NVIDIA launches new software-defined platform for autonomous driving

Shanghai (Gasgoo)- The U.S.-based chipmaker NVIDIA officially launched NVIDIA DRIVE AGX Orin, a highly advanced software-defined platform for autonomous vehicles, at the GPU Technology Conference (GTC 2019), marking the company ushering in a new era of AI computing and software for autonomous driving.

The platform is powered by a new SoC (system-on-a-chip) called Orin, which is born out of the data center and consists of 17 billion transistors, according to the chip maker's introduction.

NVIDIA said the chip is able to deliver 200 trillion operations per second, nearly seven-fold increase over the previous generation SoC Xaiver, and is engineered to process a whopping amount of applications and deep neural networks that run simultaneously in autonomous vehicles and robots.

Thanks to its compatibility with Xavier, Orin allows developers to leverage their existing development investments. Boasting great scalability, the platform contains a wide range of configurations that support Level 2+ AI-enabled assistance driving to Level 5 fully driverless operation.

The first production of the fruit of four-year R&D is expected for launch in 2022, said the company.

Another significant announcement was about the partnership forged with Chinese ride-hailing giant DiDi.

At the GTC, both parties claimed that DiDi will make use of NVIDIA GPUs and AI technology to develop autonomous driving and cloud computing solutions.

The ride-hailing giant will leverage NVIDIA® GPUs in the data center to train machine learning algorithms, and use NVIDIA DRIVE™ for inference on its Level 4 autonomous vehicles. According to NVIDIA, NVIDIA DRIVE enables data to be fused from a variety of sensors, such as cameras, lidar and radar, by virtue of numerous deep neural networks (DNN), so as to help vehicles comprehensively understand the surroundings and plan safe driving routes.

“Developing safe autonomous vehicles requires end-to-end AI, in the cloud and in the car,” said Rishi Dhall, vice president of Autonomous Vehicles at NVIDIA. “NVIDIA AI will enable DiDi to develop safer, more efficient transportation systems and deliver a broad range of cloud services.”

DiDi on August 5 claimed that its autonomous driving arm had been upgraded into an independent company focusing on R&D, production application and other business development related to autonomous driving. To speed up its self-driving R&D, it is expanding the circle of friends comprising tech leaders and OEMs (photo source: NVIDIA China's WeChat account).