消息面上，12月18日IT之家消息，据36氪援引知情人士消息称，理想汽车正式筹建硅谷的AI研发中心， 将负责智能化技术研发，已在数月前开启人员招聘。在该中心筹建之前，理想在北美有一个小型研发团 队，支持芯片研发及其他AI相关工作。而这次的动作，旨在将硅谷团队升级为一个真正的研发中心。 （原标题：港股异动 | 理想汽车-W(02015)尾盘涨近4% 据报公司正在筹建硅谷AI研发中心 聚焦智能化 技术） 智通财经APP获悉，理想汽车-W(02015)尾盘涨近4%，截至发稿，涨3.65%，报65.35港元，成交额6.87 亿港元。 报道称，该中心此次扩建主要面向辅助驾驶领域并希望招募"具备前沿AI背景"的高端人才。除硅谷AI研 发中心外，理想首座海外研发中心位于德国慕尼黑。该中心今年1月开业，负责前瞻造型设计、功率半 导体、智能底盘和电力驱动的下一代技术预研。理想汽车在国内的两座研发中心则分别位于北京、上 海，主要负责核心技术突破及整车研发。 ...

Core Insights - The global AI server shipments are expected to grow by over 20% year-on-year in 2026, driven by increased capital expenditure from North American CSPs and the rise of sovereign cloud initiatives [1][2] - The competition in the AI chip market is intensifying, with companies like AMD and various Chinese firms enhancing their self-developed ASIC capabilities, challenging NVIDIA's dominance [1][2] Group 1: AI Chip and Cooling Technologies - The thermal design power (TDP) of AI chips is projected to rise from 700W for NVIDIA's H100 and H200 to over 1,000W for upcoming models, necessitating liquid cooling systems in server cabinets, with a forecasted penetration rate of 47% for liquid cooling in AI chips by 2026 [2] - Microsoft is introducing new microfluidic cooling technologies for next-generation chip packaging, while the market is expected to transition from liquid-to-air (L2A) to liquid-to-liquid (L2L) cooling designs [2] Group 2: Memory and Data Transfer Innovations - HBM and optical communication technologies are becoming critical for overcoming bandwidth limitations in AI computing, with HBM4 expected to enhance I/O bandwidth and local bandwidth for AI chips [3][4] - The introduction of 800G/1.6T pluggable optical modules is underway, with expectations for higher bandwidth SiPh/CPO platforms to be integrated into AI switches starting in 2026 [4] Group 3: NAND Flash and Storage Solutions - NAND Flash suppliers are accelerating the development of specialized solutions to address the performance gap in AI training and inference workloads, including storage-class memory (SCM) SSDs and Nearline QLC SSDs [5][6] - QLC technology is anticipated to achieve a 30% market penetration in enterprise SSDs by 2026, significantly reducing the cost of storing large AI datasets [6] Group 4: Energy Storage Systems - AI data centers are evolving towards large-scale clusters, with energy storage systems transitioning from emergency backup to core energy solutions, expected to grow from 15.7 GWh in 2024 to 216.8 GWh by 2030, reflecting a CAGR of 46.1% [7] - North America is projected to become the largest market for AI data center energy storage, driven by major cloud providers [7] Group 5: Power Infrastructure and Semiconductor Demand - Data centers are shifting to 800V HVDC architectures to enhance efficiency and reliability, with third-generation semiconductors (SiC/GaN) expected to penetrate 17% of data center power supply by 2026 [8] Group 6: Advanced Semiconductor Technologies - The transition to 2nm GAAFET technology is underway, emphasizing higher transistor density and heterogeneous integration to meet the demands of AI applications [9] Group 7: Humanoid Robots and Market Growth - The global shipment of humanoid robots is projected to increase by over 700% in 2026, focusing on AI adaptability and application-specific designs [10][11] Group 8: Display Technology Advancements - OLED technology is set to accelerate in laptops, with Apple expected to introduce OLED panels in MacBook Pro by 2026, leading to a significant increase in OLED penetration in the laptop market [12][13] Group 9: Autonomous Driving and Robotaxi Expansion - The penetration rate of L2 and above advanced driver-assistance systems is expected to exceed 40% by 2026, with Robotaxi services expanding globally beyond just China and the US [15]

