QUALCOMM Incorporated (NASDAQ:QCOM) Arete Tech Conference December 6, 2022 11:00 AM ET
Company Participants
Nakul Duggal – SVP and General Manager
Mauricio Lopez – VP of IR
Conference Call Participants
Brett Simpson – Arete Research
Blair Botha – Arete Research
Brett Simpson
Okay. Hi, everyone. It’s Brett Simpson at Arete again, thanks for dialing in everyone. So it’s a real pleasure to welcome Nakul Duggal, to our conference today. If anyone is in doubt that Qualcomm is no longer just a smartphone chip maker, we have Nakul here to set the record straight.
I think Nakul’s been over 25 years with Qualcomm, which I think probably means you have a few patents on their Qualcomm patent wall over the years, Nakul.
Nakul Duggal
I do. Yes, thank you.
Brett Simpson
But, thanks for coming on today. And I know you’ve been traveling, so we appreciate you dialing in.
Nakul Duggal
Not a problem. Thank you.
Brett Simpson
So I guess we also have Mauricio Lopez.
Nakul Duggal
There’s a little bit of echo.
Brett Simpson
We have Mauricio on the line as well, who you all know is Head of IR and so we’ll just get straight into it. Thanks for coming on Board Mauricio.
Mauricio Lopez
Yes, happy to be here.
Brett Simpson
So I guess a good place to start our discussion is maybe just taking us back to the Qualcomm Autos Day Nakul, and maybe just summarizing some of the key messages from that event. And maybe also recap a little bit the journey that’s taken Qualcomm to a 30 billion pipeline today. If that’s good, maybe we can start there. Yes.
Nakul Duggal
Yes, absolutely. Thank you. Thank you, Brett, again, for having me. So good morning, good afternoon, everybody.
And you know, I think the Auto Investor Day, I think the information is all online, we shared a lot of detail on September 22. The big story for us in automotive is that it’s a category, it’s a vertical that we’ve actually been focused on and investing in for a long period of time. This isn’t something that happened overnight, it has taken really a long patient approach. We’ve obviously been investing in telematics and the connectivity category for a number of years, more than a decade.
And over the last decade or so we started to focus on the digital cockpit space, which really helped us transition into being recognized by pretty much every automaker today, as a supplier of application processors and application software into the auto industry and also created the platform, created the credibility that we needed to become an heiress silicon supplier. We have obviously since then acquired the Arriver stack assets from Veoneer, which are now integrated into the company.
We have won quite a bit of business in ADAS both on the silicon side and the stack side. And really, if you think about the overall strategy of the company, we have systematically been adding to our portfolio at the semiconductor level at the software level, in some cases at the full stack level, and also starting to go build a services business, which we encompass under the Snapdragon digital chassis.
So, automotive is all long term bets. These are all relationships that are global in nature that are multigenerational, they are more and more becoming directly influenced and directly set up with automakers. And we’ve been able to cross about $30 billion in design and pipeline numbers as far as our overall scale of business goes. We shared the revenue numbers for the next five years and the next 10 years, about $4 billion by ’26, about $9 million by ‘30. So overall feel very good about the business, relationships, the portfolio.
Question-and-Answer Session
Q – Brett Simpson
Let’s talk about the future of the of the car industry. And it’s maybe safe to say we’re heading into this decade where the car architecture is going to completely get overhauled. The way you build cars are going to change. And from a high level, do you think carmakers are ready for this change or not? And it feels a little bit like this sort of feature phone to smartphone transition. And I know a lot of car makers don’t like that reference, but how do you perceive things playing out in terms of this sort of big sort of transition and technology into the car?
Nakul Duggal
Yes, look, I think it is nothing like the feature phone the smartphone transition for a couple of reasons. First of all, I think car makers comes in all shapes and sizes. There are large incumbent carmakers that have been around for a long period of time, are going to be around for a long period of time just because of the scale at which they operate. There are a number of new automakers that have been only around for the last three or four years that have a clean sheet of paper to start off from.
And in terms of the capabilities and the optimization that automakers have been through over the last decade or so, some are much better position that embracing technology than others. But I would say really, every automaker is going through a big shift in their EE architecture by virtue of the fact that they have to start to think about electrification very seriously. And as that happens, there is a big opportunity to be able to simplify the car architecture to move away from Legacy architectures that were heavily Tier 1 driven, heavily MCU driven to more modern architectures that have bigger processors, much more software centric, a lot more integration.
