Nvidia's Tegra X1 Powers Smart Cars At the Consumer Electronics Show (CES) last week, Nvidia released demos of two computing platforms, Drive CX and Drive PX, for cars. These computing platforms are part of the venture called Nvidia Drive and are powered by the Tegra X1 mobile chip. Nvidia Drive PX is for the actual driving side of the smart car equation, while Drive CX deals with digital systems on the dashboard and elsewhere in the vehicle. Drive PX is an auto-pilot system, which aids in assisted parking and auto-valet feature. It is an image processing platform and it will have the capability to work with inputs from up to 12 cameras. All these cameras will work in tandem and provide improved accuracy in the auto-pilot function of the car. Nvidia also released a recording of a demo drive of a car with Drive PX that was able to find a space in a parking garage traveling at a speed of 8 kmph. Drive PX can also differentiate between vehicles and can also identify an ambulance from a van. The famous graphics card manufacturer believes that this sort of situational awareness will help prevent accidents. Meanwhile Drive CX has been described as the digital cockpit computer, and it will be the lifeline of a car infotainment system. It will facilitate digital instrument clusters, navigation, in-car entertainment and also monitor the driver. Using surround vision, it will provide the drivers a 360 degree view of the car in real time. The digital smart mirrors will then be devoid of problems like blind spots. Drive CX will also be the lifeline of a car infotainment system. It will be able to power four full-HD screens simultaneously or even two 4K screens. Drive CX will also be able to recognize speech and faces of it users. Nvidia senior manager for CUDA Libraries and Algorithms Jonathan Cohen, said, Because this is a neural network, weve shown it lots of examples of these 40 mile-per-hour limited speed signs. Somehow it has learned to recognize the speed sign even though it hasnt seen these signs before. This is the power of neural networks and the power of deep learning.