Eric Walz | Jan 07, 2022 10:00 AM WST

JiDU Automotive, the new electric vehicle venture formed by China's internet technology giant Baidu Inc and Chinese automaker Geely, will use the powerful NVIDIA DRIVE Orin system-on-a-chip (SoC) in its first model to support L4 autonomous driving capabilities, the company announced at CES in Las Vegas this week.

JiDU's first L4 self-driving vehicle will be unveiled at the Beijing Auto Show in April, followed by mass production and delivery in 2023.

Baidu says the JiDU intelligent autonomous driving system powered by NVIDIA DRIVE Orin will be able to handle all driving scenarios, from highways to busy urban streets. The JiDU EV will also constantly optimize its autonomous driving capabilities to get better over time. The autonomous driving system's software and algorithms were jointly developed by JiDU and Baidu.

"More and more established automakers and start-ups alike are turning to NVIDIA DRIVE Orin for building their software-defined cars with advanced AI capabilities," said Tong Liu, General Manager of NVIDIA's automotive business in China. "Working closely with Baidu and JiDU, we're helping bring to the market a new generation of vehicles that will offer consumers a safe and intelligent driving experience."

JiDU is a new smart electric vehicle company backed by Baidu and Chinese automaker Geely, which is the parent company of Swedish automaker Volvo Cars. The company was officially established in March 2021. JiDU is committed to become one of the world's top smart car and autonomous driving developers.

In April, JiDU's chief executive Xia Yiping told Reuters that the company aims to invest 50 billion yuan ($7.7 billion) into producing autonomous cars over the next five years.

JiDU's goal is to create and design a "human-centric, smart robotic vehicle" powered by advanced AI technology and the ability to learn. The company has accomplished a great deal in under one year.

"Since our formation last year, JiDU has begun mass production software R&D verification based on the software-integrated simulation prototype SIMUCar, bringing innovation and high efficiency in the field of intelligent development," said JiDU CEO Yiping Xia.

JiDU's "SIMUCar" has successfully demonstrated intelligent autonomous driving in both urban and high-speed areas.

In June of last year, JiDU announced it hired Frank Wu, formerly at Cadillac, to lead its design studio in China. Prior to joining JiDU, Wu spent seven years at General Motors, where he led the design of the Cadillac CT6.

In addition to its L4 autonomous driving capabilities, JiDU's forthcoming EV will be able to recognize a driver's voice. The smart EV will also be able to learn over time based on a driver's habits.

"The JiDU vehicle has been designed to operate under the concepts of free movement, natural communication, and self-improvement," said Robin Li, Baidu co-founder and CEO.

The free movement he is referring to is the L4 autonomous driving capability.

In addition to the NVIDIA-powered autonomous driving system in JiDU's first vehicle, the intelligent cockpit system of the EV will be based on the 4th generation of Qualcomm's Snapdragon Automotive Cockpit Platform. It will also be combined with software jointly developed by JiDU and Baidu.

"In 2023, JiDU's first mass-produced product will surely become the benchmark for smart cars," said Xia.

Baidu is one of China's leading developers of autonomous driving technology. The company is only one in China to rank among Guidehouse Insights' global automated driving leaders for two consecutive years. As part of Baidu's Apollo autonomous driving platform, the company's fleet has accumulated over 13 million miles of L4 autonomous driving.

Baidu has 13,000 AI patents, ranking first in the number of AI patent applications in China. Baidu is also a leader in deep-learning applications with 3,400 intelligent driving patents, according to the company. Many of these advanced technologies will be offered in JiDU's vehicles.

NVIDIA DRIVE Orin is specifically designed for software-defined cars, enabling continuous over-the-air software updates throughout the car's entire service life. The Orin SoC can perform 254 trillion operations per second(254 TOPS) to support autonomous driving functions, as well as powering instrument clusters, infotainment systems and passenger interaction using AI.

DRIVE Orin will be available for automakers starting later this year.

In June 2021, Volvo announced that its new XC90 electric SUV will use NVIDIA's Orin SoC to power its autonomous driving and safety systems.

Autonomous truck developer TuSimple, also uses the NVIDIA DRIVE Orin SoC to power its advanced "autonomous domain controller" (ADC) for TuSimple's Level 4 autonomous trucking applications.

Baidu is often referred to as the "Google of China", has been working to deploy commercial autonomous driving technology at scale as part of its open Apollo platform since 2017. Apollo is designed to speed up the development of self-driving technology through collaboration with industry partners.

Since launching in 2017, Apollo's development work has grown significantly. Baidu is now working with over 200 industry partners, including automakers BMW, Daimler, Volkswagen and hardware manufacturers Nvidia, Intel Corp and Qualcomm.

The Apollo autonomous driving platform is often referred to as the "Android of the Automotive Industry." It's grown to become the largest open-source autonomous driving platform in the world, according to Baidu.

Baidu also plans to work with other automakers in the future to install its autonomous driving system in a higher number of vehicles at the factory. The company's senior corporate vice president, Li Zhenyu, said last year that it plans to have its Apollo intelligent driving solutions available in more than one million vehicles in the next three to five years.

https://m.futurecar.com/5113/The-First-Model-From-Baidus-New-EV-Brand-JiDU-to-Use-the-NVIDIA-DRIVE-Orin-System-on-a-Chip-to-Support-L4-Autonomous-Driving

What is interesting is that the Americans cannot do this alone, meaning the chip development. The American chip company had to collaborate with the Chinese manufacturer, because the car maker will tell Qualcomm and Intel what they need that chip to do, and then Qualcomm and Intel would have to design it. Sounds like this is the process to make it work.

That seems to be how the autonomous driving is coming along in China, with a lot of inputs from American chip makers. An old Financial Times article for reference.