Qualcomm’s Role in Super-Chips, AI, and Helping Spur Android OS

We chat with the company's general manager of automotive about its outsized role in the swiftly evolving vehicle electrical architecture landscape.

Writer
ManufacturerPhotographer
nakul duggal headshot 3

At the 2025 IAA Munich auto show, we had a chance to sit down with Nakul Duggal, Qualcomm's group general manager, automotive and industrial & embedded IoT. He's the man steering the direction of the massive global technology company's efforts in automotive electronics, driver-assistance systems, AI, and everything else that’s so rapidly changing automobility as we known it.

MotorTrend: Tesla famously built its own silicon chips to run the software it wrote to run its cars. At the Shanghai auto show, Chinese OEMs Nio and Xpeng announced they’d designed their own “super chips.” Does this trend threaten Qualcomm?

Duggal: “I haven't seen the impact yet. If you think about the change in car architecture over the last decade or so, we've gone from a [distributed] architecture to central compute. What that really means is the ability to deploy very complex, advanced semiconductors, running complex custom software—not generic software, on that silicon for specific vehicles that are designed for very custom, well-defined workloads.” He went on to note that very few companies have the capability to design and deploy their own silicon, software, and applications, noting that It’s not coincidental that the Chinese OEMs designing their own chips have roots in mobile telecommunications—like Qualcomm.

“I think for us the big win is that this transition from a distributed to a central-compute architecture really consolidates the complexity of the car into some larger, more heavyweight, more high-performance SoC [system on chip]—something that we are really good at.”

MT: Is the future super chips, or will you always offer a range of chip sizes and capabilities?

Duggal: “You have to, right? This is not a one-size-fits-all market. It's a very competitive market, many tiers, many regions, many generations of cars that you support, all at the same time. So you have to meet the customers where they're at, and so you need a sophisticated, deep portfolio.”

MT: Are you able to use chips from the other markets you serve in automotive?

Duggal: “We are fortunate at Qualcomm, because we operate across many markets. As we decided to get seriously into automotive, we decided to design technology for the car in a broad manner—everything from camera technology, AI technology, safety technology. It all needed to work reliably in a much harsher environment.” He went on to note that WiFi and modem chips do the same thing for any market. These just need to be ruggedized for automotive quality and robustness to support longer life cycles.

“For other domains, you need to build a custom chip: Safety, automated driving, cockpit. These need an underlying safety fabric. These chips take advantage of the technology portfolio that we have but are uniquely built for the car.” Qualcomm works with 75 OEMs, receiving feedback from everywhere, so it’s well qualified to consider cost pressures, AI tiering, affordability, software support, generation-over-generation support, compatibility between tiers, ADAS and infotainment, as regional differences. “What do you do with Google in the U.S.? What do you do with the China ecosystem? You can only deal with this complexity when you have the right foundation.”

MT: What are the challenges as we transition to zonal and central compute?

Duggal: “Dealing with complexity from a safety, a software, a single point of failure perspective. It takes a couple of generations, but we have built five generations of silicon in nine years. So you just have to keep building and learn fast. You need to own a lot of technology.” A significant challenge has been the transition in onboard networking, from “MOST bus, FlexRay, CAN, LIN, and now things have started to move towards Ethernet for any traffic you are carrying. Then there are display interfaces, camera interfaces—those are the three main traffic bearers inside the vehicle.”

Duggal then admitted that Qualcomm doesn’t build networking chips, but other suppliers do, and these are key to developing zonal architecture.

Shifting to Agentic AI, another competitor here at the IAA is offering an agent similar to the one Qualcomm announced with Google, but partnered with Microsoft. Is this going to unfold like VHS/Beta or Apple/Android?

Duggal: “I don't think you want to think of it like that. Think of the cockpit as something that hosts many agents. The partnership that we've announced is to host the Google agent for things that Google is able to offer to you. If you had to host for a very specific set of services, an agent that deals with Outlook that can be hosted as well. Just go to Azure, work out a deal with them.” He assured us that there’s none of the inherent format differentiation that distinguishes PC/Mac or Apple/Android that would exclude users of one from accessing the other. Then he dropped this nugget:

“About 10 years or so, we actually brought to Google's attention that there's a big opportunity in actually delivering Android and various other Google type services into the vehicle. Prior to that, they were looking at maps and things like that. But this whole idea of the operating system inside a car to go drive the cockpit was something that was not top of mind.” He noted that the Google partnership has been terrific, because together they’ve been able to better tailor Google’s broad suite of offerings contextually adapted for in-car use.

“I have a phone. This is a foldable phone. The app knows it's a foldable phone, and it renders content accordingly. The car is the same way. If I have an app that understands the layout of the car, it's going to render the right way, and I'm going to have a certain experience.” He went on to outline the ways AI is helping render that experience in the car, noting the strides that language models have made in allowing cars to perceive languages and accents, unlocking the ability to control many features and functions with diminishing frustration.

And as for why Google, not Microsoft, “The reason behind partnerships with the companies like Google and Google Cloud Platform is we want to be able to bring to the platforms that we are building these types of capabilities that allow the car to become as relevant to the consumer as their smartphone is, as their PC is, as their tablet is, because the car is a much more complex product than some of these other products.” Amen.

I started critiquing cars at age 5 by bumming rides home from church in other parishioners’ new cars. At 16 I started running parts for an Oldsmobile dealership and got hooked on the car biz. Engineering seemed the best way to make a living in it, so with two mechanical engineering degrees I joined Chrysler to work on the Neon, LH cars, and 2nd-gen minivans. Then a friend mentioned an opening for a technical editor at another car magazine, and I did the car-biz equivalent of running off to join the circus. I loved that job too until the phone rang again with what turned out to be an even better opportunity with Motor Trend. It’s nearly impossible to imagine an even better job, but I still answer the phone…

Read More

Share

You May Also Like

Related MotorTrend Content: World | Business | Tech | Sports | Health | Entertainment