The Inquirer-Home

Nvidia takes on Raspberry Pi with the Jetson TK1 mini supercomputer

Will power Audi’s self-driving cars
Wed Mar 26 2014, 14:04

NVIDIA HAS UNVEILED what it claims is "the world's first mobile supercomputer", a development kit powered by a Tegra K1 chip.

Dubbed the Jetson TK1, the kit is built for embedded systems to aid the development of computers attempting to simulate human recognition of physical objects, such as robots and self-driving cars.

Speaking at the GPU Technology Conference (GTC) on Tuesday, Nvidia co-founder and CEO Jen Hsun Huang described it as "the world's tiniest little supercomputer", noting that it's capable of running anything the Geforce GTX Titan Z graphics card can run, but at a slower pace.

With a total performance of 326 GFLOPS, the Jetson TK1 should be more powerful than the Raspberry Pi board, which delivers just 24 GFLOPS, but will retail for much more, costing $192 in the US - a number that matches the number of cores in the Tegra K1 processor that Nvidia launched at CES in Las Vegas in January.

Described by the company as a "super chip" that can bridge the gap between mobile computing and supercomputing, the Nvidia Tegra K1, which replaces the Tegra 4, is based on the firm's Kepler GPU architecture.

The firm boasted at CES that the chip will be capable of bringing next-generation PC gaming to mobile devices, and Nvidia claimed that it will be able to match the PS4 and Xbox One consoles' graphics performance.

Designed from the ground up for CUDA, which now has more than 100,000 developers, the Jetson TK1 Developer Kit includes the programming tools required by software developers to develop and deploy compute-intensive systems quickly, Nvidia claimed.

"The Jetson TK1 also comes with this new SDK called Vision Works. Stacked onto CUDA, it comes with a whole bunch of primitives whether it's recognising corners or detecting edges, or it could be classifying objects. Parameters are loaded into this Vision Works primitives system and all of a sudden it recognises objects," Huang said.

"On top of it, there's simple pipe lines we've created for you in sample code so that it helps you get started on what a structure for motion algorithm, object detection, object tracking algorithms would look like and on top of that you could develop your own application."

Nvidia also expects the Jetson TK1 to be able to operate in the sub-10 Watt market for applications that previously consumed 100 Watts or more.

Nvidia Jetson Tegra K1 launch with Audi self driving car

During his keynote, Huang welcomed Audi on stage to show off a self-piloted car powered by the Jetson TK1, which it uses to process the video coming from its various cameras in real-time.

Audi didn't project a date for the self driving car to arrive in showrooms, but said it expects the technology to be ready for the road within five years. µ


Share this:

blog comments powered by Disqus
Subscribe to INQ newsletters

Sign up for INQbot – a weekly roundup of the best from the INQ

Existing User
Please fill in the field below to receive your profile link.
Sign-up for the INQBot weekly newsletter
Click here
INQ Poll

Microsoft Windows 10 poll

Which feature of Windows 10 are you most excited about?