NVIDIA Pegasus: Making Level 5 autonomous vehicles possible

NVIDIA Pegasus
  • Save
IA - WoW

NVIDIA unveiled the world’s first artificial intelligence computer called Pegasus, designed to drive robotaxis autonomously. Pegasus extends the NVIDIA DRIVE PX AI platform which can handle Level 5 autonomous vehicles. Pegasus can deliver over 320 trillion operations per second. This in comparison, is 10 times faster than the performance of its predecessor, the NVIDIA DRIVE PX 2.

Pegasus creates the possibility of a fully operational vehicle, without a driver. That means, a car without a steering wheel, pedals and side mirrors. Also, no car-like interiors. Just a clean and airy space for occupants, much like the Audi Aicon which also features Level 5 autonomous driving.

In contrast, the latest Audi A8 features Level 3 autonomous driving. This will also help drivers recover precious time which is usually lost when driving or parking, making it possible for occupants to do carry out other agendas while Pegasus gets them to their destination.

NVIDIA now has 225 partners working with them on the NVIDIA DRIVE PX platform as of now. But, 25 of them are developing fully autonomous robotaxis using NVIDIA CUDA GPUs. Now, their trunks are the cars small data centers, which are loaded with racks of computers with NVIDIA GPU servers which run deep learning, computer vision and parallel computer algorithms.

READ ALSO  [CES 2016] Acer Aspire V Nitro Black Edition gaming notebook gets Intel RealSense

But, their size, cost and power demands make it impractical for moving into production.

With the new software added into future taxis, this could result in the reduction of accidents inflicted by physical drivers and less congestion.

Also, with the implementation of Pegasus, hailing taxis will be even more efficient which does help save precious time.

Also, having Pegasus implemented means that mobility is now possible for the disabled and elderly.

What do you think of this new technology? Would it make your life a lot easier by having driverless cars?

Source: NVIDIA

Copy link