Core Insights - Huawei Euler concept stock rose by 0.74%, ranking fifth among concept sectors, with 22 stocks increasing in value [1] - The leading gainers included *ST Dongtong with a 20% limit up, Aerospace Development also hitting the limit up, and Qingyun Technology, Chengmai Technology, and Dongfang Guoxin with increases of 10.76%, 3.78%, and 3.68% respectively [1] - The sector experienced a net outflow of 356 million yuan, with 18 stocks seeing net inflows, and 10 stocks receiving over 30 million yuan in net inflows [2] Sector Performance - The top-performing concept sectors included Hainan Free Trade Zone with a 2.20% increase and Salt Lake Lithium Extraction with a 1.43% increase, while Silicon Energy and Pre-made Dishes saw declines of 2.66% and 2.37% respectively [2] - Huawei Euler concept was among the top gainers, while the dairy industry faced a decline of 2.27% [2] Fund Flow Analysis - Key stocks with significant net inflows included Donghua Software with a net inflow of 158.23 million yuan, followed by Dongfang Guoxin and Tuo Wei Information with inflows of 133.80 million yuan and 130.78 million yuan respectively [3] - The net inflow ratios for stocks like Kela Software, Qingyun Technology, and Yingfang Software were 12.71%, 11.49%, and 11.27% respectively, indicating strong interest from major funds [3] Stock Specifics - Notable stocks in the Huawei Euler concept included Donghua Software with a 3.68% increase and a turnover rate of 4.70%, and Qingyun Technology with a significant 10.76% increase and a turnover rate of 13.12% [4] - Conversely, stocks like Dahong Technology and China Software experienced declines of 3.16% and 2.86% respectively, with significant net outflows [5]

Core Insights - The article discusses the challenges enterprises face in digital transformation amidst rapid AI technology adoption, emphasizing the shift from "whether to transform" to "how, when, and if they can transform" [1][4] - It highlights the structural mismatch between the disruptive nature of technological iterations and the gradual development of enterprises, leading to a complex set of challenges [2][4] Group 1: Challenges in Digital Transformation - Enterprises are experiencing a fundamental contradiction between rapid technological advancements and their gradual IT architecture evolution, resulting in compounded challenges [2] - The conflict between IT departments and business units is intensifying, as business diversification demands high flexibility from IT systems, while IT teams prioritize stability and simplified management [4] - Three core challenges are identified: compatibility with historical investments while embracing AI innovation, simplifying management while supporting business diversification, and ensuring system stability while enabling rapid iteration [4] Group 2: Proposed Solutions - Qingyun Technology proposes a "gradual innovation" approach with the launch of the AI Infra 3.0 architecture, focusing on "restructuring and unifying" to bridge the gap between historical and future needs [4][5] - The AI Infra 3.0 architecture consists of four layers, including a foundational operating system, a unified scheduling layer, a capabilities layer covering various technologies, and an open layer for ecosystem customization [5] - The architecture aims to provide full-stack capabilities, on-demand expansion, standardized delivery, and evolution, allowing enterprises to introduce AI capabilities in a phased manner without overhauling existing systems [5] Group 3: The Role of AI Infrastructure - The role of AI infrastructure is evolving from a mere tool to a foundational strategy for enterprises, with a focus on effectively utilizing AI models rather than merely selecting the best ones [6] - AI Infra 3.0 is positioned as a pathway for enterprises to connect with the AI era, built on over a decade of technological accumulation and deep insights into customer needs and industry trends [6] - The article reflects the sentiment of many enterprises that are cautious about AI transformation, fearing the loss of past investments while trying to seize future opportunities [6]

