" 「老机套新壳」与「全线降配」背后的无奈选择。 " 作者丨 杨依婷 编辑丨 包永刚 "现在做产品规划，最大的风险不是判断错方向，而是判断刚做完就失效。"一位头部手机厂商的产品规划 师这样形容最近几个月的状态。 在手机行业，原本按季度、按年度推进的节奏，正在被一次次突如其来的涨价打碎，所有人都被迫进入一 场没有暂停键的拉锯战。 如果说未发布的新机还能通过调整配置和定价勉强应对，那么 真正让厂商进退失据的，是那些已经上市、 价格却难以回调的老产品 。 尽管有业内人士提及，少数巨头可以通过集团内部的" 动态总包 "库存管理模式来缓冲压力，即所有项目 共享库存。 但对于大部分品牌来说，"怎么把老产品的价格涨上去，是现在最核心的问题，这直接决定还能不能赚 钱。"为了对冲存储成本上行，部分厂商开始用"老机套新壳"的方式，寻找价格调整的空间。本质上，这 "不做"，正在成为消费电子中小厂商对抗成本侵蚀的现实选择。 01 消费存储市场剪刀差：端侧需求激增，上游减产 近两年，随着AI逐渐向手机、电脑、穿戴设备等消费电子终端渗透，存储需求发生了显著的升级。 长期以来，Windows 11的最低存储配置要求是64GB SSD和4G ...

Key Points - Regulatory actions intensify as X platform's French headquarters is raided, with Elon Musk facing a summons for a hearing in April 2026 regarding algorithm misuse and data handling issues [2][3] - Bosch initiates layoffs in China, affecting nearly 200 employees, particularly in fuel vehicle projects, with compensation reported as N+4, indicating a relatively generous severance package [4][5] - Honor denies allegations of copying Apple's iPhone design for its Power2 model, asserting that mature designs lead to industry consensus [7][8] - Gao Xin Retail, the parent company of RT-Mart, refutes claims regarding its CEO being taken by police, stating the reports are false [10][11] - Baidu Baike wins a lawsuit against a competitor for unfair competition, resulting in a compensation of 15 million yuan [12] - Cambricon Technologies experiences a significant stock drop, losing over 700 billion yuan in market value, prompting a public statement addressing market concerns [13][14] - Zeekr's 8X model faces leaks disrupting its release schedule, with the company emphasizing the importance of maintaining product release order [23][24] - Beijing Human Shape Robotics secures over 700 million yuan in funding to enhance its humanoid robot capabilities, with the launch of its Tian Gong 3.0 model imminent [17] - Xpeng Motors merges its autonomous driving and smart cockpit departments into a new General Intelligence Center to enhance AI capabilities [19] - Mijia Ice City is reportedly planning to build a theme park, with job postings indicating a significant investment in this venture [20][21] - Intel is set to re-enter Apple's supply chain for low-end processors, marking a shift in Apple's long-standing reliance on TSMC for chip manufacturing [40][41]

Group 1 - The core viewpoint of the article revolves around the competitive dynamics in the autonomous driving sector, highlighting how Company Y secured a partnership with a new force in the automotive industry by underbidding competitors [1][2] - Company W initially believed it had secured a deal with the new force after a lengthy negotiation process, only to be outbid by Company Y, which offered a significantly lower price of 20 million [2] - The article suggests that Company Y's strategy is not primarily profit-driven but rather aimed at enhancing its reputation ahead of an upcoming IPO, indicating a shift in focus from immediate financial gain to long-term brand positioning [3] Group 2 - The departure of a key executive, L, from Company B is noted, with insights into the challenges faced by executives in the current market environment, suggesting that the role of VP has become precarious [5] - Company A is undergoing significant internal changes, including the departure of core executives and restructuring efforts, which are indicative of broader trends in the high-end electric vehicle market [6][7] - Company Q's handling of underperforming executives is described as unusually lenient, with a five-month transition period for a poorly performing manager, raising questions about the effectiveness of its management practices [8] Group 3 - Company S is experiencing internal coordination issues among its top executives, leading to confusion and inefficiencies in communication and reporting structures [9][10] - Company E's aggressive cost-cutting measures have resulted in employee dissatisfaction, with reports of reduced benefits and high turnover rates, indicating potential long-term impacts on workforce morale [11] - The article discusses the phenomenon of key technical personnel leaving companies after completing significant projects, suggesting that the pressure and demands of end-to-end technology development may lead to high turnover rates in the industry [15][16]