But the fact of the matter is that every automakers journey through the transition is different, because every automaker has a variety of different constraints and complexities that they are dealing with some more than others. And so the ones that are having I would — are doing it a bit more deliberately and with a little bit more articulation are the ones that have a lot more complexity to manage versus others who, for example, might not have to worry about a combustion engine business, they may simply be electrification focus. So, it’s a whole lot more complex than what I can maybe compare it too in terms of any other technology transition that I’ve seen.
Brett Simpson
So let’s maybe start discussing the autonomous opportunity for Qualcomm. And I think the way you’ve positioned yourself here, you haven’t talked about robot taxis or trucking or you’ve kind of been quite specific, it’s sort of level two to level three, it’s passenger cars, and you have Arriver as a sort of building up that stack.
Can you talk a little bit about that transition from ADAS to level 2 plus? What does that look like, from a value perspective? How do you think about the business model that sits around Arriver? How do you how do you get paid for that software stack?
Nakul Duggal
So I think, maybe just for your audience, I think we’ve been investing in the ADAS business in the stack business, specifically, for about seven years or so now, with something that we’ve been focusing on in terms of doing a lot of core R&D, building a lot of core IP in that space. And also building silicon that understands and implements the foundational IP that is needed for AI for computer vision for being able to deal with the concurrency is with the thermal characteristics of what it needs to build a fulfilled ADAS system. So this work has been going on for quite some time.
We first started off with getting into the ADAS silicon business, because that was something that we could step into much more assuredly. The decision that we had in front of us about four years ago was do we organically build a stack? Or do we consider acquiring a stack to help leapfrog and accelerate our plans?
On the silicon side, I think we’ve actually done quite well because we had, in our first generation fairly modest ambitions. We won the GM business, we won the great Ford Motors business. Since then, from the silicon side, we’ve added a number of different automakers to the portfolio. But what we did was we acquired the Arriver asset, which initially started off as a commercial relationship and is now fully integrated into Qualcomm.
It brought us about 1100 people. It brought us four generations of computer vision deployed across a variety of different automakers, including Mercedes, including Subaru and many others. And that has been our starting point in terms of the integration of computer vision directly into Snapdragon, what we call the Snapdragon ride vision product. We are getting ready for deploying this in the late ‘24, early ‘25 timeframe in commercial vehicles.
And important to keep in mind that this is a fifth generation product. So there is a tremendous amount of experience in the team. It is a safety great product. It has been deployed across multiple vehicle lines already. And we are essentially modernizing it and obviously integrating it heavily into the silicon itself.
The other area of focus that we have in ADAS in the L2 plus L3 space is our partnership with BMW, which basically allows us to be able to build the full stack with BMW that we commercialize in ‘25, and then extend over to other automakers. So the strategy that we’ve had is to focus on a category of the market that we believe is going to be the sweet spot in terms of how widely that capability gets deployed, obviously, we’re going to continue to keep investing, to keep expanding the automation levels that we want to achieve. But we want to focus on a specific target that we scale up. And our customers are on that roadmap, we clearly are defining that roadmap and marching towards it. And it’s a space that I think requires a lot of patience, it requires a lot of focus. And I think so far, very happy with how things are going.
Brett Simpson
And just on that subject, I guess the competitive dynamic here, we’ve seen Mobileye go public of late and Nvidia announced their Thor platform. And there’s a lot of talk about hypervisors. And how you integrate some of these autonomy levels into digital cockpit. And some guys are a bit more narrowly focused, some guys have got a black box approach, but can you just sort of frame the competitive landscape? And how Qualcomm’s kind of winning business here over some of these rivals, that’s been in the market for a while?
Nakul Duggal
Yes. So let me let me try and break the problem statement down into two axes. The first axis is the digital cockpit space, which is something that requires a tremendous amount of scalability, because the tearing in this market has a lot of cost sensitivity, has a lot of performance requirements. And you have to be able to deal with many different software ecosystems that are all bespoke, every automaker wants to differentiate, every region is different. China needs something different from the rest of the world. And that is something that we’ve been able to, frankly, address very well, because it’s a space we entered into about seven years ago.
