Ups and Downs of Gamers in Driverless Tech

Not long ago, some assumed graphics card maker Nvidia was set to be the dominant provider of autonomous vehicle platforms.

Recent developments have thrown shade on that optimism, although some observers believe the company might still have a bright future in the segment. Having grown into a big company on the back of selling video processors for PCs, Nvidia entered the connected car as a provider of infotainment solutions. Its most notable client was everyone’s favorite futuristic car company Tesla but this came to a crashing halt in 2017, when Tesla switched partners to Intel. The latter company became a major player in the advanced/autonomous chip space with its acquisition that year of Mobileye. The combination of Intel’s considerable hardware capabilities, not to mention its deep pockets, with Mobileye’s software has proven to be very effective.

However, Bret Kenwell, a commentator on connected car tech for TheStreet.com and content manager for FutureOfAuto.com, says Nvidia’s recent loss was not a particularly heavy blow. “It’s not instantly damaging,” he asserts. “With Tesla shipping less than 50,000 cars per quarter, losing that business isn’t that detrimental for a company like Nvidia.” Besides, Nvidia has grander ambitions than just supplying your car’s infotainment processor. It really wants to be its full-time driver too. “Every type of vehicle will eventually become autonomous,” says the company’s senior director of automotive, Danny Shapiro.

Shapiro points out that vehicles using artificially intelligent brains will improve road safety and mobility. Additionally, the vehicle-as-a-service market will also be immense, given the modest costs for users. In both instances, chips with powerful autonomous capabilities will replace human operators. “Driving the convenience of transportation up, and costs down, is a huge opportunity,” he adds.

As with any huge opportunity, though, numerous other players are also vying for it. The most prominent example in the assisted/autonomous processor space is Intel, which ponied up a vertigo-inducing $15.3Bn to own Mobileye. On the hook to make this extremely pricey asset worthwhile, Intel has been aggressive in courting clients. This paid off when it was selected in late 2017 to be the chip supplier for Alphabet’s Waymo, considered by many to be the cutting-edge developer of autonomous vehicles. This was an enormous victory in the self-driving segment.

Some observers feel that the company has a natural advantage. “Mobileye has always been the merchant supplier to beat,” says analyst Christopher Rolland of Susquehanna International Group. “[Nvidia] is the one who needs more high-volume design wins.” Ryan Shrout, of namesake analysis firm Shrout Research, has a different view. “I still put [Nvidia] as the company to beat in this space, though it’s true that other traditional technology companies are catching up,” he said. “Intel in particular is going to be a tough foe thanks to its history and [research and development] capability.”

If Nvidia’s grand technological ambitions are realized, it might just edge past Intel and other rivals. The company’s long autonomous chip development process has resulted in its latest autonomous solution, the sprawling DRIVE AGX system. This promises not only to function as the vehicle’s driver but as the brains of various complimentary in-car functionalities as well. It’s effectively a one-stop solution for driverless cars. True to Nvidia’s roots as a supplier of video cards, DRIVE AGX uses graphics processing units (GPUs) to help crunch the volumes of data needed to operate a car. This differs from the central processing units (CPUs) that have traditionally been at the heart of a PC’s functionality.

“Safe autonomous driving requires enormous compute performance on board the vehicle,” argues Nvidia’s Shapiro. “While CPU performance had steadily increased in the past, with the end of Moore’s law, CPU performance has not plateaued. The unprecedented parallel processing capabilities of GPUs deliver the computational horsepower that enables us to tackle the world’s greatest AI challenges.” Challenges such as, of course, autonomous driving.

This argument seems to have struck a chord with vehicle maker Daimler and engineering giant Bosch. Working together to develop autonomous autos, the two companies have signed up for Nvidia’s technology to power the electronic control units (ECUs) of its efforts. The duo will also use Nvidia’s artificial intelligence (AI) software to help guide the vehicles.

It’ll take more than that gig to dominate the market, though. Looking down the road, is Nvidia set to remain an important player in the autonomous processing game? “On the AV front exclusively, yes, Nvidia and Intel are set to lead the charge,” says TheStreet.com’s Kenwell. “Globally speaking though, nothing stops a competitor from rising to the challenge. Further, companies like NXP Semiconductors, Qualcomm and Broadcom are likely to benefit from the rise of connected cars, EVs and AVs. As big and great as Intel and Nvidia are, we’re still in the early innings of AVs.”

Shrout believes the future is coming very soon and Nvidia will certainly be a part of it. “Full Level 5 autonomy is likely 4-8 years away, giving us plenty of time to see what pans out,” he said. “As it stands now, I would bet on Nvidia being one of those remaining companies however.”


Leave a comment

Your email address will not be published. Required fields are marked *