Global Sources
EE Times-India
Stay in touch with EE Times India
 
EE Times-India > Controls/MCUs
 
 
Controls/MCUs  

Nvidia boasts Tflop power with Tegra X1

Posted: 06 Jan 2015     Print Version  Bookmark and Share

Keywords:Tegra K1  Tegra X1  4K  Drive CX  ADAS 

The CX marks another milestone in Nvidia's efforts to create not only chips and cards but system and sub-system products for its customers. In recent years, Nvidia has rolled out everything from its own portable games device to GPU cloud servers.

In addition to enhanced graphics for dial, Drive CX includes Surround-Vision, an undistorted top-down, 360-degree view of the car in real time that could replace a physical mirror with a digital smart mirror.

Assisted driving today

With emerging technologies such as Drive CX, Huang said cars must become "software updatable, able to learn." Huang also expects rapidly improving camera technology to replace sensors in the cars of tomorrow, creating a network of computer vision. Replacing sensors with cameras will also reduce the cost of vehicles and decrease energy consumption to make such computer vision a possibility in mainstream cars.

These software-defined vehicles will need to be connected by a supercomputer, and Nvidia hopes to heed this call with the Drive PX, an advanced driver assistance system (ADAS) development platform based on the Tegra X1.

Drive PX

Drive PX is Nvidia's sub-system Tegra product for ADAS.

Drive PX uses two X1 chips, has inputs for up to 12 high-resolution cameras, and can process up to 1.3Gpx per second. The system takes advantage of deep neural networks to do the learning Huang envisions for future cars.

According to a release and event demonstration, Drive PX can:

  • ...allow a car to discover open spaces in a crowded parking garage, park autonomously and then later return to pick up its driver when summoned from a smartphone. The deep learning capabilities of DRIVE PX enable a car to learn to differentiate various types of vehicles—for example, discerning an ambulance from a delivery van, a police car from a regular sedan, or a parked car from one about to pull into traffic.

Surround Vision

Surround Vision created a simulation of Nvidia's parking lot.

Before CES, Huang and colleagues taped a drive through Silicon Valley, then processed the video in real-time at the event. Drive PX was able to identify pedestrians, even hidden ones, and differentiate between police vehicles, trucks, and regular cars. The chip also identified different speed limit signs without input from engineers.

"Nvidia is clearly taking a strong run at the future of the car. With the Tegra X1, Nvidia wants to own the object recognition, response system, and entertainment systems, and they've optimised their chips and ecosystems to take advantage of the workload," Moorhead said.

Huang said the neural network processing of different signs, vehicles and people took less than 10 per cent of the overall horsepower of Tegra X1.

While Nvidia did not announce partnerships for this technology, Audi officials spoke at the event and have spoken at past Nvidia events. Moorhead said, "Audi will most likely be the first customer of this technology."

Nvidia did not provide shipping dates for its chip.

- Jessica Lipsky
  EE Times


 First Page Previous Page 1 • 2



Comment on "Nvidia boasts Tflop power with Tegra..."
Comments:  
*  You can enter [0] more charecters.
*Verify code:
 
 
Webinars

Seminars

Visit Asia Webinars to learn about the latest in technology and get practical design tips.

 

Go to top             Connect on Facebook      Follow us on Twitter      Follow us on Orkut

 
Back to Top