China's Auto Industry Shifts to AI-Driven Intelligence

The automotive industry in China is experiencing rapid changes, the tech trend is transforming from traditional intelligence into AI-driven intelligence, while other auto markets are still undergoing the stage of electrification.

At the end of September, the Global Intelligent Vehicle Summit (2024 GIV) is held in China’s Anhui province, with a theme of "The Development of Smart Vehicles in the Era of Artificial Intelligence".

Zhang Yongwei, Deputy Chairman and Secretary-General of the China EV100, gave a speech on the event, emphasizing that China’s auto industry is undergoing a massive technological change, moving towards AI-driven intelligence fast.

As we all know that, the auto industry has underwent electrification, then intelligence, while recently, the industry is facing a new wave of automotive AI-driven intelligence. Chinese companies, including BYD, NIO, Huawei, and Xpeng are investing billions in AI technology.

This rapid evolution in the auto industry including the need for development and manufacturing of chips, the establishment of automotive computing infrastructures, and the development of controllable operating system.

Chip Challenge

The development and manufacturing of chips is crucial for intelligent driving and smart cabin, especially the high-computing power chips. In addition, the establishment of localized supply chain for the manufacturing of large amount of chips for function and sensors are needed, as well.

Now, the products from Chinese chip suppliers are widely used in various models from main carmakers like BYD, SAIC Motor, Great Wall Motor, Dongfeng and Geely. NEV startups, such as Li Auto, NIO and Xpeng has been working on their in-house developed chips.

Automotive Computing Infrastructures

AI technology needs massive data processing and computing capacities. Data shows, in the field of end-to-end intelligent driving, the start computing power in 1 EFLOPS, while the average investment in computing power by Chinese carmakers is 3 EFLOPS. Thus, to build a large scale of computing team and facilities are crucial for carmakers.

Controllable Operating System

The necessity of developing their own operating systems and reducing dependence on external suppliers are recognized by most of the carmakers, to prevent terminal incidents.