Ladies and gentlemen, please welcome Vice President, Investor Relations, Mauricio Lopez-Hodoyan.
Good morning. Welcome to Qualcomm's 2019 Analyst Day. Thank you for joining us today. Before we get started, I'd like to thank all the Qualcomm team and executives that helped put this event together. With that, let's turn to our safe harbor statement. We will make forward-looking statements in today's program, including statements regarding our expectations, projections, and other potential future events. Actual results may differ materially from these forward-looking statements. Please see our most recent Form 10-K filed with the SEC for a description of risks and other factors which would cause actual results to differ materially from these forward-looking statements. Also, we will use non-GAAP financial measures in today's program, and you will find reconciliations to the most directly comparable GAAP financial measures in today's presentation on the investor relations page at qualcomm.com.
With that, please join me in welcoming Qualcomm's Chief Executive Officer, Steve Mollenkopf.
Ladies and gentlemen, please welcome Chief Executive Officer, Steve Mollenkopf.
Well, good morning, everyone. Thank you for coming. I think this will be a good day today. We're going to go through. You'll hear from a lot of speakers, and I look forward to telling the story. It's been a while since we've talked to all of you, and so we're going to spend a lot of time talking about 5G and the things that you would expect, and look forward to it. If we look at Qualcomm, who is Qualcomm? We're really the world's leading wireless innovator. We work on technology. We've led for 30 years. We've led every generation of cellular, and our innovations are actually core to the dramatic transformation that really started with the smartphone and will end in, I think, any industry that wants to have connected things. Where are we?
If you look at where we are as a company, we've been through a lot over the last five years. The reason that we were so interested in making sure that we had the ability to participate in the upside of 5G is what you'll hear today, really. Where are we? Where do we sit today? Number one, we have the best products and the best technology roadmap that I think we've ever had in the company history. There was a lot of opportunity to be distracted. We really weren't distracted. We were very focused on producing that. You'll hear about that today. Our IP strength is unmatched, our financial strength, really based on betting on ourselves, is also, I think, a shareholder asset that we're quite proud of.
Probably most importantly, we enter into the 5G era, which is a multi-decade era, with these assets intact. It's a real, I think, story of opportunity for the company right now. Now, why are we so excited about 5G? It's because 5G is a big deal, and not just a big deal in cellular. It's a big deal in every industry that you deal with. With 5G, it's really the technology that not only disrupts the cell phone industry, but it disrupts every industry that it's next to. It's the first time really that cellular has had this big influence, not only in its home market, but also in new markets that are around it, and they are large markets. How big?
We commissioned a study to talk about the economic impact of 5G to goods and services in 2035. The number is actually $13.2 trillion of goods and services will be enabled by 5G. That was, I think, a 12.3 billion number only a couple of years ago. It's very significant. A proxy for the value that is created by the technology is really that $13.2 trillion number. As I said, it's not just another handset upgrade cycle. There's a lot of great things that happen on the handset. You get improvements in capacity, improvements in latency, improvements in data speed, improvement in the ability to deliver unlimited data speeds or data plans. Probably most importantly, you now have the opportunity to improve the connection density and the reliability of those connections.
Now wireless connectivity is really the fundamental technology that ties a lot of other industries. It's a huge upgrade cycle, larger than any upgrade cycle we've seen in the company's history. It's happening. A lot of discussions about the 4G and 5G rates. The reality is 5G is coming along exactly as we thought it would do. In fact, it's happening faster than we saw in 4G. If you look today, there are 40 operators that have already launched 5G networks. In the first year of 4G, that was only four. It's a significant thing, and it's going to continue. You'll see, we'll spend a lot of time talking about the migration of the 5G networks, how it happens worldwide, and what you'll see, particularly with millimeter wave.
It's happening quite quickly because people see the underlying strength of the technology and the ability for economic impact, and they want to participate in it. What does Qualcomm do? At a very high level, what do we do? If you look back at the history of the company, what we do is we invest and invent the fundamental technologies that make cellular happen, then we scale that worldwide. We scale it through the standards bodies, we scale it through aligning the ecosystem of cellular, and we scale it through our product business. Today, that sets up not only the opportunity in cellular, but the opportunity in all these other industries, the industries that enable that $13.2 trillion. How do we profit? We profit really in three fundamental ways. Pretty easy, actually, if you think about it.
Number one, we make more revenue per device, either in the cost of selling the baseband device, the licensing revenue that we get, or from increased content in that same device. RF front-end, fingerprint, great examples of additional components that we sell at a systems level into the device. Second is we have a growing opportunity to leverage that same technology roadmap that we produce in the smartphone space into new industries while they're being disrupted by this change that we're enabling through the invention. Third, and finally, out in time, our belief is that the future of the cloud and AI intersects with 5G. Of course, we're going to be waiting there to make sure that it's an opportunity for our shareholders. Let me give an example. I talked about industries being disrupted by cellular.
Really, if you look, 2G, 3G, and 4G was really a story about the cellular industry, with the exception of one industry, which was automotive. Big, established industry, established players, established supply chains, and very attractive semiconductor opportunity for a lot of companies. What happened was with 4G, you started to see that supply chain and the technology systems of the car changing. First with telematics, so you connected the car to the internet. It changed who could play in the car. Great entry point for Qualcomm, became a very good business for us. Second, the user experience, the way in which you drive the car, the way in which you service the car, the way in which the consumer interacts with the car is now really being driven by the experience that you've seen in the smartphone space.
As a result, it's created an opportunity for us to not only sell modems into the cars, but now we can sell computing elements into the car as well. We have a significant infotainment business, which we'll talk a little bit about today in terms of numbers. Why did that happen? Why was that an entry point for Qualcomm? It's because the computing that you're forced to do, the power consumption, the graphics rates, the real computing IP that's required in order to enable that user experience, which comes from the smartphone space, the traditional players really weren't prepared to deal with. If you go at that industry without having the ability to spread R&D across a larger computing space and to do it focused on low power, you weren't going to be successful.
For us, what we saw was that we were able to go into that business and displace a lot of traditional players because we came at it really from the benefit of having the smartphone space. We see the same thing happening in ADAS. In ADAS, you've probably seen reports of us having self-driving cars around San Diego. That's because we can take that same system expertise that we use in the smartphone, couple it with the IP roadmap, and in particular, the low-power computing roadmap that we have and offer that to customers. That's a budding business for us, and we're excited to be able to extend that. Auto, big industry disrupted by cellular. The traditional players are at a disadvantage because they don't have the IP roadmap or the breadth of technology to compete, and it opens up an opportunity for Qualcomm.
The first industry to really have that happen, it happened actually in 3G and 4G, but will accelerate in 5G. It's a good example. You're going to see the same thing happen in many other big established industries. I'll point out, too, manufacturing and gaming. Manufacturing, the way in which you automate a factory and the benefit of connecting that wirelessly, big area of focus enabled 100% by 5G. Can't be serviced by Wi-Fi. Really, there's a lot of features that are added into 5G to allow the reliability and really the latency required to make it happen, and it's a big opportunity for Qualcomm. Similarly, gaming. The gaming industry will change from essentially a device that you're used to seeing on a console, and it'll be much more streamed. You see that in China today.
It's not an accident that we are very closely enabling them to do that through really their streaming gaming service. You see that with some of the Tencent things that we have announced. It's a perfect example of the combination of 5G and the roadmap, the low-power roadmap that we have that creates an opportunity for us to really leverage that existing IP. Finally, if you look down the road, what are you going to see in terms of the intersection between the cloud, 5G, and AI? Our belief is that what you will see is really a second generation of AI and the cloud as the industry deals with the fact that there's so much data being created in people's business lives and their personal lives.
The goal is how do I make decisions, smart decisions about what data needs to go back to the cloud and which data needs to be addressed or make decisions locally? Of course, for us, that basically means the cloud, 5G, and AI will live together at the edge of the internet. Big opportunity for Qualcomm. We think we're differentiated in terms of the way in which we go after it, and I think it's a big opportunity down the road for us to really benefit from that transition. You put it all together, what do you see? You see, I think, a big expansion of our SAM in a way that we can address it without having to do anything that I think you're unfamiliar with us doing.
Meaning our roadmap, our core competency, the technology roadmap that we have from our cellular business enables us to go after some adjacent opportunities that really grow our business. If you look, we think we have about a $65 billion SAM today. In three years, that'll be a $100 billion SAM. A lot of opportunity for us, a lot of excitement around the opportunity for us to be able to go after what 5G enables. Kind of summing it all up, I think what you have to take away is we're very confident in our future. We think we have a great opportunity. As we've talked about, 5G is going to be a big thing for us, and we have a technology roadmap that allows us to grow in adjacent businesses. You're going to hear about that today.
Let me give you a quick roadmap of kind of what you're going to see. Jim Thompson is going to come up. He's our CTO. He's going to talk about a lot of things. He'll spend an opportunity really making sure you understand the technology roadmap that we have. Not only products, but IP, and it's our ability to leverage that IP into new opportunities that I think you're going to enjoy. He'll also talk a little bit about the technology in our front end, obviously, an area of interest. Alex will talk about the licensing business. The way we think about the licensing business is that a lot of the hard work is behind us there. We think that's going to be a good solid earner for us for a long time, and I look forward to Alex having an opportunity to talk.
Cristiano will talk about the products as you would expect and the rollout of 5G. There'll be a couple of new things in there, which will be interesting. Finally, we'll end with Akash, who'll put it all together, try to wrap up the financials. All in all, it should be a good story. We're excited to tell it, and thanks for joining us today.
At Verizon, we are extremely bullish on this 5G transformation. In fact, it's the core of what we're doing with our network for our customers. Our collaboration with Qualcomm is longstanding, and it remains critically important. They're a trusted partner of ours, and we can't do 5G without Qualcomm. We need them, and we love the partnership. We now have a spec. We've been deploying on that spec, and we have over 15 markets right as of the moment with 5G mobility in them, and we'll be at over 30 markets by the end of this year. When I say 5G mobility, at Verizon, that means ultra-wideband, millimeter wave technology providing massive bandwidth and speeds. We believe that that's a key differentiator that sort of makes 5G generational. It's the G in 5G.
You can't just have incremental improvements and say that this is a generational lift. That's why millimeter wave is so important. Being first out there with millimeter wave showed us that it works, and it showed us that it works really, really well, which made us more bullish even on our deployment strategy. We're going fast, we're going hard. We're working with cities all across the country to place this equipment, and bring the benefits to consumers, to businesses, and to society.
Ladies and gentlemen, please welcome Executive Vice President, Engineering and Chief Technology Officer, Dr. James Thompson.
Hi. Thanks. A lot of you don't know me, I'm sure. I've been at Qualcomm for a long time. I started in the CDMA project. That was a very interesting experience. Lately, or in recent past, or for, let's say, the last 18 years, I've been head of engineering for QCT. About two and a half years, I've been the CTO of the company, which means I've taken over all of engineering. That's what that means. I have all of engineering at Qualcomm. I think we have a pretty interesting technology story to tell. I'm going to start off going through our technology strategy. Think of that as we have all these incredible assets from smartphone, and we're applying them to other businesses, as Steve said.
I'll explain how we do that and some of the things that we do with our products in order to make that happen. 5G. We have a 5G leadership, and the 5G leadership story is really about investing way ahead of the industry. I'll talk about that, too. Our mobile platform. There's many dimensions to our mobile platform. I'm just going to talk about a few of what I think are the most important parts. Qualcomm is really an innovation factory, and that has to do with the people, the culture, the organization, and starting with the research. We have research teams that their time horizon is about 5- 10 years. You can think of these research teams as very similar to the original CDMA project that we had. They're looking at technologies that go out fairly far.
A lot of the work that they do feeds into the standards. The culture is very similar to what it was in the CDMA days when I was involved in that project, where it's pretty academic, but it's grounded in.
In practical business considerations, practical technical considerations. It's a very, very healthy culture, I would say. We've expanded that beyond wireless, so now we have an AI research team. We've had that for quite a while. I'll talk about that. We also have had a multimedia research team for probably almost 20 years now. Very successful teams. If you look kind of as the company grew, we started QCT way back when, and QCT was pretty small in the beginning. We were doing one chip set a year. As it grew, it got much more complicated. We started out in one building, and then we were in multiple buildings, and then we were in multiple cities, and then multiple countries.
We have a very large operation in QCT, and so the culture there necessarily required that we be very focused on our customer, on-time delivery, predictable execution, which is very different than research. As that grew, we needed to, I would say, create some intermediate research groups, which we call technology teams. These technology teams, they're looking two to five years out. These teams, we have probably two dozen of these teams. Examples would be graphics, video, modem development. What we do is we have these teams that are focused on that. They work with research teams on one side and then deliver the products on the other side. It's a very healthy environment. Each one of these three phases, each has its own kind of micro-culture, I would say. It's been very successful. Okay.
Our technology portfolio is really the best in the industry. Where it comes from is the smartphone. As we were developing smartphone, we started working on smartphone 20 years ago in around the year 2000. Each one of these technologies that you see up on the screen are things that we initially started organically within the company. Now, I'll just go very briefly through it. Wireless, we're very well known for 4G, 5G, obviously. One technology I'm not going to talk about today other than right now is Wi-Fi. With Wi-Fi, we actually have a fairly long history. We have very strong business there. We're really one of two companies in the world that really drives Wi-Fi. Our Wi-Fi 6 products are just coming out. We have access points now.
Since products are just coming out, we've had a chance to benchmark ourselves against that next best competitor, and I think we're doing very well. I'm optimistic about our position in Wi-Fi 6. I'll talk on our front end. This has been a 10-year journey with some ups and downs, and I'm going to talk about that later on, because I think that requires some detail. On the application processor, you may not be aware, we're very well known for modem, but application processor, I think we're certainly the strongest company in mobile. I think we're the strongest company in the world. There's many dimensions to measure that. For example, mobile GPU, I think we have the best dedicated AI processor. DSP, I think we're the best there. Camera, video, a lot of these multimedia technologies, a very strong portfolio there.
There's other components that we do, like fingerprint would be an example. The summary here is that we've got a very, very broad and deep technology portfolio. I need to talk about AI because it's something that is affecting everything that we do. We've been working on AI for about 10 years, and we have a research team, I mentioned that before, and that research team is focused on energy efficiency or power efficiency. That's something that maybe is different. We have a very different focus than the cloud guys, for example. What we do is we have this centralized team, a research team, and actually we're getting a lot of recognition from academic conferences and so forth at this point, because I think what we're doing right now, what we started doing 10 years ago, is becoming very relevant.
If you look at how we use that research investment, it goes into all those technology teams. Virtually every technology team that we have is affected by AI. An example would be, for example, camera. With camera, I'll just give you a quick example. If you're processing an image, if you process your face, for example, you would prefer to do that differently than processing grass. If you're standing on the grass. If you can determine what is a face and what is grass, what is sky, what is whatever, then you can process things differently and end up with a better result in terms of image quality. That's just an example of how we're using it all over in our technology teams.
We have something we call Qualcomm AI Engine, which you can think of that as we're in our fourth, going on to our fifth, about to release our fifth generation. That's hardware at the base. At the base, there's CPU, GPU, and also dedicated AI processors that are super efficient. On top of that, we provide software for both our internal people to use and also our customers to use. That affects almost all of our products today. I would say probably 90% of our chipsets that we sell have some AI component to it. It's been a long-term commitment, and I think it's something that's very important to our future. Okay. Our technology strategy, like I said, is leveraging all of this technology that we've developed from the smartphone.
