more than our cost of capital.
From a collateral benefit perspective, integrations and consolidation in industry have lots of good benefits, mostly focused around cost reductions. Customers don't necessarily embrace those as well. Unless you happen to be taking out a weak player they need, generally, it's disruptive. We're taking advantage of that with many of the other ones that are going on around the world, and specifically those where there has been foreign entities involved, creating nervousness around supply. We think there's some near-term opportunities for us to outgain market share on those competitors. What is the advantage that we have? Well, I certainly talk about that size and scale. If we do more than 70 billion units this year, I don't believe anyone else in the industry will touch that, and just that scale and magnitude gives us cost advantage.
From a power perspective, I've already mentioned the breadth of the portfolio, I've already mentioned we're leaders in the front end of the technology. Our products stack up against the best in the world, and we believe that they're differentiated to help our customers get a differentiation for their products. Enabling new technologies. You will find ON Semiconductor at the forefront of all of the new technologies. If you want to talk about the virtual reality, augmented reality, drones, I mean, you name it, there is no portion that you will find in the world that does not have an ON Semiconductor part helping enable their new technologies. Broad market portfolio, 84,000 products. That's a lot.
More than any one person can keep track of, what I can tell you, we've got teams of FAEs and applications people around the world who are helping our customers design their systems faster using an ON solution with our superior technologies, and we think this is a competitive advantage that most of our competition cannot match. Lastly, on that cost structure front, just to hit that one more time, the leverage that we get with a tenth of a penny a unit in savings is just tremendous. We have the opportunity to do that. It's a part of our DNA. Each year we set out, the operations team is tasked with getting at least 6% improvement, like on like, each year they have exceeded that expectation. Not dwelling on a specific application, just giving you an example of what all this means.
We talked about ADAS, one of the fastest-growing applications in the marketplace. Everybody's focused on the image sensors, that's great. I think we've got the leading position there with the leading market share, the leading technologies. Literally, if you look at power and analog, there's a lot of components there required to make those sensor systems work. ON Semiconductor, again, has the leading product in each of those categories, which basically leverages our strength from a dollar content perspective when we win these systems. How do we provide that value? How would our customers describe us? Certainly, they see us as critical to their solutions in power analog and sensors. They come to us for that. They work with us to see how we can make their products better, at lower costs, faster time to market.
All those things I've been talking about help us do that. Providing the full solutions does matter. A lot of these customers have very complex bills of materials. They're trying to bring things together. If they can operate with fewer people, it lowers their cost of ownership also, gives them much more comfort if they can rely on those semiconductor suppliers to meet all of their needs. Generally speaking, they would prefer to have fewer, larger suppliers. Lastly, we talked about operational efficiency. It's something that we've been driving for many, many years. That enables us to remain cost competitive and still drive margins upward. What are the markets? We touched on them earlier. In automotive, we believe there's high single-digit sustainable growth. This is not driven by any growth in the number of cars being built each year.
This is all about the content drive that consumers have around advanced safety, new convenience features, more entertainment features. We've talked about some of the specifics. I won't go through all of those, we have engagement at the OEM level, as well as the tier 1 level, in all of the major customers around the world, all of their major new systems. This gives us the confidence that we will be growing there. It's not just ADAS. We'll talk some more, you'll hear more about the lighting change, the motor control changes, electrical vehicle changeovers. Industrial, kind of low to mid-single-digit growth. This is certainly a GDP-driven industry. They are making investments now to solve energy efficiency problems that they've had.
Our new expanded portfolio allows us to address all of those needs from both a power regulation and control, but also in the connectivity that they need, and the sensing that they need, so that they can have the best solutions for their customers. Lastly, communications. We talked earlier, cell phones are slowing. We do see a slower growth ahead for this market, kind of low single digits. Nonetheless, the content changes, we believe, will help us overdrive the opportunity there. Now with our portfolio, we're entering into the infrastructure place portion of this communications business much more than we have in the past. Again, let's talk about some of the proof in the strategies. This is automotive I talk about. We've more than doubled since 2012 in revenues. Organically, those numbers have been sustainably close to double digits.
We've been the fastest-growing semiconductor supplier to automotive in the last five years. We're not just present in ADAS that we talk about a lot of time. We are present in every single part of the car. I'll talk about that. In every part of the car, there's an opportunity for us to increase our dollar content. If you look at a car we talk about right here, some of the bigger trends, lighting, moving to the LED lights everywhere. There's much more semiconductor content when they make that happen. In the body and interior, there's 100 or more motors in every car. By making them more efficient, brushless DC, you get more semiconductor content, and again, more content from ON there. We've got some opportunities to remove fuse boxes and tremendous amount of cabling in the car with our new eFuse that we're talking about.
Of course, they're all putting different charging solutions into the new vehicles for the mobile devices. Active safety, of course, the largest dollar content growth that we're going to see, and we'll outline that in all of our presentations today. As I showed you before, the entire bill of materials is one ON is going after, and it presents us an opportunity in the hundreds of dollars over the next few years. Powertrain, we've traditionally been in the combustion engines. Of course, there's always a push for some new products there to get better fuel efficiency, the real game changer is in electrification. That can add another $200 of ON Semiconductor car quantity in every car. That vehicle electrification, what does that really mean? You have to manage the battery. We've got the products for that. You have to charge the battery.
We got the products for that. You have to drive the motors. We got the products for that. Then you have to split off and handle all the various loads in the car for all those other motors and those other entertainment things. We got the power for that. So we've got the complete solutions there, giving us hundreds of dollars more opportunity. Again, we'll describe that in more detail later. The industrial market, we have also grown to more than two times the size in the last four years, giving us, again, I think, a presence there as one of the fastest-growing in that marketplace. You'll hear a lot about high-performance power conversion. This is really where you're looking at the higher watt loads that ON has been developing, and Fairchild helped bring us. That is a key area for growth.
Very high ASPs, high margins for us. Again, we'll talk more about that later. Industrial automation and the Internet of Things, really the wireless connectivity standards are enabling our customers to completely automate things that were very difficult before. We see this as an opportunity to outsize the overall product growth in industrial. Talk about this power conversion. One of the things most people don't know is that ON can give you the complete solution that you need for power from the sun to the palm of your hand. We provide the high power modules that are in the kilowatts to solar inversion, to distribution systems, et cetera, and then we convert those power loads, very high power loads, down to the micro loads that you need to operate in your actual electronic systems.
This gives us a great deal of influence in working with our customers to optimize not just our product content, but their efficiencies. We see this as our differentiation capability in the market. Talk a lot about motors. Now we can say that we actually can solve the largest motor loads. We're talking kilowatt-type applications with the full power portfolio, as well as the milliwatt solutions that you see going out there. We are the motor control experts. We think this is another area that's going to be growing faster than overall in industrial. You'll hear more about that in our presentations today. Communications, we've also more than doubled our revenue in the last four years. This is an area where Fairchild did bring us quite a few solutions, particularly in fast charging, but also completing our portfolio in USB Type-C.
With the combination of more and more phones going with fast charging, wireless charging, and USB, again, we think we can outpace the overall mobile phone growth pretty dramatically. Then, as I mentioned earlier, we're now getting into the infrastructure business with more of our server power type products. We'll talk more about those. All of this, we think, gives us an opportunity to outsize that market. This is just another example. This is what a phone looks like from our perspective, about $9 of content for ON Semi. The green parts are what we had traditionally. The red is what Fairchild brought to us. You can see that portfolio certainly augmented, it did not overlap with what we had. It gives us, again, bigger opportunities per phone. What would I leave you with? We have really been a transformative company.
It's not just about multi-source components and lots of components. We have technologies our customers need and are finding critical in their marketplace. We have focused those technologies and our investments on the fastest-growing areas. Cash flow is our primary focus for the company. You're going to hear a lot about that from Bernard later. We think we do that with expanding margins, and operational improvements. Certainly, growth will be something that can supercharge that, it's not something that's required for us to make accelerated free cash flow happen. Then lastly, Fairchild, there's always risk with any kind of acquisition. This is one that we think is an absolute home run. We're well ahead of plan there.
We've found nothing but positive surprises. We're expecting that we will continue to increase the expectations we have on that. We'll share more of that with you later in the presentation. With that, I'm going to pause and take some questions now. Just remember you've got the product groups and Bernard coming.
Thank you. One thing that you didn't mention was anything on ASP. You have consolidation, you have mix, you have a lot more supply being added in the part of the market that you play in. I know historically, we talked about this a little last night at dinner, you talked about a 6% decline in ASPs. Given those three factors that we just mentioned, consolidation, et cetera, how do you take the ASP side of your equation going forward?
There's always been cycles in ASPs. They generally, though, have been very well mapped to what happens in the macro semiconductor industry. I think one of the changes that is coming is there has not been a lot of investment in the analog and power capacity around the world. There have been fewer and fewer players. Every time there's fewer players, we close off manufacturing facilities to get more efficient. The net of that is you should start reaching a point where the supply and the demand are much more in balance. When that happens, you should see a more systemic environment and where ASP will not reduce as much each year. The question I always get is when does that happen? The answer is I don't know yet, but certainly, we should be getting closer.
Thank you. Thanks for the presentation. Scott Jeffrey. Gary, looking forward. Can you give a bit of a strategic view, overview, and context of what has happened over the last 10 years? How has ON changed? How has your strategy evolved and the industry also? How does that distribute ASP going forward? Thank you.
Sure. Strategy, big thing. I will tell you that in some cases, we started very early. Automotive, for example, became a major focus for us back in 2004. We've been outsizing our internal investments on that area for some time. Again, the opportunity we saw there is the one we're talking about today, which is consumers basically wanting more electronics to make their drive experience better, and that has been playing out quite well. In other areas, we have had net changes in the strategy. Back in 2004, again, we had a very strong emphasis on computing. Computing was still a growth market, we looked at owning that entire power path, et cetera. The good news is, we are indeed the largest market share guy in computing. Unfortunately, the market's not growing as fast.
What we did about three years ago, as we recognized that was coming, is we started reusing those power resources to look at power in automobiles, as you're looking at the more complex brains and using our multi-phase capabilities there, looking at other markets in communications and in industrial the same way from a power resource perspective. We started shifting those several years ago away from computing. Consumer was never a strategy for us, I don't want to look like geniuses there. That market is not growing. The margins are quite bad. Really, the three markets we've chosen early on was automotive. We changed from the computing side to industrial over time, of course, communications has always been a strong market for the last two decades. Got a mic here, yeah.
Chris, you had talked about you were going to consolidation. Can you talk a bit about what sorts of criteria you apply when it comes to size and return? Also where is the debt level
Kind of answer those reverse order. Debt level-wise, we've been pretty open saying we're very comfortable with a two times leverage, net leverage, and that it seems to be a place that gives us an efficient return on our investment. Borrowing costs are not that great, we can do much better with that money than paying it off, when you get into that two times leverage range. Two times leverage also gives you some flexibility to amp that up or down depending on what's happening with the financial markets. We like that area, that is an area we want to get to as quickly as we can. Certainly at this stage, we're looking at if that's going to be toward the end of next year.
From a what kind of things do we look at, in each of the acquisitions we've made, and in any acquisition we would make in the future, we look at what technologies will complement what we have, make us more valuable to our customers. More specifically now more than ever, we're looking at the return piece of that equation. We've got the opportunity to increase our own earnings, with some internal work on operations, and new product technologies, giving us pretty high hurdles, looking outside for acquisitions. Certainly something we've said, greater than the cost of capital, I would say, quite a bit greater than the cost of capital right now. Yeah, again, I'll answer reverse order. There is the opportunity for us to grow much faster around phones than 2% or 4%. I believe that.
We're just, again, being very cautious on the overall phone growth rates, in the numbers we give you. We want you to leave here today feeling confident that the plan is solid and not based on wishful thinking. Relative to the Nexperia sale, there really wasn't a lot of analog content there. It was mostly small signal discrete. They certainly were trying to be quite aggressive on pricing as they were making that transition. We've actually found much more opportunity than we could ever have lost, basically because their customers in automotive and industrial are quite sensitive to quality and security of supply. We've been given opportunities for share gains, and I believe that's been showing up in that business here the last couple of quarters. The integration is going quite well. I mentioned this to some folks last night.
