QUALCOMM Incorporated (QCOM) Earnings Call Transcript & Summary

Hi. Good afternoon. I'm Samik Chatterjee. I'm the IT hardware and networking equipment analyst at JPMorgan. For the next company at our Tech Forum, I have the pleasure of hosting QUALCOMM. With us today, we have Nakul Duggal, who is the SVP and General Manager of Automotive at QUALCOMM. We also will have Mauricio from Investor Relations on the line. I do encourage investors who have the unit to send in their questions via the chat, and we will definitely ask it on your behalf. Although please keep in mind that the companies are in a quiet period, so they won't be able to comment on any specifics on the quarter itself. Nakul, thank you for attending the conference. Obviously, with the automotive focus, it makes it very timely to have you here. We did host Patrick last year on the same forum. So I would kind of hand it over to you to go through any prepared remarks you have and look forward to an interesting discussion.

Thank you very much, Samik, for having me. Happy New Year, first of all, to everybody. No, I think clearly, automotive as an industry is in a very interesting phase. I think a lot of announcements coming out of CES this week. I just wanted to inform your audience, we have our own automotive redefined event on January 26th and 27th. We decided because the event are all virtual this year to try and find ourselves some time, and so we welcome you to attend and participate in that event. Our business in automotive is continuing to expand at a very healthy trajectory. One thing that is really very -- one thing that we are seeing across the board is the pace at which new innovation, whether it is in the areas of connectivity, in the areas of cockpit, autonomy, transportation, there's just a tremendous amount of innovation, a lot of it has been changed. And unlike traditionally, where the change used to be happening in a much more controlled fashion where multiple OEMs, multiple automakers follow kind of very predefined path, changes now are happening in a fairly independent and fairly unique manner. We are seeing a number of different start-ups trying to move the needle, move very aggressively. We announced our partnership with Neo for their next-generation cockpit, that's a very different pace of change. We announced our partnership with Great Wall Motors on our next-generation Snapdragon Ride announcement. And then, of course, in the traditional way, in the large automaker ecosystem, we have shared fairly large pipeline sizes of revenue that we have booked over the next 5 to 6 years. And what that all kind of points to is the car is modernizing at a pace that it has not modernized before. And what that means is the car architecture is changing, it is creating tremendous opportunity for silicon, tremendous opportunity for cloud, for software. And we -- as we look at ourselves as a business within QUALCOMM, we are looking to see how we can innovate ourselves to start to keep up with these things. While I try and speak a little bit to the 4 or 5 different areas of focus that we have within the business unit and hopefully, I can give you a sense as to how we are positioning ourselves for what is automotive and what is next generation computation. Now then 5G obviously has a very big impact and meaning for connectivity in the future. The pace at which 5G is moving, I think, is known to everybody. We are seeing 5G get launched across multiple automakers in 2021. So we're going to start to see that ramp happen. But for us, 5G in the car is much more. The modem happens to occupy a slot in the vehicle where multifrequency GNSS for hyperfusion positioning, it's something that's sit alongside the modem. This is needed for being able to get much higher levels of accuracy, which is in turn needed for navigation and autonomy. So that becomes an important distinct advantage as you start to think about the number of subscriptions in the vehicle, the car itself is a subscriber towards the driver and the passengers, so is the enterprise. We have introduced features like dual and dual active. That is very unique in the car because these are typically handset features. But in the car, we have to support these concurrently. You actually have multiple communication channels active at the same time in the vehicle, one for the vehicle, one for the driver and the passenger, very unique application for the car. We have integrated Cellular V2X into the vehicle. We announced a partnership with [ Hongxin ] in China last year, and more announcements coming over the course of the next couple of months as C-V2X start to get launched in China. That is an endpoint that exists in the same place the modem happens. So for us, if you think about 5G in the car, it is a whole lot more than