Everything that you use in your smartphone, everything you like about your smartphone is there. We own that technology. We can do whatever we want with it. If you look, you'll see that every one of our adjacent businesses outside of mobile takes advantage of that technology, every single one. Then we also share products as well. There are products that we do for mobile that we pre-plan to be used in some of these other markets, and I'll explain that. I'm going to give you one example, and that's in automotive, which is our most longstanding adjacent business. A very simple example is telematics. With telematics, that's modems in cars, and at this point, I think in the U.S., it's more than 50% of cars are now going out the door with telematics capability.
What we do is we do a modem design, no application processor in it, a modem design, and then we pre-integrate Ethernet, for example, which is not required in mobile, but it doesn't take up much area. It's very simple to do. We pre-plan certain parts to go into the automotive market. That's one example. We can take advantage of the scale that we have in mobile and then just apply it to automotive. Infotainment, a little more complicated. Infotainment is a business that we've done extremely well in, and I believe Akash will give you some ideas about where we stand in that. We're winning almost every one at the high end. Really, it's surprising, and it's not. If you look at the strength of our technology, the application processor technology we have, and we can apply that to automotive. It's very powerful.
Other competitors in automotive don't have the scale of investment that we have in those areas. I'll give you just a quick example. In a smartphone, you're running this one display, and you might run an external display at the same time, so it's fairly limited. In an automotive product, it can be running many displays, and also the aspect ratio of the displays are much wider. In fact, you might want to put multiple displays together, so you need to synchronize those displays. Those are just little tweaks to the design that we need to make ahead of time. If you look at instrument cluster, that's the thing that tells you how fast you're going and gives you safety warnings and things like that.
There are additional requirements associated with that because in instrument cluster, you want to run a safe operating system to run it, and you want to be able to reset infotainment without affecting your instrument cluster because you don't want that to all of a sudden go away. It's very important that it's safe. It's considered safe from a technical sense. There are just simple things that we do in our design up front in order to make it safe. There's things like running multiple OSs. You have a safe OS. You have Linux, for example, on your infotainment. You're running two different operating systems. You need hardware virtualization, so you need to have that in the hardware, and then you run a hypervisor on top. That's how you do it. There are things like context switching and GPUs.
You need to be able to jump between the two without affecting the other display, especially when you have a safe display. There's a lot of things like that that we do that we build into the design up front so that it can be supported for infotainment. That's why it's very hard to compete with us, actually, in infotainment because we make these little changes, and then they go into the infotainment product. Okay. ADAS is something that we're looking out to the future. We've been working on this quite a while, and I don't think we haven't advertised it too much other than some of the demos that we've done. It's a very similar story in that we're taking assets, and we're actually sharing some products with other businesses, and we're adding the requirements in for ADAS.
I think one of the headliners for that, maybe two, one is that anybody who uses our infotainment is going to have a platform that they're familiar with that they can leverage their own internal development over to. Also, neural processing is something that AI is something that is very important to those customers, and we have a very strong position in terms of power efficiency. That's something that for automotive, you wouldn't think it, but automotive power is super important because you don't want to have any active cooling. Active cooling is super expensive. If you can just bolt it down to the chassis and suck the heat out that way, that's a much better approach to doing it. I think we got a very strong story in ADAS as well. Also, there's a lot of different sensors associated with ADAS.
We're leveraging a lot of our technology, and we're working on some of the sensors. I think very compelling story. I'm very excited about the automotive business. I'm going to talk about now our 5G leadership, and that's mostly, again, a story about us investing well ahead of the industry. I'll give you a little bit of a history lesson. We have been working on wireless for 30 years. This is something we've been doing over and over again. I'll try to explain to you how we go about doing this. It's very consistent from year to year in how we approach this. If you go back to the CDMA days. We had obviously a very good idea. In order for us to convince people, we didn't just write papers and things like that.
We had to actually design the system end-to-end, so we created this end-to-end system design capability, which is very unusual, actually. Then we had to build the system ourselves. We designed it, we had to do prototyping of the system end-to-end and test the whole system, and then demonstrate the system publicly, prove to people that it actually worked. If you think of that's our playbook. We have that exact same playbook. As we went into 3G, of course, a lot of you know DO, we did DO, which was very different. It was an IP-based network. We did the exact same playbook where we did an end-to-end system design. We prototyped the system, proved that it worked, and then brought it to the standards.
Again, in 4G, we had an internal program called Ultra Mobile Broadband, we called it internally. Again, there, that was an OFDM system, MIMO system that we then continued. We brought that to then the standards collapse, so 3GPP and 3GPP2 collapsed into one thing, and we brought that to the collapsed standard. Again, in 5G, same thing. This time, some of the requirements or what we're doing is different, very different than 4G, but there also is a lot of leverage between 4G and 5G. Leverage like OFDM, leverage like some of the MIMO concepts, but it's done very different. The focus on 5G is extremely high efficiency, opening up new spectrum, and then adding these new services that Steve talked about. Okay. We invest. This chart really is to say that we invest for the long term in wireless in particular.
We've spent $60 billion so far in R&D, which is an astounding amount when that was all added up. I have to say, I was amazed. If you look at how we organize ourselves, it's again, we have this wireless research that's done, and these teams do these end-to-end designs. If you look at, it's not just a group of people that walk into a room and go, "Okay, let's do an end-to-end design." There's a bunch of work that goes on prior to that informing that design. For example, we have decades, actually, the entire history of the company, of teams working on channel coding, for example, which is a very important part of a communication system. Also, MIMO, OFDM, or I should say OFDM, probably we've been working on that for about 20 years.
That informed our design in 4G and then also 5G. We have these, think of it, we have teams doing end-to-end system design, but we also have research initiatives going, a bunch of these research initiatives going in parallel that feed into that system design. From that system design, what we do is we prototype again, we prove the system works. We typically demo. We've done a lot of demos in 5G at Mobile World Congress. Then we take those designs into the standards. Ultimately, the standards produce a result, and then we commercialize.
I think it's probably important to understand that the fact that we commercialize after that, we create products, in part, we have an advantage on the product side because we understand so deeply the technology, but then we also feed that back into the research that's going on to keep things grounded. Anyways, it's a very virtuous cycle, I would say. Maybe one other point I'll make on this chart is that you'll notice that when 4G was released or 5G was released, it wasn't just one release. There's many, many releases. Like between 4G and 5G, there was seven releases, I believe. With 5G, my expectation is it'll be very similar to that or maybe even more. Okay. When we talk about our inventions, our innovation, our focus is very much on foundational inventions.
What I mean by that is there are things that you can do with a standard or the design of 5G, for example, that make it much better. The capacity goes way up, the data rates go way up, or there's some vector of dramatic increases, say. Then there are pieces of the standard where there are, I would call them implementation details, where there's 10 different ways you can do it, and so you can pick one of them, and it doesn't matter which way you do it. Our focus, because we're doing the system design, is really about these foundational technologies. I'll just run through a few of them. Advanced channel coding, remember I said that that's really the heritage of our company. We have some of the best channel coding people in the world that work for Qualcomm.
With 5G, we picked LDPC as the channel code. Why does that matter? It's because we need to, one, be able to handle these huge data rates and also very different packet sizes, so really giant packet sizes for a data transfer, and then for voice-over IP, very small. You have to have flexibility. You want to be able to support retransmissions if you say packets get lost. There's a lot of innovation associated with that. Massive MIMO. This is actually one of my favorites. We've been working on MIMO for a long time. We had MIMO in 4G, but what's different about 5G? It's really the massive part, that practically implementable massive MIMO. What that means is what you can do is for a given piece of spectrum, is you can get a lot more capacity from it.
Traditional 4G MIMO, what we would do is we could produce a data stream or let's say up to four data streams and send it to a particular phone at a particular time, and with only that phone. We would find another person in this audience here, and we could send data, four streams to that person, maybe three streams. Whereas with massive MIMO, it's also a multi-user. You can send four data streams here, you can send four at the same time, four data streams here, four data streams here, four data streams here, and not interfere at all with each of those different data streams. What you can do is you can create, in a sense, hotspots at individual's phone, and then you can, for the other person you're communicating with, you can null out that interference.
There's a lot of math associated with it, but it actually works quite well. It's something that will improve the capacity of these networks, and then you don't have to put in new cell sites because they'll work in the same cell sites that exist. You also can get much better coverage from it. Anyways, I think a very important technology. Scalable OFDM. OFDM is part of the physical layer, and think of it as these tones that are transmitted in the spectrum. If you want to have a very wide range of spectrum going from, say, 600 MHz all the way up to millimeter wave, you actually want different tone spacings. To do that, it's very important that you have certain numerology so that you can reuse the hardware that you have, so you don't make the modem cost super high.
You need to have the same tone spacing if you want to share spectrum with 4G, and that's a very important concept as well. That's something that's very fundamental to 5G. Millimeter wave. We've been working on millimeter wave for a fair amount of time, almost 10 years, I would say. Honestly, we started off with millimeter wave, and I was a skeptic. It actually works quite well. Some of you have probably been to our campus and done some demos. It works very well. Flexible slot structure. That's also something that's very important. When 5G was originally being proposed as the range of services we needed to support, there were a lot of proposals that said we're going to do a different system for each one of these.
What we did is said, "No, let's create this framework that allows us to multiplex all of these different types of services into one system." If you're going to do a system, what's going to get deployed is the mobile system initially, then people are going to go, "Oh, am I going to spend a bunch of money on this system for IoT? Maybe not. I'll wait." Whereas if you can put it all into one system, then it gets deployed ubiquitously. I think that's a very important concept. These are foundational implementations or inventions associated with 5G. Okay. What I want to say with this slide is we have release 15 coming out, and the releases, as I said before, are going to be going all the way out through the decade, through 2030.
Maybe even beyond, because part of what we've done with 5G is we've made it, we call it forward compatible. We're trying to make it very flexible. We learned from 4G some of the inflexibility that it had. With Release 15, you can think of Release 15 as being focused on mobile broadband. That's smartphones. You can think of that's what pays the bills for the operators right now. That's why that's being deployed so broadly, because they see a lot of opportunity in mobile broadband. Some of the subsequent releases, Release 16, will have enhancements to mobile broadband, but also there's a bunch of features associated with these adjacent markets. For example, high reliability, low latency communications for some of the factory applications that we've identified. More C-V2X, so car peer-to-peer communication with cars. We've opened up some additional spectrum in unlicensed.
Those are kind of some of the highlights. In Release 17 in about two months, that will be finalized as to what the work items are in that. This is an ongoing thing. Think of it as every year and a half or so, there'll be a new set of releases associated with 5G. Okay, I'm going to talk about the mobile platform. I'm only going to talk about four different areas. One is our modem, and that includes our front end. I'm also going to talk about camera, graphics, and AI. Our mobile platform, or I should say our modem. We used to think of the modem as our baseband and then our RF transceiver.
As 5G has come in, it's added a lot of complication, a lot of difficulty in doing design of RF components for 5G, I'll explain that a little bit later. Now when we think of modem, we no longer think of it as just the baseband and the transceiver, but we think of it all the way to the antenna. That's a very important concept. I'll talk a little bit about millimeter wave, about what that looks like, and then also our front-end and how that ties in with the baseband to improve performance. Start with the baseband. If you look at products today that are in 4G that we can compare. What I've got here is a picture of the Snapdragon X24 and the Intel modem, the XMM 7660.
If you look at it, the size of our competing product is quite large. It's a factor of four larger. The process nodes are different, so our estimate is that they use twice as many transistors to produce a product that actually has a lower spec. It has a 20% lower spec from a throughput point of view. If you measure it, where they do at their peak rate, their 16-layer, and we put ours at 16 layers, they end up, and it kind of depends, there's probably 100 different measurements that we make, but you can think of it as 10%-20% less throughput on top of the lower spec. You look at power, and this is, I think for us, the biggest advantage that we have is if you take voice power, for example, the Intel modem measure.
These are measured with products and you can just go out and buy the products yourself and make the same measurement. 50%-60% higher voice power, depending on which voice mode you use. With data, I have 27%-40% higher power. It tends to as the higher the data rate goes, the bigger the difference tends to be between. We have a very, very large power advantage. Looking over on the right side with the 5G comparison. This is us compared to the HiSilicon part that's out there right now. This is in the same process node. The 2.6x larger is in the same process node. I think they're directly comparable. They're using 2.6x as many transistors to get the same function. From a peak rate, they're 40% lower, and they don't support millimeter wave.
I think from a design implementation, from an efficiency point of view, I think we're doing very, very well. In 5G because the networks aren't mature, I don't have, yet, I don't have a lot of performance data to compare, but I expect that we'll do extremely well there too, power and performance. Okay. Why is 5G RF so much more complicated? Much more difficult? One is that there's an order of magnitude number of bands that you have to support, band combinations. When you have band combinations, you have to worry about interference between all these different transmitters and receivers. It makes it very complicated. On top of that, 5G is also much wider bandwidth. It's 100 megahertz versus 20 megahertz. That bandwidth makes a big difference.
Also, the peak to average, meaning the signals, instead of being relatively flat, it moves up and down quite a bit. There are modes of 5G that are higher transmit power than you have in 4G. That combination of things makes it very difficult to do the RF in sub-6. If you look at our position, and we think of this chart here as our position in not just our front-end, but in modem. At this point, we have every single component, so power amps, LNAs, switches, antenna tuners, and then on top of that, we have the modem itself. In each one of these component categories, at this point, we are as good or better than our competitors in every single one of these areas. I'll just give you a quick proof point.
I'm sure the filters is something we acquired through the TDK acquisition, that's something that we've been catching up on in this. This is kind of like in some sense, a weakness. If you look, band 40, for example, the blue line at the top is Qualcomm thin-film SAW. We have the two competitors below it. By the way, the higher the better. That's how you to read this chart. That's lower loss in the filter. If you look on the far right, that's band 41. These bands are on either side of the Wi-Fi two gigahertz band. You can see that the dark blue again is thin-film SAW. I put in for reference our BAW 4.0 because those are out in products today. It's better than our BAW 4.0.
Then we have the FBAR competitor, which is the bottom line there. I highlighted on the left side of the chart there because what also matters about the filter is not just it's the highest, but also it has a very steep roll-off at that particular point. We also do very well there. Anyways, I thought this was an interesting proof point. Now I want to talk about how we use our front-end as part of the system, our whole modem system. There's a lot of system optimizations that you can do if you own end-to-end. If you own baseband to antenna, there's a lot of things you could do. The first one is a fairly simple example, and that's antenna tuner. If you hold your phone, there's antennas.
You don't realize it, but there's antennas all inside the phone. You hold, and you can detune these antennas. What we do is we have a closed loop system with our baseband that makes measurements and then adjusts what's called the matching of the antenna. What it allows you to do is, on the transmit side, transmit up to twice as much power out of your phone, and then on the receive side, receive as much as twice as much power. I should say twice is probably a typical kind of number. You can do much more than that depending on the circumstance. Anyways, if you see a lot of the high-end or almost all the high-end phones use this capability at this point. Envelope tracking is another one.
We had envelope trackers for 4G, and 4G was much, much more forgiving because of the narrow bandwidth, the narrower bandwidth. Because of the wide bandwidth, you really have to do a co-design of baseband so there's a lot of digital predistortion that has to be done in the baseband to make this work, and then also the design of the characteristics of the PA, and then you have to have a very tight integration of the envelope tracker. Okay? What the envelope tracker does is it puts the PA in its optimal position or optimal bias for efficiency. That's really what it does. If you look, these are just comparisons of envelope tracking to average power tracker. Most other competitors are going to be supplying average power tracker.