Today, we have most automakers on our platform. Then you have the ADAS business starting off from basic end cap going all the way up to L2 L3. That space has obviously Mobileye it has folks like ourselves, Nvidia going after that space. And if you start to think about how that space is evolving, it’s also moving in the direction of what are basic safety requirements that are going to become fairly mandatory, fairly commonplace across most automakers. And then where do you start to get into more advanced performance requirements, which may not be something that may see full attach everywhere.
The approach that we’ve taken, and obviously, different companies will take different approaches, depending upon their portfolio, but the approach that we’ve taken is to be able to build silicone, that is dynamic in terms of whether it is — whether the use cases, ADAS or automated driving or advanced digital cockpit systems, it’s the same underlying fabric. So that gives us a tremendous amount of flexibility in terms of the silicone that we’re going to bring into play. And it gives our customers a lot of flexibility because they don’t have to rethink the approach that they have to take.
What it also allows us to do is to be able to mix and match ADAS or cockpit, especially if you want both to coexist, especially for certain tiers in the market.
We are able to take approaches where you can obviously keep the two domains completely independent, scale them independently. And that allows you to be able to deal with much higher levels of scalability, much higher levels of autonomy or cockpit experiences. And more importantly, also deal with maybe more conservative automakers that are not ready for the domains to be merged. So as I think about this, if I compare ourselves to, what Mobileye has to offer, I think there the approach becomes one of scalability, but also multi domain.
I think with other competitors, it really kind of comes down to, do you have the ability to be able to scale across so many different automakers given that these are all bespoke programs, lot of software complexity, many balls in the air all at the same time. And that makes us feel pretty good about the fact that, the approach is working. And it’s bolstered by the fact that there is a large amount of design win pipeline that we can point to, that shows us how automakers are thinking about this.
Brett Simpson
Maybe just switching gears a little bit to connectivity. And I guess when we look historically, at connectivity in the car, it’s been rudimentary, I would say, I mean, really kind of E-call, E911 type services. But what we’re getting into now with over the air software updates, and partitioning and such, the infotainment systems, et cetera. Clearly is going to be much more — a much better use case, let’s say.
Can you talk a bit about like, 5G in the car? When do you start to see this inflecting? And what does a typical connectivity platform look like by the time we get out to 2024 2025, and we’re starting to integrate some of these capabilities?
Nakul Duggal
5G is infecting already. I think most — I would say almost every new design that we’re seeing is 5G first, there may be some designs that may be 4G, because they happen to be in markets or in tiers that perhaps have kind of at a lower level of the spectrum in terms of affordability, but pretty much every new design that we see today is 5G.
5G in China is being deployed at scale. We just made an announcement with Ford recently, they deployed 5G in their F150 heavy duty trucks. So this is starting to happen really across the board. BMW, we announced earlier this year with their dual SIM dual active solutions.
I think if you think about how automakers are thinking about this connectivity in general, there is a certain quota of airtime, there is a certain investment in airtime that automakers are setting aside for housekeeping functions for the car over the air updates, diagnostics, the ability to be able to go manage the vehicle is a platform.
And then there is obviously consumer usage. And part of it may be subsidized part of it may require the automaker to be able to pass on those costs to the consumer, or allow the consumer in some cases to be able to use their own SIM card connect it to the vehicle, many different business models are possible there.
But, what we see compared to four or five years ago, attach rates are going up really high. Most of it is 5G. And but as you know, for us, this means not just the basement attach, but also the RF front end attach. The other piece that we’re seeing is automakers are now protecting for V2X, because that is becoming important to have in place because as mandates start to kick in the design has to be there has to be supported. So that is obviously a positive.
And then the attach rate of Wi Fi for hotspot usage for in car connectivity usage alongside 5G is also gaining. So for us this entire category becomes a very high value category and obviously independent of the other two domains that we are spending time on.
Brett Simpson
Interesting. And so you you’re seeing 5G in China really, really now and I guess the Attach rates in other markets, when you say most of the new designs, we’re really talking about new designs today are probably like a 2025 rollout. So maybe by middle of the decade, you’ll start to see this take effect in a material way or?
Nakul Duggal
We are actually going to start our second generation of automotive 5G in the second half of ‘24. Our first generation of 5G is actually done. We started deploying those in 2020, late 2020 and all the way through 2023, early ‘24. We’ll be doing our first generation. So, we are actually starting to move very quickly into an all 5G world.
