In the field of new energy vehicles, with the rapid popularization of technologies such as intelligent driving and intelligent cockpits, related AI chips are also gradually adopting advanced process technologies, and domestic chips are also showing strong competitiveness in this field.
Huawei's intelligent driving chip, Ascend 610, adopts a 7nm process, with an AI computing power of 200TOPS (INT8), supports multi-sensor fusion and anthropomorphic intelligent driving algorithms, and has been installed on models such as the M5 and Avita 11.
The "Dragon Eagle No. 1" launched by CoreEngine Technology, a smart technology company invested by Geely Holding Group and jointly funded by ARM China, is the first 7nm automotive-grade SoC chip in China, integrating 8.8 billion transistors and supporting multi-screen linkage and cabin-to-parking solutions. It is currently used in many models of Geely, including Lynk & Co 03, Lynk & Co 05, and Lynk & Co 08. On March 29, 2025, CoreEngine Technology officially released its 7nm automotive-grade autonomous driving chip "Xingchen-1" and the corresponding smart cockpit and smart driving full range of solutions at the Eco-Tech Day. The corresponding chips will be widely used in vehicles in 2026.
BYD has also made major breakthroughs in the semiconductor field. Its exclusive AI smart cockpit uses the latest developed BYD 9000 chip. This chip uses the leading 4nm process technology and is built on the Arm v9 architecture, providing the chip with powerful computing power. In terms of communication, the BYD 9000 chip integrates a 5G baseband and supports the latest 5G network standards, ensuring the high-speed and stable operation of the vehicle's intelligent network functions, and providing strong support for efficient data exchange between vehicles and the cloud, as well as cutting-edge technologies such as autonomous driving. Its many similarities in specifications with MediaTek's automotive-grade smart cockpit chip MT8673 have also triggered widespread speculation in the market about the cooperation between the two. But in any case, the advent of BYD 9000 chip has undoubtedly added new strong impetus to BYD's development in the field of intelligent connected vehicles.
Good news also came from NIO. Chairman Li Bin announced that the tape-out process of the world's first automotive-grade, high-performance intelligent driving chip using 5-nanometer process, NIO Shenji NX9031, was successful, marking an important step for NIO in the field of automotive chips. The chip has more than 50 billion transistors and supports 32-core CPUs.
The new generation of AI cockpit chips X10 series released by CoreDrive Technology uses 4nm advanced process. Compared with the 7nm and 5nm processes commonly used in current mainstream high-end automotive chips, 4nm has significantly improved transistor density, performance, and power consumption control. It has 40 TOPS NPU and 154 GB/s ultra-large bandwidth, supporting 7B multi-modal large model end-side deployment.
Xpeng Motors has also achieved remarkable results in chip research and development. In the Q1 2025 earnings call, Chairman He Xiaopeng revealed that the Xpeng Turing chip will be successfully taped out in 2024, with a computing power of 3-7 times that of the current mainstream car-side AI chips, and the project is currently progressing smoothly. Some models have entered the production stage in the second quarter, and Turing chip models will achieve a wider market release in the third quarter. In the future, Turing chips will also be deployed on fifth-generation robots, greatly improving the end-side computing power of robots. This AI chip, which took four years to carefully polish, uses a 7-nanometer process technology. Through the dual-NPU heterogeneous architecture design, with independent safety islands and dual ISP image processors, it can complete decision-making planning for complex urban scenes within 200 milliseconds. At night and in bad weather such as rain and fog, the target recognition accuracy rate is increased to 99.7%, providing strong technical support for the development of intelligent driving and robotics.