How Tesla is using a supercomputer to train its self-driving tech
Tesla’s supercomputer requires all this computing juice to energize self-driving cars’ interior and exterior. On the inside, Tesla requires powerful computers to operate its autonomous driving software, and on the outside, it employs massive supercomputers to equip its self-driving software.
Each Tesla vehicle possesses eight surround cameras, each operating at 36 frames per second, ascended on to furnish 360 degrees of visibility around the car at up to 250 meters of range. The car’s refined driver assistance system, imaginatively named Autopilot, collects data from all Tesla vehicles on the road and nourishes it to the software built in the car. They review this data and make projections to enhance Autopilot’sit driving performance.
At the same time, this massive chunk of information is shared on Tesla’s neural network to deduce if the driving decisions made were precise or if any data was misconstrued. In an incident where the data was indeed misidentified, it will then be continuously operated through supercomputers to change its behavior, which effectively equips the constantly evolving Autopilot A.I. model.
“We have a neural net architecture network, and we have a data set, a 1.5 petabytes data set that requires a huge amount of computing. So I wished to plug into this insane supercomputer that we are building and using now. For us, computer vision is the bread and butter of what we do and what enables Autopilot. And for that to work well, we need to master the data from the fleet, train massive neural nets and experiment a lot. So we invested a lot into the computer”, stated Karpathy in an interaction with the media.
Generally, carmakers functioning on autonomous technology will utilize an amalgamation of LiDAR, radar, and cameras to develop a well-rounded sensor fusion module, assuring that their vehicles are secure on the road. An instance is Volvo’s next-gen XC90 SUV. While utilizing only cameras to refine vision-based autonomy is hard, Tesla understands that it’s the answer to inexpensive and safe self-driving cars. The A.I. system will constantly be evolving and changing.