" 当英伟达Thor芯片延期，外部芯片供应链已变得不确定且昂 贵。自研芯片能进一步拉低成本、讲好企业的技术故事。 " 作者丨梁辰 编辑丨李雨晨 智能驾驶的算力军备竞赛始于2021年。 英伟达Orin-X芯片横空出世，254TOPS的算力吊打Mobileye Q5H芯片还有特斯拉的HW3.0。在这一年 前后，蔚小理陆续开始芯片自研之路。 在车企的叙事逻辑中，一直喜欢且坚持做"难而正确的事"——自研芯片就是重要一环。 四年之后的今天，当何小鹏在业绩会上宣称"图灵芯片将把智驾能力提升数十倍"时，李斌正默默将芯片团 队剥离体系独立求生，而李想在推动自研芯片尽早上车销售——他们争夺的早已不只是销量，而是一张通 往下一代智能驾驶世界的门票。 芯片是否有必要自研是一个被讨论了无数次的话题。在不少业内人的眼里，他们并不看好车企的自研芯 片。除了至少100万颗摊薄成本的经济账要算，使用范围覆盖约1-2代车型的"有效期"也在不断弱化芯片 自研的合理性。 另一个重要原因在于，智驾的算法和芯片团队是一对天然的"死敌"： 算法要效率，不断迭代。芯片要确 定性，最好算法能用十年 。目前主流的芯片模型是在Transformer架构下支持端 ...

Core Insights - The evolution of the new energy vehicle (NEV) industry in China is marked by significant advancements in technology, with a focus on electric vehicles transitioning from basic functionality to intelligent, autonomous systems [1][2][12] Group 1: Energy System Transformation - The core breakthrough in the next decade will shift from "charging anxiety" to "mobile power stations," driven by the commercialization of solid-state batteries with energy densities exceeding 500Wh/kg, enabling 800 km range with just 10 minutes of charging [2][4] - The development of sodium-ion and lithium-sulfur batteries is expected to reduce costs by 50% by 2030, allowing vehicles to act as nodes in a distributed energy network [2] Group 2: Vehicle Redefinition - The traditional mechanical definition of vehicles is evolving into a software and AI-defined era, with innovations such as 4D printed chassis and self-repairing materials [8][9] - The introduction of lightweight materials and advanced manufacturing techniques will significantly enhance vehicle efficiency, with expected improvements in energy consumption from 7 km per kWh to 12 km per kWh [11] Group 3: Intelligent Interaction - The integration of advanced AI technologies will transform vehicles into "space robots," capable of autonomously managing passenger comfort and safety, with systems that can predict health issues [12][14] - The establishment of intelligent road systems is projected to reduce accident rates significantly, enhancing overall traffic safety [14][16] Group 4: Industry Restructuring - The automotive market is expected to evolve into a "6+N" structure by 2030, where a few major players dominate alongside niche brands, emphasizing the importance of energy ecosystem control [17][20] - The competition landscape is shifting, with Chinese automakers establishing zero-carbon factories in response to EU carbon tariffs, reducing export costs [20] Group 5: Social and Ethical Changes - The role of vehicles is changing from private property to public assets, with models allowing shared usage for community services, significantly lowering ownership costs [22][24] - The rural market is experiencing explosive growth, with affordable electric vehicles transforming logistics and transportation for agricultural purposes [24][25] Group 6: Future Vision - By 2035, NEVs are anticipated to evolve beyond mere transportation tools, becoming integral components of urban infrastructure and rural economies [25][28] - The transition from mechanical to digital civilization signifies a profound shift in human mobility and freedom, redefining societal norms and interactions [28]

