Li Auto Inc. Announces the Adoption of NVIDIA's Next Generation Autonomous Driving Smart Chip Orin

September 24, 2020 Tiera Oliver

Li Auto, a member of China’s new energy vehicle market, announced a three-way strategic cooperation with NVIDIA, artificial intelligence computing company, and NVIDIA's Chinese partner, Huizhou Desay SV Automotive. Through this strategic cooperation, Li Auto will be the first OEM equipping its vehicles, the full-size extended-range premium smart SUV to be launched in 2022, with the NVIDIA Orin SoC chipset.

The NVIDIA Orin SoC was released in 2019 and is scheduled to be in production in 2022. According to the company, Orin uses a 7-nanometer production process to achieve a computing power of 200 TOPS, 7 times that of the previous generation, the Xavier SoC. Even with the improvement in computing performance, Orin's baseline power consumption is just 45 watts, relatively equivalent to the low power consumption of the previous generation SoC.

In addition, Li Auto will provide end users with upgradeable solutions, ranging from L2+ autonomous driving with a single Orin chip and 200 TOPS in computing power, to L4 with dual Orin chips and 400 TOPS in computing power. In the future, the use of discrete GPU can further increase the computing power up to 2000 TOPS theoretically, which provides hardware capabilities for L5 autonomous driving, according to the company.

As an automotive electronics product supplier, Desay SV will build on the Orin SoC, and provide an autonomous driving domain controller with ideal performance. On this foundation, Li Auto will develop all autonomous driving program design and algorithms, aiming to become the first new energy vehicle company to independently develop L4 autonomous driving systems in China.

Benefitting from the computing performance and extensibility of NVIDIA Orin SoC, in addition to the advanced autonomous driving domain controller developed by Desay SV, Li Auto plans to achieve the full range of autonomous driving in its next generation of vehicles.

For more information, please visit: https://www.lixiang.com/

About the Author

Tiera Oliver, edtorial intern for Embedded Computing Design, is responsible for web content edits as well as newsletter updates. She also assists in news content as far as constructing and editing stories. Before interning for ECD, Tiera had recently graduated from Northern Arizona University where she received her B.A. in journalism and political science and worked as a news reporter for the university's student led newspaper, The Lumberjack.

Follow on Twitter Follow on Linkedin Visit Website More Content by Tiera Oliver
Previous Article
New Arm Technologies Enable Safety-Capable Computing Solutions for an Autonomous Future
New Arm Technologies Enable Safety-Capable Computing Solutions for an Autonomous Future

The new suite of IP includes the Arm Cortex-A78AE CPU, Arm MaliTM-G78AE GPU, and Arm Mali-C71AE ISP, with s...

Next Article
SUBARU Selects ON Semiconductor Image Sensing Technology for its EyeSight Driver Assist Platform
SUBARU Selects ON Semiconductor Image Sensing Technology for its EyeSight Driver Assist Platform

ON Semiconductor and SUBARU redefine ADAS capability with new machine vision algorithms and sensing technol...