Key Points - Elon Musk announced that SpaceX has acquired xAI, with a new company valuation of $1.25 trillion, integrating advanced rocket and satellite technology with AI capabilities [4][5] - Lei Jun refuted claims of Xiaomi's second-hand car price collapse, highlighting that the SU7 model has the highest resale value among electric vehicles at 86.05% [7][8] - Tencent's internal communication addressed concerns over its Yuanbao red envelope campaign, asserting it does not induce sharing, despite previous actions against similar marketing strategies [10][11] - Andrew Miles, former president of Sam's Club China, has joined Metro as a consultant, bringing extensive experience from his tenure at Walmart [14][15] - Huawei achieved the top market share in China's smartphone market for January at 18.6%, driven by strong sales of the Mate 80 series [17][19] - Mercedes-Benz has initiated a price reduction of approximately 10% on select models to improve dealer liquidity [25] - Tesla announced the upcoming launch of its third-generation humanoid robot, expected to produce one million units annually by 2026 [45][46] - Samsung and SK Hynix plan to expand their advanced NAND production capacity, targeting a monthly output of 40,000 to 50,000 wafers [48][49]

Core Viewpoint - Jitu's expansion into the U.S. market is fraught with challenges, despite its strong position in Southeast Asia and plans to replicate its low-cost model against existing competitors [2][3][5][25]. Group 1: Jitu's Market Position and Strategy - Jitu has established a significant presence in Southeast Asia, with a package volume expected to exceed 30 billion by 2025, showing a year-on-year growth of 67.8% [5]. - The company achieved a net profit of approximately $8.89 million in the first half of 2025, a staggering increase of 186.6% year-on-year [5]. - Jitu's market share in Southeast Asia has reached 38%, positioning it as the leader in the region, while competitors like Shopee and Lazada have lower market shares [5][6]. Group 2: Challenges in the U.S. Market - The U.S. last-mile delivery market is dominated by major players like USPS, FedEx, and UPS, which collectively hold over 85% of the market share [18]. - New entrants such as GOFO and UniUni have intensified competition, with GOFO achieving daily order volumes of approximately 150,000 to 200,000 [19]. - Jitu's previous light-asset model may not be suitable for the U.S. market, where higher upfront investments and a more complex operational model are required [20][22]. Group 3: Growth Potential in Other Markets - Jitu is experiencing rapid growth in Latin America, with a year-on-year package volume increase of 43.6% and achieving profitability in the region for the first time [12][16]. - The company is leveraging its established infrastructure in Southeast Asia to replicate its success in Latin America, where the market dynamics are similar [13]. - In the Middle East, Jitu has maintained a stable operational base, achieving a delivery success rate of approximately 85% [16].