The top line there is max throughput, so that would be the mode, the modulation scheme that has the highest peak to average. You can see that there's a very substantial benefit we get from that. Actually, the way to read this is it's on the order of close to 1 W kind of savings. In a phone, that's a lot. If you go all the way down to the bottom, that's where you're using the lowest order modulation. You don't have the peak to average is not quite as high, so you don't get quite as much benefit, but the transmit power is so high there that you're also saving something like approaching 1 W.
Anyway, my point here is that envelope tracking for 5G, I think, is something that is very difficult to do, and it's going to provide huge benefits. Another one I'll talk about here is Smart Transmit. With Smart Transmit, we monitor the output power of every transmitter on the phone very closely. You probably are aware that the FCC has requirements about exposure to RF radiation. Based on that exposure, there are certain limits, and those limits depend on what frequency you're at. There's a fair amount of complication associated with that. The typical way of handling this in the industry is you just turn down the power, or you say it can't go above this level, so you guarantee that it stays below that exposure limit.
Well, what we do is because we track it all very closely, we can actually transmit more power at times and still meet the limit. By tracking it very closely, we can improve the overall transmit power pretty significantly at certain times. What that does is it makes your link margin much better. In other words, you can go further away from the cell and it still works. It's not just the uplink that it helps. It also helps the downlink because there's actually a lot of communication that goes back and forth when you're transmitting data, and it actually helps improve that communication, which actually improves your downlink speed. I think it's a very important technology that we have that it's hard to do, actually. I'll talk about one more case or one more system implementation, and that's in millimeter wave.
Millimeter wave we do in a module. It's connected to our modem, but the module has a phased array antenna built into it. Built onto the substrate of the phased array antenna are power amps, LNAs, and switches, and also a little power management module or chip as well. What that module then we sell, and there's different antenna configurations, and we sell that module to our customers, and they put them, in most cases, three modules, and that's kind of the typical configuration. What these antenna modules do is they can steer a beam to track the base station. That's why. Maybe I'll say not just track the base station, but they can even track multipath.
If the antennas or the base station is in that direction and then you turn and your back is to that base station, you can pick up a multipath off a wall, a post, another person even. That's why it's pretty robust, is because of this beam steering capability and the beam management. It's also very complex. Having baseband, understanding the whole system and how to manage those beams is very important. Okay, onto another topic, graphics. We sell hundreds of millions of graphics cores a year or products with graphics in them. That provides us a fairly big advantage in that our platform is so ubiquitous that all of the developers optimize their software and tools for our platform. That's a big advantage we have. What is our advantage?
It's really the most power-efficient graphics core in mobile industry, or I should say, blanket statement, most power-efficient graphics core. That's really the metric that we use. The other internal metric that we use is most area efficient as well. Those two statements are true. If you think about graphics, you can always get more graphics performance by putting down more silicon. Okay. Absolute performance is really more like a design choice. What you'll find is because of power efficiency, it ends up limiting your performance if you're in something like a smartphone, because there are thermal limits. You can only dump so much power into a smartphone. I put this plot here to help you understand that. This is a measure of the Manhattan Benchmark. Think of this as a game. The benchmark runs for about a minute.
The blue line is our performance over time, and you can see this goes out to 20 minutes. Think of this as your kids playing a game. They don't play games for 30 seconds. They're playing for 30 minutes. Think of it as our product has the performance stays. That's how we design it so that it stays flat or continues on. This is our next best competitor, and what they'll do is they'll provide a benchmark that says, "Hey, look at how good we are," and they'll be way up here. If you look at it over time, it actually goes down, and the phone itself is struggling to maintain the balance of thermal power. It moves the performance up and down and up and down.
That's a very big difference between us and our competitors, and I think that's probably the most important thing. With graphics, we can use this in not just mobile, but that power advantage also, I mentioned that in automotive. That's something that's super important even in automotive and almost every industry that we go into. With camera, we may not realize this, but we're the best image processing company in the world. If you look at cameras, there are more cameras in the world today powered by Qualcomm than any other company in the world. That's pretty shocking. That's true. Recently, Xiaomi came out with their DXOMARK score for their camera. DXOMARK is a benchmark that is a standard benchmark. What it does is it just looks at image quality plus a lot of video features, electronic image stabilization.
There's all sorts of different measurements that go into this. It is really where everyone goes to benchmark their phone or benchmark their camera, sorry. With Xiaomi, they just came out with that a couple of weeks ago. We work very closely with them, and they have the highest benchmark score in the world. Also from an image quality point of view, we spend a lot of time benchmarking our own ISP, and comparing it to everyone else's. I think we're in a very good position there. I'm going to show you just one of the measurements that are made in this DXOMARK, and it's image stabilization. Electronic image stabilization, as you're taking a video, how much does it move up and down? What we did here is you just have two phones.
One is Xiaomi, one is the next best competitor. Every other one is worse than that next best competitor. It's on a board. We're not taking the competitor and shaking it up and down and making it look bad. It's actually exactly the same movement. What you'll see is that we're walking along. You can see the other one's rocking a little bit, as you go down the stairs, you can see it's very different. This is the next best competitor. That's just an example of the kind of things that we do with camera. Anyways, I'm very proud of where we are with our camera and our camera team. Okay. Little bit more about AI. AI is going to change how networks are implemented. Along with 5G, those two things are going to change things quite a bit.
Maybe I'll just explain. If you look at way you experience AI today is mostly through the cloud. A lot of it is you click on things, you send pictures in, you do things that is monitored essentially by the cloud companies, and what they do is they try to profile you, which is how AI is primarily used today. They try to get you to buy things, advertise, things like that. That's what it's mostly used for. I think that AI is going to have the biggest impact on the edge of the network. If you think about on the edge of the network, there's all these sensors. We're digitizing the world. There's all these sensors, and like, for example, a video camera is probably the most prolific sensor that exists out there, but there's lots of different types of sensors.
What AI does super well is it can make sense of data. You got all this data coming in, and it's very good at making sense of huge amounts of data. There's no amount of coding that you could do that you could do that as well as AI. It's very powerful in that respect. Then what we believe the way things will play out is that AI has become super important on the edge. I'll just give you a quick example, like automotive. AI for autonomous driving, it's going to be something that's kind of a fundamental technology to make this stuff work. You would not, for example, do that AI processing in the cloud and then send it back to your car, because there's a latency issue there.
What you want is you want all that processing to be done on the device, in this case, the device would be the car. There's privacy reasons, performance reasons, latency reasons to do that. What you're going to find is that because of that latency to the cloud, there'll be these, what we call edge clouds. You'll start seeing cloud extending itself very close to the network to take advantage of the low latency that you have with 5G. In that case, for example, you could do computational offload, for example, that you could send from your device, not just a smartphone, but many different devices, cars, and you can do a computational offload. I'm not going to go into great detail here, we think that this is going to be a very significant driver of the future. Here we go.
We've been spending a lot of time, like I said, 10 years on AI. We have what we call in mobile our Qualcomm AI Engine. That's it. Qualcomm AI Engine. These are just to give you a sense, this is a measurement of power efficiency relative to our next best competitor. You can see there's different benchmarks here. For example, ResNet is on the left. ResNet is a very kind of heavy image classification benchmark. These are the standard benchmarks that are used. You can see our performance in power. MobileNet is really the kind of a lighter weight benchmark. You can see we do very well in terms of power, because that's always our focus. Also from an absolute performance point of view, I could show those as well. With almost all cases, we're leading as well. Okay.
Why is that important? We've developed this neural processor. It's a dedicated engine, and that dedicated engine is much more capable than a CPU, for example. A CPU is horrible at AI processing, neural processing. If you look at graphics, it's much better than a CPU. FPGAs can be much better than a CPU by a factor of 10. If you go to a dedicated engine like we've developed, it's yet another factor of 10 in performance or power. It's very important that you do these dedicated engines for neural processing. We have a product, it's called Cloud AI 100. That's one that I'm really excited about, actually. I should say, it's for inference in the cloud and inference on the edge cloud.
What we did is we took our AI Engine, specifically that neural processor that we do for mobile, and then we designed it so that it could be expanded up. We expanded that. That's what the little brain to the big brain means. Then what we did is we laid down multiple instances of that neural processor on a chip with lots of IO, and then we're working very closely with some big customers, and their feedback to us is that we're better than 2x the performance for a given number of watts. In the cloud, the watts, that matters a lot. We think we have a very strong position going into this market. Key takeaways from my talk. Scale of mobile investment gives us an unmatched technology portfolio. We effectively leverage our large-scale technology investment into other businesses.
5G leadership comes from a 30-year commitment to research. 5G strongly benefits from a modem-to-antenna design approach. There we go. We are driving the powerful intersection between 5G and AI. Thanks.
[Non-English content]
The innovation that Qualcomm brings across global communications is fantastic. I mean, Qualcomm is the front runner of innovation, leading innovation. For us, it's quite natural because we are a innovative brand. We want to provide an exceptional customer experience and change. The values of these two companies, BT and EE with Qualcomm, is a natural collaboration. One of the biggest technology challenges that we had in order to launch 5G is to set up the ecosystem. Thankfully there was a great ecosystem to bring together working with Qualcomm, who are driving the chipsets, but also the evolution of the smartphones and devices. We were really quick to engage with that, build our ecosystem. BT and EE, we are one of the first to launch, not only in Europe, but globally, definitely first in the U.K.
It's part of our overall strategy for an fantastic network experience with our customers in the consumer space or in the enterprise space.
Ladies and gentlemen, please welcome Executive Vice President and President, Qualcomm Technology Licensing, Alex Rogers.
Good morning, everybody. Thank you for being here. I could actually stay here for five seconds and leave, and simply reference back to what Jim just said. If you want to understand why the licensing business is foundationally strong, just go back and listen to what he said. That is the basis of the licensing business. It is good to be here, and for me, it's actually personally interesting because I sat in my seat over there, and I could look through the windows of the New York Times Building to the ancestral home of the Rogers family business. My father and my grandfather owned a bowling alley and a bar in the Port Authority. Unfortunately, they wouldn't let me take over the business. I did what any of you would've done. I packed up and moved to California.
That business, unfortunately, is no longer there, but you can still go bowling. We don't own it. All right. Let me talk about the licensing business. There are three key attributes. $60 billion investment. Jim talked about this a little bit. The focus is that we develop technology that enables cellular. The focus of that is systems-level innovation. We've been doing it for 30 years, and we continue to do it. We're good at translating innovation into legal property rights, into a patent portfolio. 140,000 patents and applications pending worldwide. We're good at the mechanics of getting deals done. 13 billion devices licensed. Again, the bedrock is what Jim said, enabling technology. We create enabling technology that the industry wants and that the industry uses. Let me dive into it a little bit more. Three timelines here that matter.
There's an R&D timeline, there's a standards timeline. The R&D timeline is in the blue shading. The standards timeline are the steps that you see across left to right, rising up. Those represent individual standards as they're promulgated by the standards industry, by the standards organization. There's a commercialization timeline, and each of these timelines is important to the licensing business because each of these timelines and Qualcomm's activities in these timelines creates strength in our intellectual property portfolio. R&D, and Jim mentioned this, I'll just refer back to it. It's really critical. Ongoing early investment. He described to you how the machine at Qualcomm works. Very theoretical to begin with, but very practical, okay, at the end of that R&D timeline. We create enabling technology, and we create the technology before the standards are contemplated.
Before a requirement is set, we're years ahead improving cellular systems so that we have something to bring to the standardization process. That foundational technology creates IP strength. The standards timeline. We have tremendous experience, continuity in participating in the standardization effort. There's a skill set involved in participating in the standardization effort, and we're very good at it. When you bring technology to a standard, it shouldn't be merely theoretical. We're very good at building the technology before we bring it to the standard organization. We're very good at testing and simulating and proving the technology and proving its merit. When the standards organization, when the industry adopts Qualcomm technology, it adopts technology that the industry can have confidence has merit, including as compared to other alternatives. That creates IP strength in that technology and the IP associated with that technology.
Finally, there's the commercialization timeline, and this is very important. Qualcomm is unique in the industry. We understand the technology at a very, very deep level. Then we work with the entire ecosystem to enable the rollout, the implementation of the technology. We work with mobile device manufacturers, we work with infrastructure equipment providers, and we work with operators. We work with the entire ecosystem. We work across all geographies. There are very few companies in the world that can take this technology and work as broadly as we do to implement it, that also have the depth in research and development that we do. This also creates IP strength, because we bring to the industry technology and associated intellectual property that actually works and works very well.
Our IP is foundational, it's relevant, it's adopted due to its merit, and it's useful and therefore valuable. Let me talk a little bit about culture. I think Jim said more about culture than I can say and I think gave you a great overview of Qualcomm's culture. When you walk into the front door of Qualcomm's main building, and I know a number of you have been there, this is what you see. You see a two-story wall of patents, and these are actual Qualcomm patents. It reflects our culture. It's not just a culture of innovation, but it's a culture that prizes novelty, getting out ahead, creating technology that is groundbreaking. There's a tremendous focus on being an inventor and being an inventor that matters. When, in the history of our company, we hear the words, "That problem can't be solved.
You can't commercialize that technology. It will never be successful," it's like a call to action for Qualcomm. Okay. It's a rallying cry. It's just part of the culture of the organization. Everybody's familiar with the story relating to CDMA and Dr. Jacobs and the early engineers. It happened again. If you talk to Roberto Padovani, a former CTO, had Jim's job some years ago, he'll tell you that when he first came to an industry organization to propose high data rate capability for wideband mobile telecommunications, wireless telecommunications, that is bringing the internet to cellular, his response was not just, "You can't make it work." His response that he got from the industry experts in the room was, "We don't need to do that." Think about that for a second. Wireline is good enough. I've got voice, I've got text. Cellular is never going to enable internet.
Qualcomm just kept pushing forward, and we have a similar experience that we're dealing with today, maybe not quite as dramatic as the first two, but there's skepticism about millimeter wave. Great. Good. Right up our alley. It's exactly what we want to do. We want to solve these problems. What does it mean for the licensing business? If others want to clear the field and allow us to solve these problems with groundbreaking inventions, we're happy to do it. Creates strength in our patent portfolio, and that's a good thing. We build the portfolio to tailor it to a licensing business. That involves a couple of things. One of the things is that we have very broad geographic coverage. What you see filled in in blue is the coverage geographically of our patent portfolio.
There are countries out there that don't even have a patent system, so there are some countries that aren't actually covered with patent filings. We also have a very diverse portfolio. I don't need to get into the details of that because, again, when you listen to Jim talk about what we do, you can see how much innovation we put into technology aside from standard essential technology development. It's enormous. We're constantly harvesting and managing the portfolio for efficiency, coverage, and cost effectiveness. Let me focus in on 5G for a little bit. What is leadership in IP in 5G? One of the basic points I think that you should understand certainly is our view, we think it's technically true, is that leadership in IP relevant to 5G starts with leadership in IP relevant to 4G.
As Jim mentioned, we have the same air interface. There's some of the same features in 4G that carry over into 5G. There are dramatic improvements that occur in 5G, there's a lot of commonality in technology that carries over from 4G to 5G. When we look at our portfolio, we estimate that approximately 75% of our portfolio that's relevant to 4G is also relevant to 5G. There's a tremendous amount of depth there. On top of that, Qualcomm has been extremely prolific in driving new innovation into 5G features that will provide the performance benefits, the user experiences, the efficiencies, and the gains that operators want, very specific to 5G, very specific to Release 15, Release 16. There's a tremendous amount of innovation that Qualcomm has been driving that we think is foundational, as Jim mentioned, to 5G Release 15.
