Nvidia announces supercomputer for self-driving cars at CES 2016
Nvidia announced the NVIDIA Drive PX 2, a new computer for cars and the follow-up to last year’s Drive CX.
It is designed for deployment in self-driving cars and will be the brain of the vehicle. Thanks to pascal the vehicle will map in 360 positions to determine precisely where the car and calculate distance and trajectory safely, all this is possible due to the processing power for 24 Trillion Deep Learning Operations per Second, which is equivalent to almost 150 MacBook Pro’s.
Self-driving cars use a broad spectrum of sensors to understand their surroundings. DRIVE PX 2 can process the inputs of 12 video cameras, plus lidar, radar and ultrasonic sensors. It fuses them to accurately detect objects, identify them, determine where the car is relative to the world around it, and then calculate its optimal path for safe travel.
Nvidia has created a reference platform called the “Nvidia Drivenet” and it is already testing its own self-driving cars. It has nine inception layers so that it can train itself to “perceive things out in the world.” Huang says that it took just a few months for the network to recognize objects in real time. It’s not new, but Nvidia says that it’s made huge strides in recent months. In one impressive demo, Nvidia showed how it was able to detect cars even in very snowy conditions.
Nvidia is also working with Audi to test its systems, to the point where its system was able to read German road signs “better than a human can.” Daimler, BMW, and Ford are using the system to develop their self-driving cars, as well.