在首轮问询回复中，本次募投项目研发支出资本化率为83.33%，高于公司现有同类项目资本化率 64.94%。公司表示，前次募投项目研发成果和持续技术积累为本次项目奠定良好基础，使得费用化投 入比例相对较少。 从具体测算依据来看，本次募投项目根据各项目研发内容和复杂性配置人员，计划建设3年。按相同岗 位职级人员薪酬测算开发支出，结合技术基础、客户基础等情况，参考现有同类项目7个月的资本化时 点，预计本次项目拟资本化开始时间相距研发投入6个月。经测算，费用化支出2597.04万元，资本化支 出12985.20万元，开发支出拟资本化率为83.33%。 与现有同类项目相比，本次募投项目在多方面存在差异。在项目资本性支出构成上，本次募投项目增加 了项目场地投入，因公司2023年8月退租部分办公场所，致济南地区人均场地办公面积较低。开发支出 构成上，两者均以研发人员薪酬为主，但本次募投项目所需人员岗位、级别不同，部分岗位人员月均工 资有差异。开发支出资本化时点上，本次项目略短于现有同类项目，因本次募投项目技术储备和基础充 足，且客户前景、人才储备、政策推动等条件有利。 普联软件（300996）近日发布《关于普联软件股份有限 ...

Group 1 - The core viewpoint of the article highlights that Inspur Digital Enterprise (00596) has seen a significant increase in stock price, reaching a historical high of 11.12 HKD, driven by strong performance in its cloud services business [1] - The company expects to achieve a net profit attributable to shareholders of approximately 180 to 190 million RMB for the first half of the year, primarily due to the substantial growth and profitability of its cloud services [1] - According to Zhongtai Securities, Inspur Digital Enterprise benefits from its affiliation with Inspur Group and has accumulated extensive industry experience, which supports its ongoing cloud transformation and enhances operational and profit quality [1] Group 2 - The article mentions that the National Cyberspace Administration has addressed security risks related to backdoor vulnerabilities in H20 computing chips with NVIDIA, indicating a focus on technology security [1] - Shanghai Securities suggests that the uncertain external environment and technology sanctions may accelerate the push for technological independence and supply chain self-sufficiency [1] - The State-owned Assets Supervision and Administration Commission (SASAC) has mandated that all central enterprises achieve secure and credible information system replacements by the end of 2027 [1]

Core Viewpoint - Inspur Digital Enterprise (00596) has seen a significant increase in stock price, reaching a historical high due to strong expected profit growth driven by cloud service transformation [1] Group 1: Company Performance - Inspur Digital Enterprise anticipates a net profit attributable to shareholders of approximately 180 to 190 million RMB for the first half of the year, primarily due to substantial growth and profitability in its cloud services business [1] - The company is recognized as a comprehensive ERP provider in China, leveraging its extensive industry experience and knowledge to enhance its cloud transformation efforts, leading to improved operational and profit quality [1] Group 2: Market Environment - The National Cyberspace Administration of China has raised security concerns regarding NVIDIA's H20 computing chip vulnerabilities, indicating a push towards technological self-sufficiency and supply chain control amid external uncertainties and tech sanctions [1] - The State-owned Assets Supervision and Administration Commission (SASAC) has mandated that all central enterprises achieve secure and credible information system replacements by the end of 2027, further emphasizing the need for domestic technology solutions [1]

Core Viewpoint - Kingdee has established the "Enterprise Management AI Ecological Alliance" to address the challenges of AI application for small and medium-sized enterprises (SMEs), emphasizing that AI is essential for enterprise management and a key focus for the company's future development [1][4]. Group 1: AI and Large Clients - The strategy of targeting large clients is crucial for Kingdee, as one large client is significantly more valuable than multiple small clients, given the instability and low payment capacity of SMEs [2]. - Kingdee has historically aimed to serve larger enterprises and government services rather than being labeled as a provider for SMEs [3]. - The company is poised to benefit from market and technological advantages, particularly due to policies mandating state-owned enterprises to complete domestic replacements by 2027, which could release a significant portion of the high-end ERP market [4][6]. Group 2: AI Product Development - Kingdee has launched several AI enterprise service products aimed at large clients, including the "China's first financial large model" and upgrades to its AI platform [9][10]. - Despite these efforts, Kingdee's revenue from large enterprises remains low compared to SMEs, with large enterprise cloud service revenue reported at approximately RMB 1.304 billion in 2024, while revenue from SMEs was significantly higher at RMB 21.53 billion [10][11]. Group 3: Challenges in Large Client Acquisition - Kingdee's reliance on SMEs persists, with their revenue contribution from SMEs being 259% higher than that from large enterprises, indicating a need for improvement in attracting larger clients [11]. - The primary challenge in acquiring large clients lies in Kingdee's limited customization capabilities and system functionality, which are critical for meeting the complex needs of large enterprises [12][15]. - To succeed, Kingdee must enhance its product's foundational capabilities, focusing on customization and personalized service for large enterprises, rather than solely on AI concepts [15][16]. Group 4: Future Directions - Kingdee aims to aggressively target the large enterprise market, viewing it as essential for becoming a leading PaaS provider by 2026 and a world-class SaaS provider by 2030 [14]. - The company has invested heavily in R&D, with expenses reaching RMB 1.514 billion in 2024, yet it has faced continuous losses, primarily due to the inability to penetrate the large client market effectively [17].

Group 1 - The year 2025 is anticipated to be a turning point for AI application, driven by large models, leading to new creativity across various industries [1] - Despite the increasing total data production, the growth rate of the data security market has not been ideal, with a predicted surge in demand as the risks associated with data loss become more pronounced [1][5] - The report from the National Data Resource Survey (2024) indicates that the annual data production in China will reach 41.06 zettabytes (ZB), a 25% year-on-year increase [5] Group 2 - "Small data security" is emerging as a core focus in the digital economy, where AI models can refine large datasets into high-value, concentrated "small data" [3] - The integration of large models into critical sectors like smart cities and industrial control systems poses significant risks, as unprotected models can lead to widespread operational failures [4] - Over-reliance on AI for decision-making can create a "butterfly effect," where errors in model outputs can have cascading consequences [4] Group 3 - The current state of the data security market is characterized by a lack of urgency in investment, as companies often only react to data breaches rather than proactively securing their data [5][6] - The shift towards high-density "small data" will necessitate increased investment in data security, as the consequences of data loss become more severe [5][6] - The network security industry is expected to see new growth opportunities due to the complexities introduced by AI and emerging business scenarios [6] Group 4 - The "vehicle-road-cloud integration" development path is recognized as a key infrastructure for the future, with significant market potential as government policies support investment in this area [7] - Ensuring data security compliance in this sector is crucial for maintaining traffic order and public safety [7]