just simple modem connectivity. And you can kind of tie it back to what it means for the car, what it means for its architecture, the use cases, where it fits into in transportation. The second major transition, and I'm sure we'll talk about it in the Q&A session is the cockpit transition. And the simplest way to think about the cockpit is the car used to have a variety of different individual endpoints for various functions. The car used to have post, whether it was rearview cameras or instrument clusters or the infotainment itself. All of this started to get consolidated in the last 2 or 3 years. The next generation, the next phase that we are seeing for the '24, '25 time frame is the transition towards zonal computer or domain computer, where we will start to see much higher end processors that will have safety capability, that will have heterogeneous compute. And they can take the personality of the function that they are supposed to be able to solve for with similar hardware chassis, but different software platforms that are giving personality to that underlying hardware chassis. And we are seeing a tremendous amount of innovation here. The integration of e-mirrors, intelligent backup cameras, driver monitoring, child presence detection, occupant monitoring, alongside all of the traditional things that we've started to see. We are seeing a tremendous amount of change in the software ecosystem. We've been working with Google for many years. As GaAs starts to get deployed as one of the operating systems for the vehicle in an embedded fashion, and that starts to bring unique Google services. That is implemented as a virtual machine within a cockpit. So it sits alongside many other software elements. We are seeing unique ecosystems come up in China, which we are supporting, whether it's [indiscernible], [indiscernible] or Temenos. So the cockpit is become very interesting endpoint, independent of telematics. And it's all pushing towards high performance. It's all pushing towards a lot of integration of features and capabilities and software. The third area is ADAS and autonomous driving. We made our announcement at CES last year with Snapdragon Ride. That platform has received a lot of positive success. We are engaged with many different customers, many OEMs, many Tier 1s on a variety of different designs. We announced a partnership earlier this year on the stack, which we'll talk about a lot more in detail at our event that's coming up later this month. But really, our approach in that space is that we want to create an open, programmable, scalable platform. We believe that this market is going to keep evolving over the next 10 years. And we want to be able to provide the opportunity for automakers in Tier 1s to be able to have a platform that they can own, they can program, they can bring best-in-class capability. We will provide our solution, we'll provide our approach, but it is essentially an open platform. Customers can choose to use the stack in the software that we provide with our partners or build their own. The fourth area that we announced at CES last year, that is also now picking up was Car-to-Cloud. And the simplest way to define Car-to-Cloud is, when you deploy a platform that is going to have a much longer life in the endpoint, in this case, the car, you have to plan for flexibility. You have to plan for over-the-air updates. You have to plan for performance headroom. You have to be able to make it flexible enough that the platform has value over the life of the vehicle. So we have come up with ways to essentially design our hardware and software, but that we can offer business models that allow customers to be able to unlock more functionality over the life of the vehicle. Should they choose to buy connectivity from us, we can make that available. And then of course, we'll partner with various content players, services players to be able to bundle that capability. So that is something that has picked up, and you will hear more about that later this month. The final area is Cellular V2X, which we are very proud of because this is one technology that was created for the auto ecosystem. It's a wireless technology, but it was created for the auto ecosystem. In China, it has now deployed, like I mentioned earlier, so big success there. We are working with many, many different OEMs in Tier 1s there. In the U.S., the FCC made a rolling towards the end of last year to sell aside 30 megahertz for C-V2X. That is now in discussion, and we are talking to the entire ecosystem to see how do we accelerate the transition of C-V2X into the U.S. So yes, that's a very quick summary of kind of where the business is at, how we are doing. I think the market is moving really interesting, so I look forward to the coming year.

