Core Insights - Xiaomi's Q2 2023 revenue reached 673.5 billion yuan, a year-on-year decline of 4%, while adjusted net profit increased by 147% to 51.4 billion yuan [12] - The average selling price (ASP) of Xiaomi's products rose to 1112 yuan, reflecting a year-on-year growth of 2.8% [12] Financial Performance - Q2 revenue breakdown: smartphones accounted for 366 billion yuan, IoT and lifestyle products for 223 billion yuan, internet services for 74 billion yuan, and other income for 10.6 billion yuan [12] - R&D expenditure in Q2 was 46 billion yuan, a year-on-year increase of 21% [11][12] - The gross margin for IoT and lifestyle products reached a record high of 17.6%, up 3.3 percentage points year-on-year [11] IoT and User Engagement - As of June 30, 2023, Xiaomi's IoT platform connected 655 million devices, marking a year-on-year growth of 24.2% [6] - The number of active users for Xiaomi's smart TVs exceeded 62 million in June 2023, representing an increase of over 18% year-on-year [11] Research and Development - The number of R&D personnel reached 16,834, constituting 52% of the total workforce [11][12] - Xiaomi has over 3,000 personnel focused on AI, with significant progress in developing lightweight, locally deployed AI models [11]

Core Insights - Xiaomi Group reported Q2 2023 revenue of 67.35 billion RMB, exceeding market expectations of 65.13 billion RMB, but down 4% year-on-year from 70.17 billion RMB. Net profit reached 3.67 billion RMB, up 168.5% year-on-year, surpassing the forecast of 3.16 billion RMB. Adjusted net profit was 5.14 billion RMB, a 147% increase year-on-year [1][1][1] Financial Performance - Revenue for Q2 2023 was 67.35 billion RMB, compared to 65.13 billion RMB expected by the market and 70.17 billion RMB in the same period last year, indicating a 4% year-on-year decline [1] - Net profit for Q2 2023 was 3.67 billion RMB, significantly higher than the expected 3.16 billion RMB, marking a 168.5% increase year-on-year [1] - Adjusted net profit under non-IFRS was 5.14 billion RMB, reflecting a 147% year-on-year growth [1] Mobile Business Insights - Xiaomi's mobile sales showed a recovery in Q2 despite a global decline in smartphone sales, with expectations for continued growth in Q3 [1][2] - The company emphasized focusing on internal capabilities and product quality rather than competitor actions, suggesting that new product launches by competitors would not significantly impact Xiaomi's strategy or sales expectations [2] Inventory and Cost Management - Xiaomi reported a significant reduction in both internal and channel inventory, indicating a healthy inventory status [3][4] - The company achieved a historical high in mobile gross margin during Q2, attributed to improved pricing and cost control measures [3][4] - Xiaomi's cost advantages stem from scale effects and enhanced self-research capabilities, particularly in areas like screen production and camera technology [5][6] IoT Market Performance - Xiaomi's IoT business showed strong performance in China, particularly in major appliances and tablets, with expectations for inventory adjustments in overseas markets by Q4 2023 [11][12] - The global consumer electronics market is recovering slowly, but a smaller decline is anticipated in 2024 compared to 2023 [11] Internet Services and AI Strategy - Xiaomi's internet services revenue reached a historical high in Q2, with a focus on increasing average revenue per user (ARPU) through a growing base of high-end users [12][14] - The company is investing in AI, with plans to integrate AI capabilities into products and improve internal operations, including a significant investment fund for AI development [22][24] Global Market Expansion - Xiaomi is focusing on optimizing its market structure in Europe and exploring growth opportunities in the Middle East and Latin America, with a potential strategy for entering African markets [16][17][18] - The company remains confident in its ability to rebound in the Indian market after facing challenges in Q2 [18] Cost Control and Management Efficiency - Xiaomi has made significant progress in cost reduction, with a notable decrease in operational expenses compared to the previous year, indicating effective management reforms [19][21] - The company aims to balance scale and profitability while continuing to enhance operational efficiency across all departments [19][21]