法律意见书 北京市盈科律师事务所 关于北京青云科技集团股份有限公司 2025年第二次临时股东会的 北京市朝阳区金和东路 20 号院正大中心 2 号楼 19-25 层 2025 年 11 月 法律意见书 北京市盈科律师事务所 关于北京青云科技集团股份有限公司 2025年第二次临时股东会的法律意见书 致：北京青云科技集团股份有限公司 北京市盈科律师事务所（以下简称"本所"）是具有中华人民共和国法律执 业资格的律师事务所。本所接受北京青云科技集团股份有限公司（以下简称"公 司"）的委托，就公司召开 2025 年第二次临时股东会（以下简称"本次股东会"） 的有关事宜，根据《中华人民共和国公司法》（以下简称"《公司法》"）、《中 华人民共和国证券法》（以下简称"《证券法》"）、《上市公司股东会规则》 （以下简称"《股东会规则》"）等法律、法规、规章及其他规范性文件（以下 简称"中国法律法规"，仅为本法律意见书之目的，不包括香港特别行政区、澳 门特别行政区及台湾地区的法律、法规）以及《北京青云科技集团股份有限公司 章程》（以下简称"《公司章程》"）的有关规定，出具本法律意见书。 本所及本所律师根据《证券法》《律师事务所从 ...

证券代码：688316 证券简称：青云科技 公告编号：2025-059 北京青云科技集团股份有限公司 2025年第二次临时股东会决议公告 本公司董事会及全体董事保证公告内容不存在任何虚假记载、误导性陈述或 者重大遗漏，并对其内容的真实性、准确性和完整性依法承担法律责任。 重要内容提示： 本次会议是否有被否决议案：无 一、 会议召开和出席情况 （一） 股东会召开的时间：2025 年 11 月 17 日 （二） 股东会召开的地点：北京市朝阳区酒仙桥路 6 号院 9 号楼 11 层公司会议 室 （三） 出席会议的普通股股东、特别表决权股东、恢复表决权的优先股股东及 其持有表决权数量的情况： | 1、出席会议的股东和代理人人数 | 69 | | --- | --- | | 普通股股东人数 | 69 | | 2、出席会议的股东所持有的表决权数量 | 14,452,339 | | 普通股股东所持有表决权数量 | 14,452,339 | | 3、出席会议的股东所持有表决权数量占公司表决权数量的比 | 30.1730 | | 例（%） | | | 普通股股东所持有表决权数量占公司表决权数量的比例（%） | 30.1730 ...

Group 1 - The core viewpoint of the articles highlights the rapid growth of China's AI infrastructure service market, which reached 19.87 billion yuan in the first half of 2025, a year-on-year increase of 122.4%, with projections nearing 150 billion yuan by 2029 [2] - Despite 83% of enterprises prioritizing AI as a strategic focus, the actual success rate of implementation is only 29%, indicating significant challenges in AI project execution [3] - The systemic architectural imbalance, characterized by issues in computing power supply, data governance, system collaboration, and security compliance, is identified as a root cause of AI implementation failures [3] Group 2 - The CEO of Qingyun Technology outlines three phases of digital transformation since the emergence of ChatGPT, with the first phase focusing on the scarcity of computing power as a major obstacle for AI applications [4] - The second phase sees an increase in customer willingness to experiment with AI, but diverse industry needs remain inadequately addressed [4] - The third phase marks a shift in enterprises' attitudes towards serious consideration of AI integration, facing historical IT architecture issues that lead to fragmented computing resources [5] Group 3 - A significant 53% of enterprises adopt tightly coupled AI architectures, which bind model training and inference directly to business systems, leading to challenges during the iteration phase [6] - Enterprises face a triad of core challenges: maintaining legacy IT investments while embracing AI innovation, balancing diverse business demands with simplified IT architecture, and ensuring business stability during technological iterations [6] - AI Infra is proposed as a critical engine to resolve these implementation challenges, emphasizing the need for a bridge that connects historical IT assets with future requirements [7][8] Group 4 - AI Infra is defined as a platform that can achieve cost reduction, efficiency improvement, safety, and controllability through capabilities like computing power coordination, storage innovation, architecture integration, and ecological openness [9] - The deployment of AI Infra has shown to increase AI project success rates from 29% to 78%, with a 120% improvement in return on investment [11] - The global AI Infra market is projected to exceed $80 billion by 2025, with a compound annual growth rate of 58%, indicating intense competition among domestic and international players [12] Group 5 - Domestic players focus on local pain points, while international firms emphasize technological barriers, leading to a competitive landscape characterized by full-stack, vertical technology, and ecological integration players [12][13] - Companies like Qingyun Technology and Huawei are addressing historical compatibility issues and enhancing training efficiency through their AI Infra solutions [12] - The competition has evolved from product-based to a comprehensive contest involving technology, ecology, and application scenarios, with a need for domestic firms to overcome core technology bottlenecks [15]