Great. Nakul, that was a good overview. I did want to start with the discussion around Snapdragon Ride platform, which you announced at the same -- around the same time last year. And you're kind of mentioning that you've made progress working with a lot of engagement with a few OEMs already. I do kind of although get a sense as particularly as we talk to the industry and investors, both together that the expectations for getting to level 4 or level 5 autonomy in the automotive ecosystem, the timing itself of reaching that has started to get pushed out, particularly in the investment community, which I'm closer to the expectations that you'll get autonomous driving level 4, level 5 by like 2025. I think less and less number of investors are baking that in at this point. So just wanted to understand -- I mean, obviously, you're kind of at the forefront with -- of this with your Ride platform. What are you expecting in terms of timing of in term -- in relation to the different levels of autonomy that you'll go through?

Good question. When we announced Snapdragon Ride at CES last year, we specifically mentioned we are focused on levels 2 plus. And let me explain a little bit as to what the considerations are. So I think the SAE levels as defined, obviously, they were defined a long time ago. But as you start to apply the practicality of what it takes to build a vehicle. One of the first things that you have to think about is who is in charge, who is driving the vehicle, who is controlling the vehicle at any point in time. By the way, is my connection still active? Can you hear me?

Yes. Yes.

The video was frozen. So if you think about the various tiers of autonomy, level 3 is known in where the car has taken over and the driver can take control back within a period of time. So let's say within 10 seconds, 15 seconds, the driver can take control over. Anything below level 3, so anything that starts with a level 2 or level 2 plus, the car is still -- the car is driving, but the driver is in control, the driver is responsible for the driving behavior of the vehicle. So a very important distinction because it speaks to a few different things. One is the driver's attention span, how attended the driver is, how connected the driver is into the driving behavior is directly tied to the liability. If the driver is responsible for the driving behavior of the vehicle, the driver is then is responsible for any liability that it's caused because of that action. As you start to move towards something beyond level 2 plus, as you move into level 3. The challenge in level 3 is that the idea of the car being responsible for -- but being able to transition control back to the driver is a very abstract concept because how do you then do that? What do you do in different scenarios based upon speed, based on a number of sensors you have, based upon the environment that you find yourself in. Are you in a high-speed environment? Are you in an urban environment? And it depends upon the attentiveness of the driver. This is an area of uncertainty. As you start to go to level 4, the idea now is the car has a lot more sensors. It has a lot more compute. It has a lot more redundancy and the car is able to essentially be much more in control. If something goes wrong, the car is responsible for coming to a stop or limping back home in some reduced capability, but the driver does not have to necessarily connecting, the driver can stay disengaged. The car is responsible. And then level 5, of course, is a different tier. There is no driver. It's really just the passenger. If you break these down, so in terms of what it takes in terms of the compute performance, in terms of the sensors, the redundancy and then apply it to the reality of the various driving conditions within this system to exists, what you will arrive at is there is a very distinct difference between what is a convenient feature where the driver is aided by the vehicle as the driver is in control versus where the car has now taken over and automating this functionality. The reason why the dates are not clear, the dates move around is because it has to do with a number of things. Regulation. What does it mean for the automaker to be responsible? If the automaker now has to take responsibility, in what environment would that happen? Is it only on highways? Is it on perfect streets? Which part of the world? What kind of regulations apply in all the different parts in the world? How many sensors do you need? How much redundancy do you need? What kind of compute performance do you need? So it is definitely an area where there's a lot of work going on. We are involved. Our competitors obviously very active in this space. But it is not necessarily an endpoint that you can point to and say, "Oh, by such and such time, you have conquered these problems. And at least the product will get [ entry ] ". I certainly would say that there is enough technology and enough effort being put into this space that we will start to see this transition happen. But it's going to happen in chunks of time in certain geographies within certain environment. It's not going to happen overnight where you click a switch and you can move over into a level 4 type situation. And I think level 5 is definitely much farther away because that essentially is a fully automated environment. And that, I think we are still some days away from in terms of being available in a very broad -- I don't know if I...

Yes. No, that was helpful. Let me take the investor question that came in, which reads, could you give a bit of color as to how the regulatory requirements by a Euro NCAP around driver monitoring systems impact you and what opportunity that may or may not provide?

Yes. Okay. We are seeing now with NCAP and DMS. DMS, CPS, which is child presence detection, oh, I mentioned the occupant monitoring. These are all bringing in requirements for cameras as well as other types of sensors, radar sensors, other sensors to be able to detect for presence, detect for engagement, detect for behavior. And there are 2 ways -- there are 2 opportunities. One is just stand-alone assistance, where you will have some system in the vehicle that will simply be performing this function. More and more, what we are seeing -- we will make some announcements later this month, so I will look to you to kind of look out for those. But if you think about where these functions typically belong. They can belong in the cockpit system, where it becomes essentially an extension of the cockpit. There is a camera that is sitting in the vehicle or in some designs that we have, there are multiple cameras for each seat and they are detecting for engagement, for presence, et cetera. So that is simply a software partition that is sitting on a high performance compute desk. In ADAS type environment, if you think about level 1 type system where there is a camera that is installed in the rearview mirror of the vehicle. There is an inside facing, inward facing camera that is supporting driver monitoring. And that would be another place where you could integrate and subsume your functionality. So we are starting to see these trends become quite standard. Of course, it's market by market. But both the entry tier ADAS systems as well as the cockpit systems are a great endpoints to host this functionality.