Every time we've done an acquisition, we have a team that is the integration team, that gets augmented with new players with every one, but we have a playbook. The guys have been honing that playbook now for over a decade. We know the right processes. We put those in place. One of the things that helps us with Fairchild is ironically one of the things I disliked the most, which is how long it took to close. We were working, preparing for integration for three or four months longer than we had hoped would be true. As a result of that, people hit the ground running. I can tell you we've been very pleased, from a morale perspective. They were looking for an opportunity to grow, and be meaningful in their marketplace.
I think the ON acquisition helped with that, and also have just, frankly, a sense of security, after having many months of uncertainty in the process of getting the deal done. The net of all that is the folks that we sought and pursued to stay with us are doing a great job. Hard to differentiate them from a different ON employee. All right. Thank you. I will turn it over to Mr. Klosterbauer to talk analog.
All right. Good morning, everyone. We're going to talk for a little bit about the Analog Solutions Group. Quite a bit different than a couple of years ago when we got together. The theme is really the same, energy efficient analog solutions, and you'll see that as a thread throughout the presentation this morning. Where are we right now in the analog group? You can see the mix of our products is very similar to the company overall, but it's a little bit different. We are heavily leveraged where we were in the fourth quarter, at least in automotive. Communications is strong and industrial is strong. Over 75% of our revenue is in those three key markets. I believe we're a leader in automotive power management right now, and certainly in front lighting and sensor interface. We see a lot of opportunity for growth coming in ADAS.
On the communication side, we're also a leader in power. We've got a really good wall-to-battery charging solution that a lot of people are using today. As there's more and more intelligence in the charging path, we see some opportunity for growth there. In industrial, we've been number one in ASICs and ASSPs for industrial for quite a while. Our motor control portfolio of intelligent motor control is really getting traction, and we're starting to see some growth there. Some of the new products we've brought out along communications in the IoT edge nodes, we think are going to give us some great traction. The key takeaways, again, we're going to continue to grow in these three strategic markets, as Keith mentioned. We're very well aligned with the company there. Sharp focus on margin expansion.
We've combined a number of businesses within the company, from Fairchild here, and we've identified a number of opportunities we'll talk about later that are going to give us some better than normal, I would say, gross margin and operating margin expansion. We do have some disruptive products, particularly in IoT Edge, and a lot of those products are getting traction. Our design-in funnels are double digits, our revenue, and we see that type of growth going for the foreseeable future. One of the other transitions that's become a natural part of our business is beyond just ICs. As we became a solutions group, software and modules are becoming an increasing part of a solution for our customers.
We are offering module solutions to our customers, particularly in connectivity, but also putting different software stacks, allowing our customers to have some of the firmware and application software layers that they need to create their solutions and their products faster. Strategic intent, again, market leadership and analog power management. I think that just about everything that we do in the group, regardless of what division, is around power efficiency. Either it's the lowest power product in the market or it's got the highest conversion efficiencies, and that's really a differentiator that we want in our analog group. We're going to continue to grow at least twice the market organically. We've been doing that for the last several years. We think we've got the products and the customers, and we're targeting the right applications to do that.
I mentioned earlier, differentiate through integration and power efficiency, smarter solutions in how we put these products together, and better efficiency so that we've got something that we can sell against the competition on other than price, so that we can reap some value. Expand margins, talk about that in a slide here in a little bit. Have a target model to get us back to sort of that 45% margin base that we were at before. You saw earlier we were around 39% in the fourth quarter. We think that we've got a very solid plan to get within the next couple of years back into that 45% range. Fairchild added to the strength of the team that we had. The first thing, Keith mentioned it, I got about $300 million of analog from Fairchild in the communication space. Really solid team.
They had some very good products in power conversion, extremely good PMICs, very good muxes and switches. The USB Type-C portfolio really filled out the things that we didn't have at ON at that time. A really good leadership team. The leader from Fairchild is actually on my staff, and he's running mobile for all of the analog group right now, a very senior leader. We think that was a real prize in the acquisition as far as the analog team is concerned. I mentioned it completed our USB Type-C. We got a number of customers in the industrial space, too, offline power. Power conversion and offline power point of load was an area where Fairchild had a lot of strength in the industrial side, and we're seeing some traction in that. They had a cloud computing initiative that was really in its infancy.
By pairing that with our computing strength that we already had in V-Core Power in my group, I think we've got a really good power solution going forward, not just for computing, but as Keith mentioned, for some of the other areas of the market where multi-phase power is becoming important, particularly in some of the content share we think we're going to be able to get in the cloud and server computing platforms. Again, growth in these strategic areas, very significant percentages of our revenue, and we'll talk a little bit more about each of these key applications as we go through it. We do feel like we've got very solid and identified opportunities and applications in each of these markets, not just to sustain, but to improve our share and to improve our margin position in these markets. In automotive, as Keith mentioned, we're everywhere.
We've got about $90 per vehicle right now, and you'll see some of the growth opportunities that we see going forward. This is an area where we, in analog, didn't get a lot from the Fairchild acquisition. Most of that growth that you see last year was all organic. You may remember we had to sell our ignition IGBT business so we could pick up Fairchild and get their ignition IGBT business. That was sort of a wash, but we've had really good traction and organic growth in the automotive space. Going forward, we see some real growth drivers here for us, and you can see that the growth areas in the TAM are really expanding. Some of that is around power management. We'll talk about some real opportunities. All of these electronics that are going everywhere in the vehicle need power rails.
The power management of it, in some cases, is significantly more content than the application or the processor that they're putting in the function. Sensor interfaces for all the driver assistance and all the sensors that are out there, being able to interface, clean up those signals, and make them usable. Higher speed FlexRay LVDS buffers, things like that for in-vehicle communication. SmartFET. SmartFET is an area of the market that's really been dominated by a couple of people for a long time. We've got some traction there and some really good products coming out, and we think we're going to see some better-than-market growth in that area. You can see from this chart, again, similar to Keith's, all of the areas that we have analog content in the vehicle today. The colors, we've got strong product offerings in everything there that's in blue.
In green is where we're really doubling down on our investment, that's power supplies, connectivity, and SmartFETs. Those are the areas that we think we're going to see the above average, above market growth in the next few years. The couple of areas that we did get from the Fairchild acquisition, just the SmartFETs and igniters. We think there is some opportunity for growth there. Certainly, we know those customers and that market base very well, and we'll be going after that going forward. LED lighting, Keith mentioned a couple of times. This is an area where we've made a lot of hay in the last couple of years. We're literally involved with all of the major lighting suppliers today. There's three real components of these LED lighting modules, or four if you count communication.
There's communication products, there's the power supplies, there's the pixel controller, there's motion control through motor drivers, and there's all the FETs, which come from our Power Solutions Group. Because of the integration that we have at ON, we can sit down and we can put a complete solution together for the customers in this area, and we've been very effective in that. I've seen some of the new front lighting solutions for some of the European OEMs, where the entire BOM of the headlight is ON content. It's either from the Power Solutions Group in discrete, it's got our pixel controller, it's got our power supply, and it's using our motor controller. It's really an opportunity. Our FAEs and the way we go to market, we put together reference designs like that with ON product, and we take those out to the customers.
In many cases, our customers will pick up that ON reference design and put the entire thing in the vehicle. Great opportunity for us to cross-sell across the business groups. We're also winning in rear control now. More and more rear lighting is going to LED as well as interior lighting, mood lighting, emergency alert lighting. We're seeing just a lot of opportunities in that space. I mentioned ADAS power. This is an opportunity that's probably one of the fastest, if not the fastest-growing part of the power market in the vehicle right now. As you can see, there's a vision control module, and that's what Intel calls the brain. They had a big announcement at CES. They actually announced us as their partner for the PMIC and the Vcore that goes around that brain.
That's an opportunity for us for about $8 worth of content per vision control module. As you get down into radar, LiDAR, cameras, ultrasonic, all of these products need power. As you can see, there's somewhere between $30-$45 of content per vehicle once we start getting into autonomous driving or highly assisted driving, just in power management in the car. If you remember in the earlier slide, I had $90 worth of content available. This almost gives me 50% more opportunity as we get more and more ADAS in the car. This is an area where we've got an extremely large team of people focused right now, and we're winning a lot of opportunities. Move to industrial real quickly. Mentioned for a while, we've been number one in ASICs and ASSPs.
We work with a lot of big industrial companies on their motor controls, on their HVAC systems. We've got a real strong position in medical imaging. Next generations of those types of products are driving deeper submicron technology nodes and a lot more sophisticated products. A lot of standards-based IoT out there. One of the things that we've mentioned maybe before, but our trusted design and manufacturing capability has really taken traction with some of the defense and aerospace programs in the last couple of years. While these have been a long time to money, we're going to start to see some margin contribution from some really significant programs in that area because of the work that we've done to get our Gresham and Idaho fabs certified as trusted foundries.
We think that that's going to be a real opportunity, not as much for top-line growth, but a great opportunity for margin enhancement in some of our industrial products. Some of the growth drivers in this area for us, certainly mobile medical. We're seeing more and more opportunities in the mobile medical space. Week before last at Mobile World Congress, we announced our Bluetooth Low Energy 5.0 compliant part. We have the first part that's 5.0 compliant, and we also have the lowest power Bluetooth SoC in the market today. This was developed initially for medical and mobile medical products, but we're seeing applications in automotive. We're seeing applications in a lot of different consumer and handheld types of applications, particularly with 5.0, where you're getting the longer ranges out of your Bluetooth Low Energy part and still getting the kind of power products.
This product's going to be a home run for us. Connected sensing, we have the 2.4 GHz lowest power parts out there for Zigbee, Thread. We've got a Sigfox compliant part. In every area that Sigfox has released, we've been certified, and we're seeing a lot of edge node opportunities there, too, so that's some growth for us. Of course, precision motor control and robotics and smart energy and smart breakers are getting increasingly, we've got multi-core processors to look at waves for arc fault detect, to look for anomalies in power and big circuit breaking systems. We've got a lot of IP around that space, and we've got a lot of technology in the combined company now between Fairchild and ON that gives us traction in that area.
Talk a little bit about intelligent charging now as we move sort of out of industrial and automotive into more of the commuting or communication space. We are number 1 in travel adapters. We are the first to market with a full A4WP solution with our cooperation with Qualcomm. We're seeing a lot of growth in intelligent charging, and intelligent charging is not just wireless charging, but it's also fast charging, and some of the new standards that are coming out USB PD interface into those types of products, but we think there's content gain there. Here's an area where you can see that we had growth in traditional ON, but we also got a significant shot in the arm from the Fairchild acquisition. The communication space. I guess this is the chart where you can see the shot in the arm, sorry.
Again, power management, it's a common theme. Even though we're an analog group, we're in a lot of different markets, power is really our middle name in all of these spaces. Faster charging, RF tuning, we've talked about for a little while, but we are starting to see more and more traction in the RF tuning space, particularly as we're getting ready to go to 5G and some of the carrier aggregation specs are going to get a little bit tighter. We're seeing some of the big handset makers pay more attention to closed-loop RF tuning. Our mobile solution in wall to battery, again, very much like the front lighting solution, we bring a complete solution to the customers.
Particularly plays well in China and some of the smaller mobile customers, where we can bring in and say, "Here's from the adapter all the way through to the battery, the entire charging path." You can see we have those products between ON and Fairchild now. We have two products that are just releasing, the ones that you see in blue. We've got a complete reference design that we go to the market with. The other thing that we're seeing is with the fast charging, more and more drive to getting some game changers on the adapter side. If you look at the standard cube that we all know and love, it's a 5-watt adapter, but you need more power if you're going to charge faster. It's just physics.
We just released and announced at Mobile World Congress a GaN switched adapter that will actually put out about 45 W in the same space as that 25 W adapter that is out there today. You can see with each generation, the power per cubic inch or space on that adapter is going up, and that's what's allowing the evolution on the other side to go from really a 3-amp charger at about 92% efficiency to something we think could be over 10 amps of charging at a very high, 98% efficiency. Again, we had a contact charging demonstration that we displayed at Mobile World Congress, where we were charging at 8 amps over a normal cable and got a lot of attention from some of the handset manufacturers. USB Type-C, it's going to become a very popular standard. There's a lot of noise around it right now.
We believe that we've got the most complete product set in the market for that space. We were the first one with a fully certified SuperSpeed 10 gigahertz redriver. That redriver, while that looks like it is an expensive part per node, takes about $4.50 worth of cable out of that smartphone if you really have to deliver data at that speed, and we've got some handset manufacturers that believe they do. It is a great part. Like I said, we were right on time with the market for it. We think that there is about $1.50 of content in USB Type-C per phone for us. Active cabling allows some additional growth. As you can see, there is a list of a lot of handset guys that are out there that we've got design wins for already in this space. It is not just a handset.
