Car chip "battle" Nvidia, Intel, and domestic manufacturers compete in the smart race track | CES 2025

In recent years, the automotive industry is speeding towards intelligent transformation, with in-car chips playing a decisive role as the "most powerful brain" in this process. CES has become a "stage" for showcasing cutting-edge technologies in the automotive industry. During CES 2025, Qualcomm showcased a demo car equipped with the Snapdragon Cockpit Ultimate Edition platform and Snapdragon Ride Ultimate Edition. NVIDIA introduced the dual-use chip "Thor" for cars and Siasun Robot & Automation, and announced a collaboration with Toyota to develop next-generation autonomous driving technology. Many Chinese car manufacturers and chip manufacturers participated in this year's CES. Some showcased the latest intelligent cockpits and driving functions, such as JiKe, while others showcased intelligent driving-related components, like BLACK SESAME's self-developed high computational chip and Hesai Technologys high-performance 3D LiDAR. Although delayed, NVIDIA's Thor chip arrived In September 2022, NVIDIA officially released the new generation SoC chip Thor, claiming that its single chip computational power can reach up to 2000TOPS, nearly 8 times that of the current Orin product. At this year's CES 2025, NVIDIA announced that Thor has entered mass production, with a computational power increase to 5000TOPS. NVIDIA claimed that this chip can handle complex calculations required for L4/L5 level autonomous driving, as well as in-car entertainment and smart cockpit functions. It is understood that JiKe will be one of the first companies to use the Thor chip, developing driver assistance systems for more precise vehicle and pedestrian recognition. The platform is designed by the JiKe team with 2x high computational power and low power consumption. The first mass-produced native autonomous driving car, the JiKe RT, will start deliveries in 2025. Toyota is also working on building the next generation of car platforms based on the DRIVE AGX Orin platform to achieve higher-level intelligent driving functions. Aurora, Continental, and NVIDIA announced a long-term strategic partnership to deploy NVIDIA DRIVE-supported autonomous trucks in large scale. Apart from the aforementioned companies, companies like NIO, BYD, Mercedes-Benz, Volvo, and Xiaomi are also clients of NVIDIA's Thor. Notably, Xpeng and NIO are absent from the previous list. The product manager of Xpeng P and G series mentioned in an interview that due to the delay of the Thor chip, Xpeng P7+ chose to use a configuration with dual Orin-X chips. Xpeng's self-developed Turing chip was successfully sampled in late August last year. Xpeng's founder and chairman, He Xiaopeng, claimed that the chip is designed for L4 autonomous driving cars and has AI computational power close to that of 3 mainstream intelligent driving chips one chip achieves L3+ advanced intelligent driving experience, while two chips achieve L4 autonomous driving experience. NIO's self-developed chip Shenji NX9031 is built on a 5nm technology and has been first deployed on NIO's flagship executive sedan, the ET9, which will start deliveries in March 2025. Traditional car manufacturers like BYD and Dongfeng are actively developing or investing in chip industries, covering various areas such as body control chips, intelligent cockpit chips, and autonomous driving chips. Senior researcher Jiang Han from the Pangu Think Tank mentioned in an interview with the Science and Technology Innovation Board Daily, "If there is a shortage in intelligent driving chip supply, it will affect the research and development and production plans of car manufacturers." Intel "bets" on automotive solutions At last year's CES, Intel announced its entry into the automotive chip field, focusing on intelligent cockpit chips, electric vehicle energy AI management, and an open automotive chip customization platform. At CES 2025, Intel introduced a complete vehicle technology platform, including the Adaptive Control Unit ACU U310, B-series standalone graphics card for vehicles, and based on this, introduced artificial intelligence technology, AI large model software, power management, and regional controller solutions. Intel claims that in the application of electric vehicle power systems, the ACU U310 supports more advanced algorithms, automatically adjusting voltage and control frequency according to individual driving styles and road conditions to reduce vehicle energy consumption, enhance energy efficiency, and increase energy recovery in power transmission systems. Intel has already partnered with Stellantis, and their new Gen3 Evo generation FE electric formula race cars under the Maserati team will deploy Intel's ACU chips to enhance performance and efficiency in competitive racing environments. In addition, Intel announced a collaboration with Amazon's cloud computing platform, AWS, to provide efficient and flexible cloud platforms for automakers to develop, simulate, and test vehicle systems and applications. Some investors believe that Intel's recent significant decline in performance and financial struggles will need a shift towards becoming an integrated provider of automotive hardware and software ecosystems from a pure PC chip supplier to boost market sentiment. "Intel hopes to showcase a series of products and solutions at CES 2025 to demonstrate its determination in transformation to global investors." Among the domestic exhibiting companies, BLACK SESAME showcased the Huashan A2000 chip series and Wudang C1200 chip series at CES 2025. The former not only supports advanced intelligent driving scenarios but can also be applied in body intelligence and general computing, while the latter focuses on multi-domain fusion and cross-domain computing, covering various needs in the intelligent car interior like cockpit and intelligent driving. Deputy Chairman Bai Wenxi of the China Enterprise Capital Alliance believes that on one hand, with the continuous maturity and popularization of autonomous driving technology, the demand for intelligent driving will increase; on the other hand, with the continuous progress and innovation of chip technology, there will be more chips.This company has the ability and conditions to enter this field.This article is reprinted from "Cailianshe" and edited by GMTEight editor: Xu Wenqiang.