This previous generation was probably a mix of high end 4G and 5G it also has to do with the readiness of the networks because not every single market had 5G deployed at scale. So automakers decided between the tiers of vehicles and the readiness of the market how much for you to pick how much 5G to pick but that is not changing very quickly. It is essentially falling the transmission of networks which are more and more arbitrary.
Brett Simpson
And I guess given the limited modem capable of very few people have 5G today and in quote Media Tech is probably the only one that brings to mind, maybe excellence to some extent. But I mean, essentially your market share your implied market share in this auto space for connectivity is going to be very high.
Nakul Duggal
I think it’s a market that we have been involved in from the very beginning, we were involved in really creating this market. So, and it’s also a market where customers trust the experience that they’ve had the credibility that we have built. It’s a market that your product really has to work over very large set of networks, you need a lot of experience in dealing with complex problems. So yes, I mean, it certainly helps to have the history.
Brett Simpson
Maybe if we switch gears and talk a little bit more about digital cockpit, where the digital chassis, I think, as you lay out a lot? Can you talk a little bit about the business model for Qualcomm and we hear a lot about car to cloud, maybe you could explain what it is because it doesn’t feel like that’s been sort of laid out for investors in terms of like, what does this mean? And if we are talking about services, business models, how does Qualcomm insert themselves in that value chain with all the technology that you’ve got it’s relevant for that type of domain?
Nakul Duggal
Yes, that’s a very good question. I think the approach that we’ve taken is, first of all, it’s important to keep in mind that this space has traditionally been extremely fragmented, there is no one size fits all across automakers. Also, the approach that different automakers have taken is also unique. Some want to be able to bring more well defined ecosystems into the vehicle. Others want to be able to bring in or build their own ecosystems put together a variety of different solutions. Yet others in different parts of the world will pick and choose from building their own app stores, bringing their own search bringing their own content.
The approach that we have taken first of all is, excuse me to be to be very flexible in terms of how we support these different software ecosystems. Secondly, what we have done is as part of car to cloud, we have connected the platform via API’s to our cloud platforms that allow us where the business model fits in where the automakers are supportive, where we can actually offer additional services beyond simply enabling the chipset and the platform software.
So what might some examples be, there are a number of examples in terms of managing performance for the chipset over its life, where you start off with a certain performance, and then you upgrade over its life. So there are certain customers where we are able to have business models that allow us to go to that.
We are working with in some in some commercial opportunities with car to cloud, where we are offering fleet management services that are built on top of our infotainment systems.
We are adding cloud API’s to our solutions that allow us to offer data analytics. And we’ll share more at CES coming up in the next month or so in terms of providing automakers the capability to understand what type of usage is going on what features are being used? And what more could be used or what could be turned off? Where could you invest more versus less? So, this is one of those things where you have to be able to have a number of different enablers and hooks in place to be able to create those services. And then the opportunity there is to then figure out how to monetize each of these. But there is not a one size fits all strategy that’s going to work the same exact way across multiple automakers.
Brett Simpson
And then from a digital cockpit perspective, can you maybe talk a little bit about the market share ambitions for Qualcomm? Because I guess, we’ve seen a lot of wins in this space for Qualcomm and just keen to understand like the competitive dynamic and how you feel your position over the next sort of four or five years as business becomes a much bigger business?
Nakul Duggal
I think we feel good about the business because we have invested in the roadmap, first of all. We’ve brought for generation of silicon in about six years. It’s a highly scalable portfolio. We are now building custom chips, we are building chips that have multiple personalities. So like I said earlier, these are chips that could do ADAS that could also do cockpit solutions. We are very broad in the software ecosystem that we support the virtualization that we support. We offer a tremendous amount of integration of functionality. So the focus is always on making sure that we are able to optimize the overall bill of material of the system.
The chips are all safety grade chips that have SLD subsystem, so I kind of feel like we’ve earned the right to be a very strong supplier in this space by essentially investing and delivering, and customers are very satisfied with what they’ve seen. And they keep coming back. So I kind of feel like this is one of those spaces where automakers are realizing that there is no reason to change out your supplier every three years. Instead, the focus needs to be on building a strong foundation, evolving software, looking for more differentiation. And we’ve been able to find that stability in the foundation that we’ve built with the digital cockpit business.
