Nvidia is growing an automotive technology company. Pictured are its autonomous vehicle test cars at the company’s auto garage in Santa Clara, California, on June 5, 2023.
Bloomberg | Bloomberg | Getty Images
Nvidia revealed plans on Monday to test a robotaxi service with a partner as early as 2027, underscoring the chipmaker’s ambition to become a major player in the world of self-driving cars.
The service would be offered with a partner and would employ cars with “Level 4” driving, meaning they will be able to drive without human intervention in predefined regions, Nvidia officials said during a self-driving demonstration in San Francisco last month. The company declined to specify where it would operate and who its partner would be.
“We will probably start with limited availability, but work with the partner so that we can gain a foothold,” Xinzhou Wu, Nvidia’s vice president of automotive, said at the event.
Since 2015, Nvidia has offered chips and other technology for cars under the Drive brand, but that’s only a small part of the company’s business. Automotive and robotics chips accounted for just $592 million in sales in the quarter ended October, or about 1% of Nvidia’s total revenue. Nvidia announced a robotaxi partnership with Uber in October.
The chipmaker said in December that it had developed software capable of powering a self-driving car and that Mercedes-Benz models going on sale in late 2026 will be able to use Nvidia’s technology to navigate cities like San Francisco.
Self-driving cars remain one of the main areas where Nvidia can show growth outside of AI infrastructure. CEO Jensen Huang said robotics – including self-driving cars – is the company’s second-largest growth category after artificial intelligence.
“We imagine that one day, a billion cars on the road will all be autonomous,” Huang said during a launch event Monday at the CES conference in Las Vegas. “You can either make it a robo-taxi that you orchestrate and rent to someone, or you can own it.”
In addition to the chips built into self-driving cars, Nvidia sells access to its famous AI chips as well as its simulation software to automakers so they can train autonomous models and develop technologies.
Nvidia says automakers can use its Drive AGX Thor automotive computer, which costs about $3,500 per chip, to save on research and development costs and bring self-driving features to market more quickly. Nvidia said it was working with automakers to refine its technology, such as determining how hard the car accelerates, for specific vehicles.
“Some people say, ‘Hey, I need your help to train and optimize my software on your chip, but I’ll do the simulation myself,'” said Ali Kani, general manager of Nvidia’s automotive platform.
Automakers, like Mercedes-Benz, want to perfect Nvidia’s technology, market it as part of their in-car experience, and sell it as part of or alongside a new car.
Robotaxis have gained momentum over the past year, led by The alphabet Waymo, which operates a driverless commercial taxi service in five U.S. markets, including San Francisco.
Nvidia’s robotaxi announcement says it is targeting autonomous fleets in addition to personal cars that consumers might purchase.
Driving an Nvidia robotaxi
In December, Nvidia offered journalists and analysts a one-hour ride through San Francisco in a 2026 Mercedes-Benz CLA sedan.
While my car had a Mercedes-Benz employed safety driver behind the wheel, the driver said the car was driving alone for 90% of the trip.
The journey was uneventful. San Francisco is a tough city to drive in, with big hills, frequent red lights, and trucks unloading in the middle of the streets, but I never felt stressed and was able to concentrate on a conversation.
But there was one major problem: The driver took control in a sticky situation where two buses and a self-driving Waymo were trying to pass on a four-lane road with street parking and trucks on both sides unloading goods. The driver had to reverse and wait for the traffic jam to clear.
Nvidia declared my drive to be “Level 2 Plus Plus”, meaning its technology has similar features to that of Tesla Fully autonomous driving mode. Nvidia-powered cars like the Mercedes-Benz will have increasing autonomous driving capabilities, but the responsibility for keeping everyone safe still falls on the driver, who must be attentive at all times.
The chipmaker said the system would eventually be capable of “parking-to-park,” meaning it could move from one starting parking space to another, but Mercedes-Benz CLA cars won’t have that feature to boot.
“Any parking situation that you feel is intimidating, this car will solve it for you,” Mercedes-Benz Group CEO Ola Källenius said at an Nvidia event on Monday.
The Mercedes-Benz model presented by Nvidia was launched last year in Europe, but it will launch in the United States this year, Kani said.
Mercedes-Benz cars have been rolled out with lane keeping and driver assistance features to help drivers stay in their lane, Kani said. The vehicles received a lane change function through a software update and will benefit from hands-free highway driving, city driving and park-to-park features this year.
Nvidia said it uses two AI systems in Drive-powered cars to ensure safety. The car primarily runs on an “end-to-end” system, called a visual language model, which uses AI to decipher visual sensors and plot a path.
The chipmaker said it also built a second safety-focused “stack,” which uses strict rules — like stopping at a stop sign — to take over when the AI doesn’t know what to do.
Nvidia nevertheless hopes that recent advances in generative AI, driven by the company’s graphics processing units, will enable self-driving algorithms to become more capable. Nvidia is targeting 2028 for point-to-point autonomous driving features in consumer cars. Ultimately, Nvidia said it wanted to make the car itself feel like a real driver that users can just talk to.
“With transformers and generative AI, we can do so much more,” Wu said.
