Group 1 - The core viewpoint of the report indicates that after a period of consolidation, the Hong Kong stock market is expected to reach new highs, supported by factors such as a potential resolution in the trade war, improvement in Sino-US relations, and a restart of the US interest rate cut cycle [2][4]. - The Hang Seng Index experienced a "V" shaped trading pattern, closing at 25,524 points, down 37 points or 0.15%, with a notable increase in trading volume, particularly in the Stock Connect segment [2][3]. - The report highlights that the Northbound capital flow remains strong, with a net inflow of HKD 12.7 billion, marking a 37.5% increase from the previous day, and a total net inflow exceeding HKD 45 billion over the last four trading days [3]. Group 2 - The report discusses the upcoming review of the Hang Seng Index, with significant increases in the average market capitalization thresholds for inclusion and exclusion, rising to HKD 7.33 billion and HKD 4.63 billion respectively, reflecting increases of 21.6% and 17.8% [6]. - It is anticipated that 20 non A+H stocks and 5 new A+H stocks will become new additions to the Stock Connect, with 22 non A+H stocks expected to qualify based on average market capitalization and turnover rates [6][8]. - The report identifies specific stocks that are particularly noteworthy for potential inclusion in the Stock Connect, including Nanshan Aluminum International (2610.HK) and KANAT Optics (2276.HK), among others [8].

Group 1 - The US-China economic talks were held in Stockholm, focusing on trade relations and macroeconomic policies, with a consensus to extend the suspension of certain tariffs for 90 days [2] - Huawei has returned to the top 100 in the Fortune Global 500 list, while Nvidia achieved the highest net profit margin [4] - Guangzhou's GDP growth for the first half of the year was reported at 3.8%, which is below the provincial and national averages [11] Group 2 - Apple announced the closure of its first store in China, located in Dalian, due to the exit of multiple retailers from the shopping center [25] - Polestar's sales in China have drastically declined, with only 6 units sold in June, indicating a near halt in business operations [26] - Alibaba's Qwen3-30B-A3B model has launched a new version, enhancing long text understanding capabilities [27]

格隆汇7月29日丨激智科技(300566.SZ)在互动平台表示，公司与主流眼镜厂商有研发合作，但无产品用 于已经发售的产品中。公司与阿里无业务往来及合作。 ...

Investment Rating - The report maintains a "Positive" investment rating for the industry [9] Core Viewpoints - The report highlights the ongoing growth of the AI industry in China, particularly following the WAIC 2025 conference, which showcased significant advancements in AI technology and applications [8][53] - It emphasizes the potential of domestic AI models and related infrastructure, suggesting a focus on domestic model manufacturers, AI intelligent entities, and the domestic computing power supply chain [8][53] Summary by Sections Market Performance - The computer sector experienced a slight increase of 1.59% last week, ranking 25th among major industries in the Yangtze River region, with a market share of 8.31% in total trading volume [2][6][16] Key Events - The WAIC 2025 conference took place, featuring over 800 companies and showcasing more than 3,000 cutting-edge exhibits, marking the largest scale in its history [8][40] - The National Data Bureau held a press conference regarding the progress of the data element comprehensive pilot zone construction, indicating a push towards data marketization [30][35] Recommendations - The report recommends focusing on: 1. Domestic large model manufacturers 2. AI intelligent entity-related stocks 3. Domestic computing power supply chain 4. Domestic computing infrastructure-related stocks 5. High-quality data-related stocks [8][53] Industry Developments - The report notes the continuous evolution of the open-source ecosystem in China, with significant projects and collaborations emerging from the WAIC conference [22][29] - It highlights the increasing importance of data elements and their marketization, with pilot zones established in multiple provinces to explore data value realization [30][36]

Investment Rating - The report maintains a strong positive outlook on the AI application-driven demand for computing power, indicating a sustained high growth trajectory in the industry [2]. Core Insights - The computing power industry is experiencing a resurgence in interest, particularly in the domestic market, as the supply chain begins to address critical bottlenecks such as advanced process capacity and supply of high-bandwidth memory (HBM) [2][3]. - The report highlights the ongoing tight balance in domestic computing power supply, with expectations that key challenges will gradually be overcome, leading to continuous positive news for the industry [2]. - The report suggests that the second half of the year and into next year will present significant investment opportunities in domestic computing power [2]. Summary by Sections Domestic Developments - Domestic AI applications are launching new products, with notable advancements such as Alibaba's Tongyi Qianwen 3 achieving significant breakthroughs in programming capabilities, comparable to top-tier closed-source models [3]. - Tencent's CodeBuddy IDE is set to enhance software development efficiency by allowing users to create applications through natural language dialogue, eliminating the need for coding knowledge [3]. Overseas Developments - The report recommends focusing on high-end PCB, optical modules, and server ODM sectors within the overseas computing power supply chain, anticipating sustained high demand and profitability [3]. - The supply of key materials for AI servers, such as fiberglass cloth and M8 CCL materials, is expected to remain tight, leading to potential price and profit margin increases [3]. Investment Recommendations - Key domestic computing power hardware companies, such as Cambricon and SMIC, are highlighted as potential investment opportunities [4]. - The report also suggests monitoring traditional analog chips, which are benefiting from domestic substitution and local demand, with wafer fab utilization rates remaining high and terminal prices increasing by 10-20% [6]. - The report identifies potential investment opportunities in the Apple supply chain, particularly with the upcoming iPhone 17 and 18 models, which may lead to upward revisions in market expectations [9].