Core Insights - QingCloud Technology officially launched AI Infra 3.0, emphasizing "All in One, One for AI" as its strategic core, aiming to transform the barriers of AI implementation into growth momentum for enterprises [1][2] Group 1: Challenges in Digital Transformation - Enterprises face three core challenges during digital transformation: balancing historical IT investments with AI innovation, simplifying management while meeting diverse business needs, and ensuring business stability alongside rapid technological iteration [2] - The CEO of QingCloud, Lin Yuan, highlighted the mismatch between the disruptive nature of technology iteration and the gradual nature of enterprise development as the essence of the pain points in digital transformation [2] Group 2: Key Features of AI Infra 3.0 - AI Infra 3.0 possesses four key characteristics: full-stack capability, on-demand scalability, standardized delivery, and smooth evolution [3][4][5][6] - The unified architecture provides a stable, scalable, and highly compatible technical foundation for enterprises, integrating virtualization, cloud, cloud-native, and AI computing capabilities [7] Group 3: Core Values of AI Infra 3.0 - The architecture creates four core values for enterprises: investment assurance (75% cost savings), transformation assurance (100% smooth upgrades), operational assurance (over 70% improvement in management efficiency), and innovation assurance (evolutionary architecture meeting future AI needs) [8] Group 4: Industry Applications - In the education sector, QingCloud provides GPU computing, model inference services, and high-performance computing resources, addressing the core challenges of diverse deployment and usage of computing resources in universities [9] - In smart manufacturing, the unified architecture enhances IT response speed by 75% and significantly shortens the cycle from AI model development to production [9] - In the media industry, the open architecture allows for rapid compatibility with various hardware, improving deployment efficiency and addressing the challenges of scattered computing resource management [9] Group 5: Ecosystem Collaboration - QingCloud aims to create a win-win ecosystem in the AI era by collaborating with partners, emphasizing standardization and openness throughout the design, development, testing, and delivery processes [11] - The AI Infra 3.0 architecture supports heterogeneous needs and helps partners meet customer demands through its open and plug-in architecture [11] Group 6: Partner Benefits - Numerous partners have benefited from the AI Infra 3.0 architecture, including a pharmaceutical research group that quickly launched medical AI analysis capabilities and an operator that achieved integrated deployment and management of AI solutions [12]

Core Viewpoint - Qingyun Technology's stock has shown significant volatility, with an 85.64% increase year-to-date, but recent trends indicate a decline in the short term [1][2]. Group 1: Stock Performance - On November 11, Qingyun Technology's stock rose by 2.03%, reaching 62.19 CNY per share, with a trading volume of 21.11 million CNY and a turnover rate of 0.72%, resulting in a total market capitalization of 2.979 billion CNY [1]. - Year-to-date, the stock has increased by 85.64%, but it has experienced a 1.61% decline over the last five trading days, a 2.54% increase over the last 20 days, and an 18.76% decline over the last 60 days [1]. - The company has appeared on the "Dragon and Tiger List" 14 times this year, with the most recent appearance on July 23, where it recorded a net purchase of 97.91 million CNY [1]. Group 2: Financial Performance - For the period from January to September 2025, Qingyun Technology reported a revenue of 165 million CNY, reflecting a year-on-year decrease of 22.83%, while the net profit attributable to shareholders was -49.99 million CNY, showing a year-on-year increase of 20.35% [2]. - As of September 30, the number of shareholders decreased by 9.81% to 11,100, while the average circulating shares per person increased by 47.76% to 4,314 shares [2]. Group 3: Business Overview - Qingyun Technology, established on April 17, 2012, and listed on March 16, 2021, is headquartered in Beijing and primarily provides cloud computing products and services [1]. - The company's main revenue sources are cloud services (66.23%), cloud products (33.44%), and other supplementary services (0.33%) [1]. - The company belongs to the computer-IT services sector and is associated with various concepts, including AI, integrated machines, computing power, DeepSeek, and Xinchuang [2].

