Follow
Subscribe

Nvidia is partnering with Toyota on future autonomous car development

Home > News > Content

【Summary】GPU giant Nvidia announced today that it will partner with Japan auto veteran Toyota on developing Toyota’s next-generation autonomous driving systems.

Original Claire    May 11, 2017 11:05 AM PT
Nvidia is partnering with Toyota on future autonomous car development

Nvidia invited members of the global media yesterday, including reporters from Futurecar, to the company's keynote presentation at the 2017 GPU Technology Convention, presented by CEO Jensen Huang. The CEO announced some of Nvidia's future plans and new products during the keynote.

Among the big news of the day was that the GPU giant announced that it will be partnering with Japan auto veteran Toyota on developing a cutting edge, next-generation autonomous driving system.

Nvidia announced that the automaker will use its DRIVE PX AI vehicle processor platform to power its self-driving car. Toyota plans to have the car available in the next few years.

"Toyota has worked on autonomous driving technologies for over 20 years with the aim of reducing traffic fatalities to zero as an ultimate goal, achieving smoother traffic, and providing mobility for all," said Ken Koibuchi, executive general manager at Toyota. "Through this collaboration, we intend to accelerate the development of autonomous driving systems that are even more safe and capable."


Meanwhile, Nvidia's stock price surged 18%, with its first quarter earnings doubling to $507 million compared with last year's $208 million in the same period. Analysts are noting that the strong performance can be attributed to increasing needs for GPUs that can handle deep learning problems like autonomous driving and speech recognition. 

Why GPU matters in autonomous car system?

Nvidia introduced its first GPU (Graphical Processing Unit) in 1999, named the "GeForce 256", as "the world's first GPU", which sparked the growth of PC gaming market. In recent years, artificial intelligence and deep-learning are growing technologies and widely applied in many areas of the tech industry. Enabling machines to learn by themselves involves complex mathematical calculations that require a high-level of data processing ability.  A GPU working as the brain of a system in computers, robots, or self-driving cars, is more efficient when using machine learning to perceive and understand the world.

A self-driving vehicle requires a robust supercomputer onboard to process and interpret the data collected from all the various sensors on the car. Nvidia's DRIVE PX processor is a palm-size module that delivers 30 trillion deep learning operations per second. It can process large amounts of data in real-time from an autonomous vehicle's cameras, lidar, radar, and other sensors. The DRIVE PX AI will use artificial intelligence to interpret the 360-degree environment surrounding the car, localize itself on an HD map, and predict potential hazards while driving.


gtc17-auto-trucks-3.jpg

When powered by an AI processor reaching 30 trillion operations per second, a self-driving car can reach "level-4 autonomy", which is a car that can drive on its own without human intervention. Additionally, Nvidia's DRIVE PX processor receives updates over the air, so that it can learn by itself and get smarter over time.

Nvidia's self-driving dream

The DRIVE PX system that Nvidia prides itself on has been utilized by a growing number of automakers. Electric carmaker Tesla is installing this system in all of its vehicles, and hopes to achieve a more mature self-driving capability. In early March, Bosch partnered with Nvidia to sell the driver-assist platform to automakers. Meanwhile, Nvidia is also working with Baidu to create an open cloud-based computing platform for autonomous cars.

In December, Nvidia applied, and was granted an autonomous vehicle testing permit in California, joining the big-leagues in the development of driverless cars. In the future, the company estimated there will be 15 million driverless cars on the road, creating a revenue of $2 billion for Nvidia. Instead of programming a set of fixed instructions into its autonomous driving software, the Drive PX  is not dependent on detecting road markings. The processor will use machine learning to analyze real-life road situations and improve its capabilities over time.

"We envision a future society where autonomous vehicles whisk people safely and comfortably around beautiful cities," said Jensen Huang, chief executive officer and founder of NVIDIA. "The development of a self-driving car is one of the greatest technical challenges that's ever been tackled."


Prev                  Next
Writer's other posts
Comments:
    Related Content