Got it. Okay. I mean, it sounds like the content opportunity is fairly significant. I don't know if there's a way to quantify it. When you think about, let's say, the automotive market 5 years from today, where today, you have about $640 million of revenue from the automotive group. 5 years from today, what's your expectation in terms of what could that content opportunity for QUALCOMM looks like across all these kind of technologies that might get adopted or use cases that might be relevant for autos in 5 years from now?

What we have shared publicly at Analyst Day in 2019, Samik, was a $1.5 billion number, which I think was in 2024. [ Let me check ] from that. But we are still committed to that number. Like I was saying in my earlier remarks, I think the big shift that we are seeing is just the amount of content, silicon wise and software wise that has to go into each of these endpoints. And then, more importantly, the attach rate of the functionality to the vehicle sold because these are, in some cases, becoming regulatory features. In some cases, they are becoming competitive differentiators. And in other cases, they are really becoming a foundation of the architecture in the chassis of the car. You need to have these features for the car to really be relevant, so you combine all 3 of those and you essentially see a multiplier of high attach rate by a lot more content with each endpoint. And our strategy has always been, we are a connectivity and compute platform provider. On the connectivity side, we look at every possible wireless connectivity that is possible Wi-Fi, Bluetooth, 5G, RF front end, Cellular V2X, GNSS. And you start to pull all of that together in a concentrated way in a platform, that calls for a lot of complexity, a lot of high-value content. Similarly, on the compute side, the performance that is now required in these systems as you are consolidating from a functionality, you need to have much higher performance purposes. So we are starting to move more and more towards the premium tier of our portfolio, which I think you are familiar with. We have a similar premium tier on the automotive side. There is a lot more concentration there. And then in many cases, there are multiple of these SoCs that have to get deployed because of the way the architecture is shipped. So hopefully, that gives you some color to that.

Got it. So just taking kind of the current business today, you have about $640 million of revenue. Can you break that down between kind of how much is telematics? How much is infotainment? And again, if I like look at the $1.5 billion number, what is the expectation? And what -- in relation to what that mix looks like between telematics, infotainment and hopefully, ADAS and kind of other functionalities starting to creep in and add to that as well?

Yes. So we don't break these down. So far, the disclosure we have provided is simply a breakdown between mobile, RF front end kind of in auto. I can't take it further down within auto. But what I can share with you is, we've been in the telematics business for a long time. It's a business that we've had for a good 10-plus years. And the market is now starting to growth rapidly. You will start to see 5G starting this year. You're going to start to see a lot of the RF front end attached. So the value of the telematics endpoint is going to start to expand. And it is a pretty mature part of our business because we've been in it for some time, so that is going to lead. And then the cockpit business, which we started off more recently in 2016, that is also on a very nice growth curve because of the attach rate comment that I made earlier. So I would expect that as you get over the next 4 to 5 years, they both start to become fairly strong businesses for us. But at this point in time, one of them is a little bit ahead of the other, just because of where we were in time. And then ADAS is going to start to kick in in the '22 time frame, '22, '23 time frame.

Got it. Okay. Helpful. I would remind investors on the line here that if anyone has any questions, please feel free to, again, pop it into the chat, and I'll ask it on your behalf. Nakul, I did want to kind of then separate out the discussion on telematics and infotainment with you here going forward. Telematics, you mentioned you've been in the business for a long time, but where are we in terms of attach rate? And kind of where are your expectations in terms of the industry getting to a more like a full attach rate on telematics units? Also, what's the content opportunity in going from like 4G connectivity there to 5G?

