NVIDIA CEO Huang Renxun announced a new generation of highly advanced software definition platforms for autonomous driving and robotics, nVIDIA DRIVE AGX Orin, integrating 17 billion transistors, at the GTC China 2019 conference in Suzhou on December 18. With the new NVIDIA GPU and 12-core ARM CPU, the performance is seven times that of the previous generation of Xavier.
As nVIDIA’s next-generation system-level chip, orin chips consist of 17 billion transistors, a four-year effort by the NVIDIA team. The Orin system-level chip integrates NVIDIA’s next-generation GPU architecture and Arm Hercules CPU core as well as a new deep learning and computer vision accelerator that can run 200 trillion calculations per second. That’s almost seven times the performance of NVIDIA’s previous generation Ofsxavier system-level chips.
Orin handles a wide range of applications and deep neural networks that run simultaneously in self-driving cars and robots, and meets system safety standards such as ISO 26262 ASIL-D.
As a software-defining platform, DRIVE AGX Orin is able to empower a compatible architecture platform for fully self-driving car development from L2 to L5 level, enabling OEMs to develop a large and complex software portfolio.
Because both Orin and Xavier can be programmed through open CUDA, TensorRT APIs, and various libraries, developers can use products across multiple generations after a one-time investment.
“Creating a safe self-driving car is perhaps the biggest computing challenge facing today’s society,” said Huang Renxun, founder and chief executive of NVIDIA. With the investment required to achieve self-driving cars growing exponentially, scalable, programmable, software-defined AI platforms like Orin are essential to complex development tasks. “
According to NVIDIA, the DRIVE AGX Orin chip will be in production in 2022.