UPDATED 23:30 EST / JANUARY 05 2016

NEWS

Nvidia debuts Drive PX 2, DriveNet: ‘Supercomputer’ and deep learning platform for self-driving cars | #CES2016

At the Consumer Electronics Show (CES), currently taking place in Las Vegas this week, the focus is predominantly on automotive technology, including tech to enhance the in-car experience and autonomous vehicle systems.

One year after announcing its Drive CX and first-gen Drive PX computational platform, Nvidia Corp. took to the stage at CES earlier this week to show off its next-gen Drive PX 2 – a ‘supercomputer’ and deep-learning platform for self-driving cars.

The liquid cooled Nvidia Drive PX 2 consists of 12 CPU cores and four GPUs, packing 8 teraflops of processing power. With more than 10 times the computational power of the first-gen Drive PX, the Drive PX 2 is as powerful as six Titan X graphics cards, or the equivalent to that of 150 MacBook Pros.

According to Nvidia, the Drive PX 2 can deliver up to 24 trillion deep learning operations per second, enabling it to rapidly learn how to deal with commonplace driving challenges, such road debris, erratic human drivers, construction zones, and more. Problem areas that prove challenging for traditional computer vision techniques, such as bad weather like rain, snow and fog, and poor lighting conditions like sunrise, sunset and nighttime, are all addressed through deep-learning.

Nvidia Drive PX 2 computational platform for self-driving cars

Nvidia Drive PX 2 computational platform for self-driving cars

Drive PX 2 combines data from 12 video cameras and lidar, radar and ultrasonic sensors to detect and identify objects and plot a safe route for the car, but all the processing power needs software to back it up. That comes in the form of Nvidia’s DriveWorks software and Nvidia DriveNet, a deep neural network.

DriveNet contains the equivalent of 37 million neurons and has been trained to recognize 120 million objects to date and will continue learning and adding objects over time.

Nvidia is providing automakers with an end-to-end platform, allowing them to train a neural network and deploy it to autonomous vehicles in rapid time.

According to Nvidia, the combination of its GPUs and CUDA processing technology can speed up deep learning processes by as much as 20 to 40 times, dramatically shortening deep learning training cycles. “What would have taken a year now took a week. What took a month took a day,” said Jen-Hsun Huang, Nvidia co-founder and CEO.

Volvo will be the first automaker to ship Drive PX 2 in 100 upcoming test vehicles while Audi, BMW, Mercedez-Benz and Daimler are already carrying out tests.

Images via Nvidia

A message from John Furrier, co-founder of SiliconANGLE:

Your vote of support is important to us and it helps us keep the content FREE.

One click below supports our mission to provide free, deep, and relevant content.  

Join our community on YouTube

Join the community that includes more than 15,000 #CubeAlumni experts, including Amazon.com CEO Andy Jassy, Dell Technologies founder and CEO Michael Dell, Intel CEO Pat Gelsinger, and many more luminaries and experts.

“TheCUBE is an important partner to the industry. You guys really are a part of our events and we really appreciate you coming and I know people appreciate the content you create as well” – Andy Jassy

THANK YOU