We have automotive customers that are designing in our USB Type-C. We've got computing customers that are designing in our USB Type-C. We think that there is going to be a number of other consumer types of applications where USB Type-C is going to be the product of choice going forward. Talk a little bit now about computing, because Keith mentioned, in the past, we had a major focus in computing. As we've seen with computing, we got a nice shot in the arm from Skylake with the increase in content. We are up to about $2.65 a notebook, $0.60-$0.70 per desktop on that platform. How are we leveraging that into some growth and some better margin opportunities going forward? Well, really, the concept there of multi-phase power and coming up with a Vcore solution goes across a number of markets today.
As the computing power gets more and more intensive in things like mobile and automotive, they want multi-phase charging instead of a simple PMIC. Certainly, the cloud space is an opportunity for us with the acquisition of Fairchild and having the point of loads and some of the power end solutions that we have some design-ins with, and then developing our own digital controller to go with that so we can bring a complete solution to the market. This is where we are using some of those resources Keith mentioned about in the computing side today. We are not going to ignore computing. We've got some really good computing customers. We still generate a lot of revenue there. The competitive landscape in that space continues to get smaller, so we think that there is still room there, but not a lot of growth, and it is never going to be a real high-margin business.
By taking that multi-phase and expanding it into the cloud, you can see from $2.65 opportunity in a desktop, we've got about $50 worth of opportunity on Purley. It is a lot larger content. In automotive, as I mentioned earlier, it is about $8, and that is the PMIC and about 5 DrMOS that go with each one. Looks like a very similar architecture to a PC, but it has ASIL level D functional safety. There is a much higher barrier to entry in that space, and we think the margin profile will hold up there a lot longer. In mobile, we are seeing some really high-end mobile devices that want 30 watts, 40 amps of charge to light up some of these really bright displays, to do some of the other projector modes, some of the functions in handsets that some of the companies are looking at for differentiation.
We believe that we can generate that type of power with some of our multi-phase PMICs that we're developing for handsets. Again, we've had an opportunity to leverage this capability and this skill set, I think, into some higher margin and certainly higher dollar content applications. When we were here two years ago, we talked about how we were going from discrete parts to ASSPs, some full customs, and multi-module solutions. As I mentioned today, now we're going into higher-level modules, and we're developing software application and firmware for some of these things. We've got product lines now where we have multiple devices. The only differentiation is the software stack or the application layer that we put on for different part numbers.
Certainly allows us to leverage our hardware R&D dollars better, as we go forward, it's changed the mix of our engineering capabilities to more and more software in some of these applications. I'll talk about gross margin improvement plans. Certainly, the top side there is product margin. There are some opportunities. There are a number of products that came over to my shop from the Fairchild space where they were doing 100% of the back end externally. Certainly, our back ends at ON are world-class, and we've identified some opportunities for some margin expansion there. We're also looking at our internal products as well, and looking at some new innovations around tests and validating product to make sure we can get quality product to our customer but not spend as much money in test. Test is a very non-value-added function to our customers.
They want good product, but whether I spend one second testing it or 10 seconds testing it doesn't matter to them. We've actually got even some innovations and some patents around how we think we're going to be able to reduce the test time significantly and help our margins. Certainly, portfolio management is part of it. As we've gotten larger and larger, we've got more opportunity to maybe do some portfolio pruning, some price increases. Then, of course, in the new products that we're releasing, we are seeing the opportunity to hold margins longer and keep them at above corporate average margins for longer of their life cycle. There's some factory rationalization. I'm sure Bernard will talk about some of that.
As we move some of these products from six inch to eight inch and into some of our larger back-end facilities, we're going to see some more opportunities there. On the operating expense, we certainly don't want to give that all back away. We actually think that we can do better on the lower side of the chart than we do on the top as far as expansion. There's some synergies that we got with the Fairchild acquisition. There are some opportunities to reassign and reassess where we had some of our R&D teams and how we could go forward and get more value, more return on investment for our design time. There's some direct headcount growth to some of our low-cost design centers.
We've got design centers in just about every geography of the world now. We are allocating the growth or the need for additional headcount in some of our Eastern European and Asian design centers. Closer collaboration with our customers. More home runs, fewer strikeouts on products that we develop improves our return on investment as well. We think all of those things together will not only allow us to keep that 600 basis points, but actually add to that on the operating margin line. In summary, we're seeing a ton of growth in automotive and industrial. You've seen some of the exciting products we think we have there as well as in communications. Reusing technology and IP across the fast-moving markets. The capabilities that we develop in power management really do scale across the different areas that we have.
The RF communication links we can use in multiple types of products. We're seeing a lot of cross-selling, even within my group, as well as within Bill and Taner's groups, across the company with some of our products. Some disruptive technologies, as I mentioned, we're going to differentiate on power. We're going to differentiate on power and the integration solution, what we put in our SoCs. We're going to do that by staying close to these world-class customers that we have and making sure that our solution is the right solution for their next product. We'll execute the margin expansion, and we'll synergize the portfolio with the other groups. As I mentioned, having a complete solution for our customers really does add value. When I can go in with Bill's product and ESD protection when I'm selling USB Type-C, that makes it a better solution.
When I can interface my power rails to Taner's imagers for the cameras, that makes it a better solution for our customer. Anything that's easier for them to get their products to market faster, and they're sure that it's going to work, makes us a better supplier for them. That's all. Thank you.
All right. Good morning. My name is Taner Ozcelik. I head up the Image Sensor Group. I'm going to talk about some exciting stuff this morning. Our purpose at ISG is to create sensors that can see better than the human eye. That particular statement that I just said has implications onto a lot of the core characteristics of the sensor design, including low light performance, speed, Dynamic range, zoom, field of view. What do I mean by that? Our eyes, for instance, cannot see in 1,000 frames a second, but our machine vision sensors can actually do that and analyze that turbine for wobble, which is basically sometimes the difference between a greatly functioning turbine engine for a long time versus a crash that's unpredicted.
The second one is we can't see in low light, in pitch dark conditions, in extreme low light conditions like moonless, starless nights with overcast. The sensors that we have in our cars right now are able to do exactly that and save lives. We don't have wide field of view, but the sensors that we're designing and the lenses that we're putting around the sensors actually can see things like a fisheye can. It all boils down to two things. One is human safety, and the other one is productivity. I'm going to focus the rest of the presentation where those two hyper-growth markets that I'm going to identify matter the most for those two components that I just mentioned and where our vision matters the most. Number one, ADAS and autonomous driving. Number two, machine vision itself for industrial applications.
Those are two things that I'm going to talk about. They happen to be double-digit growth markets for us and have been like that for a while. For the rest of the presentation, there are four things that I want you to take away. Number one, we are a leader in automotive image sensor market, which is a market expected to grow about to 24% annually in the next five years. Second, we have optimized our portfolio with 80% exposure approximately to automotive and industrial. Third, we have expanded our margins, as you know, from 2014 by about four points to now, and we plan to expand another nine points until 2020 by basically utilizing mix shift, like Bob mentioned in his business, as well as cost reductions coming through yield improvement and supplier diversification. The fourth is we decided to get into radar.
I'll talk to you about how we're going to extend our leadership in ADAS and autonomous driving with sensor fusion. We've always wanted to be and had a goal to be number one in mission-critical applications. Why? It provides us the best gross margins and the best pricing power. We are number one in automotive. We have about 50% market share. Our cameras, our sensors are implemented across the car inside and outside, including in-cabin driver monitoring systems, mirror replacement for improving the aesthetics as well as fuel economy of the car, as well as surround view and, of course, ADAS. In industrial, we're number one in many segments, many sub-segments of industrial, including mainstream machine vision, IP security cameras, scanning, and specialty industry, which includes the military and scientific applications. Give you an example.
Our sensors are used in all of the flat panel inspection devices that are utilized in the latest smartphones from Apple and Samsung of the world. They're used in ITS systems or intelligent transport systems, where you're able to now, of course, detect the license plates of any vehicles that are going through toll booths, but also look at the number of axles and count them and so you can charge the right toll as the car goes through the toll booth. In space, we're number one. Our sensors are in Mars Rover, in Juno Orbiter, which is on its mission to Jupiter, as you know.
We also have put our satellite sensors into a company called Planet Labs, what I call a constellation of 88 satellites, each of which are a backpack size, weighs about four and a half kilograms, and they just launched it in February and added to their fleet, which is now 149 satellites. These satellites can survey the entire Earth every day, and they're on a mission as a company to what I call democratization of data science and satellite data gathering. They sell their data, for instance, to economists and companies for things like tracking the activity on shipping docks in certain regions of the world so they can predict economic activity in that region better. In medical, we're number one. We have 41% share. We're in X-rays, in portable diagnostics, as well as digital radiography.
Another interesting market that's important for us is the pill cams, which we have a leadership position with our customer, Medtronic, who has the leadership position in this market, including their competitors nowadays as well. That market itself is expected to grow to about tens of millions of units for doing endoscopy and colonoscopy in the next five years. It's going to be a nice growth market for us as well. In ISG, as I said, 82% comes from industrial and automotive, 45% automotive, 37% industrial. Specialty is about 18%. Our annualized revenues on a Q4 run rate was $685 million with gross margin of 33%. We closed the year at total of $717 million of revenue.
In automotive, as I said, we have the market and the technology leadership, and we have, as you'll I think attest to as I finish the presentation, the broadest product portfolio as well as the customer base in the world. In industrial, we're number one, but not only we're number one in certain segments, we're actually growing the market by twice, even though it's growing in mid to high single digits in certain segments. We have the technology leadership with high speed CMOS, electron multiplying CCD as well as what's called TEC packaging or thermoelectric cooling, which is becoming quite important. That's a great example of, for instance, us innovating at the system level where we're taking what they used to do at the module level, cooling thermally and electrically.
Bring it all the way down to a package level so we can gain more pricing power through that. In specialty, which is basically a segment for us now where we have narrowed the focus down to in consumer, we've decided to focus on areas where we think we're absolutely good at and have an established market position. For instance, global shutter sensors. We've been selling global shutter sensors into a lot of markets, and they happen to be quite important for two areas that we think are hyper-growth. Number one, augmented reality and virtual reality, where we have garnered a lot of marquee customers, including Facebook, Oculus VR headset, which has now two of our global shutter sensors in it, as well as drones. For instance, part of DJI, which is a Chinese company you might know, has about 50% of the entire drone market.
There's some nascent application that I'll show in IoT there as well. Our strategic intent is fourfold. Number one, we'll continue to do what we're good at, and that happens to be mission-critical applications I mentioned in automotive and industrial. Number two, we're adding radar capabilities so we can lead the next frontier in autonomous driving with sensor fusion. We plan to expand our margins through OpEx rationalization and operational improvements, and our 2020 target model is $850 million and 42% gross margin. We do believe we can get there confidently. If you look at the growth opportunities, of course, automotive still presents its largest growth opportunity. It's growing north of 20% year-over-year, expect to grow another 20% for the next five years. We have outgrown the market for the last seven years. That's how we increased our share systematically.
Major growth drivers are basically in ADAS and NCAP and autonomous driving. NCAP stands for New Car Assessment Program, for those who don't know. It's basically a rating system for the active safety for cars, and all three drive huge growth. Industrial, we're number one in certain segments, as I mentioned before, including scanning specialty segments, as well as mainstream machine vision and IP security. We have the broadest product portfolio in industrial. We sell sensors that are anywhere from $1 to tens of thousands of dollars and anything in between. That's a one-stop shop for our customers. In specialty, we are focusing on what we're good at and what we believe are going to be the growth markets that I outlined earlier. Let's look at the automotive market.
The car sales are growing at basically the global economy level, which is shown in the orange line as you see there. The light blue, I guess, at the top is showing attach rates. They're exploding. We have about 1.2 cameras right now per car, going to about three and a half cameras per car by 2020. The market is tripling itself from $600 million-$800 million. A lot of the growth, as you see here also in this chart, is coming from ADAS. Viewing is also growing because the viewing cameras, whether it's rear view and surround view, are getting installed into the smaller vehicles as part of the mandate and also, of course, to drive for people to be able to just have that nice convenience feature that the surround view cameras provide. There's also some nascent but fast growth in two areas.