Brett Simpson
I wanted to shift gears a little bit and talk about the pipeline. It’s obviously doubled in the last 12 or so months. Can you maybe help set the framework in terms of the pipe — typically, design wins last how long? How long does the pipeline sort of like playout for when you look at that 30 billion number, and how much contractual commitment is behind those pipeline numbers that we see?
Nakul Duggal
So everything that we include in the pipeline is contractually committed, these are all based upon supply agreements that are in place. In terms of the timeline, I would say, most of it is through this decade, there is very little that moves into the next decade. And I would say most of it is probably through ’28 ‘29 type timeframe. That’s probably, yes, I think that would probably be an accurate representation.
Now, obviously, the size of the pipeline changes depending upon the complexity of the content, we are doing a whole lot more ADAS, we are doing higher end ADAS, we are going higher end cockpit systems, and that obviously, is making the pipeline much more biased towards fairly a lot of ADAS content.
Brett Simpson
And when you look at that pipeline, would you say the majority of it was connectivity? Or is it more digital cockpit, is the autonomous the largest part, any way you can sort of frame how that might split out would be would be very helpful?
Nakul Duggal
I would say, in order, I would say autonomous cockpit, and then connectivity.
Brett Simpson
Got it. In terms of size, okay.
Nakul Duggal
And then one of the, there is a fusion happening of autonomous and cockpit in some systems. So those also start to blend.
Brett Simpson
Yes. And that’s where you get there, the virtualization, I guess, as well. One of the areas that you didn’t mention in your pipeline today is this sort of architecture shift towards zonal controllers? Can you maybe just sort of lay that out in terms of the vision around this opportunity? How big than do you think this architecture changes with carmakers?
Are they committed to it? I guess we look at a lot of the traditional 32-bit microcontroller type features today. And we can see why this is going to consolidate into zones, if you like, but sort of help frame this for us in terms of where this development might be and how nailed down do you think the transition to zone is going to going to be?
Nakul Duggal
Yes, I think a two-part answer, I think the first part would be, there is certainly a transition away from smaller microcontrollers, to, let’s say an Uber controller that is able to integrate a lot more functionality just because the complexities higher, and it’s better to be able to centralize that.
I think on the zonal controllers, I think it depends upon the automaker and their car architectures, automakers that have more monolithic architectures that are, you know, that have lesser amount of SKU complexity. I think it’s easier to get to a zonal controller architecture because you can essentially embrace that and start to go deploy solutions that way. I think automakers that have a lot more SKU complexity, lot more architectural complexity. I think it’s a bit harder, because it really comes down to what functions you can centralize in zones. What do you still have to have as passed through and transported all the way to a central controller?
So I think different automakers are going to experiment with what is the right timeline for them to really make a full transition towards zonal controllers. I do think that when you pick electrified architectures in certain tier of vehicles or automakers that have a fairly straightforward architecture in place. I think that’s where it’s all controllers will perhaps take off sooner.
Brett Simpson
Interesting. And so maybe that’s more second half of the decade when this becomes more of an opportunity. And to what extent do you think Qualcomm can address this?
Nakul Duggal
Yes, it’s certainly something that we are always looking at. And, for us the very start from is what is the differentiated IP that you need to have in place? What is the complexity that you have to go design for? And then finally, what is the size of the business that we can really go after. It’s an area that we are obviously looking at very carefully. And right now, there is no zonal controller in the pipeline.
Brett Simpson
Okay. And I wanted to get your perspective on this whole kind of software defined vehicle that gets talked a lot about in the industry? And what does this really mean for chip makers in your mind, I mean, we hear a lot about a lot of things in the car over the decades have just been patched on. And now we’ve got to sort of start again and rewrite code in a different way for the industry. But can you just sort of help sort of parse that out? And what would the implications be for somebody like Qualcomm?
Nakul Duggal
Yes. So I think — there are at least two things that kind of occurred to me when I think about SUV. I think first, first of all, automakers have not had a standard blueprint around which they develop software. Everybody is either relied on Tier 1s, or done something in house or some combination thereof. But there is no one standard way of doing any of this.