Let me go over this chart that Jim went over. I can't come close to describing the particular areas of technology and their usefulness, why they're important, the way Jim did. I want to emphasize from a licensing perspective, that this provides strength to the licensing business, because it's, again, the R&D efforts that are years ahead of standardization activity, enabling fundamental benefits for 5G. Massive throughput, spectral efficiency, network capacity, low latency, ultra-low latency, and tremendous reliability. Having IP associated with these fundamental pillars of 5G is very important. Again, it's very important to the ongoing strength of the licensing business. It's a continuous process. It's important that, for example, every data packet that's going to be transmitted in the 5G system has Qualcomm IP because of the channel coding that we put into it.
It's very important, for example, for the licensing business, that we solve problems to enable millimeter wave to be made mobile. It adds value to the patent portfolio. It continues to be valuable, and we see a long runway going forward. As you move from 4G to 5G Release 15, we're pushing into finalization of the specification for Release 16, the second release of 5G, and we're continuing this never-ending process that Qualcomm has been engaged in of driving innovation to drive improvements, in this case, dramatic improvements, into the cellular industry, and then in this case, in particular with Release 16, into other industries that are going to adopt cellular. Aside from enhancing and improving the mobile broadband experience, Release 16 is directed towards very specific improvements that will enable cellular in other industries.
Qualcomm has again, been extremely prolific in innovation in this space. It's not a recent activity on our part. In these fundamental innovations that enable Release 16 and will enable technology in the cellular space to move into other industries, we've been working for years to develop technical solutions and developing IP around these technical solutions. I could call Jim back up here, and he can go through each one of these for you in much more detail than I can, but I'm just going to touch on them very briefly. High-precision positioning. Extremely important in the automotive industry. The level of precision is mind-blowing. Okay? The specs are really tight. The technology is difficult to do. We've been working on this for years, and we brought to bear R&D and tested our technology to make this possible. Unlicensed spectrum, really important to operators.
Okay. Bringing cellular technology from licensed spectrum that operators own into unlicensed spectrum, and doing it in a way that doesn't disrupt the current technologies existing in the unlicensed spectrum space. We've been working on this for years. Okay. We're able to enable the industry to pick this up. Increases capacity for operators and creates opportunity for private networks to adopt cellular. There are a number of different things that we've done, that we've been working on for years, and we've been very prolific in pushing technology into the second release of 5G. Again, that is a strength of the patent portfolio going forward and fundamental to the licensing business. Increasing complexity is good for Qualcomm. The more complex, the better off we are because we solve complex problems. That's good for the licensing business as well.
The goal is to not simply bring technology to the industry and create a standard. The goal is to then drive the technology out into the world and actually build optimized mobile devices and optimized networks that actually deliver the performance that everybody expects, that users expect. In order to deliver that optimized performance, it requires implementation innovation. Jim talked about a couple of these things. You have to optimize these devices for power efficiency. They're mobile devices. They're battery operated. You have to have innovation around that to ensure that you're continuing to drive commercial success, and Qualcomm is very good at doing that. Again, we don't just bring technology to the industry and ask the industry to adopt it. We build devices. We test them. We conduct simulations. These are proven.
Then when the technology makes its way into the standard and when the technology makes its way out into the market, we continuously work on optimizing the technology to ensure peak performance. That's good for the licensing business because it creates more depth and breadth in our portfolio. We have valuable implementation patents, and that process is a continuing process. On this side of the talk I'm going to give, this sums it up. Decades of R&D and standards leadership. Fundamental foundational systems-level innovations and IP across all of these generations of standards. Continuing very robustly, very active innovation in driving 5G, Release 15, Release 16, and moving forward, and continuing very actively to drive innovation and implementation, which is good for the portfolio, good for the licensing business. Let me talk about a timeline.
For the last few years, maybe the last four years or so, there's headline news about this timeline, and the headline news has always been the challenges that we've faced. I'm going to step down and step back from that headline, which you're all very familiar with, and I'm going to talk about the story. I want to at least remind you of the story underneath the headline. There are two critical stories about execution, continued execution. One is executing on innovation, continuing innovation, and you heard that from Jim. Never losing sight of the fact that we're a technology company, and that our goal is to develop enabling technology, and we continued to do it, okay? Never lost sight of that fact, and we never lost the execution edge. Then continued execution and licensing. We actually continued to get to agreements.
When you look back towards the front end of this timeline, coming out of a China resolution, we've signed up over 150 Chinese patent license agreements. After we announced, or after we disclosed our rates for 5G, we've now signed up in the last two years, I actually think the disclosure occurred two years ago today, over 75 5G agreements. Third, before, during, and after the FTC trial, we've signed what we refer to as anchor agreements, important agreements. If you look at what we've managed to do under this timeline, under the headline news, if you look at what we've managed to do in this timeframe, and you compare our execution in signing up 5G agreements to the same timeframe when we first disclosed rates for 4G, the score is 75 to seven.
That's how many we had signed up by the end of that two-year timeframe in 4G. We've continued to execute. Key takeaways. Unmatched patent portfolio value. Continuous R&D and standards leadership. System-level innovations and foundational IP and critically fundamental innovations and inventions enabling 5G. Continuing strong execution. 75+ agreements to date and multi-year anchor agreements recently signed. Thanks very much.
Qualcomm and ASUS have a long history of strategic partnership that is built on the shared passion and persistence for creating the world's most incredible technology. Over the years, this partnership has produced a wide range of impressive devices, including tablets, wearables, always-connected PCs, flagship smartphones, and most recently, the world's most powerful and authentic gaming smartphone, the ROG Phone II. 5G will bring both evolutionary and revolutionary changes to how we interact with our devices, platforms, services, and environment. Our collaboration with Qualcomm on 5G gives us access to the most advanced, reliable, and competitive connectivity.
We will now take a short break and be back in just a few minutes. Thank you.
[Break]
We need an approach to 5G that will fuel invention and transform our industries and our communities for decades to come. We are working with Qualcomm to bring 5G capabilities to more devices and create new experiences for consumers.
Partners like Qualcomm is really key to us on this journey. They help us to solve complex problems and bring amazing new experiences to our users. By working together, Lenovo, as the number 1 PC manufacturer in the world, and Qualcomm, as a leader in computing and connectivity, together, we're able to meet the needs quicker than by trying to do it on our own.
Now we're working with partners on a new class of incredible devices built on the Snapdragon 8cx. All of these devices will be able to take advantage of Windows 10 Enterprise security, privacy, and ease of management. With Qualcomm and the Snapdragon 8cx platform, we can really move the industry forward.
5G absolutely promises to transform the way that we live, work, play, and learn. It can impact every industry, every vertical, our global economy, and frankly, our global society. You know, a year ago, we would have been talking about 5G in the future. Now we're able to use the present tense.
Thank you all of our partners for the hard work and for the belief that we can bring innovation and excitement again to the mobile industry. Congratulations.
Ladies and gentlemen, please welcome Qualcomm President Cristiano Amon.
Good morning, everyone. Just ask you to be patient with me. I kind of lost my voice going party after the earnings call. Actually, even though I wanted to go party, we're busy working on the RF front end. I'm just recovering from a cold. I'm going to try to do my best. Hopefully, you can hear me okay. Have a good set of information to tell you today. We're excited to have this analyst day we haven't had in a long time, hopefully will give you an idea how we're thinking about the company in the future. Everything that we have been busy and focused on the past two years, putting together and preparing, and looking into this transformation, not only of the industry with 5G, but really a transformation of Qualcomm.
When we look of the opportunities ahead, it's really one of the great set of opportunities we had in the history of our company, as Steve outlined. I will break the presentation today in three parts. How are we executing in our core business and what are the opportunities coming to us from 5G and the smartphone transition. How we're building on our adjacent platform and business, which is part of the company growth and diversification of our revenues. Then we're going to give you an overview also on some of the long-term bets we're investing right now to think also what we're going to be doing beyond 5G. I will start the first part of the presentation about 5G. If anything, I ask you to be patient with me because there's so many things I want you to understand about this transition.
You hear a lot about 5G in general and how broad it is, how fast it is, and then what it takes to be a leader in 5G. I hope my job will be probably walk you through each one of those steps. First, I'm going to start with 2019, this year, that is ending right now, just to show a comparison. I would really like to show that comparison because I remember in the 4G transition, we had four operators that launched. It was the three operators in Korea and one operator in the U.S. with Verizon Wireless, three OEMs, LG, Samsung, and HTC. At the time, granted, it was right before the smartphone transition, I had everybody on their BlackBerry sending messages saying, "Who needs 100 MB per second?
There's really no use case for this. We look at 5G, the reality is completely different. As of right now, there's 40 operators that launch. There's over 40 OEMs. Those operators that indicated that they're building 5G have about 2.4 billion subscribers that they currently serve. The potential for faster transition is much higher. I also want to point out a couple of things that you can see in this map. Geographies such as China and Europe, unlike the 4G transition, they're now coming online at the same time. That will change also the dynamics and the scale of the transition. If you look at the mobile ecosystem today, China has a big component of the mobile ecosystem. That's going to accelerate the transition as well.
The point is when we look at the scales, also to highlight how 5G is going to be deployed. You look in 2019, you have millimeter wave in the U.S. By 2020, a couple of other things happen. More markets and operators come with 5G, but also millimeter wave happens in markets such as South Korea and Japan. With that, we're going to start to see the real deployment of how 5G was designed to be implemented. If we keep going, by 2021, we'll see millimeter wave now coming to a number of other markets. I'm going to highlight Italy and Germany, Hong Kong, Singapore. As of today, we're working with China Mobile on 100 verticals for industrial applications, many of which with millimeter wave. Our expectation is by 2021, all the developing economies will have millimeter wave as well.
That's how we think about the expansion of 5G around the globe. It's fair to expect by 2020, you're going to have all metropolitan areas with 5G coverage at least with the mid-band or sub-6 as well as refarmed bands. That's going to be a part of the conversation today. 2021, you'll see millimeter wave expanding beyond Japan, Korea, and United States markets. I also remind you, I like to go back in history. I actually had the opportunity to work in Drive Test on the Japanese PHS system, the cordless telephone system. Japan, for example, the base stations for PHS were 100 meters apart with fiber backbone. That is perfect. It's like Lego build-out for millimeter wave. Expect to see the rollout of millimeter wave to actually be quite successful in dense markets such as Japan and Korea.
With that, I want to tell you how we think 5G networks are going to be built. There is a lot of confusion in the industry today when there's reference to 5G sub-6 or millimeter wave. This is not like CDMA and WCDMA. Those are more like different choices that operators are making. 5G has been designed so you have a combination of all of those spectrum that you have available for you. The way you should think about it is you're going to have those new spectrum that are being allocated, the mid-band in the case of 3.5, that's going to get deployed on top of existing microsites. That's what we call the sub-6. Then the high-band millimeter wave gets deployed as well. Then the existing 4G spectrum gets refarmed to 5G with a feature called Dynamic Spectrum Sharing.
Dynamic spectrum sharing is going to be a feature that we're going to hear a lot as we go to 2020, as you allow the refarming. Unlike the densification of the network in 4G. They happen for capacity. Millimeter wave, which does require new sites or a more dense network, is really for not only capacity, but performance. Some of the new use cases will require Millimeter wave. Let me just walk into some metrics of how the networks get deployed, and then I want to walk into the phase approach. If you look how easy it is to deploy some of those technologies, and I want to use this opportunity to challenge the false premise that Millimeter wave is really difficult to deploy.
If you look at a city like San Francisco, existing cell sites, because remember, many of the dense areas have been through a process, a 4G capacity build, and that creates more sites. If you look at existing sites in San Francisco, you can basically get millimeter wave coverage with about 65% reliability, which is pretty good, especially as being a technology that's designed for the dense areas. When you apply, for example, in the case of Germany, Frankfurt, existing sites, you apply the new band of sub-6, you have 78%. Once you start refarming the spectrum with dynamic spectrum sharing, you get to 96% coverage reliability.
The way you should be thinking about this is every operator that is deploying 5G will make use of all of their spectral assets to refarming with DSS, plus the new bands, and deploy millimeter wave when that spectrum becomes available. We're really excited about the ability to do this. When you look at millimeter wave in itself, the performance is really what makes millimeter wave shine. There's some stats now. We have Signals Research, its third-party assessment of millimeter wave, and I wanted to have that information here because a couple things that were said about millimeter wave, that you don't have the ability to have an all-day battery life. 14 hours of usage of millimeter wave meeting power and thermal requirements, 10x the peak performance of LTE, and then you get no line of sights with 200+ Mb per second.
It's a significant improvement in technology and the enhanced mobile broadband, but in the bandwidth and the latency will really allow new use cases. Certain use cases in the smartphone will not be possible without millimeter wave. I have some examples for you in terms of compelling use cases. One thing that we've been saying since the beginning of this 5G transition, in the same way that 4G changed the music industry. I'm sure nobody here, if there is, will be an exception, take CDs to their car today. People just stream music. No matter where you are, you have enough 4G coverage and performance to stream music. That's going to be the issue with video. Not only with the ability to consume video streaming in high resolution, but the ability to upload video. It's going to change video creation.
It's not far-fetched for you to imagine that today YouTubers will be broadcasters. It's a big change for the broadcaster industry, especially when you think about sporting events. High-density venue, how you're going to consume media and how you're going to generate media, you need millimeter wave. You need that for bandwidth. The other use case that I like to display, and I think all thanks to Epic Games' Fortnite, the industry understood that mobile gaming is a much larger addressable market than console. The advent of cloud-based gaming, millimeter wave will allow you to actually, and especially form factor changes with bigger screen devices, like foldable devices.
We're starting to see, in the case of Republic of Gamers, in the Tencent device that we work with, Tencent in China, special devices for gaming, millimeter wave will enable for you to play mainstream console gamings into a mobile device. The transformation of the enterprise is a big one. It's going to be a big topic of this conversation, especially as we think about new business. Millimeter wave in an enterprise alongside Wi-Fi 6 indoor build-out not only will enable a lot of the industrial cases, and you can only do it, manufacturing robot, for example, you cannot do it with sub-6 in terms of amount of data and the latency and the mission-critical requirements. Also the ability to remove data from hard drives and move it to the cloud.
Millimeter wave bandwidth and latency allow you to actually move storage to the cloud, really feeding into the strategy of many of the big cloud companies, like Microsoft Azure, as you think about the potential of OneDrive and other services going forward. Millimeter wave, an essential part of 5G. It is happening, and it's going to happen across multiple geographies, and it's going to be fundamental to drive new services. With that, I want to go over build-outs. It won't take a long time to do, but I have two purposes in walking you to this build-out of 5G networks. One is to show how 5G networks are being built right now in a phase approach. The other one is to also raise awareness about what competitive leadership means. A lot of people say, "I have a 5G chip." I've seen that in the 4G era.
When we started with 4G, there's a lot of people have chips, for some reason, they never get to commercialize and scale because the feature set keeps changing, and keeps changing very fast. This is where we really focus on when we think about the build-out of networks. You have an LTE network today, and you're starting building in 2019, 5G. You have this alphabet soup of NSA, SA, FDD, TDD. You start building the higher bands of 5G using the existing 4G cores in a non-standalone mode. The next step, as you start to bring some of the mid-bands and you basically have a lower 5G bands, also the lower bands with FDD, and you use a technology called DSS to refarm. It really works like magic because both systems use OFDMA modulation.