证券代码：688316 证券简称：青云科技 北京青云科技集团股份有限公司 2025 年第二次临时股东会会议资料 2025 年 11 月 1 / 14 | 2025 | 年第二次临时股东会会议议程 | 5 | | --- | --- | --- | | | 议案一、《关于购买股权资产暨与关联人共同投资的议案》 | 7 | | | 议案二、《关于变更公司名称、注册资本及修订<公司章程>并办理变更登记的议案》8 | | | | 议案三、《关于修订公司部分管理制度的议案》 9 | | | | 议案四、《关于调整担保相关事项的议案》 10 | | | | 议案五、《关于续聘公司 2025 年度审计机构的议案》 11 | | | | 议案六、《关于增补非独立董事的议案》 14 | | 北京青云科技集团股份有限公司 2025 年第二次临时股东会会议须知 为了维护全体股东的合法权益，确保股东会的正常秩序和议事效率，保证大 会的顺利进行，根据《中华人民共和国公司法》《中华人民共和国证券法》《上 市公司股东会规则》以及《北京青云科技集团股份有限公司章程》《北京青云科 技集团股份有限公司股东会议事规则》等相关规定，特制定 2025 ...

Core Insights - Qingyun Technology Group reported a revenue of 165 million yuan for the first three quarters of 2025, with a year-on-year reduction in net profit loss [1] - In Q3 2025, revenue increased by 21.96% and gross profit grew by 37.28%, with a gross margin increase of 3.93 percentage points compared to Q2 [1] - The company launched over 35 new projects in various emerging industries, including low-altitude economy, embodied intelligence, drug development, material science, autonomous driving, and chip design [1] Financial Performance - Revenue for the first three quarters of 2025 reached 165 million yuan [1] - Net profit attributable to shareholders showed a reduction in loss compared to the previous year [1] - Q3 2025 saw a 21.96% increase in revenue and a 37.28% increase in gross profit compared to Q2 2025 [1] AI Development and Market Position - Qingyun's AI computing platform experienced explosive growth in user numbers, increasing over tenfold year-on-year [1] - The company introduced nearly 30 mainstream large models to create a "computing power + model" full-stack ecosystem, aiming to lower the barriers for enterprises to adopt AI [1] - Qingyun Technology is committed to making each project a benchmark for AI computing applications [1] Strategic Collaborations - The company maintains strategic partnerships with several domestic chip companies, launching large model services based on Huawei's Ascend chips [2] - Collaborations with Northern Computing Network and Shudao Intelligent Computing have been established to achieve deep integration from hardware adaptation to computing power operation [2] - Qingyun joined China Telecom's "AI + Ecological Going Global Alliance" to promote its mature computing services internationally [2]

Mergers and Acquisitions - Qingyun Technology plans to acquire a 25% stake in Zhongtong Tianhong for 37.5 million yuan [1] Earnings Disclosure - Wuliangye reported a 65.62% year-on-year decline in net profit for Q3, with revenue down 52.66% to 8.174 billion yuan [2] - BYD's Q3 net profit decreased by 32.60%, with revenue at 194.985 billion yuan, a 3.05% decline [3] - Changjiang Electric's Q3 net profit fell by 9.16%, with revenue at 29.044 billion yuan, down 7.86% [4] - Weichai Power experienced a 29.49% increase in net profit for Q3, with revenue at 57.419 billion yuan, up 16.08% [5] Shareholding Changes - Yangfan New Materials announced that shareholder Ningbo Xinfan Investment plans to reduce its stake by up to 1%, totaling no more than 2.3475 million shares [6] - Ruisheng Intelligent's controlling shareholder and vice president plan to reduce their holdings by up to 3%, totaling no more than 1.8727 million shares [7] Risk Matters - Guodun Quantum confirmed that there are no undisclosed significant information or ongoing major asset restructuring activities [8]