So there is, in some sense, a lack of a standard in terms of how do you do software development. And it resembles some of the challenges some of the complexities that exist in the enterprise software space as well, when you had a lot of different enterprises that were trying to go to the cloud, but had a lot of on-prem or a lot of local implementations, they were all very bespoke.
So, there is the need to kind of move towards the standard. And then I think the other piece is the cloud is becoming more and more important in terms of how next generation development is being done for the car, because you have to be able to have some level of parity between what is happening in the cloud versus what’s happening in the edge.
So, right now, at least the focus is on how do you create some orchestration between doing development in the cloud, doing some level of virtual platform development, and then have the ability to be able to implement that on a bench system, and then hopefully implement the same exact thing in a car.
But the level of complexity is very high, because the car in some sense is not like any other ad system. The level of complexity, the safety requirements, the real time nature, the heterogeneous nature of the subsystems involve, the various supply chain dependencies. So I think frameworks being created, and there are architecture being created in terms of how to go simplify, it’s going to happen in phases.
And I think it’s going to require a lot of careful planning by automakers in terms of how to orchestrate this. But it is not necessarily, I would say, a single approach or a single template that you can just go apply across any automaker, because it really comes down to making the transition from the car architecture that you have, to the new car architecture that you wish to get to. And it’s not like there is a common car architecture that every automaker is going to arrive at simultaneously.
So, for us, it is actually an opportunity to really be able to get involved in a lot of upper layer software that sits on top of silicon and even cement for the partnerships that we have with automakers.
Brett Simpson
Interesting. And on that theme, is there a monetization point for Qualcomm if — you mentioned V2X maybe gets built in but not launched or the Arriver software model. How do you get paid for 5G over the life of the car, is there ways where Qualcomm maybe looks at charging customers or the fee structure in this industry slightly differently than we’ve seen traditionally from Qualcomm?
Nakul Duggal
I think one of the reasons to focus on software so much, is exactly what you just said, Brett, which is that, you have to be build the chassis — build a digital chassis that is accessible through API’s, you have to be to have silicon that has more headroom. Built a lot of software products that have to evolve over their life, the stack is a great example. It will evolve over its life, you will keep adding more features into the same silicon. And you can charge multiple times over the life of that stack.
V2X is another great example where it sits dormant until the requirements get turned on. So, the approach that we’re taking is to be able to build this platform, such that it has the hooks in place for us to be able to do more with it. And it is something that will continuously evolve. Because the next generation silicon we’re building is going to have a lot more performance. Not everything is going to be needed on day one.
But as the platform evolves, automakers will need to do more with it. And that creates the opportunity, whether it is something that turns into a subscription model, or another feature that is deployed that you charge for once, or engineering services, whatever that might look like.
But the relationship that we are building with the platform that we are deploying is no longer really a one shot hardware relationship. We are building relationships that continue over the life of the platform.
Brett Simpson
Okay, that’s clear. Nakul I think this is probably a good time to transition to investor Q&A. So I just wanted to bring in my colleague Blair, to help manage through that process. And Mauricio is also online. So, Blair, over to you with the Q&A.
Blair Botha
Yes, sure. Hi, Nakul. We’ve just got a question here on mapping for ADAS, and also around the collection of the data there. The question is, how are you doing this? And what is your strategy here? And then are there any problems I guess, with the sharing of the mapping data?
Nakul Duggal
I’m sorry, I may have missed the first part of the question. How are you doing this?
Blair Botha
For ADAS, and you’re mapping for ADAS and the collection of data there?
Nakul Duggal
Yes. So mapping for ADAS is a category that we’ve actually been focused on for quite some time, we’ve been focused on localization, we’ve been focused on making sure that we are able to localize ourselves in a grid, and then use that localization for adding additional levels of accuracy.
There is a large amount of data that is available through a number of sources through partners’ sources through automakers. It really depends upon what level of automation, whether you’re talking about L2 or beyond L2 plus L3.
The type of data that we rely on is speed data, road curvature data, really be able to get a horizon in terms of being able to manage the overall performance, the overall driving experience better. And we have between our between systems that we have built ourselves that we are further investing in and partner systems, we are able to build that into our stack. It’s a category that we are continuing to invest in. And we will share more with you in terms of what that starts to look like as a product in the coming weeks and months. Nothing additional to share right now.