Core Insights - The article discusses the challenges and advancements in the robotics industry, particularly focusing on the transition from machine learning models to real-world applications, emphasizing the importance of stability and reliability in deployment [2][3][31]. Group 1: Structural Challenges in Robotics - The robotics industry faces three structural barriers: morphological fragmentation, data cost and coverage, and deployment system issues, which hinder the generalization of robotic strategies [2][3]. - Morphological fragmentation leads to difficulties in data sharing among different robotic forms, requiring retraining when switching platforms [2]. - The high cost and scarcity of real-world robotic data limit the ability to cover long-range tasks and complex interactions, making cross-morphology generalization challenging [2][3]. Group 2: Research and Development of Being-H0.5 - The research team led by Lu Zongqing introduced the Being-H0.5 model, which aims to create a more stable and generalizable robotic control strategy by leveraging human-centric data and addressing the aforementioned barriers [3][4]. - The model utilizes a unified state-action space to overcome inconsistencies in action definitions across different hardware, facilitating knowledge sharing and transfer [4][24]. - The UniHand-2.0 dataset, comprising over 35,000 hours of data, serves as a foundation for training, integrating human hand operation data and robotic manipulation data to enhance the model's performance [13][23]. Group 3: Experimental Results and Performance - Experimental results indicate that while specialist models perform slightly better, the generalist Being-H0.5 model shows comparable performance, particularly in long-horizon and bimanual tasks, which are critical for assessing deployment stability [9][10]. - In real robot experiments, the generalist model demonstrated significant improvements in long-horizon and bimanual tasks, highlighting its potential for stable deployment in complex environments [9][10]. - The model achieved an average success rate of 98.9% on the LIBERO benchmark and 53.9% on the RoboCasa benchmark, showcasing its robustness in both simulated and real-world scenarios [14][15]. Group 4: Deployment Mechanisms and Stability - The introduction of mechanisms like MPG (Motion Policy Generation) and UAC (Unified Action Control) is crucial for ensuring the stability of the model during deployment, particularly for long-range and bimanual tasks [17][18]. - The absence of these mechanisms leads to significant performance degradation, emphasizing the importance of stability in real-world applications [17][18]. - The research highlights that achieving reliable deployment requires addressing both action distribution constraints and asynchronous control issues [33]. Group 5: Implications for Future Robotics - The findings suggest that cross-morphology unified action learning is feasible, allowing multiple robots to share the same strategy without extensive retraining [30]. - Human hand video and action data are essential for developing generalist models, providing a natural action prior that enhances generalization and transferability across different robotic forms [30]. - The work underscores the need for a comprehensive approach that integrates data, alignment, generation, and deployment stability to advance the field of general robotic intelligence [30].

有业内人士猜测，千问 APP或将会延续"会办事"的思路，推出AI买年货、奶茶咖啡、电影票、飞机票、酒 店、景点门票等功能，并通过请客的方式直接免单。 No 千问春节大请客! 吃喝玩乐,免单不停!！！ 2月6日见 淘宝 闪购 飞猪 大麦 高德 盒马 " 真金白银的投入，能否培养用户「有事找AI」的习惯？ " 编辑丨李希 2月2日，千问APP宣布投入30亿启动"春节请客计划"，以免单形式请全国人民在春节期间吃喝玩乐，感 受AI时代的全新生活方式。淘宝闪购、飞猪、大麦、盒马、天猫超市、支付宝等阿里生态业务也将加入千 问春节攻势。 据悉，此次千问 30亿春节请客计划，在阿里历史上的春节活动中投入最大，在今年春节大厂AI大战中投 入金额也最高。 据千问 APP官微，"千问春节请客计划"将于2月6日正式上线，强调"吃喝玩乐 免单不停"，还有大额现金 红包，但官方尚未透露具体的活动形式。 1月15日，千问APP宣布已接入淘宝闪购、支付宝、淘宝、飞猪、高德等阿里生态场景，测试AI购物功 能。千问APP内部人士透露："半个月前的发布，其实就是在为春节攻势做准备。市面上大部分AI助手还 停留在聊天阶段，而千问希望在今年春节期间， ...

Key Points - Major international companies are reportedly planning to lay off 30,000 employees due to funding difficulties in AI data center expansions, with Oracle considering selling parts of its business [4][5] - Wenta Technology expects a significant loss of 90 to 135 billion yuan in 2025, primarily due to issues with its core asset, Nexperia Semiconductor, which is under temporary control restrictions [8][9] - Tencent has appointed AI expert Pang Tianyu as the Chief Research Scientist for its AI division, focusing on multimodal reinforcement learning [12][13] - Cambricon has achieved its first annual profit in ten years, with a projected revenue increase of 410.87% to 496.02% in 2025 [16][17] - GAC Group anticipates a loss of 80 to 90 billion yuan in 2025, with a 14% decline in vehicle sales [20][21] - ByteDance's Hongguo short drama app has surpassed 100 million daily active users within three years of its launch [20] - Apple has announced unprecedented innovations for this year, following a record quarterly revenue of $143.76 billion [40][41] - OpenAI is reportedly preparing for an IPO in Q4 2023, aiming for a valuation of approximately $500 billion [56][57] - Waymo plans to raise about $16 billion in a new funding round, which would value the company at nearly $110 billion [44][45] - The global memory chip market is facing supply challenges, with major manufacturers tightening order approvals to prevent excessive stockpiling by clients [52][53]