Yes. So on the attach rate, there are a number of different reports out there. I think I've seen numbers where 2/3 of all vehicles have connectivity by '22 or '23. What we are seeing, just to give you a sense, is pretty much every automaker that we work with in all markets are now planning for a connected car solution. There is almost nobody that I can think of that is not planning a connected car solution. So let me break that down a little bit more. The markets that may not be focused on connected car might be maybe some commercial segments in the low end, maybe some emerging market opportunities and maybe some vehicles that are below a certain threshold, the economy vehicle, et cetera. But I would say even there, Samik, in the last year or so, even in those tiers, we are seeing a very different shift happening because we have a major opportunity in terms of increased productivity, in terms of bringing in a lot more value in the commercial segment. We are seeing a lot of competitiveness in the economy segment. So in terms of the opportunities we are seeing, we are really seeing a very high attach rate of connectivity of cars. Your second question was on content. With 5G and really with premium 4G, the one decision that we made a couple of years ago that is now coming into place, we've brought in our RF front end business, so all those qualified RF front ends are now getting added to the portfolio. Also, features like C-V2X are kicking in and then I mentioned advanced GNSS. So all of that is creating an opportunity where we were going from a simple modem in the basic 4G time frame to not only a more advanced modem with 5G or with LTE advance, but also a lot of other content that surrounds that modem opportunity.

Nakul, I've had like 3 questions come in over a short period. So I'll try to see if I can hit some of them. So first one, given Apple, Google, Amazon all have talked about and invested in automotive in some way, whether it be hardware or software, how integral is QUALCOMM to get their targets in the sector? I don't know if you have any comments on that.

Yes. I mean I can speak a little bit to what we are seeing and what partnerships we have. I think, clearly, Google and Apple with their -- in their auto and car play efforts, that's really an adjunct to what we do. It's really just a broadened smartphone. With Google, we have a partnership with them on the Google Automotive Services, so we work with them very closely on bringing in that automotive environment, the embedded form of Android that works for auto into the vehicle, and that's an area that we've been working with them on for many years. With Amazon, of course, a very important partner for us. And they are also focused quite heavily in this space across a variety of different areas and we are in discussions with them across many different opportunities. The way that maybe I would answer the question is the car is quite different from a typical consumer device, right? The level of complexity in the car, as I highlighted in my remarks, is quite a bit different. The cockpit space is -- it's really a platform to be able to host a lot of different functionality. And that, I think, creates the opportunity for us to be able to bring in a lot of our high performance solution, a lot of our software, but also work very closely with all of the partners that you mentioned in the car environment. And I think that is really how it is going to be because the car really is a bit of a different platform than a phone and it has to support many different use cases, many different types of customers and currently in some cases. And it is a combination of an automotive use case, consumer use case and enterprise use case all at the same time. So I think there is a lot of opportunity in my mind for like and work with these partnerships.

Okay. Moving to the next one. How does the development of zonal architecture affect QUALCOMM's effort or impact QUALCOMM? Like will QUALCOMM make domain controllers themselves as automotive -- automobiles move towards zonal architecture in the 2024, '25 time frame?

Yes. So I think please look for more information on this at our event that is coming up on the 26th and the 27th. But maybe just to address it very briefly, this transition that is happening, what it is doing is making the hardware chassis out to the SoCs. A bit more reusable across what actually is a different application in the vehicle. You'll get some consistency there. And then software is how you differentiate between the hardware chassis. But given the experience that we've built over the years in telematics, in infotainment, in cockpit, now in ADAS, I think that is what we're going to bring to look forward. That is how we can . The short answer is yes, we're going to be very much part of the transition towards zonal.

Okay. In the last few seconds I have here, the last question. Can you kind of quantify QUALCOMM's market share in automotive? And is -- are you competing with likes of MediaTek, like your competitors in smartphone? Are you competing with those companies in this space as well?

I think each domain is a bit unique and a bit different. So I think the competitors are -- we have different types of competitors in different spaces. I think telematics, like I said, we've been present there for many years. But there are traditional competitors in smartphone that are obviously also looking to expand into auto. I think the cockpit space was a little different. I think it's a combination of traditional silicon suppliers and [ newer ] suppliers, that are present. And then, of course, in the ADAS space, also quite similar to the infotainment space. And with the added differentiation of integrated stacks being available in some cases versus open platform and others. I think each category is quite different than -- but, yes and there is a lot of competition. A lot of players out here.

Great. Nakul, that's all the time we have. And thanks a lot for taking the time to attend the conference, and thank you for a great discussion.

Thank you. Thank you very much, Samik.

Thank you. Bye.