Group 1 - The core point of the news is that Leapmotor is in talks with Huawei for potential collaboration, focusing on smart driving solutions, vehicle ecosystem, or supply chain technology [2] - Leapmotor has achieved significant sales growth, becoming the top-selling new car manufacturer in China, with July deliveries reaching a record high of 50,129 units, a year-on-year increase of over 126% [2] - The collaboration with Huawei could enhance Leapmotor's brand value, particularly in key technology areas such as smart cockpit and autonomous driving [2] Group 2 - Leapmotor adheres to a "fully self-research" strategy, with its smart driving system, Leapmotor Pilot, utilizing a "pure vision + millimeter-wave radar" perception solution, emphasizing cost-effectiveness by avoiding high-cost lidar [3] - The Leapmotor Pilot system has reached version 3.0, featuring core functions such as ACC, LCC, ALC, APA, highway NOA, and city NAC, with some models offering optional lidar [3] - Leapmotor's smart driving features primarily address existing issues but still lag behind leading manufacturers like Huawei, as it heavily relies on high-precision maps and cannot achieve map-free autonomous driving [3]

在广东深圳，诞生了华为、比亚迪、大疆等众多创新型企业。它们不仅在国内市场表现突出，还凭借实 力不断开拓国际市场。这些企业的一个"秘密武器"，就是国际标准。据深圳市市场监督管理局统计，截 至2024年底，深圳市企事业单位累计参与研制国际标准3379项，主要涉及光伏、无人机、基因检测、石 墨烯等多个新兴产业领域。（深圳特区报） ...

华为版本CUDA，全面开源开放！ 最新消息，华为宣布为其昇腾AI GPU开源其CANN软件工具包。 红交 发自 凹非寺 量子位 | 公众号 QbitAI 华为轮值董事长徐直军在主题演讲中强调：华为AI战略的核心是算力，并坚持昇腾硬件变现。 基于这一背景下，徐直军在会上宣布华为昇腾硬件使能CANN全面开源开放，Mind系列应用使能套件及工具链全面开源，支持用户自主的深 度挖潜和自定义开发，加速广大开发者的创新步伐， 让昇腾更好用、更易用 。 CANN，神经网络计算架构，提供多层编程接口，帮助用户构建针对华为昇腾的AI应用程序。 它是一个由多种技能栈和算子加速库等组成的软件生态系统，换句话说，它就像是华为版本的CUDA，为GPU提供相同的接口。 巧合的是，同一天，一传奇GPU大佬创业公司终于浮出水面，他们不搞消费级GPU，而是做CUDA类似软件生态系统。 因为一旦开发者想要迁移到其他产品。那么就需要重写代码，使用生态相对不成熟的替代库，同时也会失去英伟达围绕CUDA建立起来的庞大 技术社区的支持。 之前有项目尝试将CUDA功能（通过转换层）引入其他 GPU供应商，但由于英伟达阻拦，这些项目大多未能成功。2024年 ...

Core Insights - Huawei is committed to deepening its strategic focus on the healthcare industry, emphasizing systematic approaches to digital transformation [1][4] - The healthcare sector faces three main challenges in its digital transformation: innovation in technology architecture, transformation of service systems, and conversion of talent skills [2][4] Digital Transformation Challenges - **Technological Architecture Innovation**: Traditional IT architectures are inadequate for the high demands of AI in terms of computing power, data, and real-time processing, leading to data silos [2][4] - **Service System Transformation**: There is a need for integrated services covering the entire digital transformation lifecycle, requiring a robust digital platform to connect online and offline services [2][4] - **Talent Skill Conversion**: Both users and technology providers must shift from training single-skilled personnel to developing more versatile talent, focusing on enhancing AI literacy and skills [2][4] Methodology for Digital Transformation - Huawei has developed a "Three Layers, Five Stages, Eight Steps" methodology for implementing digital transformation, which includes redefining intelligent business, AI development and delivery, and continuous operation of intelligent applications [6][10] - The methodology emphasizes collaboration with partners to create industry solutions, leveraging Huawei's ICT infrastructure [6][10] Case Study: Ruijin Hospital - Ruijin Hospital has been actively pursuing digital transformation in its pathology department since 2021, collaborating with Huawei to publish the first "Digital Intelligent Pathology Construction White Paper" in 2023 [7][8] - The RuiPath pathology model, developed in partnership with Huawei, was recognized at the "AI for Good" global summit [8] Strategic Goals - Huawei aims to bridge the "digital divide" in healthcare by making digital transformation accessible to organizations of all sizes, from large hospitals to community clinics [9][12] - The "Partner + Huawei" model allows various healthcare institutions to leverage Huawei's technology and solutions for their digital transformation [9][12] Service System Upgrades - Huawei is transitioning its service model from traditional fault repair and project delivery to a comprehensive service covering the entire digital journey, including planning, construction, operation, and iteration [10][11] - The introduction of the "Excellence Navigation Plan" provides dedicated architects to assist large institutions in their digital transformation [10][11] Collaborative Ecosystem - Huawei is building a collaborative ecosystem with partners, transforming them from mere technical executors to co-innovators, enhancing the overall service delivery in the healthcare sector [11][12] - The O3 partner service platform integrates Huawei's delivery experience and enables deep collaboration with partners [11][12] Practical Applications - Huawei has successfully implemented various digital solutions in healthcare, such as the "Digital Pathology Integrated Machine" for grassroots medical institutions and the "Ascend + DeepSeek" integrated solution for smart healthcare applications [13][14] - The collaboration with Shenzhen Nanshan People's Hospital has led to the deployment of the world's first 50G PON technology in healthcare, showcasing Huawei's commitment to advancing digital transformation in the industry [14]

