Qualcomm: Ushering in a New Era of Mobility

Disclaimer: The Futurum Tech Webcast is for information and entertainment purposes only. Over the course of this webcast, we may talk about companies that are publicly traded, and we may even reference that fact and their equity share price, but please do not take anything that we say as a recommendation about what you should do with your investment dollars. We are not investment advisors, and we ask that you do not treat us as such.

Transcript:

Daniel Newman: Nakul, it’s great to see you here at CES 2024. How have you been?

Nakul Duggal: Great. Good to see you again, Dan.

Daniel Newman: Yeah, it’s good to have you here. It’s always good to catch up with you. CES, huge moment in every automakers world right now. You see pretty much what, 130,000 people converge and if not, the top thing that people want to see, it’s what’s going on in automotive and I would say the trend line, AI, AI, AI, AI. But in automotive you’re talking about autonomy, you’re talking about infotainment, you’re talking about digital cockpit. AI is going to change at all. So this year has to be another big year for Qualcomm. Tell me a little bit about what the company is talking about here this year at CES in terms of automotive.

Nakul Duggal: Yeah, I mean, CES has been an automotive show for us for a number of years. This year it’s Snapdragon digital chassis with a GenAI twist, and that’s what you’re going to see a lot of. But no, I think for the company, AI has obviously become very central to a lot of things that are going on, PCs, phones, and absolutely automotive. We’ve been on a great journey on the automotive side of the house because we’ve been building this business, this platform for a number of years. And automotive is as an industry, going through a fairly complex set of changes because the pace of innovation, especially as you look at how China is moving, especially as you look at how the EV transition is happening, there is just a lot that automakers have to accomplish to be able to be competitive. And for that they need platforms, they need partners, they need the ability to be able to move fast. And I think we’ve built a platform that allows them to be able to go do that. And we are seeing that in our numbers. We are seeing that in the wins that we are getting. We are seeing that in the trust that our customers are showing to us.

Daniel Newman: Yeah. So this year, you’re sitting in the studio with me. Last year we did an interview. It was you and me, sat outside of this Snapdragon demo vehicle. Very cool futuristic vehicle. I think you and I have had a chance to sit in it a few times around the world now. We’ve sat here, we did it in Germany, and every time we sit down, things have proliferated and moved forward a little bit.

Nakul Duggal: Yeah.

Daniel Newman: What’s kind of progress looking like in this industry over the last, I feel like about two years ago you really, two, maybe two and a half, I think you’d probably say longer, but the last couple of years have been huge for Qualcomm. We saw your design pipeline go from a little over 10 billion to almost $30 billion. I’ve seen design wins from just about every major OEM, strength in China, strength here domestically. But what are the key technology drivers that you say that are creating this momentum and what’s kind of happening even in this last year that’s really notable?

Nakul Duggal: Well, I mean looking back, it is pretty straightforward and logical, but the strategy that we put out for ourselves was we’re a technology company and we develop a lot of different technologies. Some of these technologies are embodied into products that can be used as is by a car. Others need to actually be adapted for the car. So we took a big step back and we said, what’s our starting point? Where do we really need to start? And it all came kind of down to the car is going to change as a platform, it’s going to be a mobility platform, it’s going to have a lot of different use cases. It’s going to be innovated upon by many different customers in various parts of the world. So we have to build something that is very versatile, scales up, still embodies all of the key technology that we have, but it has to be something that has to work for the automotive industry.

And there are two things that are very important, safety and quality. And safety and quality, if you don’t have that, this is not a business that you really want to be thinking about very seriously. So we made those decisions four or five years ago and the semiconductor crisis hit, it became very clear to the automakers that they had to make sure they had partners who they could rely on, not just for technology and support and features, but they’re there when they need them. We have prioritized automotive all the way up to the top in terms of which industries we focus on. So today we find ourselves in this situation where we’re basically focused on two major areas of differentiation. One is connectivity, the other is compute.

And if you think of those two key blocks, computation is becoming key for the car because it is becoming this product that keeps absorbing more and more technology because the use cases are endless. And of course connectivity is something that you absolutely need. On top of that, we’ve been building a ton of software. We’ve been building a ton of applications. For example, we are building an ADAS stack. Qualcomm is getting into the business of developing an autonomous driving stack, a safety stack, something that you would say, what’s a smartphone company doing building a stack?