Core Insights - The World Artificial Intelligence Conference (WAIC) 2025 emphasizes the transformation of AI from a technological marvel to a tool that enhances social efficiency and public welfare, with a focus on making AI accessible and beneficial for everyone [1][2]. Group 1: AI Applications and Innovations - The WAIC showcased over 800 companies and featured more than 40 large models, 50 AI terminals, and 60 intelligent robots, marking the largest scale in its history [2]. - Key technological advancements include Alibaba's Qwen3 model supporting real-time switching in 119 languages, SenseTime's "Riri New" V6.5 model for complex travel planning, and Tencent's immersive 3D world model [2]. - In the cultural and public welfare sectors, AI technologies are being applied to projects like the digitalization of ancient maps and solutions for visually impaired individuals, demonstrating the humanistic aspect of technology [4]. Group 2: Education and Industrial Applications - In education, companies like iFLYTEK and Whale Robot are using AI to enhance learning experiences and cultivate AI thinking among youth [5]. - Industrial applications are highlighted by Baidu's intelligent computing system and Tencent's AI productivity platform, showcasing AI's evolution from a supportive tool to a productivity engine [5]. Group 3: Forum Discussions and Industry Collaboration - The conference featured forums focusing on technological breakthroughs and ecosystem building, with companies like China Telecom and Yili Group presenting their AI innovations and forming industry alliances [5]. - The "AI for SME" forum introduced solutions aimed at alleviating the challenges faced by small and medium enterprises, addressing issues related to computing power and funding [5]. Group 4: AI Experience in Urban Settings - WAIC introduced a three-dimensional experience network that extends AI engagement from the exhibition hall to urban environments, enhancing public interaction with technology [6][7][8]. - Activities include robot bartenders, smart buses, and immersive experiences that allow participants to engage with AI in various contexts, from cultural to industrial applications [6][7][8]. - The event aims to demonstrate that AI is not just a distant concept but is actively reshaping daily life and infrastructure, making technology more human-centric and accessible [8].

Summary of Key Points from Conference Call Records Industry Overview - The computer sector's holdings are near historical lows at approximately 2.0%, compared to a peak of 6% during the 2018 cloud computing boom, indicating potential for growth [1][3] - The 2025 World Artificial Intelligence Conference (WAIC) has a total area exceeding 70,000 square meters, with significant proposals from Premier Li Qiang regarding AI accessibility and governance [1][4] Key Insights and Arguments - The AI sector is expected to see several catalytic events, including the WAIC, the anticipated release of GT5 in early August, and the Rtwo or V4 models in the latter half of Q3, which will drive AI technology development and application [1][5] - Recommended AI application stocks include Kingdee Digital Enterprise in Hong Kong and several A-share companies such as Jingjie, Hand, Saiyi, Kingsoft, and Hongsoft, all of which have stable fundamentals and positive Q2 earnings expectations [1][6] - China aims to enhance AI chip independence through improved single-card capabilities and distributed network construction, exemplified by Huawei's Cloudmagic 384 [1][7] Technological Developments - Huawei's Cloudmetric 384 features high bandwidth and low latency designs, significantly improving multi-expert inference efficiency, equipped with 384 cards, achieving a single card BF16 computing power of 800 TFLOPS, surpassing NVIDIA A100 [1][8][9] - The AI chip sector is witnessing significant advancements, with the VICO2.25 showcasing multiple AI chips, including the 910C chip with BF16 computing power nearing 800 TFLOPS and a memory capacity of 128GB [1][10] Liquid Cooling Technology - Liquid cooling technology is highlighted as a key feature at the Wetalk 2025 conference, with the Chinese liquid cooling server market projected to reach approximately 11.5 billion in 2024 and expand to 68 billion by 2027, reflecting a compound annual growth rate of over 45% [2][12] AI Applications and Innovations - The 2025 AI conference showcased over 40 large models, with significant updates from major companies like Alibaba, SenseTime, and China Mobile, focusing on model upgrades and applications [1][20] - Kingsoft's WPS AI 3.0 version emphasizes AI PPT and personal knowledge base features, enhancing user experience through improved document interaction and real-time modifications [1][29][30] Market Trends and Future Outlook - The AI sector is transitioning into a multi-agent collaboration phase, where complex task planning and iterative reflection are becoming standard practices [1][37] - The evolution of enterprise AI is expected to significantly alter business models, moving towards performance-based pricing structures and revenue-sharing models [1][38] Recommendations for Investment - When selecting ToB AI software companies, it is advisable to focus on firms with stable business development and significant future growth potential, such as Kingdee, Maifushi, Jingjie, and Xunfei Medical [1][39]

