According to the Star Daily on December 28 (Reporter Chen Junqing), another automotive company has begun self-research on chips.

Recently, Great Wall Motor signed a strategic cooperation agreement with Nanjing Jiangbei New Area, and the RISC-V automotive chip design enterprise Nanjing Zijing Semiconductor Co., Ltd. (hereinafter referred to as: Zijing Semiconductor) cultivated by Great Wall Motor settled in Nanjing Jiangbei New Area.

Zijing Semiconductor was established on November 29, 2024, and is a company focused on RISC-V automotive-grade chip design and development.

Products include automotive-grade chips based on open-source RISC-V.

According to data from Cailianshe Venture Capital Pass, Zijing Semiconductor was jointly funded by Tianjin Great Wall Investment Co., Ltd., under Great Wall Motor, former Chief Engineer of Great Wall Motor's EE architecture, Cao Changfeng, Nanjing Sujixin Enterprise Management Partnership (Limited Partnership), and Jiefa Semiconductor (Shanghai) Co., Ltd., with Cao Changfeng serving as Chairman. The above shareholders hold shares in the order of 41.0958%, 34.0405%, 20.7542%, and 4.1096%.

In terms of products, the Zijing M100, as the company's first star product, is the first domestically designed automotive-grade MCU chip based on the open-source RISC-V core, which was lit up in September this year.

According to reports from the Star Daily, the aforementioned chip adopts a modular design, has a reconfigurable core, provides faster processing speed and less time consumption, while meeting the functional safety ASIL-B level requirements and the ISO21434 network information security standards.

The Zijing M100 is aimed at body control, capable of handling multiple systems and meeting the diverse needs of different vehicle models and architectures, making it relatively flexible in application. It is reported that Great Wall Motor plans to widely equip the Zijing M100 chip in various models, with an expected installation volume of no less than 2.5 million units over the next five years.

Great Wall Motor Chairman Wei Jianjun mentioned in a Weibo post in September this year that the first open-source RISC-V automotive chip cultivated by China, the Zijing M100, has been successfully illuminated, and this small chip is of great significance.

▍Domestic car manufacturers are accelerating their layout in automotive chips.

Reporters from the Star have noted that besides Great Wall Motor, many domestic car manufacturers are also investing in chip development.

Among the new car-making forces, NIO and Xpeng announced the successful tape-out of their self-developed chips in July and August this year, respectively.

Specifically, NIO's intelligent driving chip "Shenji NX9031" claims to be the industry’s first high-end intelligent driving chip manufactured using 5nm automotive process technology, which is expected to be installed in the NIO ET9 model to be launched in the first quarter of next year.

Xpeng's "Turing chip" is claimed to be the world's first AI chip applied simultaneously in AI cars, AI robots, and flying cars, targeting the L4 autonomous driving sector.

As for traditional car manufacturers, SAIC Group, Dongfeng Automobile, Geely Automobile, and Great Wall Motor are beginning to invest in the chip industry on a large scale, with product layouts covering body control chips, intelligent cockpit chips, and autonomous driving chips.

For example, SAIC Group is increasing its layout in the automotive chip field by establishing industrial funds with multiple parties, investing in chip companies, and partnering with industry giants to establish joint ventures. In recent years, SAIC Group has continuously invested in chip companies such as ChuanTu Microelectronics, Shangyang Tong, and Chipsea Technology.

BYD completed its strategic investment in Topline in February 2021. In the following three years, BYD invested in nearly 80 companies, of which nearly one-third are related to the chip semiconductor field, covering segments such as AI chips, intelligent driving chips, Silicon Carbide epitaxial chips, and semiconductor devices.

At BYD's 2023 annual Shareholder meeting, Wang Chuanfu, the chairman and president of BYD Company Limited, stated that in the future, there will be an investment of 100 billion yuan in the intelligent driving field, focusing on the research and development of intelligent driving technologies, including generative AI and large models.

In March 2024, Geely's subsidiary, Core Technology, launched the high-level driving SoC chip—AD1000, which can fully meet the intelligent driving needs from L2 to L4 levels. Earlier, the first self-developed 7-nanometer automotive-grade SoC chip "Dragon Eagle 1" by Core Technology was released in December 2021, and as of December 2023, the shipment volume has exceeded 200,000 pieces.

▍Opportunities and challenges coexist for automotive companies developing their own chips.

Regarding the phenomenon of automotive companies increasingly developing their own chips, many industry professionals have expressed their views.

The new generation of automotive companies places more emphasis on the layout of AI-related chips. Xiaopeng Motors founder He Xiaopeng stated during the tenth anniversary event of Xiaopeng Motors in August this year that companies aspiring to make a mark in AI may have non-generic chips, such as proprietary AI chips like the Xiaopeng Turing chip.

NIO founder and chairman Li Bin also mentioned at the NIO IN 2024 event held at the end of July this year that NIO still insists on developing its own chips, mainly because it clearly recognizes that intelligent electric vehicles will once again become the pinnacle of technological innovation, and AI will be the core foundational capability of intelligent electric vehicle companies.

However, some industry insiders indicate that the investment in chips is too large, and the input-output often does not match, making self-developed chips not cost-effective.

Chen Shujie, Vice President of Chipstar Technology, stated during the 2024 World New Energy Vehicle Conference in September this year that the R&D costs of automotive chips are relatively high, with the total R&D investment for a large SOC automotive chip possibly requiring 1 billion yuan, and the total shipment volume needs to reach tens of millions for the company to achieve profitability, so profitability is not easy.

Chen Shujie believes that "if car companies want to self-develop chips, they must meet two conditions: first, the shipment volume must be sufficiently large; second, there must be a strong chip design foundation and software capability, otherwise the input-output will not match."

Regarding the industry impact, Zou Guangcai, Deputy Secretary-General of the China Automotive Chip Industry Innovation Strategy Alliance, stated in a media interview that if car companies self-develop chips, it will to some extent break the industry division of labor and compress the living space of component suppliers. In the future, the issues surrounding the division of labor mechanism in the automotive chip industry need to be jointly addressed by the upstream and downstream.

The 2024 annual "Research Report on the Evolution Trend of Integrated Software and Hardware for Autonomous Driving" published by Chentao Capital and others shows that integrated software and hardware autonomous driving solutions can become a trend. On one hand, this is because they can achieve higher performance, lower power consumption, lower latency, and closer integration; more importantly, they can bring significant cost advantages to enterprises.

Regarding the issue of large investment in self-developed chips by car companies and the difficulty of achieving a proportional input-output relationship, the research report states that in the case of a 7nm process and high-performance SoC of over 100+ TOPS, the R&D cost exceeds 0.1 billion dollars (including labor costs, wafer processing fees, packaging and testing fees, IP licensing fees, etc.). Liu Yudong, Managing Director of Chentao Capital, publicly stated that from the perspective of economic considerations for car companies, if the shipment volume of self-developed chips is less than 1 million pieces, it may be very difficult to balance input and output.