One is in-cabin for driver monitoring, which is part of ADAS, but it could also be part of convenience in the car, and also mirror replacement. We can actually view now the surroundings of the car, whether it's the rear view mirror or side view mirrors, better than the human eye can, and enable this through a lot of technologies that we have, for instance, with LED flicker mitigation. It's a nice market for us. Our market share in ADAS is 70%, and in total in automotive is about 50%. To be exact, it's 50.3. We've grown our market share since last year by about 2%. We closed the year at $295 million of revenue, so approximately $300 million. We think we're going to get another 20-plus% growth this year and exit the year at about $400 million of run rate in Q4. So far so good.
A leading electrical vehicle manufacturer has some hardware out, as you might know, and they're also going for a mainstream deployment later this year, which will use the same system. We now have 9 of our cameras, basically the entire camera architecture that they have in their cars. 2 for front view, 2 for forward-looking side cameras, 2 for rear-looking side cameras, 1 for rear view, 2 in-cabin, so they look at the driver, and 1 mirror replacement. Our customers are essentially who's who in the automotive industry. We have a marquee customer list. There is not a single, I think, major automotive OEM or tier 1 that is not our customer. Some of these customers are buying in the tens of millions of dollars a year from us.
Of course, there's been a lot of attention since the camera deployment is growing and growing, the market is growing in automotive, a lot of companies are trying to get in. TSR, which is Techno Systems Research, which is basically the most reputable image sensor market research company that particularly tracks automotive image sensors, has just issued their report in 2017 in February. They predict based on 2015 to 2017, and based on all the designs that we have in the pipeline, that every one of our competitors, including Sony, Toshiba, Pixelplus, and others, have lost share except for OmniVision, which basically is gaining 0.4%. In the same time period, we have gained 3%. Bob has talked about this. Keith mentioned it. I'm sure Bill is going to talk about it also.
One of the major reasons that we are successful in this business is that we now offer the entire electronics MoM In the module in cameras. In addition to image sensors, we have the ISPs, of course, the in-vehicle networking devices that come from Bob's team, power solutions also from Bob's team, signal and protection devices, as well as storage devices from Bill's team. We now can go to market as a unified solution. Why is it important? Because it can enable us to build the best system solutions and optimize them in power and size and cost, and customers love that. Second, of course, is that we can go to market faster and enable them to go to market faster, which is also what they like. 50% share in a market brings you a lot of power with the ecosystem.
We have a vast network of customers and partners that we work with. We work with anybody in the lens manufacturing that is a major provider in discrete ISP, SoC providers, synergies manufacturers, as well as independent software houses. For instance, the SoC providers like Mobileye and NVIDIA have selected our sensors for their platforms as sensor of choice in the 1 meg space and in the 2 meg space, and we continue to work with them quite closely to do that. Tier 1s. Before even tier 1s will see the RFQs, sometimes OEMs come to us and show their sensor architecture to verify what they're planning is implementable. Especially with now radar, we have a better position, and therefore we get their RFQs before even the tier 1 sees those. We continue to innovate, and we continue to innovate with our ecosystem.
One example I'm going to give you about that is where the viewing and the sensing now is merging and the lines between the two are blurring. You want to be able to use the cameras that are around the car to be able to do sensing as well. You don't want to have two cameras doing separate things. Why do you normally have two cameras? Because each of those cameras are typically done for two different reasons. One is the human vision, so you want to display the result, and the other one is the machine vision. You want to just detect an object. They happen to use different color filters. One is a clear pixel, the other one is a color pixel. How do you merge them two?
We have a technology called Clarity+ now we're deploying in all of the major SoC manufacturers that pre-IP to insert in their processors so they can process that and have one single camera to do ADAS and viewing. What is the road ahead of us? What's next? As foggy as this picture might be, the answer actually lies in this picture. As you probably saw on Monday and announced on Monday, we decided to get into radar. We acquired the millimeter wave assets from IBM with a lot of nice patent portfolio to come with it and people that have had two decades of experience in 77 gigahertz radar deployment. We're delighted to have them on board and looking forward to expanding our portfolio.
With that, we're able to go from nine sensors that I showed you to 19 transceivers, 19 sensors in total with 10 radar transceivers. It's a huge portfolio expansion operation. Let me show you the numbers. If you look at level 1 and level 2, which is on the left, all the way up to level 4, which is basically the first instance of being able to drive the cars together, our SAM now goes from $25-$200 per car. That's through radar expansion as well as the camera data rate expansion and all the components that are around the module in addition to sensors that I do with the components that Bob and Bill does. There are three reasons why we think we're in a better position to lead this market.
Number one, we can reduce the energy by basically going to a single pre-processing step through a single pre-processor versus two, typically what these systems use. We can eliminate the cabling to have one cable to go from the sensor complex to the central processor, which actually has huge implications in size and weight, which is obviously a big deal as I think Keith mentioned earlier. We're able to do better sensing because imaging can give the path of interest to radar, and radar can give the region of interest to imaging. As a result, of course, radar, as I showed you in the image, is able to see in poor visibility conditions. It can detect the range, and it can have a better accuracy in doing so.
I think this is going to be a really substantial growth area for us, and we plan to lead it. Now, how do you become a market leader? You can't do that if you don't have the best-in-class technology. We have world's best in automotive pixel technology, which is the reason that we have had a chance to lead the industry and garnered 70% share in ADAS, as I mentioned before. We have processing technologies in LED and image processing, like I mentioned before. We own our own CFA and microlens factory, which also is important because then you can innovate at the system level within the sensor stack. We have a package technology called IBGA. It stands for Imaging Ball Grid Array, which we had invented about seven years ago, and we have patented that quite thoroughly, and that's now the de facto standard in automotive imaging.
We have the widest dynamic range going all the way up to 140 dB. Stacking enables some new frontiers for us because we can now miniaturize the cameras, and we have this already in-house and deployed in different places in the car. We have the global shutter technology. One last thing is, I think we kind of surprised the industry with this also that we have put ASIL compliance circuitry, complex ASIL compliance circuitry on the sensor itself. All of our new products in ADAS has that technology. We also are putting cybersecurity, because as the cars are getting connected and are able to drive themselves, it's important to secure it properly. We're investing big time there. I think we have at least two, maybe three years of leadership there in the industry. We have five sustainable competitive advantages.
We have a very strong design win pipeline in the order of north of $1 billion. We have three-year visibility, and that, of course, as the design cycles in automotive are long, as you know. We have a robust technology lead, as I mentioned. We have the complete solution, as I showed you. What's more, we're able to produce twice the products per year and have the largest automotive team in imaging in the world. You might have seen also the announcement a couple of days ago, injected another 50 people in the U.K. who have decades of experience in designing image sensors. We're proud to have them on board as well. Finally, we think we're going to lead the sensor fusion with radar and sensor fusion, as I mentioned before, and that's a major focus for us going forward. Industrial. Industrial vision segment is pretty interesting.
It has about $600 million of market size, going up to about $800 million. It's growing in mid to high single digits. The growth in that market was about 6.5%, but we were able to outpace the growth in that segment and achieve 12% instead of 6.5%. We have both CCD and CMOS, which none of our competitors do. They happen to be important to have both because they have very different characteristics. CCD is still the best in uniformity as well as performance in extreme low light conditions. PYTHON CMOS is very good industry leader in high-speed CMOS design, and they cover the whole gamut in terms of pricing strategies. These cameras are used in a lot of things, but there are two fundamental growth drivers in this market.
Number one, the Industry 4.0 initiative, which a lot of companies are shifting to. That has to do with robotic factory automation, which is powered by a lot of cameras, plus the artificial intelligence now that's enabled through the cloud. They're all coming together and changing the way that we manufacture. Second is an initiative which is tied to the first one in China, called Made in China by 2025. That's basically an initiative, in a nutshell, to change or convert the Chinese economy from mostly a manufacturing economy today to a service economy. For instance, Terry Gou, who is the CEO of Foxconn, has stated publicly that they will automate their factories by about 70% by 2025. The huge growth.
We are the leader in machine vision, as I said, in the mainstream segment. We have a huge force in China, so we're capitalizing on that. The other thing I want to point out as part of the mission critical application portfolio we have is we've been working with a company called ARRI. You may or may not know this, but if you go to movies and if you watch the credits, you'll probably see a lot more of it these days. ARRI, which is a camera company with our sensors called Alexa, has won the Emmy Award in 2013. In 2016, they have won the Academy Award for scientific and technology category. These cameras, Alexa cameras, are so popular. They're in, of course, major producers and directors like Oliver Stone and George Lucas, but also getting into some new producers like Netflix and Amazons of the world.
They're used to shoot a lot of the Academy Award-winning movies themselves, like The Life of Pi, The Revenant, and Birdman. I encourage you to watch the credits and you'll see how prevalent these cameras have become. We're proud of the fact that now the industry thinks digital is better than film. That's one reason that we're excited about it. Specialty, again, we've narrowed it down to two things. One is global shutter, and the other one is depth sensing. Global shutter is a way to essentially capture the image at the same time, not do it row by row, which is rolling shutters do, and it has artifacts. Depth sensing is important because then you're able to see with a single sensor, and you don't have to deploy two sensors to do stereo vision, and we can now enable that with a single sensor.
We're in a lot of the fast growth areas that utilize this technology. We have a leadership position in that. In global shutter, for instance, our efficiency is in these sensors about four to five times our competitors. Facebook, as I mentioned, Leap Motion, all of these companies that are doing VR headsets like Valve are utilizing our sensors, and in drones with DJI. We're also seeing some new uptick in wearable computing, eyeglasses. You might know Snap Spectacles as well as some IoT devices like vacuum cleaners that are connected to the internet, as well as refrigerators that want to put four cameras inside so you can see what's in there when you are in the grocery shop as you're coming home and you stop there. There's some growth there. Margin. As I said, we've improved our margin by about 4 points since 2014.
We have another huge opportunity in front of us to improve by about 900 basis points. There are three areas that's going to come from. Number one, we have managed to, I think, mostly dual source almost all components of our supply chain. front end, back end, probe, final test, et cetera. That's, I think, a good progress, and we're going to benefit from that. Second, we're transitioning to lower cost fabs as well as to lower process, lower geometries, and smaller process nodes. For instance, going from 150 nanometers and 200 millimeter or eight-inch wafers to 110. Also to 65 nanometer and 300 millimeter wafers. There's a significant opportunity for us in front of us for yield improvements.
We are doing a lot of things there, including fab process improvement, as well as binning, which, of course, a lot of the industry in the processor side has adopted already successfully. Reducing our test times, which is also a pretty big low-hanging fruit for us to improve the manufacturing side. On product development, we have a lot of die shrinks in the plan as well as BOM cost reductions, and these things include, for instance, reducing the cost of the protective film or of course the package or substrate that goes into the entire image sensor product itself. Finally, in OpEx, leverage is a big deal in imaging. You can design a really good image sensor and utilize it everywhere. That has been one of our reasons for success in the past.
We will be doubling down there and we will, of course, be tightening our controls on OpEx and rationalize it across the whole group. With that, I think we'll be able to hit 900 basis points and go up to 42%, as I said earlier. In summary, we have a leadership position in both automotive and industrial. We have 50% share in automotive. We have the best technology and the portfolio itself. We have a very robust design win pipeline north of $1 billion, as I said. We have the broadest range of customers and partner relationships, and we're investing in radar for sensor fusion. In industrial, we have number one position in several sub-segments. We're growing above the market. We have the top to bottom portfolios that I mentioned earlier, $1 to $10,000.
We have also a very, very robust design win pipeline in that segment. With that, I think we'll be able to get to 42% by 2020. That will come through, of course, cost reductions and strong OpEx control for the operating margins. What's better, I think with us, as opposed to my competitors, say Sony or OmniVision, we're part of a manufacturing company. We're going to be tapping into the vast manufacturing scale and the logistics networks that Schromm and his team have set up for us. That's all I have. Thank you.
Thanks, Dana. We will take a 15-minute break and reconvene at 9:45 A.M. Mountain Time. Thanks. All right, folks, let's get started. Let me welcome Bill Hall to the podium.
Sure.
All right. Good morning.
Good morning.
