The robotaxis are being seen as the next big thing in the automotive market that would put an end to private ownership of cars, but the technological hurdles in this direction are yet to be cleared. It is known that the computational requirements of robotaxis are enormous — perceiving the world through high-resolution, 360-degree surround cameras and lidars, localizing, tracking vehicles and people around the car, and planning a safe and comfortable path to the destination. All this processing must be done with multiple levels of redundancy to ensure the highest level of safety.
Nvidia claims to have found an answer to this problem by launching what it calls the world’s first artificial intelligence computer designed to drive fully autonomous robotaxis. The new system has been codenamed Pegasus and it extends the company’s DRIVE PX AI computing platform to handle Level 5 driverless vehicles. The company claims that Pegasus gives more than 10 times the performance of its predecessor, Drive PX 2, delivering over 320 trillion operations per second. The company also informed that more than 25 companies are developing Level 5 Robotaxis on NVIDIA CUDA GPUs.
Pegasus is powered by four high-performance AI processors. It couples two of NVIDIA’s Xavier system-on-a-chip processors — featuring an embedded GPU based on the NVIDIA Volta architecture — with two next-generation discrete GPUs with hardware created for accelerating deep learning and computer vision algorithms. The system will provide the enormous computational capability for fully autonomous vehicles in a computer the size of a license plate, drastically reducing energy consumption and cost.
Intel has been making strides towards autonomous cars, through intelligent partnerships and delivering superior products the company has managed to emerge as a dominant player in the automotive ecosystem. The rival NVIDIA, who has the early mover advantage, through this new product has cleared that the company is not ready to cede its place so easily. Pegasus will be available to NVIDIA automotive partners in the second half of 2018.