Group 1 - SoftBank Group plans to lead a group of Indian startup founders to Silicon Valley next month to explore artificial intelligence technology, aiming to inject more AI elements into its investment portfolio [1] - Sumer Juneja, head of investments for SoftBank Vision Fund in Europe, the Middle East, Africa, and India, stated that the company is assisting its portfolio companies in adopting AI technology and has arranged for founders to engage with top companies in the field [1] - SoftBank intends to arrange for up to 20 Indian startup founders to participate in the trip, with investments in companies such as Oyo, Ola, and Swiggy [1] Group 2 - SoftBank has been active in the venture capital space but paused investments due to investor disappointment with loss-making startups, resuming investments in the second quarter of this year [1] - The company currently has over $40 billion in cash and is preparing for the IPO of chip design company Arm to raise additional funds [1] - Juneja emphasized the importance of having a top-tier technology team and being in an industry that can leverage AI to enhance business model efficiency when considering new investments [1] Group 3 - The rise of ChatGPT has led investors to accelerate support for AI companies, with Juneja warning that many industries, including SaaS, could see a significant number of companies become redundant if they fail to quickly utilize AI technology [1] - The biggest challenge today is how to fully leverage AI technology [1] - SoftBank has invested over $140 billion in startups since the establishment of the Vision Fund in 2017, supporting the development of new technologies like AI [2]

Core Viewpoint - ARM's dominant position in the smartphone processor market poses both a significant asset and a challenge for its upcoming IPO, which aims for a valuation of $60 billion [1] Group 1: Market Position and Challenges - ARM holds a near-monopoly in the mobile and consumer electronics markets, but this dominance may limit future growth opportunities [1] - The company has not penetrated the hottest segment of the chip market for AI models, where NVIDIA is the leader [1] - ARM's relationship with Apple has become complex, with Apple being a major customer but not prominently featured in ARM's IPO documents [2] Group 2: Revenue and Business Model - ARM's revenue model relies heavily on licensing fees, with a 2.7% royalty rate translating to $0.11 per chip sold, leading to limited pricing power [3] - Despite a stable revenue stream from older products, ARM's overall revenue declined by 1% to $2.7 billion last year [3] Group 3: Future Growth Areas - ARM is exploring new markets such as automotive and cloud computing, where it currently holds 41% of the automotive market and 10% of the $18 billion cloud processor market [5][6] - The automotive sector has seen a 36% increase in licensing revenue, indicating potential for growth [6] - ARM's technology plays a supportive role in AI, with partnerships in the autonomous vehicle sector and collaborations with cloud giants [5][6]

【#京东外卖员被曝换新工服##知情人士回应京东外卖员换新工服#】京东近日官宣推出独立外卖App， 同时，一款形似"赛车服"的京东外卖骑手服装照片开始在社交平台流传，引发网友猜测：继饿了么更 名"淘宝闪购"并更换橙黑配色工服后，京东是否也将更换工服？记者从接近京东人士处获悉，网传京东 外卖骑手新工服并非官方发布版本，仅是内测版的其中一款。（智通财经） 京东也要跟上这波换新潮流了，工作服大变天 .... ...

Core Insights - Nvidia is set to release its quarterly earnings report, with a focus on whether the strong demand for AI products can offset the decline in global computer hardware sales [1] - Nvidia previously projected a revenue growth of approximately two-thirds and a threefold increase in earnings per share for the second fiscal quarter ending July 30, 2024, driven by robust demand for its AI processors [1] - The current supply of Nvidia's H100 AI processors meets only half of the market demand, with expected shipments for next year projected to reach between 1.5 million to 2 million units, a significant increase from the anticipated 500,000 units this year [1] Industry Trends - The demand for Nvidia's AI chips is impacting the broader computing device market, as major buyers are investing heavily in AI while cutting back on investments in the general server market [1] - Foxconn, the world's largest electronics contract manufacturer, anticipates strong demand for AI servers in the coming years, although it warns of a decline in overall server market revenue this year [1] - Lenovo, the leading global PC manufacturer, reported an 8% year-over-year revenue decline due to weak demand from cloud service providers for servers and a shortage of AI processors [1] Cloud Service Providers - Lenovo's CEO noted that cloud service providers are shifting their demand from traditional computers to AI servers, but the supply of AI servers is constrained by GPU shortages [2] - TSMC, the largest chip foundry and exclusive manufacturer of Nvidia's AI processors, expects demand for AI server chips to grow nearly 50% annually over the next five years, but this growth is not enough to offset the downward pressure from the global tech downturn [2] - Major cloud service providers like Microsoft, Amazon, and Google are focusing on building their own AI infrastructure, facing challenges due to the high costs associated with upgrading every component in AI servers [2] Market Projections - TrendForce forecasts a 6% decline in global server shipments this year, with a moderate recovery of 2% to 3% growth expected in 2024 [5] - Companies like Meta and Microsoft are reducing server procurement to invest in AI hardware, with Meta cutting over 10% of its server purchases [5] - Analysts highlight that supply chain bottlenecks, including shortages in advanced packaging and high-bandwidth memory, are limiting the production of AI hardware products [5] Beneficiaries of AI Demand - Foxconn is positioned to benefit from the shift towards AI servers, as it provides a full range of services from components to final assembly, and its subsidiary is the exclusive supplier of Nvidia GPU modules [6] - AI orders have accounted for 50% of revenue for WiWynn, a subsidiary of Foxconn's competitor Wistron, more than doubling from the previous year [6] - PCB manufacturers are also expected to see significant growth, with the revenue share from AI servers projected to rise from less than 3% this year to 38%, as AI servers require seven times more PCBs than general servers [6]