Well, we are no longer just a smartphone company. We are also going kind of in that direction. We are engaged with various ecosystem partners from SaaS, you’ll hear about Salesforce, you’ll hear about something we’re doing with J.P. Morgan to be able to go open up the platform for SaaS partners and automakers to be able to bring differentiators onto the platform. The cockpit is changing completely. You will see from us how GenAI is actually already on our platforms for 2024. We have silicon that is in commercial vehicles, in production vehicles in 24 that is ready for GenAI. It’s a software update and that’s done. So for us, this is one of those things where you really have to be able to look at all of the assets that you have in place, have a long-term strategy, and then have the customer base trust you to be able to go implement that platform.

Daniel Newman: First of all, I’m glad you pointed out the diversification. It’s important you reiterate that often. I can say as an analyst, I’ve been talking to both the street and to the market about Qualcomm’s diversification and I think the message is starting to land. I think just last week I was on CNBC and I was talking to them. They were actually asking about your XR and the innovation, which I by the way, expect some of that XR to find its way into the CAR. Like what I did there? I practiced that one earlier.

Nakul Duggal: Very clever.

Daniel Newman: Yeah, it was good. But I expect to see that all these technologies start to find, you could start to see some of the AI, PC capabilities becoming part of the GenAI inside the car because an NPU is going to find its way into the automobile and you’re going to start to see, and then you’re talking about SaaS and applications, the world’s converging. And I think that’s, you guys call it like the connected intelligent edge or I’ve heard Cristiano refer to it that way. It’s all coming together. And so I’m glad you caught that. I’m glad you reiterated that because it’s really important. But I also wanted you to reiterate a little bit about that GenAI. I heard you say that.

Nakul Duggal: Yeah.

Daniel Newman: Most popular trend in 2023.

Nakul Duggal: Yeah.

Daniel Newman: In fact, I didn’t have to talk about anything else to the media the entire year. Thank you. But you said GenAI is finding its way into the digital cockpit. Can you just talk a little bit more about what that experience may look like Nakul?

Nakul Duggal: Let’s go back a couple of years. When we started to think about designing compute for the car, ADAS was very important for us. So we said, okay, you need to be able to build AI for ADAS and should we build two completely different pieces of silicon or should we just build one common piece of silicon? We can use it for cockpit, we can use it for ADAS. We went with the latter decision. So the platforms that we built for the car are not separate between what you put into ADAS versus what you put in the cockpit. What that allows us to do is to essentially have the hardware, the headroom that you need to be able to bring in something like GenAI, it’s already built into a cockpit solutions. It has been for like three years now.

So you mentioned the NPU. We over dimension automotive silicon because A, we have to be able to support a much longer life cycle. Things will change, new features will come in that’s going to need more performance. But very importantly, the NPU that you highlight and other AI-related blocks, we already have that in the silicon. It’s already there. In cars that will be rolling out in 24, those cars have that capability built in.

So then the question is, okay, how do you use GenAI in the car? And I’m sure you’ve covered this across so many different companies in so many different use cases, but the way that I think about the car is as follows. It requires more than one person to be using it. The people that use the car could be different people, there could be multiple people present. The context in the car is very important because it is obviously a device that takes you from A to B. So that becomes a necessary input into what we’re going to go do with GenAI. The use cases are very broad. It’s not a productivity use case. It’s not a use case that is tied to the things that you might consider as, oh, I’m drafting an email and please comment. It’s not those things. You will see a few things that are super interesting.

For example, we have virtualized the service manual for the vehicle. It is indexed through GenAI and you can have a conversation with your car through a network that is deployed in the vehicle. We’ve taken LlaMA 7 billion, that’s running locally and you can now have a conversation and you can ask the vehicle questions about giving context. Hey, I see a diagnostic code, I see this alarm light turn on. What does it mean? And you get a response that is picked up all locally. There is no cloud involved in something like this. That’s one simple example.

Daniel Newman: How much time I tried to find the temperature, VMS, my daughter’s car because every car is so different. I mean just this weekend I was literally in the menus for like, and I was going to YouTube trying to find demos and it was like even by year it was different. So you’re saying you could basically go to it, use natural language processing, talk to the car, say, I see the sensor. Tell me what it means.

Nakul Duggal: Dan, the part that is very interesting and all of us can relate to this because we all experience these things every day. The idea of being able to understand how a consumer experiences the car as a product from the edge is something that automakers have always struggled with. What do people like, what do they not like? What can we change? If we can change it, how do we go change it? And one thing that is changing, first of all, because cars are becoming so software-first, so digital. Secondly, there is so much compute capability available. So you’re not designing the car for just enough hardware. There is actually plenty of room to be able to do more. And then when you bring intelligence from the edge and make it available back into the cloud, and GenAI is a great tool in terms of how to go do that, it changes the way that you roll services out.

