Processing the huge volume of data from the Robocar’s array of radar, lidar, cameras and GPS sensors calls for serious computing power. Processor manufacturer NVIDIA delivers the solution.

Angus Frazer (copy) & NVIDIA (photo)

Capable of performing up to 24 trillion operations per second, the DRIVE PX2 AI supercomputer made by NVIDIA is perfect for the job. The energy-efficient NVIDIA graphic processor, which has the same processing power as 150 high-end notebooks, lends the Robocar deep learning capabilities. The upshot is that it learns more with each race.

Set up in California in 1993, NVIDIA forged its reputation by instrumentally spurring the growth of the PC gaming industry with high-end graphic processing units. The company, which is today at the forefront of AI development, has been in partnership with Audi for over a decade, with the Audi A4 already using a NVIDIA chip as early as 2007.

A Tegra K1 chip from NVIDIA is a core component of the driver assistance controller (zFAS) in the new Audi A8. Another example of cooperation between NVIDIA and the four rings is the piloted-driving Audi Q7 deep learning concept, which impressed industry experts when showcased last year at CES in Las Vegas.