凤凰网科技讯（作者/李治钦）11月21日，广州车展宝马展台举办了新世代技术媒体沟通会。宝马透 露，明年推出的新世代iX3将首发宝马独有的4695大圆柱电池，可实现900+千米的CLTC续航里程，10 分钟可补能400+千米。 宝马表示，其46毫米直径大圆柱电池，正极采用高镍三元+圆柱钢壳，实现提升能量同时稳定镍元素的 作用。负极采用硅碳负极+定制电解液，延长电池寿命。新电池能量密度较前代棱柱型电池提升20%， 搭配800V高压平台和自研碳化硅逆变器，充电性能和能后将有大幅提升。 在首发车型新世代iX3上，可实现大空间、长续航、更安全和超快充四大亮点。由于采用无模组和电池 车身一体化设计，车内纵向空间将更加宽裕；能量密度提升后，新车可实现900+千米CLTC续航里程； 4695大圆柱电池采用钢壳设计，结合底置防爆阀，可实现更好的安全性；补能方面10分钟可补能400千 米，21分钟可充电至80%。 此外，宝马还展示了新世代车型的座舱交互、驾控技术以及与Momenta合作开发的辅助驾驶技术。人机 交互方面，仪表台前方的4K投影区域，采用宝马纳米涂层玻璃，可在佩戴偏光墨镜的情况下正常使 用。辅助驾驶方面，Moment ...

Core Insights - Li Auto received the "Outstanding Contribution Award" at the 2025 World Internet Conference for its innovations in artificial intelligence and future mobility technologies [1][3] - The award highlights Li Auto's self-developed advanced driver assistance technology, which was selected from over 400 global technological achievements [1][3] Group 1: Award Recognition - The World Internet Conference aims to recognize individuals and companies that have made significant contributions to global internet development [3] - Li Auto was distinguished as a "Growth Potential" award recipient, standing out among nearly 200 applicants, indicating high recognition in the integration of AI technology and smart vehicles [3] Group 2: Technological Advancements - Li Auto has established a comprehensive capability from academic research to practical application, focusing on converting cutting-edge theories into user-perceptible technologies and products [4] - The company has made significant breakthroughs in AI products, particularly in advanced driver assistance and smart cockpit technologies, supported by a robust R&D system and substantial investment [4] - In 2024, Li Auto underwent two technological architecture transformations for its driver assistance system, transitioning from rule-based algorithms to an AI era centered on imitation learning and reinforcement learning [4] Group 3: AI Innovations - Li Auto launched the world's first VLA driver model in August, which possesses five core capabilities: spatial understanding, thinking, communication and memory, behavior, and iteration, providing a "personal driver" experience [4] - The MindGPT multimodal cognitive model, developed by Li Auto, is the first self-developed model by an automotive company to be registered under China's "Interim Measures for the Management of Generative Artificial Intelligence Services" [6] - The Li Auto assistant has evolved from a smart voice assistant to an intelligent agent, capable of tool usage, complex task completion, and memory understanding, enhancing user convenience [6] Group 4: Future Directions - Li Auto aims to continue its commitment to open collaboration and transform the values advocated by the World Internet Conference into practical pathways for the automotive industry, promoting high-quality industry development [6]

