At CES 2016, NVIDIA has announced the release of its Drive PX2 which is the company's latest offering in the autonomous vehicle hardware market. NVIDIA hopes that the Drive PX2 will provide the ability to train, optimise and develop the neural networks which will be the foundation of self-driving cars in the future.
Several earlier NVIDIA components combine to form the Drive PX2 such as DIGITS and DriveWorks. However, Drive PX 2 hardware is also the successor to the Tegra-powered Drive PX released last year with the new Drive PX 2 representing a major computational power jump with 12 CPU cores and two discrete "Pascal"-based GPUs. A GPU represents a form of accelerated computing as it uses a graphics processing unit (GPU) together with a CPU to accelerate scientific, analytics, engineering, consumer, and enterprise applications.
NVIDIA has not revealed the full specifications yet, but it has made certain details available. For example, the Drive PX2 consists of two Tegra SoCs (System on a Chip) along with two liquid cooled GPUs.
The liquid cooling consists of a large metal block with traditional copper tubing delivering liquid via an external radiator. This is required to dissipate the excess heat delivered from a total of 12 CPU cores, including eight ARM Cortex A57 cores and four "Denver" cores.
The Drive PX2 will draw a maximum of 250 watts and will offer up to 8 TFLOPS of computational horsepower and 24 trillion "deep learning operations per second." This makes the Drive PX 2 - according sato NVIDIA - 10 times faster than it's predecessor at running these specialised instructions and has nearly 4 times the computational horsepower when it comes to TLOPS.
Similar to the original Drive PX, the driving AI platform can accept and process the inputs of up to 12 video cameras. It can also handle LiDAR, RADAR, and ultrasonic sensors.