Nvidia, known for making semiconductors and video game components, has also been busy working with automakers for a decade now, powering digital dashboards and infotainment systems with its Tegra processors. Now, it has come up with its own autonomous car technology to deal with the many instances drivers face while on the road.

A lot can happen. Things can go flying at you, an animal can run towards its own death, and all of this happens too quickly to always react safely. That's where Nvidia wants to bring in its new $10,000 computer, called Drive PX. With today's technological advances, such as AI “deep learning”, speech recognition, and massive computer data centers, Nvidia considered bringing all that intelligence into a car computer for Drive PX (Wired). It's all about teaching cars how to respond to different behaviors in different situations.

By using a deep neural network via Nvidia's Tegra X1 chips, Drive PX could process millions of images by breaking them down into patterns and textures for recognition. If a situation were to occur that wasn't in the database, Drive PX would “break it down in a similar fashion, comparing its patterns to those it has stored” (Cnet).

If some of this sounds familiar to you, you might be thinking of DAVE, autonomous vehicle technology developed by DARPA. However, Nvidia's technology is 3,000 times faster than DAVE, according to company CEO, Jen-Hsun Huang. Using two Tegra X1 processors, it works by using 2.3 teraflops of processing power, and can handle input from 12 cameras at once, making it the most advanced computer platform on the planet, says Huang. The $10,000 computer is currently available for universities, automakers, and developers who want to develop their own self-driving car platforms.