What I would say that I think becomes quite important is, there are many different ways to get to localization. And I think it’s a combination of using computer vision, using crowdsourcing, using large amount of data that is actually available to be able to go create that grid that you need to have in place that is needed, depending upon the type of automation you have in place.
Blair Botha
Okay, perfect. Thank you. And then we’ve got one more quick one. And this is just around ASPs. And how to think about the trajectory of ASPs supposed to be thinking about the domain controller. And if we look, I think Brett spoke about the zonal controllers as well. So they just asking for a bit more color on how to think about ASPs in this business?
Nakul Duggal
Yes, we haven’t really broken that out more issue. Any thoughts that you have on that one? I don’t think we’ve shared any.
Mauricio Lopez
Yes, can you give a generic update in terms of content per vehicle on a base car, it might be as low as $200 per vehicle. A premium type car would be about $3,000, and that was in the Auto slide presentation. So you can reference that if someone who scored our webpage will have it there.
Blair Botha
Okay, great. Thank you. Great…
Brett Simpson
I just had a couple of follow ups, maybe one on the returns within the auto business. I think, obviously, there’s a lot of R&D reuse that goes into autos that you’ve been developing elsewhere are sharing IP, if you like, across different divisions.
And I guess you don’t have the same concentration of customers that maybe you have and like smartphones. So how should we think about well, the R&D investment that’s needed? How do you size that? And how do we think about the returns that the business can deliver over time when you scale up the auto division?
Nakul Duggal
Yes, so I think, the investments that we are making are mostly in the following areas, we are obviously missing ADAS, which is investing in both silicon and stack. We’re investing in AI and computer vision. And then, of course, we are investing in silicon development, because these are all SOCs that are very much customer associates for auto.
But keep in mind that the same technology and the same SOC is even are used in other parts of Qualcomm’s businesses as well. So it’s not like these only have one home. I think we’ve shared that we expect that the business starts to become accretive in ‘26. After most of the ADAS costs are behind us, but right now, the more that we are in is in investment mode for ADAS.
Brett Simpson
Okay, that makes sense. And you’ve mentioned your 1100 people in Arriver. Is that right?
Nakul Duggal
That’s right. Arriver brought over 1100 people. Yes.
Brett Simpson
Okay. That’s helpful. And then maybe just in terms of China. I guess when you look at the car industry, and the decoupling that’s going on, generally between the U.S. and China, a lot of the customers that you’ve been announcing the VWs et cetera. They’ve got big businesses in China.
How do you think about the auto opportunity in China specifically? And I think you’ve talked about great wall and other customers there. But is this going to be somewhere where geopolitics might have a negative impact on your ability to scale up there? Or have you seen anything there that concerns you, especially when there’s local players like Horizon or our Black Sesame or other folks that are trying to do some of the areas that you’re working on today?
Nakul Duggal
I mean, clearly China is a pretty dynamic space. And there’s a lot that needs to be considered. I can speak to the business that we are in the middle of. We have very deep partnerships with most Chinese automakers, domestic automakers, JVs, new EV companies, like I mentioned earlier, 5G V2X are areas that we have very deep partnerships on. The cockpit business, we have very deep partnerships on a number of automakers are using the Snapdragon cockpit solutions.
ADAS we are continuing to expand our footprint, since the acquisition of Arriver, we have one additional business on the Arriver platform in China. So, I kind of see this to fall, of course, there is a lot to keep watching out for. At the same time, the size of the market is so large, and the pace at which the Chinese automotive industry is modernizing is so rapid, that we really find ourselves in a position where really all of the technology and the products that we have are finding homes within our Chinese ecosystem, more and more, in fact, I think ADAS is a very important area of focus for us in calendar ‘23. Because we have now completed the acquisition, there’s a lot of opportunity for the products that we are building.
Brett Simpson
Okay, good. Well, I think we’re pretty much out of time here. So, I think we’ll call an end to the session, but great discussion, Nakul, really appreciate your time and thanks for dialing in, certainly learned a lot about the whole car space and sounds like it’s going to be a very dynamic 2023.
Nakul Duggal
Thank you very much, Brett.
Brett Simpson
And Mauricio, thanks very much for dialing in.
Mauricio Lopez
Yes, thanks, Brett.
Brett Simpson
Much appreciated. Okay, that ends the session. Thanks for everyone for dialing in.
Be the first to comment