Core Viewpoint - The Lynk & Co 10EM-P, a mid-size plug-in hybrid sedan, is set to redefine the market with its competitive pricing starting at 192,000 yuan, featuring standard all-wheel drive and lidar technology, which challenges traditional luxury car pricing strategies [1][6]. Group 1: Performance and Technology - The Lynk & Co 10EM-P is the only plug-in hybrid sedan in the 200,000 yuan range to offer standard all-wheel drive and lidar, breaking the norm that associates these features with high-end models [2]. - Built on the CMAEvo platform, the vehicle features advanced suspension systems and has demonstrated superior performance in various tests, including a 83.2 km/h in the moose test, outperforming a German luxury competitor [2]. - The car's powertrain includes a 1.5TEvo engine with a thermal efficiency of 47.26% and a low fuel consumption of 4.2L/100km, achieving a 0-100 km/h acceleration in just 5.1 seconds [3]. Group 2: Intelligent Features - The Lynk & Co 10EM-P is equipped with 29 sensors, including lidar and multiple radar systems, ensuring comprehensive monitoring for enhanced driving safety [5]. - It is the first car globally to feature the NVIDIA Thor chip, which offers 700 TOPS of computing power, enabling advanced AI-driven driving assistance capabilities [5]. - The vehicle's advanced driving assistance system, powered by the Thor chip, aims to provide a seamless driving experience across various conditions, making high-end technology accessible to a broader audience [5]. Group 3: Market Impact - The launch of the Lynk & Co 10EM-P is expected to disrupt the mid-size hybrid sedan market, setting new standards for value and technology in the 200,000 yuan segment [6]. - The vehicle's pricing strategy and feature set challenge the traditional luxury car market, potentially leading to a shift in consumer expectations regarding hybrid vehicles [4][6].

Core Viewpoint - The new generation of Zhiji LS6, a five-seat intelligent SUV, has officially started pre-sales, offering both super extended range and pure electric versions, aiming to disrupt the large SUV market with advanced technology and luxury features [1][21]. Group 1: Product Features - The new generation Zhiji LS6 features the "Lingxi" digital chassis, integrating advanced electronic architecture and top-tier hardware for enhanced driving performance, including a unique tire blowout stability control and active anti-roll technology [3][11]. - The vehicle offers a spacious interior with a high space utilization rate of 90%, innovative seating configurations, and luxury amenities such as a dual-door refrigerator and advanced sound systems [11][19]. - It is equipped with a 27.1-inch 5K cockpit screen and a 15.6-inch entertainment screen for an immersive user experience, along with upgraded safety features for adverse weather conditions [6][10]. Group 2: Pricing and Models - The new generation Zhiji LS6 is available in four models with pre-sale prices ranging from 209,900 to 289,900 yuan, with specific models priced at 209,900 yuan for the pure electric Pro and 279,900 yuan for the pure electric Ultra [4][21]. Group 3: Technological Innovations - The "Hengxing" super extended range technology, developed in collaboration with CATL, allows the LS6 to achieve a range of 1500 km, addressing common concerns about range anxiety and charging convenience [15][19]. - The vehicle features an industry-leading 66 kWh battery with a pure electric range exceeding 450 km and ultra-fast charging capabilities, allowing for a 310 km range boost in just 15 minutes [17][19]. Group 4: Market Positioning - The new generation Zhiji LS6 is positioned as a top choice in the 200,000 yuan segment for intelligent five-seat SUVs, emphasizing its electric and intelligent travel transformation [21].

Core Viewpoint - The launch of Lynk & Co's EM-P AI Plug-in Hybrid 2.0 showcases the company's commitment to integrating AI technology into its electric hybrid vehicles, enhancing performance, safety, and user experience in the mid-to-large plug-in hybrid sedan market [1][9]. Group 1: Product Features - The Lynk & Co EM-P AI Plug-in Hybrid 2.0 features a 1.5T Evo engine and DHT Evo hybrid drive system, achieving a combined power of 390 kW and torque of 755 N·m, with a 0-100 km/h acceleration time of approximately 5.1 seconds [5]. - The vehicle is equipped with a four-wheel drive system and laser radar, allowing real-time adjustment of power output based on external conditions, ensuring stability and control in various driving scenarios [5]. - The car's fuel consumption is approximately 4.2 L/100 km under CLTC conditions, balancing performance and economy [5]. Group 2: Technology Integration - The vehicle incorporates the LYNK Flyme Auto system, powered by NVIDIA Thor chip and Qianli Haohan H7 driver assistance system, enabling advanced driver assistance features and multi-scenario cruising [6]. - The AI-driven Starry AI Cloud Power 2.0 system utilizes deep learning and high-precision environmental data to optimize travel strategies across over 1,000 scenarios [3]. Group 3: User Experience - The interior offers 1.2 m² of space per passenger, with features such as ventilation, heating, and massage functions for all seats, along with a Harman Kardon 23-speaker sound system for an immersive audio experience [6]. - The EM-P AI Plug-in Hybrid 2.0 enhances the overall travel experience by combining performance, intelligence, safety, and comfort [9].