Investment Rating - The report does not explicitly state an investment rating for the industry or specific companies Core Insights - The AI application activity in the domestic market has significantly increased due to the release of new models, particularly Kimi, while overseas applications like Perplexity have shown slight growth [10] - Google reported strong FY25 Q2 earnings, with a 32% year-on-year increase in cloud platform revenue, reaching $13.6 billion, driven by AI demand [11][20] - Tesla's Q2 2025 revenue was $22.496 billion, a 12% year-on-year decline, with net profit down 23% to $1.172 billion, attributed to increased competition and policy changes [22][24] Summary by Sections Overseas Market Review - Kimi's popularity continues to rise due to new model releases [7] - Google FY25 Q2 performance exceeded expectations, with significant growth across advertising, search, and cloud services [11] - Tesla's Q2 financial results were disappointing, reflecting challenges in the automotive sector [22] AI Application Activity - The activity of chat assistant applications remains stable, with Kimi experiencing a substantial increase in engagement due to the new K2 model [10] - Alibaba's Qwen3 model has significantly enhanced long-text processing capabilities, while OpenAI has launched the ChatGPT Agent for advanced task execution [10] Google FY25 Q2 Performance - Google achieved an adjusted EPS of $2.31, surpassing market expectations, with total revenue of $81.7 billion, up from $71.3 billion year-on-year [11][12] - The company has raised its capital expenditure guidance for 2025 to $85 billion, primarily for AI infrastructure [18][21] Tesla Q2 Performance - Tesla's automotive revenue fell to $16.661 billion, a 16% decline year-on-year, with the company facing challenges from policy changes and increased competition [22][24] - Despite short-term challenges, Tesla is focusing on advancements in Robotaxi and AI technologies for long-term growth [25][27]

全球 AI 竞赛开启， 看好海内外 AI 产业链有望共振 行业研究 [行业周报 Table_ReportType] [Table_StockAndRank] 电子 投资评级 看好 上次评级 看好 [Table_ReportDate] 2025 年 7 月 27 日 [Table_Author] 莫文宇 电子行业首席分析师 执业编号：S1500522090001 邮 箱：mowenyu@cindasc.com 杨宇轩 电子行业分析师 执业编号：S1500525010001 邮箱：yangyuxuan@cindasc.com 信达证券股份有限公司 CINDA SECURITIES CO.,LTD 北京市西城区宣武门西大街甲127号金隅 大厦B座 邮编：100031 证券研究报告 [Table_Title] 全球 AI 竞赛开启，看好海内外 AI 产业 链有望共振 [Table_ReportDate] 2025 年 7 月 27 日 本期内容提要: [Table_S [Table_Summary ummary] ] 请阅读最后一页免责声明及信息披露 http://www.cindasc.com 2 ➢ 本周电子细分 ...

Core Insights - The narrative around large models has shifted towards enterprise-level AI Agents, focusing on the integration of AI into business processes and the creation of replicable, operational intelligent platforms [1] - Companies that can deliver measurable ROI through AI integration will be seen as the ultimate players in the market [1] Company Summaries 1. MaiFus (02556.HK) AI-Agentforce - MaiFus has focused on the "last mile" of enterprise AI application, emphasizing the concept of "delivery equals operation" for its AI-Agentforce platform, which highlights deployability, operability, and sustainable optimization [2] - The AI-Agentforce 2.0 integrates workflow orchestration, RAG knowledge engine, and DevOps lifecycle management, enabling efficient development and deployment of high-value AI applications [2] - The platform allows frontline staff to quickly generate and manage agents using natural language, reducing deployment barriers and accelerating AI application penetration within organizations [2][3] 2. ByteDance HiAgent - HiAgent is a highly platformized intelligent agent platform that aims to create a standardized, scalable operating system for AI agents, facilitating large-scale deployment and cross-scenario replication [4] - It features a unified agent orchestration framework that integrates a three-stage execution chain, supporting natural language, flowcharts, and API task flow construction [4] - HiAgent has been widely applied internally at ByteDance for tasks such as content review and customer service automation, and is gradually being offered as a SaaS product to external enterprises [4] 3. Dify - Dify is an active open-source intelligent agent platform that has gained traction in the GitHub community since its launch in 2023, primarily serving small and medium enterprises and AI developers [5] - The platform supports private deployment and a plugin ecosystem, allowing developers to build adaptable intelligent systems at low costs [5] - Dify is focused on creating a standardized open-source community to accelerate deployment efficiency for enterprises [5][6] Market Insights - MaiFus has chosen a challenging yet correct path by focusing on scene understanding, process re-engineering, and business closure rather than competing on computing power or model parameters [3] - HiAgent's strengths lie in its platform standardization and component-based development, which enhance system stability and reduce marginal costs for large-scale deployment [4] - Dify's lightweight platform is well-suited for sectors requiring private deployment, such as healthcare and government, due to its ease of deployment and strong controllability [6] Conclusion - The AI Agent market is diversifying, with companies like MaiFus focusing on value realization, while others like Baidu and Huawei pursue deep industry integration [7] - The ability to integrate AI with business processes and deliver measurable commercial value will determine the winners in this competitive landscape [7]