Core Viewpoint - The article emphasizes the importance of clarifying the boundaries of assisted driving capabilities in the automotive industry to enhance safety and consumer understanding [3][4]. Group 1: Misunderstanding of Assisted Driving - The CEO of Momenta, Cao Xudong, highlights that there is a significant misunderstanding among consumers regarding the capabilities of assisted driving technology, which can lead to unrealistic expectations [5][6]. - A recent incident involving a Haobo GT vehicle crashing into a stationary construction vehicle while using adaptive cruise control has raised concerns about the technology's ability to recognize static objects [6][9]. - The complexity of recognizing common obstacles, such as a 50 cm cardboard box, poses significant challenges for assisted driving systems, which often rely on high-performance sensors [6][10]. Group 2: Technical Challenges - The article discusses the technical logic behind the "counterintuitive" nature of assisted driving systems, where simple tasks for humans can be complex for machines due to their reliance on data and probability [10][11]. - The difficulty in recognizing stationary objects is attributed to the fact that many static items on the road do not require avoidance, leading to a cautious approach by assisted driving systems to prevent unnecessary braking [11][12]. - The industry is working on improving technology through multi-sensor fusion and collecting extreme case data to address these challenges and enhance the recognition capabilities of assisted driving systems [12][13]. Group 3: Industry Responsibility and Training - The industry leaders stress the need for a strong sense of responsibility in the development of assisted driving technologies, contrasting it with the more flexible nature of software products [12][13]. - Companies like Momenta are actively engaging in training sales personnel to better communicate the capabilities and limitations of assisted driving systems to consumers [13].

Core Insights - The safety of advanced driver-assistance systems (ADAS) has become a significant concern in the automotive industry following several accidents, highlighting the need for clearer communication regarding the capabilities and limitations of these technologies [2][3] Group 1: Misunderstandings in Consumer Perception - There exists a cognitive gap between consumer expectations and the actual capabilities of ADAS, leading to misconceptions about what these systems can handle, particularly in complex scenarios [3][4] - A recent incident involving a vehicle colliding with a stationary construction vehicle while using adaptive cruise control has raised questions about the system's ability to recognize static obstacles [3][5] - Consumers often believe that ADAS should easily manage common obstacles, such as large objects on the road, which reflects a widespread misunderstanding of the technology's limitations [3][6] Group 2: Technical Challenges - The detection of common obstacles, such as a 50 cm cardboard box, poses significant challenges for current ADAS technology, which relies heavily on high-performance sensors like LiDAR [4][6] - The complexity of recognizing static objects is compounded by the fact that many of these objects are part of the road environment and do not require avoidance, leading to potential safety issues if the system reacts inappropriately [6][7] - The industry is working on improving ADAS through multi-sensor fusion and extensive data collection to address these challenges and enhance the system's understanding of various scenarios [7][8] Group 3: Industry Response and Training - Companies are actively engaging in training programs for sales personnel to better communicate the capabilities and limitations of ADAS to consumers, ensuring that users have a realistic understanding of the technology [8] - The industry recognizes the need for a responsible approach to developing and deploying ADAS, emphasizing the importance of quality and safety in engineering practices [7][8]

Core Viewpoint - The article discusses the rising interest and investment in embodied intelligence within the automotive industry, highlighting key figures and funding rounds that indicate a trend towards this technology [5][6][7]. Group 1: Investment and Funding - The startup focused on embodied intelligence has attracted significant attention, securing approximately $50 million in total funding from prominent investors such as Yuanjing Capital, Sequoia Capital, and BlueRun Ventures [5][6]. - Another startup in the same field, Itstone Zhihang, raised $120 million in its angel round, setting a record for the largest angel round financing in China's embodied intelligence sector [7]. Group 2: Key Personnel and Expertise - Wang Kai, a former CTO of Li Auto and now a partner at Yuanjing Capital, is a key figure in the startup, bringing extensive experience in AI technology and engineering from his previous roles [6]. - The project also includes a core technology executive from a leading automotive company, who has a strong background in autonomous driving and has been involved in end-to-end production delivery processes [6][9]. Group 3: Industry Trends - The embodied intelligence sector is gaining traction as a new growth point for automotive companies, with many executives from the autonomous driving field transitioning to this area for entrepreneurial opportunities [8][9]. - Companies like Tesla and XPeng are viewing embodied intelligence as a significant avenue for future development, indicating a competitive landscape for talent and resources in both autonomous driving and embodied intelligence [9].