You can actually deploy an existing site, a 5G base station or eNodeB, and the devices will coexist, operators don't have to clear the spectrum. They can start building 5G on existing 4G spectrum, and those systems will coexist with DSS. The next step, you upgrade the core of the network into an SA, and then you start to bring capabilities like carrier aggregation, which basically make use of all of your spectrum assets that you have in 4G plus the 5G, so you can have wider channels as well as extend coverage. Successful roll-outs of 5G technology, you need all of the above. Especially when you think about the chipset and device development. The device, as we build the device ecosystem ahead of the network, you need to have a chipset capability that has all of those features.
They're going to be required, and especially they'll be required in 2020 as you need to bring the refarm spectrum with DSS, and you need to bring the carrier aggregation so you have the ability to use existing spectrum. That's how the networks are going to be built. We've been through the 2019 phase. 2020, you're going to hear a lot about coverage build-out using existing spectrum, hear a lot about DSS, and that's why we're excited about the speed of the 5G ramp and our position in this space. With that, I will shift the conversation to where we are in the market. The way I look at this is the 5G acceleration was a very important milestone for Qualcomm. Unlike 3G or 4G, we got the devices ahead of the network.
As devices are already ahead of the network, the best 4G phone you can buy today, even if you don't have service, is a 5G phone. Devices move on, especially on the Android ecosystem. We are being consistent and conservative in our sizing of the market. I think we said in the last earnings call, because we've been calling the market down as we get to the tail end of 4G, we're assuming that some of the drivers in the market remain constant. However, when you analyze that, I think that builds a lot of confidence in the 5G rollout because replacement rates are low today. You have 3+ year-old 4G phones. When we did the 3G to 4G migration, phones in some markets were just two years old. You have about three, in some case, some market, four, even five. Unlimited data plans.
As soon as an operator launches 5G, unlimited data plans goes into 5G. That's the lower cost per bit to deliver data, works in the operator economics. The device already switched. The 5G devices in the high-end premium tier, those are the ones range in all of the operators, especially in all developed economies. We do have new experiences of service that are coming with 5G. Consumers also responded. Korea has been an enormous success, I think exceeded all expectations. In China, with their projection of 1 million eNodeB base stations by the end of 2020, the 10 million subscribers are now 15 on the pre-launch. The projection to go in China alone, 150 million subscribers by 2020. Those give us indication that we're going to have a successful rollout of 5G. I want to point to this slide.
I like to point to the Samsung roadmap, especially on the Android ecosystem, is the number 1 player. You can see that the device ecosystem already moved, regardless of the time that it takes to build coverage. You can see not only the Galaxy Note10 5G, you see the A Series, the below that, the A90, as well as the Fold, which is their flagship product. All of those already moved to 5G. That's the consequence of having the devices ahead of the network and a benefit the operators are going to have in terms of build coverage that they can rely on features such as DSS. As we get to the end of this first session, one important topic to highlight is the changes coming with 5G. Every generation of wireless, we saw changes.
Changes in form factor devices, kind of early to predict, we do see signs of changes in the form factor as screens get a little bigger and video and gaming become mainstream for consumers. We have completely changes on how we think, for example, in the case of 4G, how we think about computing and mobile computing. The way to think about the changes coming with 5G is, a simple way to describe 5G for a mobile device is 5G is going to connect you with the hyperscaler cloud companies 100% of the time in a reliable manner. That will change the capabilities of the applications. The capability of the applications are going to be less bound by the OS and the storage and the processing power you have on your device.
If you look at the China domestic mobile market and you look at the super apps like WeChat, we see the potential for that to happen in 5G, as you're going to be always connected to the cloud, you have unlimited storage and on-demand computing. What we're going to see in the 5G era will be a consolidation of the applications given by the power and scale of the hyperscaler cloud. Less dependence on the OS. The concentration that we have on the OS with the app stores and all of the long tail of apps, that will remain, but we're really going to see the separation between the more and more of the casual apps and the apps that are going to drive the user experience, as basically the cloud and your device are 100% connected in a reliable manner.
That creates a lot of opportunities for innovation and for things to change. I'll point you to one detail. We've celebrated pretty hard at Qualcomm, even though we got unnoticed by the market. In the 5G transition, unlike 4G, Android operational systems now give developers the metric. Before, in the 4G era, is whether you have or you don't have an internet connection. It doesn't even, in some cases, doesn't even make a distinction between cellular and Wi-Fi. Now an application developer know what your bandwidth and what's your latency that you have. That allows you for you to change behavior of applications based on the bearer, creating value for the connectivity, but also improving performance. We're just at the beginning of the curve and we're excited about the opportunities that we're going to have for transformation even in the mobile space.
I think it's fair to say, if you just look at what happened to every generation of wireless, one thing that we know is you can never take anything for granted in mobile. It always changes. We're happy to be a constant in each one of those transitions as OEM change position in the market, and that's because we invest early in leadership. The last slide on this section before we go into execution of our mobile business is for us, it's not about building a chip, and we say we have a 5G chip. It's really about building an ecosystem. That's the approach that we took within 5G. You saw Qualcomm in every single pilot and trial about 5G with every operator and every infrastructure vendor. The platform of choice was Qualcomm.
Every single development of infrastructure in the labs, how the infrastructure gets developed, the other side of the radio is Qualcomm. We have basically worked across all geographies with all industries. We have been the platform of choice for this technology to be developed, and that what's give us all around leadership in the 5G transition. Starting with all the development platforms across modem and front end to building chips and to building devices and working with service providers. We're very proud of what we're able to accomplish through a lot of hard work in 2017 and 2018 to accelerate this technology. The launch in 2019 has been a busy year, bringing this to market and now ready to really achieve scale of this ecosystem in 2020.
With that, I want to tell you how we're thinking about execution on the core with the opportunity we have with 5G. The first thing is a big expansion of our SAM, as in addition to what we're doing in the core chipset, we add our front end. We see the ability to grow our SAM in mobile. We've been looking for growth in mobile now for a while, and we're very excited with this 5G transition. Basically between 2019 to 2022, we can have compounded annual growth rate of a double-digit in mobile. The execution to drive that, it's going very well. We have talked about the number of device designs that we have. 230 is not a small number. 230, I will repeat, of the 230 designs that we have, no exceptions for the 5G content, not the legacy 4G.
For the 5G content, it's Qualcomm RF front end-to-end. We look at our designs that are broken into three categories. 5G smartphones, driving not only our existing customers, the China ecosystem, into the flagship devices in the high tier. Driving 5G into other industries. With embedded modules, they're going into industrial, going to automotive, going to computing. Driving a lot of the designs for mobile fixed wireless access on the carrier case for mobile broadband and some other designs that go into new applications, which I will single that out in the next slide. We're very proud of that. The reason I want to bring this separate slide to you is, there are a lot of discussion about 5G. It's beyond phones. Phone's going to drive a lot of the volume and the growth in the company in '20s.
The easiest thing is to go upgrade the smartphones. When you think about 5G beyond phones, we became the platform of choice for all those companies. Many of those companies and services, you need a reliable solution and you need a global solution. As a consequence, we have over 30 technology suppliers that are designed specifically for the other industries, industrial IoT, automotive, enterprise, fixed wireless, and compute, as well as mixed reality devices. All those promises of 5G going into other industries. I mentioned the project that we have, for example, in China, with 100 different verticals experimenting as 5G use cases. The good thing about being early and you have a mature platform with generations ahead of our competition, that we became the platform of choice for those use cases. That will create opportunity for the business beyond mobile.
As we look to scale 5G, we've been building a comprehensive roadmap. We started in the first half of 2019 with our first Snapdragon X50 modem. We are right now working on the second generation X55 Snapdragon modem. You're going to see a lot of details in the coming weeks, as it's going to be driving a lot of the new set of devices throughout the end of the year and the beginning of 2020. In 2020, you're going to see 5G across our entire portfolio of products. Not only the 8 series, the 7 series, the 6 series integrated SoC and beyond, you'll see us driving as fast as possible the entire device ecosystem to 5G. There is, as you heard from Jim Thompson, there's no question that the benchmark for mobile SoCs is Snapdragon.
People ask me sometimes, and I think the question came up in the conversation yesterday. We have been leading in a number of technologies, and I won't list them all, but Jim highlighted a few like camera and AI and graphics. When people ask us how well are we differentiated in those areas and what leadership means. When you apply a mobile SoC into a high-definition head-mounted display for virtual reality devices, the technology needs to perform because when you're fully immersed in VR, motion to photon matters or you get nauseated. The ability for you to exercise your GPU really matters. That's why Snapdragon has been the only choice that you've seen in head-mounted display, whether it's Microsoft HoloLens or is the Facebook Oculus Quest. It's just because we have the leading performance.
We expect that to continue, especially for the capabilities that will drive 5G use cases. We'll continue to be pushing AI as a key feature on a Snapdragon AI engine. Gaming, the promise of mobile gaming and as use case is very high, as we can really expand the addressable market of gaming with cloud-based gaming, as well as having the lead camera. As we talk about Snapdragon in digital and we talk about execution in mobile, it leads us to the conversation on front end. Our strategy is working. As Jim outlined, we took the necessary steps to make sure we'll have every single piece of technology, and we have the ability at the component level to have equal or better performance than the competition, but also innovate substantially at the system level. It's been a multi-year investment for the company.
We completed the acquisition of very important asset that was filters from EPCOS TDK. We had invested in this business to increase performance so we could actually exceed the benchmark of FBAR in terms of filter performance. We have had success in 5G. What I'll show you is not all of them, but third-party report. I think IHS teardown came with devices that they tear down, and you can see kind of where we are. It's in addition to modem and transceiver. Some of those devices you can see power amplifier, our GaAs PAs in those devices. Diversity modules, envelope trackers, filters, antenna tuner, and in the case of millimeter wave, our antenna module. It's a comprehensive modem to antenna. It touch every single part. My favorite slide. It's the next slide.
The Galaxy Fold is being launched by Samsung sub-6 globally with Qualcomm in all geographies, not yet with millimeter wave. Sub-6, you don't require a lot of advanced technology. It's not as complex as millimeter wave. They chose the Qualcomm front-end components for the Fold sub-6. You can see a lot of the components that we have. We have 5G PA module, we have the DRX diversity modules, envelope trackers. We also have the 4G diversity modules, antenna tuners, aperture tuners, filters, duplexer, and extractors. That's an example of what our execution has been. We're excited about that.
Akash will give you a little bit more color, we feel that we took the right approach of make sure we had a highly differentiated solution across how we think about our test platforms, how our solution is tested and certified, how it performs better than the competition at the individual component level, and we have system-level features. In the 5G era, it's not enough to basically be thinking of the modem digital only. Successful companies in the 5G era will be addressing modem to antenna because the complexity in RF is very high and it's a Qualcomm-sized problem to solve. With that, I will shift to the next part of our presentation, talking about how we're executing beyond mobile.
I want to remind you of what our strategy has been, and that's the strategy that we put in place for about two years as we chose how to diversify our business and how to generate growth beyond mobile. The easiest way to describe this strategy is to look at this diagram, and RF front end is a great example of that. We have a strong channel. Our channel in mobile, our Snapdragon channel, is a very strong channel. It's a channel that spans not only from the Snapdragon relationship with the OEMs, but across the entire ecosystem, from carriers, from application developers, and from standards. In that channel, we continue to push the technology boundaries, and we invest in a lot of new technologies as we continue to expand content.
That's why you saw our expansion towards RF front end, and you saw our expansion into new areas, for example, as biometrics. On the right side, we look at industries that are being disrupted by mobile. The scale of mobile is enormous, and those industries get disrupted by mobile, automotive is a great example. We can go create that channel and tailor-made our products in a highly leveraged R&D way so we can execute and generate growth in our industries. With that, we've been expanding our SAM and well-positioned to address large opportunities for the company with incremental investments, basically building scale on our large R&D that drives both our licensing business and our product business. I have one diagram that hopefully explains where we're heading and how to position the different investments in business that we have set up in the company now.
If everything starts with the phone and with the Snapdragon SoC, as highlighted by Jim Thompson presentation, that will drive it in case of mobile, our 5G leadership in the 5G transition, that will hope to grow and generate a lot of earnings for the company, in the mobile market, which is our core business. As we push new technologies into that channel, going from the left to right at the bottom of this diagram, you see how we continue to add content from what we successfully did with Wi-Fi, Bluetooth, going to the RF front end, biometrics, and integrated modules.
Going up from the mobile platform, we apply that to automotive, compute, and IoT business, and we're making some new bets leveraging the technology for the evolution of the data center, the edge data center, which has new parameters of total cost of ownership and performance per watt, as well as evolving our automotive business to autonomy. That's how we structure ourselves, how we are leveraging R&D and making investments that allow many of those bets to grow accretive to QCT and diversify our business while maintaining leadership in mobile. Building on each one of those, I will start with the automotive as we look at the adjacencies. Automotive is one that we're very proud, has been a success story for Qualcomm. In addition to add that SAM to our semiconductor business, we see that we have a roadmap to get there.
Start with the consolidation of multiple ECUs in the car with a high-value SoC. We started with our telematics, and we have an opportunity on the telematics to add also connectivity on Wi-Fi and Bluetooth, but that's evolving to Cellular V2X. Cellular V2X is making good progress of being standardized. Recently in Europe, is now in a good position with the DSRC. In China, it's been standardized and been deployed. That changes safety on cars and create new opportunities to continue to evolve in the telematics. The digital cockpit transformation with infotainment was also a very good opportunity for Qualcomm, as eliminate not only multiple discrete systems into the car, but allow modern cars to deal with the fact that consumers are driving, looking at their phones. You have a much changed experience as you rethink about the digital cockpit in the era of the connected car.
We've been quite successful there with Snapdragon. From there, we see opportunities to go to different places. As we develop safety requirements, especially as we do mission-critical dashboard, we're building our ADAS platform and also cloud device management, building on the automotive. We're in the automotive business to stay. It has been a successful story for Qualcomm. This one in particular, we're going to give you a little bit more metrics for you to see how we're doing and how this business is going to evolve over time. A scorecard is really all of the designs that we have received for a platform in automotive. We're the number one in telematics. We're going to be now, in 2020, the number one in premium next-generation infotainment. 18 automakers working with our Qualcomm platform, and our pipeline is now $6.5 billion.
We had increased the pipeline of design wins within our automotive segment. The other segment is compute. I know there's a lot of skepticism because it's been one that has not yet happened before, competition in the PC space. We feel very good about where we are right now, which was about the ability to create not a second-class Windows PC. The evolution of mobile and the evolution of computing and productivity, which has been changing because of mobile devices, create a real opportunity for us to grow into this market. Anything for Qualcomm is growth, since we don't participate in the market today. We have done successfully a partnership with Microsoft.
The point that I want to make, and it's just not unique to Microsoft, but if you believe in the digital transformation of IT with the enterprise cloud, if you look many of the drivers of growth in Azure, you will see a connected PC on the other side. As you move your entire enterprise data to the cloud, as you move everything to Office 365, as you manage PCs like you manage mobile devices with Intune, as you move your hard drive to the cloud, you need a connected computing experience. The next transformation of the enterprise with the hyperscaler cloud is going to create an opportunity for more value of a connected PC experience, and that creates an opportunity for Qualcomm. I would like to summarize that with our partnership with Microsoft. As you can see in this picture is the new Microsoft Surface X.