8月6日，广东移动携手华为联合推动5G-A与AI技术跨界创新 ，正式发布了全球通AI手机。这不是简单 的一款新品手机，背后是移动AI时代服务升级新标杆——全球通尊享权益包，它体现了在先进技术的 硬核突破下，通信创新先锋们正将通信产品背后的人文关怀提升到一个新高度。 第三，"私享权益"则对应全球通用户的多个角色和生活的每一个细节。为全球通用户提供终端保险、绿 通挂号线上问诊线下陪诊，同时提供家财保障、意外保险；匹配机场与高铁休息室礼遇、国际漫游包 等。 生态协同 从"单打独斗"到"共赢共生"的模式创新 "尊享服务"的可持续，离不开产业链的协同发力。广东移动与华为等伙伴的深度合作，构建了"技术共 建+产品共创+营销联动" 的生态体系，为服务升级提供了长效支撑。 "全球通AI手机不仅仅是一个终端，也是一套个人数智化服务整体解决方案，是落实'人工智能+'行动、 推动AI赋能千家万户的重要举措，也是践行'民有所呼、我必有应'服务承诺，更好满足人民群众美好数 智生活需要的重要体现。"中国移动广东公司党委书记、董事长、总经理葛松海表示。 当下，5G-A与AI的跨域融合，为全球通AI手机以及其背后的个人数智化服务整体解决方案 ...

自华为鸿蒙系统问世以来，系统生态内App数量迅猛增长。记者从市场监管总局所属中国网络安全审查 认证和市场监管大数据中心(以下简称网数中心)获悉，为切实保护App用户隐私安全，助力鸿蒙生态持 续健康发展，网数中心积极开展鸿蒙生态移动互联网应用程序认证研究，通过对鸿蒙微内核架构、安全 访问机制等核心技术研究，深入了解鸿蒙系统权限分级体系和动态授权流程，结合鸿蒙系统"全场景、 跨终端"特点，分析在分布式环境下个人信息收集使用风险点，重点围绕跨设备数据流转过程中的隐私 泄露、不同终端间权限同步管理，以及在复杂场景下隐私政策执行不到位等问题的研究，持续优化鸿蒙 系统App的合规检测策略和认证规范。 截至目前，网数中心已向学习教育、金融保险行业的多款鸿蒙App颁发"App安全认证证书"和"金融科技 产品认证证书"。下一步，网数中心将持续深化鸿蒙生态移动互联网应用程序认证的研究工作，按照中 央网信办、工业和信息化部、公安部、市场监管总局联合开展的2025年个人信息保护系列专项行动工作 要求，为App专项治理提质增效提供有力支撑。(经济日报记者郭静原) 同时，为筑牢App安全防线，网数中心不断对移动应用监测平台升级改造，深度 ...

SiC负极电池具备更高的电池密度，可在相同物理尺寸下实现更大的电池容量，同时也能支持更薄的机身设计。首款搭载SiC电池的智能手机于2023年2月 推出，其大规模普及始于2024年下半年，以荣耀X60 Pro、Redmi K80和vivo S19等畅销机型为代表。该电池在中端价位段的普及不仅加速了该技术的市场 渗透程度，并有效缓解了消费者在屏幕使用时间延长以及高耗能应用(如端侧生成式AI)普及背景下所面临的关键痛点。 2025年5月，中国主要OEM厂商的平均电池容量均实现了两位数的同比增长，这使得6000mAh及以上电池容量的智能手机市场份额从一年前的9%跃升至 35%。截至2025年5月，OnePlus凭借其中国市场专属的Ace系列位居平均电池容量首位。同样，荣耀(Play系列)、华为(畅享系列)和vivo(S系列)等品牌也在 其专供中国市场的机型中配备了更大容量电池。 | | | 数据来源： Counterpoint Research《全球手机机型销量追踪报告》，2025年5月 根据Counterpoint Research最新的《全球手机机型销量追踪报告》，2025年5月中国市场智能手机的平均电池容量达到 ...