And you will see that there are a number of automakers that are very sophisticated technologically. They’re very sophisticated in terms of how to go build a software product. These are things that are going to come very naturally to them. They’re just going to think of the car as a digital product that has so much more potential. And then of course there are the traditional OEMs that have to figure out how to make this an advantage. We find ourselves situated very well because we are designing platforms that cater to both ends. We are learning very quickly from the customers that know how to move the needle, that know how to push the envelope in terms of how do I bring something that nobody else can bring as quickly as possible. Those are who we are catering, who we are designing our platforms for. And then of course we work with everybody else to make sure that we kind of accelerate this. So I’m super excited, especially because the timeline between when the technology can actually be converted into a product, that we have shortened. We will see that in 2024 in cars.

Daniel Newman: So here’s an opportunity for the industry. I believe the design cycle’s too long. I don’t believe people have the patience. I won’t speak to specifics Nakul, but I have a car that’s a 2022 that actually shared the electronics of the 2012 version of the same car. I love the car, I bought it, but what I’m saying is the electronics, it’s an abomination that you actually have 10-year-old and the cockpit was not updated. Now if you’re basically telling what I think I want to hear and what the market wants to hear is even as cars could minimally make maybe physical designs and people have kind of gotten used to this with maybe like the Teslas, but other OEMs are going to be able to follow that path of being able to dynamically overhaul cars without maybe having huge redesigns both inside, outside, cockpit, infotainment. Even the software of the ADAS and the programming, cars can be completely revolutionized every year, every couple of years. Is that where you see it going? Is this cycle going to get much shorter to enable manufacturers to be more innovative and more quickly?

Nakul Duggal: I mean, clearly it is not a technology problem. Mostly the problem has to do with what kind of legacy you come from, what dependencies you have built into the way that you design a platform. How separable are they? Can you build a platform that is faster moving and leave the other platform behind? Are you a big company? Are you a small company? How much sharing you did? If you look at some of the very nimble automakers that we work with, they’re moving in the cadence you’re talking about. Every two and a half years, it’s a brand new hardware upgrade and the previous generation software works on the next generation hardware, vice versa. It is looking very much like what you might expect in consumer devices.

Now obviously moving onto an EV platform makes that straightforward because it’s a cleaner architecture. It’s a newer architecture. When you’re dealing with automakers that have to manage both and that have built architectures that obviously span both types of powertrains, there is legacy, there is complexity. So look, I mean, I think what you are asking for as a consumer is not lost on any automaker. The question is how quickly and with what approach do you get there? We are seeing all kinds of approaches. We are seeing automakers cordon off certain regions where they want to be able to move faster and do unique designs. We are seeing automakers pick some tiers of their vehicles and just design them differently. But everybody’s going to have a slightly different approach to it because it is more than just technology that comes,-

Daniel Newman: And a quick kind of final question here, Nakul, is the alignment of kind of proprietary IP from the automaker and the custom sort of stack and capabilities. I call it the kind of box, it’s not a black box. You guys have a building blocks approach, enables them to keep kind of what’s unique, quickly pair it with the Snapdragon platform for automotive. Is that what you’re looking to do, is to basically enable all these OEMs to be able to have their proprietary but build quickly so that they can stay custom but not have to go through all the pain that they’ve historically gone through to be custom?

Nakul Duggal: Yeah, that’s actually a great question. A lot of the new conversations that we are having with automakers are not only what you described, which is we provide building blocks on top of which they develop theirs, but where they would like to influence our building blocks or where they would like our building blocks to be modified slightly, but just for them to be created very uniquely for their requirements. We’re getting into conversations where we are licensing some of our building blocks to automakers and some of those building blocks may actually be semiconductor agnostic because they want to be able to get a faster start. They want to be able to mix what we have with what they have and create something new from that. So the business is moving in this direction where we are becoming true partners into what it will take for this chassis to change.

Because fundamentally, keep in mind that automakers have limited amount of, the traditional automakers especially, they are learning all of these new skills very quickly, but there is still a lot of work ahead of them in terms of how to become very adept very quick. And I think that’s where we feel like we’ve been able to actually make a big impact with the strategy that we’ve had.

Daniel Newman: Nakul, thanks so much for sitting down with me.

Nakul Duggal: Absolutely. Great to see you again.

Daniel Newman: Great to see you.