NVIDIA gains approval to test self-driving cars in California

Over the last two decades NVIDIA managed to become the dominant force in the computer graphics area and last decade the company was at the helm of the GPGPU computing revolution. These days NVIDIA is increasingly referring to itself as "the AI computing company", as a lot of researchers around the world are using NVIDIA GPU based products to investigate AI, to perform machine learning tasks and to create neural networks.

One of the most visible deep learning projects is the self-driving car. NVIDIA is cooperating with lots of car makers to sell its hardware and expertise, but the company also runs its own self-driving car tests to improve its systems.

Yesterday, NVIDIA reached an important milestone as the company has been added to the California Department of Motor Vehicles' list of companies that are allowed to test self-driving cars on public roads.

As Trusted Reviews reports, California is a front-runner in encouraging self-driving car innovation and even has regulation allowing firms to test autonomous vehicles without a human behind the wheel:

Recently, a plan from California's Department of Motor Vehicles that would require manufacturers to complete a 15-point report on safety and other issues before testing self-driving vehicles faced criticism from the automotive industry.

While the new regulations have been attacked as a threat to innovation, with the Federal government set to weigh in on the matter, California has been one of the only states to encourage self-driving car innovation in recent years.

In September this year, Governor Jerry Brown signed off a new bill that permits autonomous car tests without the requirement for a human passenger to be behind the wheel.

Here you can see the NVIDIA Drive PX 2 in action, this is the self-driving vehicle platform currently in use by all new Tesla cars. In the video below, you can see the NVIDIA BB8 car navigating some roads. Based on what the company shared, the main goal here seems to train its systems in real-life conditions, so it can provide more valuable, "ready-to-use" systems to car makers.