I will start out with a warning to the people in the first row. This is my Analyst Day suit, and it's eight years old, and the buttons are a little tight on it. I know Keith has the opposite problem, anyway. I got to flip. Here we go. Okay. You've seen every one of my compatriots have a slogan, and ours is, "New powerhouse in power semiconductors." It sounds a little bit fluffy, marketing kind of stuff. If I was to pick five words that best described how my business changed in one year, and I was speaking to an adult, not using curse words, I would pick, New powerhouse in power semiconductors. If you look at the landscape a year ago, you had one really big guy in power discretes and power modules, and then a whole lot of mid-level guys.
Two mid-level guys, one bought another one. All of a sudden, you've got two guys at the upper echelon of power discretes. To try to put this in a politically correct way, is giving our customers great comfort to have another big supplier of power discretes and modules. Key takeaways. I just talked about the first one. We are now a powerhouse in power. Number 2, we take margin improvement very seriously and we put as much intellectual innovation into margin solutions as we do into new product solutions. Number 3, just like Bob and Taner before me, we're all expanding products into the same three markets. Number 4, in the past, if you look 10 years ago, we were a me-too company, and if you look 5 years ago, we were what I would call a fast follower.
We're in a much different mode, where we are innovating and actually developing disruptive technologies. In power, it's gallium nitride and silicon carbide, which by the way, you may hear me refer to the term wide bandgap, which includes both of those technologies, just as a shortcut. eFuse and Smart Passive Sensors, and I'll talk about them in a little more detail. What is our strategic intent? Quite simply, we want to narrow the gap as quickly as we can between us and Infineon as the number one supplier, and we want to continue to create more separation between us and the power suppliers that are behind us. Right now, we have good level separation. We're clearly the strong number two supplier. We want to drive more towards that number one position. We'll continue innovating products and technologies.
We'll continue fanatical cost reductions, which drive margin improvements. We have a model. Today, we run $2.5 billion of revenue at 33% margin. Our model is to get to $2.7 billion in 2020 at 38% margin. Quite frankly, the $2.7, I think, is a very easy number for us to hit. The 38, we've proven that we have done that in the past, with the past versions of this group. Now we're going to look at what does our market now look like, what does our revenue look like by market segment? You'll notice the 3 key markets, at least they're to my right, yeah, they're to your right also. Automotive, communications, and industrial is almost 70% of all of our revenue. That is where we're doing all of our product investment. Consumer and computing, we treat a little more opportunistically.
They drive a lot of volume, and they help in some of the cycles, but we're focused on new product growth in the pie chart slices on the right. Automotive is our biggest at 25% of revenue. We have over 7,000 products that are qualified for automotive. I had one of my guys go through the internet the other day to try to compare that to ST, and Infineon, and NXP, and the next best guy he could find through using the internet was at 2,700 parts that were available for the automotive market. Beyond that, it dropped pretty quickly. Again, we have a great portfolio there, probably the biggest portfolio in the world in this space of discrete power and modules. In communications, we're number one in power discretes, so the smartphone market.
The last one on there, I like to save the best for last, is industrial. A year ago or even months ago, industrial for us was 15% of our total revenue. The big benefit we got by bringing in Fairchild was we moved that number up to 20%. You will see it more graphically when I show you some revenue numbers. What was good about it is, it is all new customers for us, and it is new products for us. It is very accretive kind of thing. There is no overlap with our previous industrial business. All right. Now, we call ourselves a Power Solutions Group. Simple reason for that is 75% of our revenue comes from power products. We spend about 90% of our R&D spend on power products. If you look to the left side, you see a pretty complete sampling of the major categories.
The areas that are driving the most revenue, we are investing in the strongest, would be different voltage flavors of MOSFETs, rectifiers, various kinds of power modules, IGBTs, wide bandgap, again, which is a shortcut for saying silicon carbide and gallium nitride, and gate drivers. We are now Power Solutions Group and a strong player in that group. I am going to show you the benefits of Fairchild. It is pretty stunning, actually. If you look at this chart, I think Taner had a similar one. In this case, we show every power technology that falls into the power discrete bucket on the left side of this chart, and across the top, five major players. If you see a star, that means that you have top of the market share. Generally, we look at the top three or four players.
If you have a check mark, that means you have portfolio there, but are not in the top four players, a plus sign means you are investing heavily in that area. This is what it looked like last year. When we combine ON and Fairchild through the magic of animation, you can see, looking down the list, there is a ton of stars now on the ON side versus Infineon, who is the first column to the right. What I would contend to you is that on a macro portfolio basis, we have a much bigger portfolio or at least a bigger portfolio than Infineon. We can go win business from them. I said macro view. Let us take it down more to a micro view.
This chart focuses on IGBTs, you describe the portfolio between voltage levels and types of packages you put parts in. Here is where Infineon fits. They are widely acknowledged as a strong player in IGBTs. ON, before we purchased Fairchild, had started investing about six years ago in IGBTs. We have introduced five generations of IGBTs in that timeframe. You see we have made a lot of progress catching up with Infineon. You throw in the Fairchild acquisition, we now have better IGBT coverage than Infineon does. I will go to something that may be more representative of the power market, because it is a much bigger chunk of the revenue, look at MOSFETs. Again, voltage levels are down the left side, packages across the top. You also see some TAM data.
High voltage is about 20% of the market, low and medium voltage are each around 40% of the market. Here's Infineon, widely acknowledged as a leader in MOSFETs. Pretty good portfolio. This was ON before we purchased Fairchild. We had focused in the low and medium voltage area and done a pretty good job. This is where you add in Fairchild. All of a sudden, we're covering the whole spectrum. Again, I would contend that even at a detailed level, we have a wider portfolio than Infineon. A chart I don't have in here, just due to time constraints, is I could show you many power efficiency curves and curves that come from our customers, who tell us that our latest trench MOSFET technologies and IGBT technologies across various figures of merit for efficiency is better than Infineon.
What I'm telling you is we have the portfolio that matches Infineon, we have performance that matches them, and we're climbing on them in market share. Hopefully, you get my point on Infineon. All right. Again, like Bob and Tanner, I think Keith had this, we're focused on three segments. In automotive, for my group, it's definitely the HEV, EV category where we're focused. Traditionally, we've been really strong in body, I'll talk about that in a sec. In industrial, the three big trends that we're developing products around, one is alternative energy. You see this driven largely in China due to pollution in China, the government keeps pumping more and more dollars into solar energy, opening up big opportunities for us. We actively sell products to that market today in the form of modules. Next one is industrial motors.
Interesting statistic I came across a year or so ago was that 35% of the electricity that supplies the world's needs for electricity goes into industrial motors. In contrast to that, lighting is only 18%. It's the single biggest category, electrical motors. It's critical that as a civilization, that we improve the energy efficiency of these motors. Third on the list is cloud power. I think Bob talked about it a little bit in his presentation. Again, a lot of us think of the cloud as, if I look up, there's all these ones and zeros up there somewhere. In reality, it's big rooms filled with servers and disk farms and telecom and datacom equipment. They all require high power and very precise, clean power, and power that is uninterruptible and so forth. A lot of our energy goes into parts for that market.
The interesting thing is, all three of those industrial markets, EV and body, you're generally using the same technologies, just slightly different topologies, and circuit implementations. Last one's communications. Big one for us is smartphones, wearables, and adapters. We play pretty big in power adapters. Let's go in a little deeper with some of these. In automotive, if you look at the chart on the left-hand side, I think Keith actually stole my thunder a little bit. If you went back to revenue in 2012, we've actually doubled organically in 2012 to 2015, our revenue. We did that because we decided in 2012 to focus on automotive. Last year, we grew organically 11%. If you throw Fairchild into the mix, it's like another 30% growth. Now it's our biggest segment. It's over $600 million a year in revenue.
The key systems we're interested in are powertrain. As I mentioned earlier, EV and HEV is where we're focused. Interestingly, a lot of people think that that is the extent of the electronics that drive a hybrid car or a EV car. No, there's also onboard chargers. There's auxiliary pumps. You used to run an air conditioner off a fan belt, you now run it directly from the battery. There's also things like charging stations in your home. All of these require very high power conversion systems. I'll talk about the technologies on the next page. Body and lighting. Traditionally, body motors is where we've derived most of our revenue. As Keith said, there's over 100 different little motors in your car that do a variety of things, open doors, let windows roll up and down, and so forth.
We also are getting into some disruptive areas, such as occupancy detection with our Smart Passive Sensors, a new cabling architecture we call ring architecture with our eFuse technology. I'll talk about that later. In safety and ADAS, we've long been strong in power steering. A lot of that strength comes from the Fairchild side. Big place for us is cameras. As Taner pointed out, we have a lot of products that attach to his image sensors to form a camera module. I'll move to the next page. The future of automotive. We see the TAM growth in the areas we're investing being around 9% CAGR from 2017 to 2020. You'd have to do the math yourself there. I don't show it to you. Again, we intend to play big in HEV and EV.
The key technologies there today are IGBTs that are in Power Integrated Modules. We also see coming down the line, super junction, wide bandgap, silicon carbide, and gallium nitride. Other than the modules, the other things are the switching elements that switch the power. Autonomous driving, big buzzword. The real benefit is it is driving ADAS adoption very quickly, and development of new technologies for ADAS. Again, our major play there is supporting Taner's products. Each of his cameras or image sensors needs an EEPROM to store calibration data. They also need very clean power rails, and we provide that with our advanced LDOs. I'll talk a little bit about the ring cabling system. A kind of unique idea. We've been working with several major automotive OEMs, companies in Michigan, if that helps. Down their value chain or supply chain.
Some of the module makers and guys like that. The problem today is most everything in a car is point-to-point wiring. If you want to protect a motor that opens up your rear tailgate, you've got to run a wire from the battery to a fuse box all the way the length of the car to the motor, and then all the way back to the battery again. That puts in a ton of wiring. We have something called eFuse, which is a self-resettable fuse. You can put it right at the point of load. The point, if you want to protect that motor in the back, you put it right next to that device.
Now, what that allows you to do is, instead of point-to-point wiring, you run kind of a bus around the car in a ring fashion, and you have little stubs off of that bus that will protect, have eFuse that will protect the various things that are at the end of those stubs. It'll pull, as Keith said, tens of meters, and hundreds of pounds of copper out of the car. Pulling weight out of a car is very critical, as it has a direct implication to miles per gallon, and particularly in hybrid cars, goes a long way. That's a disruptive technology that we're trying to drive in the industry. All right. Industrial market. We were not a strong player before the purchase of Fairchild. We did have about $200 million of revenue.
Annualized in 2016, we're now up to about $500 million of revenue per year, and that's 20% of our total business. The key areas that we are engaging in is solar power, mainly working with Chinese energy companies. We're in production today and shipping products today. Cloud power is another one, which we've talked about already. The key there is how do you get a lot of power in a small space? Power density. Then industrial motors. Got to improve the efficiency to cut down the amount of electricity that we're using for industrial motors. The TAM there is growing at about 2%. That market, the wireless market, which we're including smartphones, iPads, wearables, and adapters, is about 2% growth. The areas that I talked about earlier were solar. You can see all the key high-power technologies there.
Power Integrated Modules, IGBTs, silicon carbide, gallium nitride, and so forth. You get into cloud power, it's a lot of the same kind of things. Again, just different topologies. Cloud power, in some cases, is lesser levels of power, so you can go to low and medium voltage technologies. Industrial motors, again, very similar kind of products to the first two. I guess the good news there is if you take that stuff and you take EV, we're building portfolios of parts where the knowledge can be reused and tweaked to go into a variety of these kind of applications. In IoT, of course, it's always a great buzz phrase. We do have one area we're kind of excited about, and that's our Smart Passive Sensors. These things are basically, think of an RFID device that has four sensors built into it.
The four sensors are pressure, temperature, proximity, and moisture. That device requires no battery because, like RFID, it actually harvests energy from the device trying to read the data of the sensor. It has a lot of use in automotive. I kind of skipped past it, but like You can weave them into the fabric of a seat, a car seat, and it will detect the weight in that seat and whether that weight is a human being or a bag of groceries. Some pretty neat stuff. We think ultimately it's going to be the Internet of Things that are going to take advantage of this wireless technology that provides a lot of sensory data. Okay. Communications market. We've had good, steady, organic growth over the years. We're engaged with all the major suppliers of smartphones.