Core Viewpoint - The rapid development of China's advanced driver-assistance systems (ADAS) is marked by a 62.1% growth rate in the first half of 2025, indicating a shift from laboratory research to industrial application, despite facing challenges in safety, reliability, and standardization [1][3][5]. Technological Advancements - The growth in the ADAS market is driven by significant reductions in hardware costs and breakthroughs in algorithms, with laser radar costs decreasing by up to 1000 times over the past decade and computing power increasing by 1000 times from 2014 to 2024 [3][4]. - The transition from rule-based algorithms to data-driven models using deep learning is enhancing system performance, particularly in perception, prediction, planning, and control [3][4]. Application Scenarios - City Navigation Assistance (City NOA) is becoming a competitive focal point, utilizing multi-sensor fusion and high-precision mapping to enable automated driving in complex urban environments [4]. - Despite advancements, interoperability issues among different vehicle models and systems remain a challenge, complicating cross-model and cross-regional compatibility [4]. Safety and Reliability - Safety is a critical concern, with over 90% of traffic accidents attributed to human error, and ADAS aims to mitigate these risks [5][6]. - Current systems still face reliability issues, as evidenced by incidents where systems failed to recognize traffic signs or obstacles, highlighting the need for improved user education and clearer communication of system capabilities [6][7]. Standardization and Collaboration - The industry is experiencing fragmentation due to varying regional standards and infrastructure, which hampers widespread adoption of ADAS technologies [8]. - Experts advocate for a unified "vehicle-road-cloud" standard and data exchange protocols to facilitate interoperability and reduce redundant investments [8][9]. Industry Cooperation - Collaboration among automakers, chip manufacturers, and technology providers is essential for enhancing system performance and reducing costs, with some companies already achieving total costs for urban ADAS systems below 10,000 yuan [9]. - Regulatory oversight and industry self-regulation are crucial to ensure transparency and accountability, particularly regarding safety and system limitations [9].

Core Insights - Xiaomi Group reported a total revenue of 116 billion yuan for Q2 2025, representing a year-on-year growth of 30.5%, and a net profit of 11.9 billion yuan, up 134.2% year-on-year [1] - Adjusted net profit, based on non-IFRS measures, reached 10.8 billion yuan, marking a 75.4% increase year-on-year [1] IoT Business Performance - Xiaomi's AIoT segment showed strong growth in both domestic and international markets, benefiting from the expansion of its new retail system [3][4] - The company has shifted its focus from rapid expansion to "scaled closure" in its overseas retail strategy, aiming to open 400 to 500 new stores by the end of the year, with a potential for over 1,000 stores in the following year [4] - The growth rate in the Chinese market is higher than in overseas markets, primarily due to the rapid development of the home appliance sector [3] Mobile Business Insights - The gross margin for the mobile business saw a quarter-on-quarter decline, attributed to rising component costs and a limited number of new product launches in Q2 [5][6] - The company anticipates a recovery in gross margin by Q4 2025, coinciding with a more significant release of new products [5] R&D Investments - Xiaomi's R&D expenditure increased by over 40% year-on-year in Q2, marking the fastest growth rate in recent years, with investments spread across core technologies such as chips, AI, and the Surge OS [7][8] - The company is committed to enhancing its technological capabilities, which are deemed essential for maintaining competitiveness in the market [7][8] AIoT Margins and Market Dynamics - The IoT business's gross margin showed a year-on-year increase but faced some pressure on a quarter-on-quarter basis, influenced by promotional activities like the "618 shopping festival" [6][9] - Despite competitive pressures in the home appliance market, Xiaomi remains on track to meet its annual targets for the IoT segment [7]

Group 1 - SAIC Motor sold 338,000 vehicles in July, a year-on-year increase of 34.2%, achieving seven consecutive months of year-on-year growth in sales [2] - GAC Trumpchi's new model, the Xiangwang S9 Qiankun series, features all-wheel drive and improved off-road capabilities, with a 35% increase in all-terrain passability and a 40% reduction in slip [2] - Chery Automobile became the first Chinese brand to export over 5 million vehicles, with 119,100 vehicles exported in July and a total of 669,300 vehicles exported from January to July [2] Group 2 - Xiaomi's founder Lei Jun stated that the company has made significant investments in driver assistance technology, indicating a commitment to improving this area [2]