There's no backup plan in this one. It's just a Qualcomm Surface. There is no x86 version. We have a custom Snapdragon that we work with Microsoft, and it's branded Microsoft SQ1. Microsoft has taken responsibility for the whole experience from the application compatibility, the device, all the way to the chip. The PC industry has turned this page into the connected PC. We feel that the ability to create a thin-like category, we now match performance with the Core i5, and we exceed in some of the applications that make use of GPU and the other capabilities, especially AI, that we have into our processor. We are working to basically embed connectivity, make it seamless for connectivity to be activated in those devices and align ourselves with the message of digital transformation of the modern enterprise IT.
Our scorecard here is the number of designs that we secured to date. You can see now we had multiple commercial launches. We have dedicated enterprise-ready platforms been built, one by Samsung, one by Microsoft. There's going to be more coming. It's already switching to 5G with the Lenovo-announced 5G PC. This one, it's a good opportunity for Qualcomm. We can grow another Samsung type volume, in this opportunity with pre-material devices, very accretive to our earnings. We're optimistic about the PC transition this time. Now, as we get to the last part of our conversation of adjacents, I want you to bring you to the IoT. IoT is really about eight segments for Qualcomm. It is a good business.
The way you need to think about this business is a business that we have been leveraging our technology and products into many other industries, be able not only to diversify QCT, but generate growth. It is growing at double digits, and it's very accretive to QCT margins as well. It's another expansion of SAM for the company, leveraging our technology. I will highlight just a few things. As the first four segments, it's connectivity, small cell, fixed wireless, and industrial IoT. Couple interesting data points, 11,000 customers. During the past two years, as you know, we did not complete the acquisition of NXP, but the company move on and been actively working how to create a channel so we can repurpose our technology into many different segments. We're very happy to tell you today that we now have 11,000 customers.
For a company that in highly concentrated in mobile space, which is highly concentrated, we have tens of customers with a B2B engagement. Now, over the past two years, now we do business directly with 11,000 customers, which is a big gain in scale in an efficient manner that allow us actually to continue building this business. A lot of the earnings power of the company, in the coming years, will come from this 5G transition. This business, as we continue to grow this business, highly leverage on R&D. After the 5G transition is completed, will be a significant business for the company together with some of the long-term bets. We're very pleased with what we accomplished in this category that we call IoT today. A few highlights on connectivity. Wi-Fi 6 is a big transition for Qualcomm.
We have the first certified Wi-Fi 6 products in both the access point and the mobile devices. We're very happy with our position in retail and in enterprise. We're starting to see Wi-Fi 6 meshes coming to the market. Fixed wireless 5G CPE is a new area. 5G creates an opportunity for many operators. We announced our high-power CPE for range in rural areas. They create opportunities for operators to augment their business case. Small cell, it's something that we see a lot of traction right now, and I want to highlight this. This has a lot of strategic value. Base stations, they don't have as much units as you have in the phones market. They're very important, especially on the strategic side for modem leadership in the creation of some of the demand for devices. Two areas worth mention.
One is a lot of the private 5G industrial networks and enterprise networks. Countries like Germany pioneer with the allocation of dedicated spectrum for private 5G build-outs, is being followed by other geographies. We see our small cells being selected to build private networks in partnership with existing companies. We also design our small cells in all of the big infrastructure vendors, with the exception of one. There is this transformation, or transition is a better word, of the industry to flexible RAN architectures and Open RAN. The very first Open RAN public networks being built in the industry is Rakuten in Japan, has been built with the Qualcomm small cells for 5G. In industrial IoT, we see that an opportunity, very defensible business, global business across industrial handhelds, retail transformation, robotics, and surveillance camera. Those are the four segments.
The other four segments of this IoT is what are we doing in wearables across watches and trackers, voice and music. Think about the technology. We can provide the technology on the Android ecosystem that can compete with the Apple AirPods, is our high-fidelity, high-performance earbuds, and we start to see some OEMs now designing together with their smartphone. We've been very happy with the R&D that we did in this area. We have consumer IoT and mixed reality devices. Beyond that, we are making some long-term bets for the company. I want to highlight two. First, it continues to expand our SAM. You see the opportunity here for Cloud AI and for autonomy. I want to start with the Cloud AI.
The way for you to think about the cloud today, you have on-device AI, I think Jim Thompson, in his presentation, talked a lot about that, and you have the cloud. 5G not only brings those two together but also moves the cloud closer. If you look at companies that have been well-positioned, I think, for example, YouTube with all their CDNs for video, it put them in a very good position for Google, to provide gaming streaming. You're also going to see, as a lot of the IT infrastructure goes to the cloud, you're going to see the build-out of the edge clouds, and it is going to be a necessity for industrial and some of the consumer applications.
As we look at the opportunity of expansion of the SAM within the data center, we've been working to develop this product, which we announced, and is the Cloud AI 100. It's really addressing those industries who want optimized design, high performance, and low power for the automotive industry, the data center, the 5G edge, and 5G infrastructure. We see now four type of applications. Application in infrastructure, as you look to virtualize some of the RAN in some of the controllers. You see this as a very good opportunity to replace CPUs and FPGAs. We see opportunities within the data center for inference processing. We see opportunities for highly efficient computation in the automotive sector, especially for autonomy. We see this for 5G edge with mobile operators and cloud providers.
That's the card that we're now shipping to many of our customers, that they're evaluating our data center product. It's built on seven nanometer. It's designed really to handle inferencing workloads. We feel we're very well-positioned, especially as you think about the equation of performance per watt. Versus the existing incumbents, it's over 2x the performance and 2x the performance per watt, especially as you think of total cost of ownership. You can see one of the benchmarks of our PCI card versus competitive PCI cards. We're very excited about that. That's a long-term bet for the company, but it shows the power of the technology and the IP that we create for mobile, how can that actually scale to many industries and allow the company to generate growth in a highly leveraged manner.
The other bet I want to share with you today is the autonomy. When you think autonomy, the larger portion of the market is really highway autopilot. Thinking about going level 3, 3+ , level 4. We have a basket of technologies, going from multiple processing elements in sensors, plus the ability to have an SoC that will meet safety standards, which is in part of our development as we build our infotainment business. We're making a lot of progress. The initial feedback has been great. Unfortunately, there's not so much I can announce this, but I highly encourage you to watch what we're going to do at CES. I think you're going to see substantial progress in how we had turned our autonomy research into products and how that's going to turn into design. That's another new bet for the company.
We're very excited about it. As we get to the end of my presentation, we talk a lot about 5G being transformative for many industries. It goes beyond phones. As it transforms many of those industries, it really creates a very big tailwind for many of those Qualcomm adjacent and growth bets. It feeds in each one of those, creating new use cases across automotive, across compute, across our IoT segments, and some of the long-term bets, basically making what we do more valuable, more defensible, and accelerate the opportunity for growth. We also have more news to come. In the first week of December, in our tech summit, we're going to announce the new family of Snapdragon products. Our next flagship with a 5G modem will be announced, and you're going to see those in devices coming in 2020.
In CES, you'll hear a lot about what we've been doing in ADAS. Really excited about this opportunity we have ahead for the company. It's one of the best opportunities we had in the history of Qualcomm. What I want you to take away, when you think about this analyst day and what we're doing with our product business, is we now hedge our bets on 5G. Doesn't matter who wins. If Apple wins, if Samsung wins, if a China ecosystem wins, we're well-positioned. We have a global customer footprint for 5G with multi-year agreements across multiple customers. We feel very confident about our position in 5G. 5G is accelerating across Sub-6, DSS, which is going to push the boundaries of performance and capabilities on modems, and as well as millimeter wave. Modem to antenna is real.
Performance is significant when you design as a system, and the designs that you see on the front end, I understand that that's an area that many of you will face with skepticism, but we're very happy we accomplished today. We actually have conservative targets, but we feel very good about the potential, and we're going to continue to make progress in RF technology, designing modem and RF as one in the 5G era. We have become the platform of choice for other industries as 5G go beyond phones, and we're well-positioned to expand and diversify and have a much more diversified company after the 5G transition. It's really a significant opportunity ahead. Thank you for your time. Thank you for being here with us today. Thank you very much.
We'd like to invite everybody to please join us for lunch downstairs. We will resume immediately following lunch. Thank you.
[Break]
China Mobile has maintained a very good partnership with Qualcomm, which result in significant achievement in driving 5G development. China Mobile looks forward to working closely with Qualcomm to boost the device innovation. We hope to strengthen joint efforts on technology and standard to ensure the continuous optimization and evolution of 5G. Moreover, we will extend the collaboration onto the study of customer demand, user experience, and the new applications to build the new 5G ecosystem.
For Bosch, 5G is a key technology for the digital transformation of our business. Bosch is a big company. We are active in many different vertical industries, automotive, manufacturing, agriculture, building automation, and so on. In all these domains, 5G may have the potential to transform and even to disrupt industries. Now we are very happy and excited about the collaboration with Qualcomm because it combines the best of the two worlds. On the one hand, Qualcomm is one of the leading 5G technology providers, and on the other hand, Bosch is an industrial giant. In order to unlock the potential of 5G in different vertical industries, we need this very close collaboration because one player alone cannot really do that.
For the first time ever in a long, long time, we're very, very excited about what 5G can really bring to society. I think at Rakuten, our approach is not to only focus about what 5G means, but also fundamentally transforming the underlying network architecture. Probably one of the most exciting projects that we are undertaking is the work that we are doing today with Qualcomm, which is fundamental creation of infrastructure. Not a lot of people are really aware that Rakuten new 5G base stations are actually being created in collaboration with Qualcomm. This is the world's first Open RAN architecture for 5G. That is incredible. Never happened before. We think millimeter wave is a vital part of the ecosystem, and that's largely why our relationship with Qualcomm is so impactful.
I think there is an unbelievable wealth of opportunities that millimeter wave could bring for enterprises and industries.
Ladies and gentlemen, please welcome Executive Vice President and Chief Financial Officer, Akash Palkhiwala.
Good afternoon, everyone, thanks for coming to our Analyst Day. Thanks for coming back after lunch. I was a little worried that after the really good presentations in the morning, no one was going to come back for the financial discussion. It's good to see you all back here and interested in the financial information. It gives me job security. I'm very excited to be here in my new role, representing Qualcomm, and had an opportunity to meet several of you yesterday and looking forward to meeting more as opportunities present themselves. We started working on 5G a long time ago, and it's great to now see networks being launched, devices being deployed, and what that brings to us in terms of growth. You've heard a lot about that from the speakers before me.
What I'm going to focus on is how do you put a financial framework around the data that was shared previously. That will be the focus of my presentation. All right. This is effectively the three set of topics I'll cover. I'll go through the revenue forecast and growth opportunity, things we're doing on the margin expansion side, and then capital allocation, and then address capital return and M&A as a part of that discussion. With the investments that we've made over the last couple of years, we have really strong growth vectors in place with 5G, RF front end, adjacent markets, and then QTL as well. I'll go through each one of those, and I'll start with the growth discussion first. Think of our business in four parts, and the first two effectively represent QCT as you know it.
You have the mobile handset platform and the adjacent platforms. What is included in the mobile handset platforms is the core chipset, and I'll provide more details on that. That's the platform that will benefit from the transition to 5G. There's the RF front end and fingerprint. The total revenues for that segment is $11.3 billion in fiscal 2019. Going over to adjacents, we'll talk through automotive, compute, and IoT. The total revenue for that segment in fiscal 2019 was $3.4 billion. You'll see that we've included RFFE connectivity, mobile connectivity, and fingerprint in the mobile handset platform. We just feel like, especially RF front end, the growth of that market is so closely tied to our mobile chipset business, that that's the right way of thinking about RF front end going forward. Long-term opportunities, I'll briefly touch on ADAS and cloud edge AI.
I know the rest of the presenters have talked through it already. I'll give a quick financial lens on how we think about it. Then finally, I'll finish talking about licensing, just the overall status of the program and how to think about 5G impact. As you know, QTL revenues in fiscal 2019, $4.6 billion. Okay. For mobile handsets, of course, the key driver for us, both for the QTL and the QCT chip business, is the market. The key assumptions as we look at the market, we are forecasting a market that's consistent with its current form, both in kind of the total scale of the handset market and the mix of OEMs across the board. We're assuming that market carries forward. We don't have a market growth assumption built into our plan.
Within that, the way we thought about the 5G forecast, we mixed a combination of tops-down analysis and bottom-up assumption. Tops-down, we looked at how the transition from 3G to 4G happened, and what could we learn from that. We applied that to the transition from 4G to 5G. There are two key differences. The first difference is the availability of chipsets across tiers in 5G. When 4G was launched, we had a premium-tier chip and premium-tier devices only in the first year. As 5G, you're well aware, we have chips across multiple tiers all available in 2019, late in 2020. The second difference is China. In 4G, China started two years after the rest of the market. In 5G, we're expecting China to really not just start at the same time, lead in some cases, the 5G deployment. You have that deployment curve pulled in.
When you take a combination of those factors, we are forecasting the market to go from the 200 million midpoint that we gave for 2020 to greater than 750 million in 2022. In order to test these numbers, we also looked at the bottom sub plans of our OEMs, which operators are deploying in what scale, both Jim and Steve and Cristiano talked through the scale of the deployments, so we considered that, and the design pipeline for each of our OEMs. We've been working with them, and we have a pretty good idea of when they're going to launch their phones. This is the baseline assumption for our financial plan in terms of 5G devices. It's obvious, but just to clarify, we expect this transition to continue, so as you go beyond 2022, we'll see the additional remaining devices that are 4G transition over as well. Okay.
Talking about core chipsets. When we say core chipset, what does that mean? It means our modem, application processor, transceiver, PMIC, and connectivity chip. It's the first set of chips that we've been selling for a significant period of time to our customers. That SAM, we expect to grow from $26 billion-$ 35 billion at a CAGR of 10%. If you think about the key driver for the CAGR, 5G is very complex. It's not just as you heard, it's not just sub-6 and millimeter wave. There's TDD and FDD, there is NSA and SA, there's carrier aggregation, there's DSS, and then you have to do all those things at the same time in a device. There's tremendous complexity. It plays to our strength, and that's one of the key drivers of the growth in the market.
As you look at the bar for calendar 2022, you'll see that about two-thirds of the bar is 5G. While there is a certain curve for unit penetration, when you look at the SAM, the $ SAM, we expect a much quicker acceleration to 5G, which, given our position in the technology, it allows us to participate in it in a good way. In terms of our traction, we've quoted this number before. We have over 230 design wins across all devices in 5G, which is a great starting point as we look at this ramp going forward and the transition happening and our ability to participate in it. The last point for this transition, as we've announced before, we have a multi-year deal with Apple. When those products get launched, we will have an incremental benefit, as well as revenue scales up.
Overall, when you think about this growth opportunity, it's in two parts. We grow with the SAM, and on top of that, we have growth driven by gaining share at Apple when they launch phones with our chipsets. On this page, we'll talk through RF front-end. When you look at a similar metric for RF front-end, we see a growth from $13 billion- $18 billion at a CAGR of 12%. If you look at the technology or content mix, we expect the growth to be all driven by 5G. As devices move from 4G to 5G, we'll continue to have 4G content in those devices, but the growth is all driven by 5G content.
As we have discussed earlier through the day, we have a very strong portfolio and strong position, especially in the 5G content, it'll allow us to participate in that upside. In addition to that, we also, through dynamic spectrum sharing, expect 4G bands to be used in 5G mode as well. As that happens, more content moves from 4G to 5G and increases the 5G portion of the pie. It sets us up in a really nice position to further expand our position in terms of share. From a design win perspective, the same metric as we quoted on the handset side, we have over 230 design wins for 5G chipsets, and virtually all of them have our RFFE. We already have a strong design win pipeline that we're going to benefit from.