2016, you see a big drop or a big increase due to Fairchild. Again, we're up to about $600 million of yearly business. Main areas for us are smartphones, and within smartphones, cameras, and biometrics all use image sensors, so we get the attached products. High-speed serial interfaces like USB, HDMI, MIPI interfaces, they all require protection, ESD protection. We have devices there. Of course, battery life is a key thing. You want to be able to charge your batteries and not have them explode on you. So we have battery FETs that actually assist in that area. Wearables. Key there is size, really, and I'll talk about size on the next chart. Then in charging, basically there's multiple categories. Fast charging, wireless charging. Small form factors are the key there. Again, it uses a lot of MOSFETs and typical type power conversion type circuitry.
The growth of wireless is smaller than the two other areas I showed you. We're looking at about 2% growth between now and 2020, and I'm talking about CAGR. Key area is the camera module. Again, attach rate with Taner. Charging, you see FETs, more of the low to medium voltage FETs. You'll see some wide bandgap because in the future wide bandgap is good because it provides power conversion. Because it switches really fast, that allows you to use smaller transformers and capacitors and things like that around it. You can put in the size of a 10-watt adapter, 45 watts of power. It's a game-changing technology in adapters. Then smaller form factors, which I mentioned. It's a key competency of onsemi. It's been that way for the 10 years that I've worked here.
Couple of key ones would be we have the world's smallest MOSFET, we have the world's smallest double EEPROM, we have the world's smallest LDO, and we have the world's smallest package for small signal diodes and transistors, and that's the X4DFN. That one is interesting. It's actually encapsulated on the wafer. The size of it, if you look at the cross-section area of an X4DFN part, it's roughly the same area of the cross-section of two human hairs. That's how small it is. It's almost undetectable with the human eye. I'll get into I think I have two more slides here, and let you get to the fun with Bernard. Let's talk about GaN and silicon carbide. First off, we have both technologies. One of the great synergies with Fairchild, we were working with GaN, they were working with silicon carbide.
We have a GaN Cascode transistor in the market today. We have a silicon carbide diode in the market today. Next quarter, we'll have a silicon carbide FET in the market. We're in all those areas. Two years ago, we were more bullish on GaN. I think there was a lot of hype about GaN being the better of the technologies. Quite frankly, because silicon carbide's been around longer, it is actually beginning to be adopted, and we see faster growth there than we do in gallium nitride. Still investing in both, but we're gonna see most of our growth in the next three years come out of silicon carbide. Benefits. I talked about smaller form factors. Another neat one is in automotive, wide bandgap materials can tolerate very high temperatures, you can actually build a car that doesn't have a liquid cooling system in it.
That would remove quite a bit of weight. All right. Second to last slide here. I want to talk about margin improvement. I think I just showed you about seven slides on new product strategy. We spend about 75% of our engineering spend on new products and about 25% on margin improvement. There's 2 categories of margin improvement. The lower one you see is what we call evolutionary. That's the stuff that pretty much the whole industry knows how to do. The typical go to larger wafers or do die shrinks, go from gold to copper. That helps us with the every year offsetting the degradation in ASP. If you want to grow margin, you have to go to revolutionary type margin improvement plans. We do some things that are fairly unique. A simple one to understand is our high-density lead frames.
We have a factory in China that can build a billion units a week of surface mount devices. We happen to use 24-row lead frames, whereas most of our competitors use 10-row, which basically means for every 10 products they make in the same period of time, we can make 24. A huge benefit in throughput and cost reduction. Rapid test metallization. This is a new concept where you actually have a metallization layer that connects all the die on the wafer so you can test them in parallel, greatly shortening the test time. When you're done, you just etch out that metal layer because it's no longer of use to you. The last one I'll talk about is the bottom one, it says plasma die singulation.
This is a technology that only ON Semiconductor has, we've had it for a while, it's in fact been in full production for two years. Typically, when you want to singulate a die, you see that picture of the wafer there all the way to the left. You use a mechanical saw, you cut it vertically, you cut it horizontally, you pull a rectangular die out. I should say this. With a saw, you have to have a lot of space between die for that saw to be able to cut down through that gap. That gap is called the scribe street. When you etch it, you can go to a much smaller scribe street because you're doing it chemically. The other benefit is if you etch it, you don't have to do rectangles.
With a saw, you have to do rectangles. We can now do hexagonal shaped die or any shaped die. Benefit to hexagonal is a lot of times we don't use the space in the corner of a rectangular die. You're saving all that die space. These kind of savings can give us over 20% more die on a wafer, which is significant. It's really the evolutionary stuff keeps pace with the ASP reductions. A lot of it driven by our operations group. The revolutionary stuff is what enables us to raise our margins. Our plan is to go from 33%-38% over the next three years. Summary, we're a leader in power semiconductor market. It's really us and Infineon at the top with a pretty big gap between the next guys. Our portfolio is as strong as Infineon.
Our performance parameters are better than Infineon. We're gaining market share on Infineon. Margin, as I just showed you, is critical to us. We're investing heavily in power technologies, but also in disruptive technologies. Finally, as Bob and Tanner said, we all work really well together in terms of our products. I have products built around Bob's ASICs, and I have products built around Tanner's image sensors. I think I am done. I think now there's a question and answer period.
We have a Q&A period for the business units. Bob, Bill, and Tanner.
Okay. How about you?
Thanks for the presentations. Mark Lipacis from Jefferies. One question for Bob, one for Tanner. Bob, I think investors group the analog companies into groups by gross margins, and there's kind of like a group in the 30%-40% range and a group in the maybe the 50%-60% range. I was wondering if you could give us a view of the competitive landscape. Your bogey is 45% gross margins. Do you compete against these guys in the 50% and 60% range? Do you have aspirations to get your margins into that bucket? That's for you. Then for Tanner, you start off your presentation talking about making sensors that can see better than the human eye.
I think you covered that, but just to help me internalize it, if you're just displaying that to the human eye, then maybe you don't get to realize the benefits. Maybe you could give us a concrete example of how a sensor that can see better than the human eye, how that manifests in an application in a car that makes it safer. Thank you.
Sure. I'll Is this on? I'll answer the first question first. On the analog side, we certainly see ourselves competing against the guys that are in that higher echelon of margin profile. We don't believe that 45% is the end game. That's sort of what we think is a realistic expectation between now and 2020. We have a lot of very high volume products that we're continuing to sell that are sort of a headwind against getting higher than that in the near term. A lot of the products that we're introducing into the market, we're introducing head-to-head against that group that you would say is more in that 48% to 55%, 60% kind of market range. We have a lot of products in that category. Realistically, again, we think that 45 by 2020 is achievable, but I don't think that's our end game.
All right. Let's dissect the question to what I had also mentioned earlier. If you look at what I mean by better than human eye, there are five things that I mentioned. One is speed, the other one is low light, third is dynamic range, and fourth is zoom, and fifth is field of view. Take low light, for instance. We have sensors right now that can see in extremely dark conditions around the car, where you don't have a lot of lighting, and extremely bright conditions. I'm talking about conditions, for instance, what happened in the Tesla crash and beyond. You personally probably realize this when you drive every day. There's certain times that one second or a fraction of a second that you can't see anything.
The sun is in your eyes, or even if you just put the shade down, it's still too bright for you to see below the shade and on the road. That's dynamic range. Very low light to very high light. We have sensors now that can see better than the human eye. 140 DB is way better than human eye. Take speed. A human eye can discern about 30 frames a second, approximately. We now have sensors basically go up at 90 frames a second, these are getting deployed today. We have machine vision sensors that go up to 1,000 frames a second. I think some of that, like hundreds of frames a second, will find its way in automotive too. The reason for that is obvious, but I'll say it.
You can go faster and brake sooner if you have a faster sensor, right? That also means saving lives. Field of view, a human eye can only see basically 140, 150 in a wide angle. We have sensors with particular technology around the sensor itself, but also lens, that can see much wider angles. You can see around the car better. You don't have to deploy as many of them to stitch together, and so forth. Those are the examples I would give in that car application that you were asking about.
Thanks for the presentation this morning. Harlan Sur, J.P. Morgan. In addition to, I think you've conveyed to us the breadth of the product portfolio, but you've also stressed integration potential as a kind of a key differentiator. I think one of the things that Keith and Bernard have talked about over the last few quarters, is the success with the Power Integrated Modules, especially in solar and some of the heavy industrial motor segments of the market. Maybe for those of us that don't quite understand what these power modules are, give us a quick overview of that. More importantly, what is the dollar content capture relative to just selling discrete components and the potential margin profile? Is this new product segment one of the faster-growing segments within your product portfolio?
Okay. I'll try to remember all those questions and maybe like Keith, I'll go in reverse order.
Yeah.
The profile, is definitely growing very, very fast for us largely because of the pollution issues in China. The first market that we've really gotten into quickly is into solar applications. The ASP of those devices, they're somewhat simplistic Power Integrated Modules, and they're in the $30, $40 range. A Power Integrated Module basically is either IGBTs or MOSFETs, some switching technology that's in a half bridge or full bridge topology, has some diodes sprinkled in there. Generally, what differentiates them from other modules is that they take very high levels of current and voltage. They can go up to 1,700 volts. They can handle up to 300 amps of current. You're talking about a lot of power.
We happen to, I would say, hate to use the word luck, but there was a neat intersection of, we had the right IGBT technology and the right PIMs at exactly. A customer who has been a traditionally strong customer for us in China, that all at the same time, that customer needed to rapidly get a lot of PIMs out of us to meet Chinese government orders. That's been a great opportunity for us. We intend to expand more into motor control and then more into automotive. That's kind of the landscape of the PIM. The highest priced PIMs I've seen, you get into HEV and EV, and you're looking at $200 PIMs.
Great. Thank you. Just one follow-up question. In trying to tie some of the themes that you've been talking here at the product level to the trends in the end markets, one of the things in the cloud and data center segment of the market that is going to happen either this year or next year is the move to 48-volt power in the data center. I'm just wondering, do you think it's going to be a trend that starts to fire this year, and how does the ON team benefit from this?
Well, I think that both Bill and my groups benefit from that because of some of the higher voltage MOSFET and higher voltage, higher current carrying capability products that we have. Again, one of the areas that our teams work very closely together is whether it's a point of load that I develop for the cloud guys with all of the intelligence, or whether it's a discrete or a PIM type of product. We work together on the same MOSFET technologies, and we work with our customers very carefully. I don't think it's going to be this year, at least not on any sort of major scale. I think we're well positioned for that ramp and that transition when it does come.
Thank you.
I think another additional answer to that question is that we have everything from discrete PIMs, I mean, discrete MOSFETs, all the way up to Bob's high-end SmartFETs and DrMOS and all that kind of stuff. We give our customers a variety of ways to create the same kind of circuit, basically.
Thank you.
Hi, guys, up here, Ross Seymore from Deutsche Bank. Taner, a question for you. Actually, two questions, but they're both on competition. First, in your lead in the automotive space, 50% market share, et cetera, for a long time, your competition was focusing on the consumer side and handsets, and you guys, for one reason or another, started focusing on auto much, much earlier. You have a dominant position now. Have you seen the design win share change as they started to focus away from the handsets as that market slows? I would assume they're looking at autos as well. You wouldn't see it in the market share revenue today, but you might see it in the design activity. That's the first question. The second one is on the RF, or excuse me, on the radar side of things.
Just talk about the competitive advantage you might have, because there's some pretty strong incumbents already in that space. How does adding that capability, how can you differentiate versus the Infineons, the NXPs, STs, ADIs, et cetera?
That's right. To start with the first one, yes. We track this very, very closely. That's why I mentioned that our design win pipeline has increased. In fact, it has increased dramatically. There's several reasons for that. I think our stickiness is mattering more and more. We measure this, as I said, in the company quite closely. It's over $1 billion. We have three-year visibility. Automotive is a huge, long design cycle, as you know. Yes, competitors are coming in, but this stickiness and long design cycles are actually working to our favor. That's why I showed the partner ecosystem, because we have a huge footprint in the whole partner ecosystem and are in these reference designs from Mobileye and NVIDIAs of the world, as well as lens manufacturers, ISP providers, SerDes providers. I mean, TI is building cameras with our stuff in them.
Imagine that as a competitor of our company. That's working to our favor. The other thing is that even though the competitors are coming in from a different space and they're focusing on it, their still number one priority is mobile. If anything happened, capacity shortage, whatever, or design resources, Sony is still focused on Apple and the rest of the mobile industry. OmniVision, the same way, because they earn their money there. I hope you got this through the presentation, but imaging is one of these markets that is incredibly open to innovation. It is not a commodity, and it's nowhere near being a commodity at all. That's why I mentioned and opened the speech with the stuff that we work on and why being better than the human eye matters a lot.
