From its roots as a provider of graphics processing power for video games, Nvidia has risen to become one of the most important Silicon Valley companies developing technology for self-driving vehicles. This was underscored recently when the company announced further collaborations with two major OEMs—Volvo and Volkswagen—and top global automotive supplier ZF. Together, the announcements signal that automakers are using Nvidia’s technology for a broad range of autonomous systems, some of which are tilting from research-and-development projects into firm production plans.
Chief among them: Volvo will use Nvidia’s Drive PX computing platform, pictured above, to develop advanced software and artificial-intelligence applications for autonomous vehicles. The two companies had already collaborated on the XC90 SUVs used in Volvo’s Drive Me project, and this announcement entrenches their relationship as Volvo plans to launch autonomous vehicles by 2021.
“On the platform side, Nvidia is clearly dominating
in the development and early-implementation phases
of autonomous vehicles.” – Mike Ramsey, Gartner
Speaking at the Automobil Elektronik Kongress in Germany, Nvidia chief executive officer Jensen Huang further announced the company will continue developing deep-learning technology with Volkswagen at the automaker’s Silicon Valley laboratory and that it will deepen its partnership with global supplier ZF Group.
The announcements come at a time when competitors such as Intel are investing billions in self-driving technology and the race to provide chips, platforms, and artificial intelligence is ramping up. Tesla Motors was the first to use the Drive PX in production vehicles, and Nvidia has a long-standing relationship with Audi. More recently, it has inked deals with Toyota, Mercedes-Benz, and global supplier Bosch.
“On the platform side, Nvidia is clearly dominating in the development and early-implementation phases of autonomous vehicles based on the announcements that have come so far,” says Mike Ramsey, research director at Gartner, a global technology consulting company. “These recent announcements extend their lead.”
At Volvo, Nvidia will be developing both hardware and software while working with the Swedish automaker’s top supplier, Autoliv, as well as Zenuity, a new subsidiary of both companies formed earlier this year to focus on software development. Some of the work will remain proprietary to Volvo, while in some cases, Autoliv will be free to market products to other automakers.
“This is significant, because a lot of the development work we’re doing can be leveraged by other automakers,” said Danny Shapiro, senior director of Nvidia’s automotive business unit. “You’re starting to see the industry welcome these type of collaborations.”
“We need to be a systems supplier, and we need
to understand the whole system to understand
the functionalities of the future.”
– Stefan Sommer, ZF
Look no further than ZF, which also announced a deal with German lighting and sensor component supplier Hella. The companies plan to work together on camera and radar ventures that could help OEMs meet upcoming European New Car Assessment Program (NCAP) standards being developed for the mass deployment of semi-autonomous and autonomous vehicles. Nvidia is now a piece of that puzzle. Drive PX will underpin the efforts of ZF and Hella to create artificial-intelligence software that optimizes camera and radar data for autonomous or semi-autonomous functions. Stefan Sommer, CEO of ZF Group, says perfecting the combination of that artificial-intelligence information with mechanical systems is vital.
“We need to be a systems supplier, and we need to understand the whole system to understand the functionalities of the future and see business models that we can sell with software solutions,” Sommer said. “By combining our technology and understanding it with each other, we are much faster and have a benefit in terms of having a promising technology other competitors do not have.”
With Volkswagen, the fruits of Nvidia’s collaboration are a bit further down the road. The two companies are working at VW’s Silicon Valley Data Lab to explore how vehicular information stored in data centers can be gleaned for insights, such as how traffic flow can be improved in particular cities. They’re using artificial intelligence in data centers instead of in the cars themselves, but in many ways, Nvidia’s Shapiro said, that’s just as important.
“AI is having a transformative effect on the auto industry in general, and a lot of this means focus on the Drive PX in the car, and running deep-learning algorithms in the vehicle, but there are lots of other implications for vehicles and connected cars,” he said.