Qualcomm edges Nvidia, Mobileye in competition to infuse tech into next-generation vehicles

Giant chipmakers such as Qualcomm, Nvidia and Mobileye have become linchpins in the auto industry's efforts to transform vehicles into rolling supercomputers.

The three companies are locked in a fierce competition to provide the building blocks and the smarts to power a new generation of vehicles.

As cars increasingly become defined by software — capable of getting over-the-air updates and toggling computing workloads between on-board systems and the cloud — the early advantage tilts toward Qualcomm.

The wireless backbone that Qualcomm developed decades ago for mobile devices differentiates it from competitors in the automotive realm, and it's providing an essential link between those on-board and off-board systems, said Steve McDowell, chief analyst and founder of NAND Research.

"Qualcomm's secret sauce here, from a technology perspective, is the connectivity piece," he said.

Connectivity has provided the company a tail wind. Qualcomm reported $811 million in second-quarter automotive revenue, an 87 percent year-over-year increase versus the same time frame last year.

The company's rivals have reported more modest results.

Nvidia ranks as a $3 trillion company, with more than $30 billion in revenue reported Aug. 28 in its latest quarterly results. But it reported automotive-specific revenue of $346 million, 37 percent ahead of the same period in 2023 — robust growth — but less than half of Qualcomm's growth rate.

Mobileye, a prime supplier of chips and driver-assist technology, reported revenue of $439 million in the second quarter, a 3 percent decrease from the second of quarter 2023. Mobileye reduced expectations for the remainder of 2024, citing China's sluggish economy.

Direct comparisons of the three companies can be complicated. There's commonality in that all provide chips and software for vehicles. But they specialize in distinct aspects of in-car technology — infotainment, driver-assist and digital cockpits prominently among them — and their offerings are often aimed at different tiers of the market.

"They all play within similar spaces, but their approaches come from very different places," said Phil Amsrud, associate director at S&P Global Mobility who focuses on semiconductor research and automated driving technologies.

Qualcomm is poised for a steep growth curve, according to analysts. The company claims a $45 billion automotive pipeline, a result of the "pretty shrewd" way it has built its automotive portfolio and balanced short- and long-term plans, Amsrud said.

Much of the company's momentum comes from its ability to package as few or as many features into its "digital chassis" as an automaker desires, and provide turnkey systems rather than discrete features.

"They can do the heavy lifting for customers," McDowell said. "They said, 'We can drop this in your lap, and you can customize your digital cockpit relatively quickly.' "

Starting with the modems that powered General Motors' OnStar communications system in 2002, Qualcomm has plotted an ever-broadening strategy to supply automakers with semiconductors that underpin technology-rich in-cabin experiences — and more.

The company works with "all major" global automakers, a spokesperson said, including GM, BMW and Volvo. The revamped Volvo XC90 is expected to roll out with Qualcomm's Snapdragon digital chassis on Sept. 4. More than 350 million vehicles worldwide contain Qualcomm technology, the company said.

Qualcomm's momentum comes as no surprise to Nakul Duggal, general manager for the company's automotive, industrial and cloud businesses.

"It's exciting in some ways, but we said we'd do all these things," he said. "We're doing what we said we'd do."

Connectivity was the starting point. Duggal foresees the industry's move away from dozens of electronic control units toward centralized computing as Qualcomm's next key growth driver.

The company introduced its Snapdragon Ride Flex system on chips co-developed with Bosch in January. It is designed to let car companies run both infotainment and critical safety functions such as driver-assist features on a single chip. China has emerged as a key market and business battleground.

That's where Nvidia notably expanded a partnership with BYD involving its next-generation Drive Thor chips this year. Qualcomm anticipates strong business there, too. Its Flex architecture helps Chinese automakers compete at entry-level pricing yet choose a centralized platform that can scale globally.

"There's all this cost pressure in China," Duggal said. "The price point and architecture of your silicon makes a huge difference in whether you get to compete."

On the driver-assist front, Qualcomm, Nvidia and Mobileye have all taken different lanes.

Nvidia emphasized building technology for sophisticated driver-assist systems and fully autonomous driving. It has trickled that technology downward into more basic systems.

Mobileye started with basic driver-assist functions, developing chips tailored for computer vision technology, and has pushed into more advanced driver-assist systems and full autonomy.

Focused on its digital chassis and cockpit technology, Qualcomm was late to develop driver-assist features, not starting in earnest until the acquisition of its Arriver software unit from Veoneer in 2022.

The rollout of the more sophisticated driver-assist systems has proved challenging in personally owned vehicles, and automakers have recalibrated. They've shifted some attention from autonomy and higher-end driver-assist functions toward user experiences in the cabin — a move that has played into Qualcomm's traditional strengths.

McDowell cautioned the market could shift again, perhaps toward artificial intelligence offerings in vehicles. AI advances have made Nvidia a household name. In that sense, the second-quarter earnings reports might look more like a snapshot in time rather than a sign of long-term trajectories.

"But Qualcomm's AI is nothing to sneeze at," McDowell said.

In May, Qualcomm revealed a partnership with Microsoft in which it will supply chips that can run AI tasks for use in the company's Surface-branded laptops and tablets.

Such uses are straightforward enough, Duggal said, but the more tantalizing opportunities he foresees with AI deployed in automobiles sit behind the scenes and provide insights for automakers.

Automakers can understand not just questions that owners are asking AI systems about the car, but the context in which the inquiries are made.

For example, if a driver asks about a particular warning light, the automaker can gain an understanding as to whether the information provided is helpful, the time and location of the inquiry, and what type of charging station may have been most recently visited, then use that data to detect patterns across models or fleets.

Should the zeitgeist change and AI fizzle, however, Qualcomm is still well positioned for automotive success, Amsrud said.

"They have a strategy that can incorporate AI where it makes sense, but they're not overly dependent on AI driving everything," he said. "If the world decides that AI is not quite ready for prime time, they're OK."