Core Viewpoint - The UK Competition and Markets Authority (CMA) has approved Broadcom's acquisition of VMware after a second-phase investigation, indicating that the deal will not harm competition in the computer server market [1] Group 1: Acquisition Details - Broadcom announced its intention to acquire VMware for approximately $61 billion, marking one of the largest technology deals in history [1] - The acquisition aims to strengthen Broadcom's position as a formidable competitor in the software sector [1] Group 2: Regulatory Investigation - The CMA initially expressed concerns in March about potential negative impacts on market competition, specifically that the deal could make computer servers more expensive for UK businesses [1] - Following a thorough investigation, the CMA concluded that the acquisition would not weaken competition in the computer server market [1] Group 3: Timeline and Extensions - Broadcom and VMware originally planned to complete the transaction by May 26 of this year, but they agreed to extend the deadline by one year to await the outcome of regulatory reviews [1]

2023世界机器人大会期间，小米集团高级副总裁曾学忠指出，仿人机就像人一样，会实现人生的三级 跳。在幼儿期，会学习各种各样的动作，掌握最初级的操作能力；到少年期，开始学习知识，掌握领域 的知识图谱；到了中年期，历经各种场景的实践，它就会最终沉淀为人生的智慧。 第三，长期来看，小米希望以自身为示范，带动仿人机器人的广泛应用，拓展机器人在3C和汽车制造 场景之外更多场景中的应用和价值实现。 据曾学忠介绍，遵循这一规律，目前小米正在积极推进仿人机器人在自有制造系统中的分阶段落地。具 体而言，小米对机器人产业的布局，将会分为三步走： 第一步，短期内，小米的仿人机器人将扎根某一个场景，实现多个功能做产品技术的验证，进行原型机 的开发。在这个过程当中不断迭代所需要的硬件、硬件系统、传感器系统、感知控制系统。 第二，到了中期，小米的仿人机器人将融入小米制造、智能制造多个场景来做产业验证，进行规模化的 验证，这一过程中，小米会搭建泛化平台，提升部署效率，实现仿人机器人在制造场景当中的数据积累 和模型迭代。 ...

在今年的一季度财报中，小米表示，2023年4月，小米集团正式组建AI实验室大模型团队。目前小米AI 领域相关研发人员超1200人。 近日，小米的大语言模型MiLM-6B首次现身C-Eval、CMMLU两大AI模型评测榜单。 资料显示，MiLM-6B是由小米开发的一个大规模预训练语言模型，参数规模为64亿。截至当前， MiLM-6B在C-Eval总榜单排名第10、同参数量级排名第1，在CMMLU中文向大模型排名第1。 ...

Core Points - The California Institute of Technology (Caltech) has reached a "tentative settlement agreement" with Apple and Qualcomm regarding a patent infringement lawsuit related to wireless communication chips used in electronic devices [1] - In a previous trial in 2020, Caltech won a judgment requiring Apple to pay $837.8 million and Qualcomm to pay $270 million, totaling over $1.1 billion in damages, but this ruling was overturned by an appellate court in 2021 [2] - The retrial originally scheduled for June 2023 was indefinitely postponed, and Caltech has also reached a settlement with Samsung Electronics regarding similar patent infringement claims [3] Summary by Sections Lawsuit Background - Caltech filed a lawsuit against Apple and Qualcomm in 2016, claiming that wireless communication chips used in millions of Apple devices infringed on its patents [1] - Other companies, including Microsoft, Dell, HP, and Samsung, have also faced similar lawsuits from Caltech for patent infringement [3] Legal Proceedings - The initial ruling in 2020 favored Caltech, leading to significant financial penalties for Apple and Qualcomm, but the appellate court later deemed the damages legally unsound, prompting a retrial [2] - The court has requested further status reports from the parties involved by August 18 [1] Current Developments - The details of the tentative settlement agreement between Caltech, Apple, and Qualcomm remain undisclosed, and it is unclear if both companies are included in the settlement [1] - Ongoing litigation continues between Caltech and Microsoft, HP, and Dell regarding patent infringement [4]