Group 1 - The Robot ETF (159770) experienced a decline of 1.22% on August 7, with a trading volume of 240 million yuan and a premium rate of 0.07% [1] - The ETF achieved a cumulative increase of 6.85% over three consecutive trading days from August 4 to August 6, with a total trading volume exceeding 1.2 billion yuan, ranking first among similar products in the Shenzhen market [1] - As of August 6, the latest scale of the ETF reached 6.665 billion yuan, also ranking first among similar products in the Shenzhen market [1] Group 2 - Among the constituent stocks, Jingpin Special Equipment rose over 8%, while other stocks like Jingye Intelligent, Huadong CNC, and Tianzhihang-U also saw gains, while Robot fell over 5% [2] - The Robot ETF closely tracks the CSI Robot Index, with industry allocations including manufacturing and information transmission, software, and IT services, featuring major holdings such as Huichuan Technology, iFlytek, and Stone Technology [2] - A recent paper published in "Nature Communications" reported the development of a programmable flexible folding robot capable of object manipulation and surface movement, which could enhance applications in autonomous systems, haptic displays, and smart healthcare [2] Group 3 - Recent domestic and international industrial catalysts have led to an increase in participants in the sector, with companies like Huawei, ByteDance, BYD, Xiaomi, and GAC intensifying their focus on embodied intelligence, while overseas companies like Tesla and FigureAI are accelerating commercialization [3] - The emergence of companies like DeepSeek in artificial intelligence is driving the development of general-purpose robotic models, facilitating the realization of embodied intelligence in humanoid robots, marking a phase of diverse innovation in the humanoid robot industry [3] - The commercialization of humanoid robots is becoming a clear trend in industrial applications, with recommendations to pay attention to domestic component manufacturers benefiting from this trend and to monitor upcoming events related to the humanoid robot industry chain [3]

本系列工作第一作者张泽宇，中国人民大学博士生，研究方向为大语言模型智能体的记忆机制和个性 化；谭浩然，中国人民大学硕士生，研究方向为大语言模型智能体。陈旭，中国人民大学预聘副教授， 研究方向包括大语言模型，信息检索等。 近期，基于大语言模型的智能体（LLM-based agent）在学术界和工业界中引起了广泛关注。对于智 能体而言，记忆（Memory）是其中的重要能力，承担了记录过往信息和外部知识的功能，对于提高智 能体的个性化等能力至关重要。中国人民大学高瓴人工智能学院与华为诺亚方舟实验室聚焦大语言模型 智能体的记忆能力，在该领域的研究早期，形成了一套完整的包括综述论文、数据集和工具包的研究体 系，致力于推动该领域的发展。 智能体记忆机制的早期综述 （TOIS'25） 在 2024 年 4 月，团队完成了早期的关于智能体记忆机制的综述。该综述从不同角度对智能体的记忆 进行了全面讨论。该综述讨论了「什么是智能体的记忆」和「为什么智能体需要记忆」，总结回顾了 「如何实现智能体的记忆」和「如何评测智能体的记忆能力」，归纳整理了「记忆增强的智能体应 用」，并提出当前工作存在的局限性和未来方向。通过该综述，团队希望能 ...

Core Insights - The exploration towards Artificial General Intelligence (AGI) highlights embodied intelligence as a key direction, focusing on the interaction and adaptation of intelligent agents within physical environments [1] - The development of embodied intelligence is marked by the evolution of technology from low-level perception to high-level task understanding and generalization [6][9] Industry Analysis - In the past two years, numerous star teams in the field of embodied intelligence have emerged, establishing valuable companies such as Xinghaitu, Galaxy General, and Zhujidongli, transitioning from laboratories to commercial and industrial applications [3] - Major domestic companies like Huawei, JD, Tencent, Ant Group, and Xiaomi are actively investing and collaborating to build an ecosystem for embodied intelligence, while international players like Tesla and investment firms support advancements in autonomous driving and warehouse robotics [5] Technological Evolution - The evolution of embodied intelligence technology has progressed through several stages: - The first stage focused on grasp pose detection, which struggled with complex tasks due to a lack of context modeling [6] - The second stage involved behavior cloning, allowing robots to learn from expert demonstrations but revealing weaknesses in generalization and performance in multi-target scenarios [6] - The third stage introduced Diffusion Policy methods, enhancing stability and generalization through sequence modeling [7] - The fourth stage, emerging in 2025, explores the integration of VLA models with reinforcement learning and tactile sensing to overcome current limitations [8] Product Development and Market Growth - The advancements in embodied intelligence have led to the development of various products, including humanoid robots, robotic arms, and quadrupedal robots, serving industries such as manufacturing, home services, and healthcare [9] - The demand for engineering and system capabilities is increasing as the industry shifts from research to deployment, necessitating higher engineering skills [13] Educational Initiatives - A comprehensive curriculum has been developed to assist learners in mastering the full spectrum of embodied intelligence algorithms, covering topics from basic tasks to advanced models like VLA and its integrations [9][13]