Group 1 - The core point of the article is the departure of Qiu Shanxiang, the general manager of Volkswagen Import Cars, primarily for personal career development reasons [2] - Qiu Shanxiang's tenure lasted just over a year, having taken the position in August 2024, and he is succeeded by Zhou Mi from within the company [2] - Volkswagen Import Cars is responsible for the sales of imported Volkswagen models in China, with the Touareg currently being the main sales model [2] Group 2 - In 2025, the sales volume of the Touareg in the domestic market was 3,624 units, representing a year-on-year decline of 43.96% [2] - The domestic import passenger car sales reached 440,000 units in the first 11 months of 2025, down 29.7% year-on-year, marking the fourth consecutive year of decline in imported car sales [4] - The peak sales of domestic imported passenger cars occurred in 2014 at 1.26 million units, followed by a downward trend, dropping below 1 million units in 2020 and further declining to 705,000 units in 2024 [5] Group 3 - Volkswagen Group's global CEO, Herbert Diess, emphasized the commitment to the "In China, for China" strategy, focusing on deepening engagement in the Chinese market and accelerating local R&D [5] - In 2026, Volkswagen Group plans to launch over 20 pure electric, plug-in hybrid, and range-extended models in China, marking the most intensive product rollout in its history in the market [5] - In 2025, Volkswagen Group achieved global sales of 8.98 million units, with over 2.69 million units delivered in the Chinese market, maintaining its position as the top foreign car manufacturer in China [5]

Core Viewpoint - The article emphasizes that the launch of the "Zhenwu 810E" chip is just the beginning, and the core lies in the collaborative transformation of the "Tongyun Ge" ecosystem, which integrates chips, cloud platforms, and AI models to create a synergistic effect that exceeds the sum of its parts [5][16]. Group 1: Chip Development and Performance - The "Zhenwu 810E" chip is designed to meet the specific needs of AI cloud applications, featuring a high bandwidth memory of 96GB HBM2e and a peak interconnect bandwidth of 700GB/s, which supports large-scale AI model training and inference [6][7]. - The chip's architecture allows for efficient multi-card interconnection, with PCIe 5.0×16 interfaces providing double the bandwidth of PCIe 4.0, thus reducing data transfer bottlenecks [7]. - The "Zhenwu 810E" has a low power consumption of 400W, enabling denser cabinet deployments and improving energy efficiency, which lowers operational costs for data centers [7][8]. Group 2: Strategic Positioning and Ecosystem - Alibaba's strategy involves a dual closed-loop system that integrates storage and computing, with the "Zhenwu 810E" chip forming part of a comprehensive self-developed AI ecosystem that includes the "Zhenyue 510 SSD" and "Yitian 710 CPU" [11][12]. - The collaboration between the chip, cloud services, and AI models allows Alibaba to create a tightly coupled architecture that enhances performance and reduces inefficiencies associated with cross-architecture adaptations [14]. - The "Tongyun Ge" ecosystem positions Alibaba to compete effectively against global tech giants, as it has developed a full-stack AI capability that integrates hardware and software solutions [16][24]. Group 3: Market Applications and Competitive Edge - The "Zhenwu 810E" chip is particularly suited for industries such as automotive and finance, where there is a high demand for AI training performance and low latency [21][22]. - Alibaba's solutions have been successfully implemented across various sectors, serving over 400 industry-leading clients, which demonstrates the practical application and scalability of its AI technology [23]. - The article notes that Alibaba's AI capabilities have reached a level that allows it to be compared with leading global players like NVIDIA, Oracle, and OpenAI, thus reshaping its valuation logic in the market [16][24].