Core Viewpoint - The discussion around VLA (Vision-Language-Action model) is intensifying, with contrasting opinions on its short-term feasibility and potential impact on the automotive industry [2][12]. Group 1: VLA Technology and Development - The Li Auto i8 is the first vehicle to feature the VLA driver model, positioning it as a key selling point [2]. - Bosch's president for intelligent driving in China, Wu Yongqiao, expressed skepticism about the short-term implementation of VLA, citing challenges in multi-modal data acquisition and training [2][12]. - VLA is seen as an "intelligent enhanced version" of end-to-end systems, aiming for a more human-like driving experience [2][5]. Group 2: Comparison of Driving Technologies - There are two main types of end-to-end technology: modular end-to-end and one-stage end-to-end, with the latter being more advanced and efficient [3][4]. - The one-stage end-to-end model simplifies the process by directly mapping sensor data to control commands, reducing information loss between modules [3][4]. - VLA is expected to outperform traditional end-to-end models by integrating multi-modal capabilities and enhancing decision-making in complex scenarios [5][6]. Group 3: Challenges and Requirements for VLA - The successful implementation of VLA relies on breakthroughs in three key areas: cross-modal feature alignment, world model construction, and dynamic knowledge base integration [7][8]. - Current automotive chips are not designed for AI large models, leading to performance limitations in real-time decision-making [9][11]. - The industry is experiencing a "chip power battle," with companies like Tesla and Li Auto developing their own high-performance AI chips to meet VLA's requirements [11][12]. Group 4: Future Outlook and Timeline - Some industry experts believe 2025 could be a pivotal year for VLA technology, while others suggest it may take 3-5 years for widespread adoption [12][13]. - Initial applications of VLA are expected to be in controlled environments, with broader capabilities emerging as chip technology advances [14]. - Long-term projections indicate that advancements in AI chip technology and multi-modal alignment could lead to significant breakthroughs in VLA deployment by 2030 [14][15].

Group 1 - The Lynk & Co 10 EM-P has been launched for pre-sale, with a price range of 192,000 to 222,000 yuan, positioning it as a mid-to-large intelligent electric hybrid four-wheel drive sedan [1] - The vehicle is equipped with a new generation 1.5T Evo hybrid engine, achieving the industry's highest thermal efficiency of 47.26%, and features a hybrid asynchronous motor for rapid response with a 10ms engagement time [1] - The system delivers a maximum power of 390 kW and a maximum torque of 755 N·m, highlighting its performance capabilities [1] Group 2 - The entire lineup is equipped with NVIDIA Thor chips and lidar technology, marking the first use of a super hybrid battery [3] - The vehicle offers a pure electric range of 240 km under CLTC standards, with a fast charging time of less than 15 minutes from 30% to 80% [3]

Group 1: New Vehicle Pre-sales - The all-new Tank 500 has commenced pre-sales with a price range of 360,000 to 388,800 yuan, featuring two models: Hi4-Z and Hi4-T [1][3] - Dongfeng Fengshen L8 has also started pre-sales, priced between 129,900 to 149,900 yuan, offering four models [2][16] - Lynk & Co 10EM-P has opened pre-sales with a price range of 192,000 to 222,000 yuan, featuring three models [2][16] Group 2: Vehicle Specifications and Features - The Tank 500's dimensions are 5078 × 1934 × 1905 mm with a wheelbase of 2850 mm, inspired by Chinese classical architecture and available in a new color, Dunhuang Green [5] - The Tank 500 is equipped with Coffee OS 3 smart cockpit system, featuring multiple screens including a 15.6-inch central display and a 17.3-inch rear entertainment screen [7] - The Dongfeng Fengshen L8 measures 4762 × 1910 × 1677 mm with a wheelbase of 2825 mm, offering various color options and a unique 4.5L cold and warm box in the rear [13][15] - The Lynk & Co 10EM-P has dimensions of 5050 × 1966 × 1487 mm and a wheelbase of 3005 mm, featuring high-end materials and advanced technology [18][21] Group 3: Advanced Technology and Performance - The Tank 500 features the third-generation intelligent driving system Coffee Pilot Ultra, with 27 sensors and 20 active safety functions, enhancing its off-road capabilities [9] - The Dongfeng Fengshen L8 includes a 1.5T hybrid engine paired with a 30.3 kWh battery, achieving a rapid charge from 20% to 80% in 18 minutes [15] - The Lynk & Co 10EM-P is equipped with the NVIDIA Thor chip and advanced driving assistance features, boasting a maximum power of 390 kW and high thermal efficiency [19][21] Group 4: Tesla's Strategic Shift - Elon Musk confirmed the dissolution of the Dojo supercomputer team, stating that Tesla will focus on AI5 and AI6 chips, indicating a strategic shift in resource allocation [4][22] - The decision to discontinue the Dojo project reflects Tesla's need to streamline its AI chip development efforts, which could impact its future market positioning [24]