As long as you can keep the pace and innovate better and focus on it like crazy, maniacally, I think we'll continue to pull away. One strategy there we did was with radar. Now, second question, why are we getting into radar? We believe, again, as I said, to be able to drive the cars autonomously, you need multiple sensors. Imaging is not going to be enough. In fact, LiDAR will be there as well, and even including far infrared imaging, which I also showed on one of the slides, to detect human beings like pedestrians, as well as animals on the road, right? In radar, there are incumbents, like you said, but if you believe in the story of sensor fusion, which I think almost everybody believes in, it's going to matter how you fuse them.
None of these competitors that we have, NXPs of the world, other players have imaging. If you believe in sensor fusion and a lot of processing or pre-processing has to be done near the sensor so you can cut down on other things like I showed, weight of the cable, size of the module, the number of processors you need to have, which means lower cost and lower power, as I explained, and also fuse them so that you can use data from one to benefit the other one and vice versa, you're going to do a better job, right? Our 50% market share gives us a footprint with all the ecosystem partners right now, and believe me, they're all looking into radar or developing radar. Give you an example, publicly informed Delphi, huge on radar, huge on imaging. They're huge on Mobileye too, and they're basically fusing.
It's just one of the examples of a customer that we'll be working with together. We believe there's huge potential room for strategy and getting ourselves into radar with definitive architectural changes where we can combine the chips differently and have cost reduction, weight reduction, power reduction.
This is Steve Smigie, Raymond James. Two questions, one for Bill. In your comparison relative to Infineon, do you sense customers will start to view you two guys as the main players and others will have a harder time competing, or is it still relatively competitive there? Then for Bob, I was hoping you could comment a little bit on the lighting market for automotive. A number of your competitors talk about opportunities in that market. Do you feel with your suite, you'll be the dominant player even with the competition, or is the market just expanding so quickly there's expanding pie for everybody? Thanks.
My answer to your question about would our customers be consolidating around us and Infineon as their two major suppliers. When we first looked at this thing, Infineon as market share here, us plus Fairchild are here, and everybody else is down here. My thoughts were that it was going to be a clear Infineon and ON would be the two top guys, and all these little guys were going to lose share to us. What we found, which surprised me, even before the purchase went through and then after the purchase went through, was that customers are actually looking for another competitor that they could leverage against Infineon. They've become, I guess, too dominant and wanted to have an alternative to Infineon in the marketplace.
Hi, Vijay from Mizuho.
Sorry.
I'm sorry.
Can you repeat your question again?
It was just on the LED lighting space.
Oh, okay. Yeah. On LED lighting, we do think we're going to continue to be, if not the top guy, one of the top two guys. Again, a little bit like in Tanner's space. We're currently working with so many of the front lighting guys today, geographically all over the world. We know, for example, that our next generation products are already designed in to the top automotive lighting guys for their next generation products. As I mentioned, it's not just my power buck boosts and pixel controllers, but it's the FETs from Bill's group, and it's really a complete lighting solution. We know we're going to maintain share, certainly, or gain share in the next generation.
There are a lot of players that have certainly turned their eye to that space, and I think that while it will continue to grow, I think we'll have to stay on top of it. We're already introducing our fifth-generation products to some of our early adopter customers, and we're working with them on their next platforms beyond the one that's not even in cars yet. I think we've got a good roadmap and a good technology path. We've got good teaching customers that are helping us stay ahead of some of our competitors. Certainly, the pie is too big there for the competition not to be putting a lot of energy there.
Great. Thanks. Vijay from Mizuho. Just a question for Taner. Taner, you mentioned you had a millimeter wave radar acquisition. When do you see the first camera plus radar products hit the market? You talked about processor players. Who are the big processor players you're working with? Are there Intel, NVIDIA, Mobileye, and when do you see those products come out? One last question for the group. You guys talked about gross margin expansion in all the segments. How much of that is dependent on revenue versus product refinements? If you have revenues flattish, can you still drive margin improvements across all the segments? Thanks.
All right. I'll answer my questions first. What was the first question? I forgot. Yeah, remind me.
Just when can we look at millimeter wave radar on the market?
Radar. Yeah. 2019 is the timeframe for that. 2019 is the first timeframe that you'll see a radar product from us in the market that will be fused with image sensors. Of course, we're selling image sensors that we're going to be preparing the processors especially, the co-processors that we design, which is sitting right next to the ISP. Sorry, sensor in the form of an ISP to be able to enable the radar fusion. 2019. Are we working with all the associate providers? Yes. With all of the ones that you mentioned, Mobileye, NVIDIA, Intel. Mobileye, of course, we're in production already, have been for forever, since the beginning. We are in their next generation designs with the 2 meg, which is the next big frontier, with NCAP driving that in a big way with a wide field of view.
NVIDIA, we're in their platform. They have chosen us as a sensor of choice also for 2 meg. I think the first production with NVIDIA systems will be later this year. Intel, we're also partnering with them, and there's some developments that we're doing with them in Europe. We're engaged everywhere.
Thanks.
I'll take a shot at your second question, which I think was, do we rely heavily on market cycles to improve margins? Of course, there's no doubt that in an upmarket, factories get more utilized and our margins get better. Our strategy going into every year is not looking at, well, what's the market going to do, and planning around that in terms of cost control. Our operations group, with Mr. Schromm over here, they have a goal to take out 6% per year no matter what the market is doing, and that compensates an average of 6% per year of ASP reduction. My guys who tend to be working more of the revolutionary projects, they're trying to get another 4% per year, regardless of what the market is doing, in order for us to grow margin.
Yeah, it becomes a little harder in a down market, and it becomes a lot easier in an up market. We don't plan or strategize or depend on the market to hit these kind of plans.
I'd say from my space, it's very similar. We've got very specific margin enhancement programs that are targeted around. If you looked at the revenue growth numbers that we put up on our slides, they're significantly below some of the TAM and SAM growth in some of the markets that we targeted. We're really not counting on wild optimism in the space of revenue growth to get to these gross margin improvements. We think that specific programs with specific. Some of them are down to 10 basis points from this.
Yeah.
These are programs that we measure our teams against every month and make sure that we're on track for. Those things will happen. If we get a little tailwind from the market, that just makes things easier.
Kind of a follow-on answer. I know you asked follow-on questions, I guess I can do a follow-on answer. We have a group within my organization that's just chartered with cost reduction. Right now, we're running about seven major programs, okay, aimed at specific technologies. Within each program, there's anywhere from 10 to 20 projects of things they're doing to try to improve that. It's a big deal. It's not just five product engineers in a corner trying to get rid of dogs or something like that.
By the way, I'll just attest that this thing that Bill mentioned is so good that Dave is copying it exactly as it is. It's another testimony to the success of the program.
Sure. Thanks a lot.
Thanks for the presentation, guys. This is David Williams on for Christopher Rolland of Susquehanna. Can we talk a little bit more about the wireless charging opportunity? There's increasing expectations in the market of the biggest handset player kind of adopting a solution, maybe in the near term. If this happens, how do you believe that would influence the ecosystem in wireless charging? Would your expectations be for other smartphone makers to follow suit in the near term? Secondly, on that opportunity, I believe you guys have historically focused on magnetic resonance rather than magnetic induction. What's your current view of which way the market's heading between those two technologies?
Certainly, if the world's largest handset maker does have wireless charging in their next product, it will accelerate adoption, I think across the board. I think that people have been kind of waiting for a stimulus like that in wireless charging. There are still a number of standards out there, and it may be induction. We still believe that there's some room in induction charging, but we believe that resonant charging is a better solution long term, and that the ecosystem will migrate to that. In response to that, what we're doing, though, is we're introducing some tri-mode products right now that will work both with induction as well as magnetic resonance. Because we think that there will be infrastructure put in place if our friends in Cupertino do put in wireless charging, and it's induction.
There will be a lot of infrastructure put in place to support that, and we'll want to be able to take advantage of that as well. That said, we are starting to see some pickup from some other people on resonance charging. We've gotten some new board members recently, or some new members recently in the AirFuel Alliance that I think are an indication that there are some people that are starting to get on board with resonance as well. To your point, our roadmap has gone from being sort of singularly focused around resonant to acknowledging that we need to have a tri-mode product in the market. Particularly for automotive, because they want a solution that works for everybody. We're working with some of the automotive guys on tri-mode transmit solutions as well.
Any more questions for the team? Thank you, guys.
Thanks.
Now it's time for the most anticipated presentation of the day.
Yeah.
Hello.
Sorry.
Here.
Thank you, Brad. You're putting the pressure on. Appreciate that. I'll keep this short and sweet. Key message, very simply, free cash flow, free cash flow, free cash flow. We're really intensely focusing on free cash flow. We are going to drive that through operating leverage, and I'll show you now a lot more details on how we intend to do that. I'm going to start with the four key takeaways, so you have the answer before you look at the details. We are providing a target model with a 40% gross margin and a 19% operating margin. As was mentioned throughout the presentations and in the question and answer session, the assumptions that we put in there for revenue, we believe are pretty modest.
We think in our internal plans are to do better than that, but we wanted to be grounded on something that was credible and that we could achieve. I'll go into more detail into the model. Number two, free cash flow growth. That's the key message. All of this model at the end translates into how much cash we can generate, and we are very excited. At the bottom, you can see in green, we expect to generate $900 million or in excess of $900 million of cash by 2020. Mentioned it already. A lot of it is without revenue growth. We'll show you the details. We can achieve a lot of margin and operating leverage expansion without revenue growth, but obviously, we will pursue the revenue growth actively to make it even better. I'm going to show some increased synergy targets for Fairchild.
As was mentioned by Keith Jackson in his presentation, we are extremely happy with how Fairchild is going, and we are to a point where we can show some raising of our synergy targets that are reflected in the target model. Fourth, after we generate all of this cash, the question is, what do we do with all of this cash? I will show you some more details on that. Fundamentally, initially, our goal is to use the incremental cash flow we generate to delever to about two times net, and then we should be able to get back to a shareholder return policy. Again, these are the different topics I will cover. First, Fairchild synergies. We have talked about those in relation to our 2015 base.
Our numbers were originally $150 million, then we increased it to $160 million when we announced the deal as an exiting velocity for 2017, going to $200 million for 2018, and going to $225 million for 2019. We have upped that now to an exiting velocity of $180 million at the end of 2017, $220 million at the end of 2018, and $245 million at the end of 2019. Most of the improvement is coming on the COGS side, not that much on the OpEx side. On the OpEx side, we are very happy that we have been able to accelerate those and do them earlier than we thought, that the number is about the same. The incremental amounts come from more synergies that we can get on the COGS side. The target model is based on $5.6 billion of revenue, which is approximately a 3% CAGR. I will show some detail on that in the next chart.
Drives to 40% gross margin, get some operating leverage on operating expenses down to 21%, and it translates into a 19% operating margin. By the year 2020, that represents a non-GAAP EPS of $2.00, and a free cash flow in excess of $900 million. Now I am going to go into some of the assumptions and the transitions as to where we go from where we are right now to this target model. As was discussed earlier, here are our assumptions that went into the model for the end market growth. Automotive being the strongest in the high single digits, 7%-9%. Industrial, a little bit more GDP-like at 3%-5%. Communications at 2%-4%. The few end markets we have focused the least on, we are expecting those to decline over the next 3-4 years.
Bottom line, when we look at this, it gives us about a 3% CAGR, and that assumes an industry CAGR in that 1.5%-2%. So it is, again, fairly modest industry assumptions. We think we will outgrow that. Obviously, our internal plans are to do much stronger than that. Gross margin. By the way, one of the things I failed to mention on the previous chart, all of our comparisons are against Q4 2016 annualized, which is the first time we have Fairchild as a full quarter so that we can make apples-to-apples comparisons. Gross margin, we are in the 35% range, so we are actually 35.2% in our last printed quarter. We intend to grow it to 40%. Here you can see which ones are the main contributing factors. Manufacturing efficiencies, and you heard some of those mentioned by BU peers today.
It includes insourcing, but it also includes test time reductions and some of the revolutionary changes that Bill talked about or some of the yield improvements that Taner talked about. In general, we should expect about 80 basis points coming from manufacturing efficiencies. Mix, we should get about 90 basis points, and I'll show some charts in a second. Fall-through, indeed, with a 3% fall-through, we still get 150 basis points on the incremental revenue. Basically, what it says is, if there is no revenue at all, we should still be able to get to 38.5% gross margin with absolutely zero growth. Obviously, the flip side is the same. If the revenue is higher than 3%, we should be able to outgrow the 40% or achieve the 40% earlier. Manufacturing consolidation.