As you look at this SAM, our target for share in this is greater than 20%. We feel like with the transition happening to 5G over the next year or so, it creates a chance for us to go achieve that and then become a much stronger player in the RF front-end ecosystem. When you take those two things together, the RF front-end growth and the chipset growth, and look at it, we wanted to bring it together and reiterate the metric we had given before, that as we go from 4G to 5G, we expect content per device to grow by 1.5x. The 1.5x growth includes sub-6 and millimeter wave for RF front-end and for the core chipset.
The way to think about this metric is, if you take a 4G premium tier device with our 800 tier chipset to a 5G premium tier device with our 800 tier chipset, like for like device, we expect our revenue opportunity to grow by 1.5x. As 5G goes into lower tier devices, that carries over. The same increase in opportunity would apply to the 700 and 600 tier as well. Turning to adjacent platforms. This slide shows a revenue breakdown of the three categories we are including in the adjacent platforms. We have auto, compute, and IoT. I'll just quickly talk through each one of them. SAM for auto means telematics, connectivity, and infotainment. We'll talk about ADAS separately later in the deck. This is an existing business that we have.
I have a slide on auto, so I'll wait for that slide to give more details. The second is compute. Cristiano talked about the very significant opportunity we have in the compute market. From a financial perspective, this is clearly a large profit pool that we are not participating in, and it's an area where we can reuse technology from mobile, so at very low incremental cost, participate in a big profit pool. We'll look for the opportunity to do that. In terms of our financial assumptions, we are not assuming a big win in compute and the guidance we're giving. We're just excited about it as an upside opportunity. For IoT, we've split the IoT market here, our revenues into non-cellular and cellular.
Clearly with 5G coming in, it's going to be a big advantage for us as we focus on the cellular IoT market in addition to the non-cellular. I have a slide later in the deck on that as well, so I'll talk about that a little more. There are two key points for us as we look at these adjacent markets. First is these markets are demanding technology from mobile, and so we can leverage our R&D base with limited incremental cost, bring the technology to these markets, and grow our revenues. The second thing I would say is because of that strategy, these markets are margin accretive to QCT. We're able to address these markets in a margin accretive fashion. Okay. Transitioning to auto. Just to reiterate, for us, the SAM here is telematics and connectivity and then infotainment.
The way we think about telematics SAM, we started with 3G and 4G chipsets in cars. Now we are going to 5G. We can bring in RF front-end now that we have that asset that can attach to cars as well. Jim talked about C-V2X, Cellular V2X. As that technology comes in, we'll be able to address that on top of those. What happens is what used to be a chipset sale suddenly became a telematics platform where we can add more content on top of it. Connectivity includes Wi-Fi and Bluetooth, and we've been in that business for a while, and we got a portion of that business through the CSR acquisition. The light blue is infotainment, and I think Cristiano went through some significant detail on what the infotainment revenue includes.
We feel like there's a very big SAM that's available to us, and we have all the right technologies to go access it. In the middle, we are showing the design win pipeline. You've seen this number before, $6.5 billion. What we tried to do in this presentation is take one step forward and help convert it into what it means for revenues. In fiscal 2019, we had over $600 million in revenues in auto, and we are forecasting that goes up to $1.5 billion in five years. Of that $1.5 billion, approximately 75% of the revenue is related to designs we've already won. There's obviously high predictability in those revenue streams, and the remaining 25% is based on designs that we'll win. Also, as you think about beyond 2024, this is not the peak year for us in auto.
As we have the strong design win pipeline, we continue to have a steep growth profile beyond 2024 as well. We really like the growth profile of that business because it complements the 5G growth profile in mobile and adds more tailwinds to that growth. From a financial perspective, the other thing that's very interesting about this opportunity is, and I think this was discussed earlier, is we're reusing technology and chips created for mobile, and we are planning them in such a way that they can be used in auto in later stages. It just allows us to really leverage the R&D, and then there's some auto specific software work and other work done that allows us to bring it to the auto market. From a margin perspective, as this business grows, it'll look very interesting to us.
Turning to IoT, Cristiano mentioned our total customer base here. We have over 11,000 customers, and the revenue is split across a lot of different segments, geographies, and different types of customers. There's really good diversification within this pie from a customer perspective. Then as you break down the CAGR on the market and you look at cellular versus non-cellular, for the areas we are addressing, there is significantly higher growth available in the cellular part of the market. That is our opportunity to go and participate in that, use 5G as the lever to increase our presence and grow revenues. As we think about this market, we think we have an opportunity to grow with the market, and that's our financial planning assumption. Turning to long-term growth opportunities. Same premise as you've seen before.
We're going after the ADAS market, and we are addressing the cloud edge AI market, and we are reusing all the technologies available from mobile and chips available from mobile to go do that. Again, while these are new markets for us, we are leveraging so much technology. For ADAS, we are leveraging our channel as well. It puts us in a pretty good position to be able to access the market. One interesting metric that I wanted to leave with you is to address this total SAM of $18 billion, we are investing an incremental $200 million. It's a very small incremental investment to address this SAM. This $200 million is already in our spend run rate. This is not something that's incremental to what we've guided before. It's within our spend run rate.
It's an example of how we can really leverage the R&D in mobile to address larger markets. Turning to QTL. As Alex mentioned, we've signed over 75 licensees for 5G, and then we have some multi-year anchor agreements around which the program will evolve. We're in a pretty good spot and very good start on the 5G licensing program. From a guidance perspective, we've provided guidance for the December quarter. We've provided guidance for the March quarter. I think you have the actuals for September. You should be able to use those numbers and the seasonality around those numbers to forecast QTL revenue for fiscal 2020 and really use that as the baseline going forward. There are two upside opportunities to that revenue base. First is Huawei. When that gets resolved, it'll get added on top of that forecast. Second is 5G benefit.
We're not planning for any 5G benefit in QTL at this point. There are two ways in which you could see significant benefits could accrue to QTL. First is replacement rates. When we went from 3G to 4G, and I think Cristiano mentioned this as well, there was a significant increase in the replacement rates. If we could see a very similar phenomena with 5G, and that would increase the size of the market, benefit QTL. The second benefit is as devices at the mid and low tier, those users upgrade their devices, and they buy more expensive devices to take advantage of what 5G offers. That will be another incremental benefit to QTL. Those two would be upside benefits to the program. Moving over to operating expenses and leverage. This is obviously an area of key focus for us.
We want to realize operating leverage and realize increase in margins as we go forward. If you look at the cost reduction plan that we just concluded of $1 billion, the way we approach the plan is we cut things that were non-core to us. Then we focused on driving SG&A efficiencies. We tried to do this in a way such that we can preserve the investments we needed for RF front-end and 5G and execute on those things. We're pretty proud of how over the last couple of years we've been able to execute on RF front-end 5G while reducing our cost structure. Going forward, our focus will be continue on to look at SG&A efficiencies, focus on R&D productivities, then really be selective about any new investments that we make.
Any new investments we'd make would be consistent with the philosophy we've outlined of being able to leverage the mobile R&D base into new markets. Transitioning to capital structure. We're very happy, obviously, with how we've done with our capital return program. Over the last three years, we have returned $36 billion to the shareholders, $10 billion in dividends, and $26 billion in buybacks. When we started our incremental buyback program about 15 months ago, since then, we've bought $23 billion stock back at a price of $65. With that, we retired 22% of the shares outstanding. That program has worked out extremely well for us. We're very happy about how that played out.
During this time, these three years, we invested $17 billion in R&D, a lot of it was around the concepts Jim outlined in early investment, kind of making 5G happen, taking a systems approach, driving chipsets and royalties licensing at the end. Looking forward, we're focused on strong balance sheet, investment-grade rating. Obviously, keeping our balance sheet strong is extremely important to us. We will grow dividends over time. From a buyback perspective, our baseline will be anti-dilutive buybacks. Opportunistically, we will consider buybacks on top of it. That's kind of our overall capital structure policy going forward, very happy with how things have played out over the recent past. From an acquisition perspective, we've done 21 acquisitions over the last five years. As you can see, most of these acquisitions were much smaller.
We've done two large acquisitions in the five years, the last one being acquiring the EPCOS asset from TDK, which we recently completed the transaction. One of the things we're proud about is how we were able to acquire that asset and a couple other companies. Together, we were able to build a portfolio of RF front-end technologies that complemented our internal R&D investment. We feel pretty good about the portfolio that we created partially through M&A. Looking forward, we'll keep looking for tuck-in acquisitions, buying small technology teams, buying assets that fit with our existing strategy, and help us execute on it. That'd be the primary focus. Opportunistically, we'll evaluate other M&A as well. Going to our three-year financial targets, and what this slide really is a summary of what we've said on all the other pages.
That's how you should think about these numbers. One of the key things for us is as we grow going forward, a lot of the growth vectors are driven by QCT. You will see a shift in the mix of businesses between QCT and QTL, where QCT will become a larger portion of our profits. From a revenue growth perspective, when you take the SAM growth of our existing markets, again, I'm excluding the longer-term bets. Our existing markets, it's around 10%. From a revenue perspective, we plan to grow with the SAM, plus have Apple revenues on top of it. Right? That's how, if you think about the longer-term modeling, it should be the SAM-related growth, plus the Apple-related revenues.
From a QTL perspective, I think I already outlined the way to think about the business going forward is the baseline profile opportunity from Huawei resolution, and then opportunity from 5G on top of it. From an operating margin perspective, QCT, we think we have an opportunity to grow over 20%. We are very optimistic about where this takes us, our existing growth vectors. Then as some of these large markets materialize, we would have an ability to grow beyond that target. From QTL perspective, we're guiding 70% operating margin, and you should think of a band around that number, where lower than 70% is effectively reflective of the current run rate that we've provided to you. Then as we resolve Huawei, we would be higher than the 70%. Just maybe summarize the key takeaways from my presentation.
We're assuming a mature handset market in existing macroeconomic conditions. We see an accelerated move transition to 5G happening versus a transition to 4G. On the mobile handset platform side, we feel like we're in a very strong position with RF front end and the 4G to 5G transition, the chipset opportunity for us. Beyond mobile, we have growth vectors in all the markets we are addressing, including auto, IoT, compute, and then the longer-term opportunities, ADAS and AI. Finally, QTL, we have over 75 licenses completed for 5G, and that'll serve as a base for the expansion of the licensing program. When you look at this, we feel like we're in a very strong position to grow earnings faster than revenue and hit the targets that we've outlined. That concludes my presentation. Thanks for coming here today.
With that, I'll call the Qualcomm management team to come back on for Q&A.
Okay. Who is first?
Hi. Hello? Okay. Hi, it's Mike Walkley with Canaccord Genuity. I know there'll be a lot of questions about RF. I might just switch gears here. Just on QTL, just how should we think about maybe the next steps or the timing on the FTC case, the appeal case with the Ninth Circuit Court? How might the business be contained given your 75 plus agreements and licensing deals, your recent deals with Apple and Samsung? Are those deals set no matter what the Ninth Circuit rules, just how do we frame about the risk of this case and the timing?
I'll start, Mike, and then turn it over to Alex. In terms of timing, so as you know, we have the stay in place, unanimous panel from the Ninth Circuit ruled that we were entitled to a stay, which has obviously been extremely satisfying to us. The timing currently is that oral arguments on the merits should take place in February, perhaps March. All depends on the court scheduling and their dockets. It won't be earlier than February, and it is possible that it'll move into March. We'll do the argument then, and again, based on the court's history and scheduling, it looks like it would take anywhere from, I don't know, seven months to a year and a half before they make their ruling. That sounds too specific than I want to be, but that's a range I've basically been given by several people.
As you know, Alex and his team have continued to actually all through the FTC case trial, they were continuing to, as we showed you on the timeline, negotiate license agreements. I'll turn it over to him to fill in the rest of that answer.
Yeah. I think two things, Mike. First, tremendous amount of confidence in our position on the appeal. The second thing is, I think you should just consider the fact that as we were entering into these agreements, including some of the key agreements that you mentioned, we had a very clear understanding of what the FTC issues were. The parties entered into these agreements with those in mind. Both parties understood what these issues were, and I think that kind of the best way to think about it is we're smart enough to understand how to resolve and how to get to long-term deals, taking those issues into account.
I think just to be clear, on picking up on Alex's point, we have and continue to view the FTC's complaint as without merit entirely. Nothing about it we think was meritorious, neither from a legal perspective or from a factual support for the legal perspective. That will remain our position until we are successful here. What Alex was essentially saying is even if you look at some of the things they've accused us of, we're dealing with those effectively.
Thank you. Hi, Chris Caso from Raymond James. I have two questions regarding millimeter wave. One, if you could address the incremental cost, and I think this is one of the concerns of putting millimeter wave solutions on, what's the incremental cost of the phone? Will that be prohibitive, make the phones too expensive? Secondly, from a Qualcomm perspective, it sounds like you've got very good penetration on that and the sub-6 in the next generation. What's the competitive moat that you have, such as you have these designs, you'll hold on to them as you go on to the next generation?
Yeah, happy to address that question. Look, when you add millimeter wave to the phone, you increase the number of antennas, and it has been an increase in bill of materials, and we see an increase in the overall cost. However, having said that, we're bringing millimeter wave to all of our tiers, including we have now our 700 tiers designed with millimeter wave. When we look at some of the price point of the devices, especially as the technology is getting scale, we've been positively surprised by the fact that it's proven to be viable. I expect cost to go down as we go to next generation modules. Like any technology, you have a technology premium, but you have efficiency gains and ability to reduce the number of modules per device.
The metric will be our 700-tier device coming with millimeter wave within 2020 in the first half. I think your second question was about a competitive landscape. If I understand, you want a competitive landscape in millimeter wave, sub-6, or both?
In both. How do you hang on to the designs that you have? What's the advantage for Qualcomm compared to-
On RF, modem, or both?
On RF.
RF. All right. We have three vectors of differentiation. I think some I will put in the technology category, another one is in the scale category. Let me start with the technology category first. We've been very focused on 5G, and I think as Jim Thompson pointed out in his presentation, we worked very hard to make sure we could win the performance comparison at the individual component level, whether it was a GaAs PA, a filter in the low band, a filter in the high band. We feel now we have a very competitive solution at the individual component level. As a matter of fact, we are winning some discretes, especially what we call agnostic, on a HiSilicon platform or even in an Exynos platform because of our performance at component level. That's how we could compete with the existing players.
What we actually done, and I'll say the technology differentiation, is really implement capabilities at the system level, and I'll walk you through a few examples. One is this technology that we call Smart Transmit, that basically take measurements at the millisecond level and basically adjust power. We're seeing delta of multiple DBs is the difference between have coverage or not having coverage. Almost in the days of feature phones, whether you have a good antenna or you don't have an antenna, and a good antenna, and you have significant performance at the edge of the cell. The other capabilities that we have is how efficient we can make envelope tracking in 5G versus APT. The other driver of differentiation is what I call the scale model. In the 4G era, we were not yet ready to enter the RF space.
We never get much credit for, but our test platforms end up having an influence about market share in some of those players. When we develop a chip, a baseband chip, and that chip, we have samples that go to our customers. We build mobile test platforms, and we build reference designs. Before we have commercial software on the baseband, we take this into every geography, across every infrastructure vendor, every software version, whether it's a Huawei base station, ZTE, Datang, Samsung, Ericsson, Nokia, you name it, and across multiple operators' implementation, and then we have a commercial software on the modem. We, in the 4G era, will do this with our test platforms, and we'll pick who goes into the board, whether it's Avago goes or Murata goes, Skyworks goes, or Qorvo goes. Our test platforms are only built with a Qualcomm front end.