Core Viewpoint - Pure GPU can achieve basic functions for low-level autonomous driving but has significant shortcomings in processing speed and energy consumption, making it unsuitable for higher-level autonomous driving needs [4][40]. Group 1: GPU Limitations - Pure GPU can handle certain parallel computing tasks required for autonomous driving, such as sensor data fusion and image recognition, but is primarily designed for graphics rendering, leading to limitations [4][6]. - Early autonomous driving tests using pure GPU solutions, like the NVIDIA GTX 1080, showed a detection delay of approximately 80 milliseconds, which poses safety risks at high speeds [5]. - The data processing capacity for L4 autonomous vehicles generates about 5-10GB of data per second, requiring multiple GPUs to work together, which increases power consumption significantly [6][9]. Group 2: NPU and TPU Advantages - NPU is specifically designed for neural network computations, featuring a large number of MAC (Multiply-Accumulate) units, which optimize matrix multiplication and accumulation operations [10][19]. - TPU, developed by Google, utilizes a pulsed array architecture that enhances data reuse and reduces external memory access, achieving higher efficiency in large matrix operations compared to GPU [12][19]. - NPU and TPU architectures are more efficient for neural network inference, with NPU showing a significant reduction in energy consumption compared to GPU [36][40]. Group 3: Cost and Efficiency Comparison - In terms of energy efficiency, NPU's performance is 2.5 to 5 times better than that of GPU, with lower power consumption for equivalent AI computing power [36][40]. - The cost of NPU solutions is significantly lower than pure GPU solutions, with NPU hardware costs being only 12.5% to 40% of those for pure GPU setups [37][40]. - For example, achieving 144 TOPS of AI computing power with a pure GPU solution requires multiple GPUs, leading to a total cost of around $4000, while a solution with NPU can cost about $500 [37][40]. Group 4: Hybrid Solutions - NVIDIA's Thor chip integrates both GPU and NPU to leverage their strengths, allowing for efficient task division and compatibility with existing software, thus reducing development time and costs [33][40]. - The collaboration between GPU and NPU in autonomous driving systems enhances overall efficiency by avoiding frequent data transfers between different chips, resulting in a 40% efficiency improvement [33][40]. - The future trend in autonomous driving technology is expected to favor hybrid solutions that combine NPU and GPU capabilities to meet the demands of high-level autonomous driving while maintaining cost-effectiveness [40].

Core Viewpoint - The article discusses the challenges faced by Li Auto due to delays in the delivery of NVIDIA's Thor chips, which are critical for the company's new range of vehicles. The delays have prompted Li Auto to accelerate its own chip development efforts. Group 1: Chip Delivery Issues - Li Auto's suppliers were notified that the launch of the upgraded L series range was postponed from March to May due to delays in the delivery of NVIDIA's Thor chips [1] - The initial mass production timeline for the Thor chip was promised for the end of 2024, but this has been pushed back multiple times [1] - The first batch of Thor chips delivered by NVIDIA faced significant engineering and design issues, leading to a reduction in promised performance from 700 TOPS to below 500 TOPS [1] Group 2: Self-Development of Chips - In response to the delays, Li Auto is accelerating its self-developed chip project, aiming for delivery in the first quarter of next year [2] - The cost of the first self-developed chips for Li Auto and its competitors is estimated to be between 300 million to 400 million USD, with ongoing investments for the second chip development [2] - Li Auto is exploring the application of large model technology in vehicles, with a focus on improving feedback and adjustment cycles through self-developed chips [2] Group 3: Challenges with NVIDIA - Li Auto faced unreasonable demands from NVIDIA during the testing of the T chip, which included a refusal to provide a list of issues and solutions [3] - NVIDIA's contract terms with Li Auto were described as lacking penalties for delays and containing unfair clauses [3] - Despite these challenges, Li Auto has successfully launched a fully functional intelligent driving system and is progressing with the upcoming VLA model [3]