Since we acquired Fairchild Semiconductor, we have been talking about the fact that we are going to be doing some consolidations, and that should give us about 120 basis points of the transition to get to the 40%. Last but not least, we have included a modest amount of divestitures. It's less than $100 million, but that should give us also an incremental 40 basis points of help. We're obviously looking at more, if we can do more, but right now that's what we have played into our model. On the manufacturing consolidation front, currently we have nine major fabs. Four are eight-inch and five are six-inch. We are going to continue moving over time to more eight-inch and consolidating our six-inch networks. We think that just putting numbers on the fab consolidation, it's about a $60 million opportunity that we should be starting to get by 2019.
On the back end, manufacturing, it is an extremely efficient operation. We produce 70 billion units a year. Fairchild Semiconductor had been outsourcing some of their materials and, as a result, we have nice opportunities in bringing that in, and the savings from that should be about $20 million just on the pure insourcing. Plus, as we talked about, there's other operational savings that we are actively working on. On the mix front, already a pretty sizable amount of our revenues are in the three end markets we like, which typically have above corporate average gross margins. We still expect to continue growing that. We are at 70% today in terms of automotive, industrial, and communications, and that should be growing to about 82% by 2020, with consumer and computing shrinking down to about 18%.
We did the granular calculation at the detail level, and that yielded about 90 basis points of contribution to the improvement in gross margin and operating margin. On the operating margin front, we exited Q4 2016 at 12.9%. Our target is 19%, we need about a little bit in excess of 6% to achieve our target. I already talked about the details. About 480 basis points come from the gross margin front, we still expect to obtain some leverage on our operating expenses or operating leverage. About 60 basis points come from the revenue growth, the 3% that we put in there. About 70 basis points are the conclusion of the different synergy projects we have in place for the Fairchild Semiconductor acquisition. Operating expenses.
Our model there is, obviously, we do have synergies, but also our model is that we expect them to grow at about half of the pace of revenue growth. It's modulated based on what happens with revenue, and we should still obtain a fall-through or a leverage as a result of growing at about half the pace of the revenue growth. As you can see on the historical, we do have a good track record of having reduced that from about 24% to already 22%, and we'll keep going down to 21% over the next several years. If you translate that into what is the free cash flow model, the operating cash flow, we expect that to be in the 21%-23% range. By 2020, interest expense is expected to be in the $100 million-$110 million. Obviously, there is an external factor there.
Depends what happen with interest rates, and that could swing the numbers up or down, but we think that that's a pretty reasonable assumption. Our cash tax rate, we expect by that time to be around 12%. I have another chart on that one. We'll keep our capital expenditures in the same 67% model that we have had. We believe that's a very good and sound model. It allows us to source equipment, in many cases, for the front end on the used market and get another 10 or 15 years of mileage on that equipment, and therefore, it's a pretty efficient way to do it. We also have our Fujitsu joint venture, in which we can get the increased capacity at a pretty cost-effective way. We expect capital expenditures to remain in that 6%-7%. Right now, P&L and CapEx are pretty much the same.
The depreciation is expected to follow through and be the same case. When you get all of that translates into a free cash flow yield of 15%-17% of revenue. We're really on track. We have always talked about the fact that we expect to do about $500 million of free cash flow in 2017, substantially up from where we were in past years. With the Fairchild synergies, with the operating leverage, with all the operating improvements And the disciplined capital spending, we believe we are on track to go much higher and to achieve that target of $900 million by 2020.
Addressing the question about really what's the revenue going to do and what it's not going to do, we decided to add a chart that shows really sensitivity to a 2020 model for different levels of revenue, all the way from 0%-6%. Obviously, the middle one is the one that we chose as the model, but here you can pick your flavor and see the different metrics and what it does at different revenue levels. You can see that with no revenue growth, you still achieve in excess of $1.50 and 38.5% gross margin with $734 million of free cash flow in 2020. On the flip side, you can go to much more rarefied levels if you go to 6%. When we look at that, we feel pretty good about the model.
We feel pretty good that we have a plan that's grounded on a very good set of actions and plans, that does not only depend on what happens with the external market. We have talked about aggressively de-levering the balance sheet. These are the towers of the maturities we have right now. We did already redeem a convert instrument in the month of January that reduced our gross debt. Obviously, didn't do too much on our net debt since we had to use cash. We intend, as we have said multiple times, to get down to two times net leverage. We also have, just to be on the safe side, a $600 million untapped, undrawn revolver to be used whenever we need it. We like to operate around $500 million of cash to be really in a safe place. Capital deployment.
We expect to achieve that two times leverage by the middle or second half of 2018. After which, we expect to go back to returning cash to shareholders. The flavor of which we have still to decide whether it's share buybacks or a dividend, but it's in the plans. As Keith mentioned, we are not averse to acquisitions, but if we do acquisitions, they do have to have a very compelling financial and strategic intent. Last but not least, we do have a sufficient amount of cash in the U.S. or access thereof. We have, on point number two, about $1.8 billion of offshore cash can be repatriated because of all the NOLs that we have and the tax structure that we have. Talking about that on a tax front, we are running about 10% in 2017 on a cash tax basis.
We'll be going to about 12% for the years subsequent. I get a lot of questions about where is our position with the impending or possible changes in tax and tax code. In general, we support it. In general, we think it's something that will be good for the industry and good for us. We are a net exporter out of the U.S., we think that we are in a position to really take good advantage of any reform that can happen. If I summarize what I've presented, really the key point is a steep acceleration in free cash flow, going to that $900 million target for 2020. Getting a lot of margin expansion through synergies and operational improvements. Not depending that much on revenue growth to achieve the targets, but obviously we'll pursue it should help. We have raised the Fairchild synergy target.
We believe that we're now in a position to recognize a little bit more towards the achievement of that target. We will be aggressively de-levering, then follow that with shareholder returns. Short and sweet, that's what I have. Now I'll open it up for questions and answers, not only for me, but for any others that you might have. Yes, Ross.
Hi, Bernard. Thanks for the presentation. Two quick ones. First one is the same one I asked Keith. ASPs, how does that play into the gross margin targets you have, especially with the mix change and consolidation you guys made?
The basic assumption is still the same 6%. We didn't play on anything different, and we have the same operational, normal day-to-day cost reductions that Bill's team put in place to offset that. All of the other are complementary additional programs that will yield margin improvements. We didn't assume that consolidation or anything would give us more pricing power. Obviously, if it comes, great. It'll be better.
The second one, which hopefully is just as quick. On the CapEx to revenue side, the 6%-7%. If you're consolidating fabs in the future, and I know that's a few years out, does that change the 6%-7%?
We don't believe so. We do have still needs to fuel capacity. To your point about the 6% or the ASP decline, even though the top line doesn't go up, the units go up in the same 6%-10% proportion, even without growth in revenue.
Hi. Yes.
Yes, Craig.
Craig Hettenbach with Morgan Stanley. Just a question on the computing consumer outlook. How much is that a reflection of just you have a much bigger opportunity set with Fairchild now, and you're looking to be more selective versus just how you're viewing those end markets?
In general, we view those end markets are being mature. The margin profile is not great. In that sense, yes, we'll be focusing less on those, and that might mean that we may be a bit more selective. We are going to pursue areas like a server. That's an area that is something that we think is worthwhile. In general terms, our R&D dollars are going to the other three areas, and therefore, we don't expect to have a significant push in consumer and computing. Yes, Steve.
Yep. Bernard, on the upside case for revenue is the, say, 6%, is that 3.5% GDP? Is that what gets you there? Is the right way to think about revenue opportunity maybe two times GDP or something like that? You've got a lot of drivers, but still seems like it takes a big macro to help.
That's a tough one to answer. We have tried to do correlations between GDP and end markets, and indeed, around three or so is where things are not that good. Probably something in that neighborhood could drive a market to be in that three and a half could drive top line to be much faster. It's difficult. Yes.
Hey, Bernard. Michael McConnell, Pacific Crest.
Yes, Mike.
On Fairchild, with the higher synergies, one thing we haven't heard today was the plan for their distribution network. My understanding was you guys were looking at maybe consolidating that. How should we think about that? Would that drive more upside to the synergies, and would that be more top line or margins? Thank you.
I think it will be more on the top line. Our sales force and our sales team are actively working on doing a consolidation of distribution. Fairchild had greater than 60% of their revenues going through disti. We had more like 50 something. We think partnering with distributors now is a very good thing. With $3 billion of our revenue flowing through a disti channel, we have become a very meaningful partner with distribution, and we're getting a bigger share of mind, and hopefully it translates into a bigger share of wallet. We think it's more increased top line than really cost savings. There might be some minor by having less distributors to deal with, but in general terms, it's more the opportunity to partner up to get more design win activity and more top line. Yes, Chris.
Doesn't look like it's on. Oh, it's on now.
Now it's on.
Just two questions. One is within the 3% growth rate that you talked about, how much is portfolio pruning contributing to that? In other words, with the bigger scale from Fairchild, you could be a little more selective. Maybe talk a little bit about that and how that affects the overall growth rate. Also, in your presentation, you talked about that you hadn't quite made a decision about use of free cash flow after you hit the 2.0 debt targets. If you could give some background about where your head is right now, and where you might prioritize between buyback dividends and then debt reduction below two times.
Yes. First question.
On portfolio pruning.
On the portfolio pruning, yes. By virtue of assuming a decline in some of the end market revenue, we are de facto including some pruning. We also played in a small set of divestitures in the numbers, that's also played in. It's more a function of that, of really focusing intensely on growing those end markets that we believe in and that have better gross margins and refocusing the others. In relation to the dividend versus share buyback, that's more a discussion with our Board and understanding we do have a share buyback program that we got approval for $1 billion, and we have used about 300 and something of that. We have an avenue to immediately go back to that.
Going below two times net leverage, we will look at that. In general terms, I think it's probably better use of our money to return it to shareholders than to go below two. We think that two is a level that we feel comfortable at and gives us the flexibility to do up or down if we need to.
Hi, Shawn Harrison, Longbow. Two questions. If my math is right, you had about $2.5 billion of debt exiting the year and about $1 billion of run rate EBITDA. I'd expect probably at least $500 million of free cash flow or more this year is what you've offered up. Why wouldn't you be at two times net leverage entering 2018, and you could start next year with return to cash to shareholders? The second being just the upside synergy target. Where are we at right now? Where should we be exiting the March quarter in terms of Fairchild synergies just on a run rate?
On the first question, our total gross debt at the end of the year was $3.6 billion, not $3.5.
Two and a half net, my bad. Two and a half net debt.
Yes. Well, we should be able to get earlier, and obviously as soon as we get there is when we'll reinstitute our buyback programs. Right now, we say it's somewhere in 2018. The second question was?
Where will you be at in terms of Fairchild synergies exiting the March quarter on an annualized run rate basis?
We are probably a little bit still ahead of the plan. Remember we said that we were going to get about $160 million. Now we're saying $180 million by the end of 2018, of which $130 is OPEX. We're probably already something in that $140 or so at the end of Q1, approximately. Any further questions?
Great. Sorry if I missed this, just in terms of bringing Fairchild stuff in-house, I think one of the reasons they had outsourced some stuff or left it outside was they were trying to smooth out margin over the course of economic cycles. I was just curious, is the opportunity better to bring that in-house and that's the decision, or how should I be thinking about the cycle versus actual margin?
We think our model, which right now our model has been 80/20, still allows for that to buffer yourself in cycles up and down. Fairchild actually had moved 50% of their back end outside. The main reason is that they didn't have the size and scale, or they closed down their finance factory, where they had a lot of that back end and outsourced that because they didn't have enough volume. We have $70 billion that we flow through our factories. In that sense, we have typically much better cost than what the sub cons have. We don't have to pay them the profit, and we can pocket it ourselves. Using that 20/80, it still allows for that management of the cycle and moving capacity in and out depending on where you are in the cycle, but not 50/50.
Any comments?
In the past year, we talked about returning cash back through dividends as well, and that 80% of free cash flow. Post 2018, is there a possibility of getting a chunk of that through dividend?
Yes, it is a very real possibility. It still has to be decided with the board, but it is a possibility, a very real one.