Tech giant Nvidia has launched a new ‘in-car supercomputer’ at CES 2016 called Drive PX 2 that uses artificial intelligence to tackle the complexities of making truly self-driving cars work.
Offering 360-degree situational awareness, Nvidia says the highly advanced GPU (graphics processing unit) benefits from advances both in supercomputing and deep learning.
We are leveraging these,” said Nvidia co-founder and CEO Jen-Hsun Huang, “to create the brain of future autonomous vehicles that will be continuously alert, and eventually achieve superhuman levels of situational awareness.
The Drive PX2 can process inputs from 12 video cameras, lidar, radar and ultrasonic sensors, all at once. It then blends them to identify objects, work out where the car is relative to everything around it, and then calculate the exact driving route.
It can carry out 24 trillion deep learning operations per second; it’s equivalent to having 150 MacBook Pros as your car’s brain, says Nvidia.
This allows it to quickly work out surprising drive challenges such as:
- Unexpected road debris
- Erratic drivers
- Poor weather
- Difficult light (such as sunrise)
- Extreme darkness
The latest Drive PX 2 system is ten times as powerful as last year’s first-gen Drive PX system; this is now being used by more than 50 car makers, suppliers, developers and research institutes to work on autonomous cars.
The new Drive PX 2 system will be available in late autumn 2016 – and the first car maker to use it? Volvo, which will roll out the new in-car artificial intelligence supercomputer in its self-driving XC90s early next year as part of a public trial in Gothenburg.