When an OEM gets a commercial software chipset, our front end is tested and certified. He can build that with different solutions, but he has to test himself. I think that gain on scale and time to market is also big differentiation on Qualcomm as well because now the name of the game is a little different since we have a 5G front-end solution across the entire chain.
It's Rod Hall with Goldman Sachs. Thanks. Two questions. One would be regarding the millimeter wave unit volume. I don't think we question whether you have a competitive advantage in millimeter wave. It's more when do these units ramp? I wonder if you could help us understand within the 200 million units next year and the 450 the year after, then maybe even the 750, if we're getting really greedy about data. If you could help us understand how these units ramp through time. That's question number 1. Question number 2 relates to 5G penetration in total. The penetration rates that we calculate based on not just your numbers but industry numbers, suggest that 5G by 2021 will be well over 10%, 12%, 13% penetration of the smartphone installed user base.
That number is significantly higher than what we would have calculated the same year for 4G. We know that conditions are different and so on, and as you pointed out, Akash, the chip availability is different, but it's still a much different number, and even if we back out China, penetration's still a lot higher in the rest of the world than it would have been at 4G at the same time. I'd just like to get a little bit more color on why that's the case.
For sure. I'll maybe start with your second question. As I said in my comments, the biggest driver in our minds, and there are two drivers, the biggest driver is China. It's not just the China market. The intensity comes from the China OEMs and the OEMs wanting to take the 5G phones that they make in China and really export it in different markets. They see it as an opportunity to be there first and gain share from other players. The China phenomena is not just about the China market, it's about Chinese OEMs as well. The second is you have phones across multiple tiers, and maybe that also speaks to the question that Cristiano answered earlier, is we'll be able to go down the price premium curve much faster because you don't just have the 800-tier chips, you have 700 and 600 as well.
It gives the OEM options to be able to hit aggressive price points with 5G, which will drive adoption rates. Those are the two factors I'd look at. From a forecast breakdown perspective, we're not giving that insight at this point in terms of what's millimeter wave versus sub-6. The way I would recommend you think about it is think about the key markets, and Cristiano outlined which markets are going sub-6 and which markets are going millimeter wave. U.S., Japan, Korea in 2020, those are the three key markets that will have millimeter wave deployed. You would expect premium-tier phones, and these are mostly premium-tier markets, most of them to have millimeter wave capability. Maybe that's a way to think about scale.
Thank you. It's Matt Ramsay from Cowen. I'm not sure if this is for Akash or Cristiano, but you walk through the slides, and you guys lay out a pretty compelling case of 50% content per device. I think, back into some numbers in my head, I think it was a 20% growth rate in automotive. Some pretty big numbers around edge computing and auto. Yet SAM growth at 10% seems pretty conservative to me. Are there things on the negative side that we should think about that are depressing that SAM growth rate? Maybe if my assumptions are correct that things could be better than that, what keeps QCT op margins at 20%? Why can't they be higher? Thanks.
Maybe from a SAM growth perspective, that weighted average of 10% is a combination of 10% growth for the chipset market, the core chipset. We had 12% for RF front end, and then we have the adjacents growing in the 10% range as well. Really that 10% was just math based on the growth rates we have in those three markets. We are not including, as I mentioned during my prepared remarks as well, we are not including the upside opportunity that we could have if we had a home run win in compute or we had a significant win in ADAS, which has a timeline to it, by the way, and then also in AI. Those are incremental things from a revenue and margin perspective.
Hi, over here. It's Ross Seymore from Deutsche Bank. Akash, another one for you and kind of following up on Matt's question on the EBT side of things. Can you just talk us a little bit about some of the moving parts to get to that 20% or higher? Is it mix? Is it the adjacencies coming in with higher margins, same thing with RF? Then on the other side of the equation, but also sticking with EBT is on the QTL side, why is that only 70% when I think historically it used to be in the mid to upper 80s for your operating margin in that segment?
Sure. On the QCT side, I'd say that maybe the three vectors that are most influential in getting to the 20% margin target is the 1.5x premium on a like for like basis that we talked about, which includes core chipset and RF front-end premium for 5G. The third factor would be Apple when that business comes in. Obviously it'll be incremental to our margin directly because we've already done the investments, in the process of doing the investments required to scale them. Those would be the factors that improve the QCT margins, and that probably matter the most in getting the margin targets. On QTL, you have to recall that versus a few years ago, the scale of the business is a little different. I think that impacts the operating margins a bit.
In terms of R&D, we've been investing on 5G and really making sure that we have a very strong portfolio. That impacts the margin some as well.
Thank you. Stacy Rasgon at Bernstein. I have two questions on chipsets. First, how much of that 10% SAM growth for the core chipsets from 2019 to 2022 is simply because you weren't selling chips to Apple in 2019 and you presumably will be selling chips to Apple in 2022?
That's like to like, it includes Apple in both the 2019 and the 2022 SAMs. The growth is not being driven by including Apple in that.
Is there Huawei in that SAM as well or no?
Yes.
There is. Okay. I guess to follow up on Matt's question, so much of the growth, I think the market was going from almost no 5G now to 2/3 or 70% of it in 2022 is 5G, and your revenue opportunity is up 50% from 4G to 5G. I guess I'm still struggling with the 10% also. Why shouldn't it be a lot more than that? What else is going on to take it there?
Yeah. Just to confirm, the way to think about it is growth because of the SAM growth plus Apple business on top of it. What we're guiding towards is not 10% inclusive of the Apple business. It's incremental.
No, I understand that, but the 50% upside is in the core business, doesn't include Apple. That's going to be extra presumably because right now it's zero.
Yeah. As we said earlier, our assumption is that the market structure remains the same. Our share, our position remains the same as we go forward, and as a result, we would grow with the market at the same rate as the market with Apple incremental on top.
10% without Apple and then throw Apple on top?
Correct.
Okay. Thank you.
Again, because of reasons like auto and RF front-end, we could do a little better than 10%, and then Apple on top.
Yeah, thanks. It's Brett Simpson at Arete Research. You talked a lot today about the adjacency opportunity in QCT, my question really is about QTL and how the non-smartphone opportunity plays out in the next couple of years. My understanding is that you've done a lot of deals with guys like Avanci, where you've got patent pools, you've got caps on things like IoT autos. Can you maybe talk a bit about how you monetize the non-smartphone part of QTL? What portion of the volume today might be non-smartphone and what sort of growth do you expect, or what's the contribution from that going forward? Thanks.
Let me address kind of just generally non-smartphone part of the market. It currently is actually a very small percentage of our overall revenue. Right now, we're not guiding anything significant happening in the non-smartphone space. With respect to these deals that we entered into with platforms, for example, such as Avanci, we actually haven't done a lot. We've done a couple and there are alternatives. For example, with respect to the Avanci 3G, 4G licensing opportunity for automotive, we actually have an automotive program currently in place. We've had a program in place for 10 years. A licensee can opt to stick with a Qualcomm license directly or opt out and take an Avanci license. That actually, the automotive licensing business has remained largely intact without a lot of difference from the Avanci platform.
There hasn't been a lot of experimentation in these other platforms. I think the overall takeaway is, right now, Akash can comment on this, we're not putting down any markers for a real shift in the revenue makeup for licensing.
Yeah. Hi. Thanks. CJ Muse with Evercore ISI. I guess, Akash, back to you. A couple questions on the margin side. I guess first off, for PBT, exclusive of QCT, QTL, I think that's been running at about a $800 million loss. What are you projecting into fiscal 2022? Then back to the PBT margins for QCT. It looks like it's an incremental PBT margin of 30, 35%. I would think that, if the chipset business were growing around 50% over the next kind of three-year type CAGR, that we would see more leverage there. Are there investments that need to be made? Is there a time period beyond fiscal 2022 where we see the real acceleration in terms of margins? Thank you.
On the first part, I think you're referring to the corporate other column in our results. There are two things to it. First is the net interest expense that we have. We've guided, during the earnings call that on a quarterly basis, that amount would be $100 million. You should expect that as kind of the run rate going forward. That's the net interest expense, so $400 million for the year. Then, for the OpEx, which is long-term investments that we're making, we expect that to be roughly flat going forward. There's no specific plan to increase the scale of the investment. You should be able to use that to model it.
The second question?
Can you repeat the second question again? Sorry.
Sorry. If you add up what you've given us in terms of adjacencies, et cetera, it looks like QCT revenues are growing roughly 50% over that three-year timeframe. I would think that the incremental PBT margins would be above 30%-35% implied in the guide for PBT margins.
Yeah. As we think about our growth profile, we don't have a specific scaling on OPEX assumed. The goal is to continue to invest judiciously and really recognize the benefit of the operating leverage as the revenue grows. We feel like having a target of greater than 20% is the right place to go for now, as different markets kick in, maybe there's an opportunity in the longer term.
Hi, Ivan Feinseth, Tigress Financial Partners. Thank you for taking my question. What were your original thoughts in the acquisition of NXP, and how do you feel like What would've been different if you were able to complete the acquisition, and what direction did you end up going without it? Also, do you feel that the government will continue to issue waivers to do business with Huawei? What would happen if they don't?
Sure. Why don't I take that? I'll do the second one first, then I'll do the first one. The second one, I think, one thing with Huawei you should remember, we have a pretty small product business with Huawei. For us, we're actually somewhat insulated from all of the news cycle with Huawei from the product side. Obviously, on the licensing side, we're trying very hard to get them and their agreement done. We tend to have less of the ups and downs based off of that. I don't think long-term, you should think of the opportunity with Huawei on the product side to be very significant. On the NXP question that you had, you should really think about the NXP acquisition in the context of when we started it, which was several years ago.
At that point, our position in the auto market was very different than what it is today. If you look at what's happened, particularly in our infotainment business, it's grown dramatically over that time period, in part because we came in with really differentiated technology, really from the roadmap that Jim Thompson described, and we were able to build that business really without some of the things that we wanted to have with the NXP acquisition. You go forward three, four years later, many of the channels or technology entry points that we were trying to get, we've been either able to do ourselves or because we're so much closer to 5G, the need to get them has probably become lessened. I think for us, we have the ability to leverage our existing roadmap.
Of course, we're looking at M&A and we're trying to get ourselves in a position where we can take advantage of that from a balance sheet perspective, as you've seen from Akash's presentation. We've been happy with the markets that we've been able to drive in the meantime since the time when we initiated that NXP acquisition.
Hi, it's Vijay from Mizuho. Just on the 5G side, you talked about 175 million-225 million units for next year and 450 million units the year after. Just wondering what's the attach rate you're assuming for RFFE and millimeter wave that you build into your core chipset, et cetera. Thanks.
Just to confirm, the forecast we gave is for the total size of the market rather than our units. I think millimeter wave, I answered Rod's question earlier, that is probably the right way to think about what the attach rate would be for millimeter wave overall for the market. We would obviously expect and hope to win the millimeter wave attach for our platforms. We feel like we are in a very strong position on millimeter wave from a competitive landscape perspective. From an RF front-end perspective, again, the forecast was the market. Obviously, all those devices will need RF front-end capability for 5G. It really comes down to our ability to win the core chipset and then offer an end-to-end advantage and be able to have a performance advantage that allows us to win the RF business as well.
Hi. Srini Pajjuri from SMBC Nikko. Thanks for the presentations. Akash, I want to go back to the margin question again. You said that, I guess for the March quarter, you're guiding mid-teens margins, operating margins for QCT.
Right.
Your longer term model is 20%+ , at the same time, you're not spending much in terms of your OpEx. I'm still confused as to why it's only 20%. I know you said plus, why not 25 or 30? Is that because your gross margins in 5G are somewhat below that of 4G, or is there anything else going on?
I think it's really we're trying to be careful with our margin target. There are a lot of variables in the 5G growth rate, right, as when it happens, how it happens, in terms of which tiers. There are several unknowns. As this plays out, we feel like having a greater than 20% margin target is a reasonable place to be right now.
Thank you. I have a question for Alex. On 5G licensing, you talked about how strong your portfolio is. Can you put that into perspective as to compare that with your 3G portfolio and 4G portfolio? The reason I'm asking this is that I've read a few times that Huawei has at least more 5G patents than you guys have. I'm just trying to understand that. What does that mean? Is it a matter of number of patents or how strong their patent portfolio is? Along the same lines, if you could talk about, since the trade war has begun, have you seen any change in behavior in China in terms of compliance? Thank you.
Okay. Taking the first one, there has been a lot of discussion about quality versus quantity, and probably a lot more on the quantity side. One of the things that I tried to do today is actually provide some insight into quality. I don't want to get into a battle on who has more patents where. I think I try to give some of the key foundational points. Early R&D, very strong foundation in 4G. I think the points that were made, not just in my presentation, but primarily in Jim's presentation, about driving some of the key fundamental features in 5G that enable the performance benefits, the benefits to operators and users, the benefits to device manufacturers that we talked about are really indicative of the heft and value of the portfolio we have going into 5G.
It's just not really worthwhile to start getting into debate about quantitative metrics, who has more here. One of the things we came across recently, and it's not with respect to the company that you just identified, is somebody that said, in a discussion, "I have a lot of patents on 5G." We started taking a look into them, and all the priority dates are after the date of the finalization of the spec. There's a lot of noise out there that I think we're trying to cut through a little bit, but it's kind of an endless debate that doesn't really go anywhere. With respect to compliance is good. Actually, we're not laying out the data on compliance. It's actually been quite good. We're not seeing any pushback on compliance relating to the overall geopolitical issue.
We obviously have the Huawei issue, but we're not seeing any pushback on compliance.
I'd just add one thing to Alex's first answer with respect to your Huawei question. As Alex said, there's a difference between quality and quantity, and there have been studies, and it's quite.
Clear that the quality of our patent portfolio, in addition to the quality of the contributions we make versus others make to the standard, in particular in 5G now, given all its complexity, it's a big difference. You can't look at numbers, because numbers can be very deceptive, deceiving, and we're confident that we have real strength in both the patents and the contributions to the standard.
We have time for one more question.
Hey, I'm Mitch Steves from RBC. I just have one just to kind of clarify on the 5G side. It looks like a lot of the growth is predicated by a pretty substantial ramp there. I guess I have two points. First is what gives you guys clarity in kind of the unit growth rate there? Essentially, what are the chances we see kind of an inventory build in China? Because it seems like that's kind of the big driver. Secondly, kind of dovetailing off a prior question, when you get to, let's say, 450 or let's say 700 million units, what's the margin trajectory look like specifically to those phones?
I think from a forecast perspective, as I said, there's a lot of things in play right now with 5G. The best we could do is look at how 4G played out and what can we learn from that. Also, as I said, there are two key differences where China is starting earlier and we have chipset across tiers. We've been talking to various customers as well, and we are aware of what their plans are of launching handsets. We're trying to make an educated estimate based on the data points we have on how the growth rate will materialize.
We got a lot of question about the 5G. Maybe I just want to reiterate one thing we said during this call, especially as you look of the projections we make of SAM growth. We had assumed in our projections that as the market is, and we've been calling down the mobile market, we're assuming the market stays as is, and our current share situation stays as is. We're assuming that we're going to have competition, the competition is going to execute. Any change in replacement rates, any change in share will change that dynamic and will be upside for QCT. I just want to probably reiterate what are our assumptions on the market on the rollout, but we'll feel good is a very high confidence number.
Okay. Well, thank you everyone for coming. I think this is going to be the year for 5G. We're excited about it, and we're very happy to be able to share the story. Thank you.