证券研究报告：计算机 | 公司点评报告 发布时间：2025-05-08 股票投资评级 买入 |首次覆盖 个股表现 2024-05 2024-07 2024-09 2024-12 2025-02 2025-05 -23% 7% 37% 67% 97% 127% 157% 187% 217% 247% 277% 汉得信息 计算机 资料来源：聚源，中邮证券研究所 公司基本情况 | 最新收盘价（元） | 19.73 | | --- | --- | | 总股本/流通股本（亿股）9.85 | / 9.48 | | 总市值/流通市值（亿元）194 | / 187 | | 52 周内最高/最低价 | 25.22 / 5.76 | | 资产负债率(%) | 18.0% | | 市盈率 | 103.84 | | 第一大股东 | 陈迪清 | 研究所 分析师:孙业亮 SAC 登记编号:S1340522110002 Email:sunyeliang@cnpsec.com 分析师:常雨婷 SAC 登记编号:S1340523080001 Email:changyuting@cnpsec.com 自主软件产品业务全年实现收入 17.74 亿 ...

Core Viewpoint - Han's Information is focusing on building autonomous AI agent solutions to promote B-end applications, aiming to enhance its digital and intelligent product offerings for large enterprise clients [1][2]. Group 1: Company Overview - Han's Information has been deeply engaged in the computer software service sector, providing digital and intelligent products and solutions to over 7,000 leading enterprises across various industries including manufacturing, finance, retail, and energy [1]. - The company has transitioned from being a pure ERP service provider to a software product vendor, establishing four main business lines and focusing on traditional ERP and ITO services for stability while driving growth through self-developed products [3]. Group 2: Product Development - At the 2025 Han's User Conference, the company launched its B-end AI application product/service series "De.Ling," which includes three major product series and one service series, aimed at supporting enterprises in building AI capability systems and upgrading their intelligence [2]. - The company plans to release version 1.6 of the Lingyuan AI platform by the end of May 2024, which will fully support MCP [2]. Group 3: Financial Performance - In 2024, the company achieved an 8.57% year-on-year revenue growth, with self-developed software product revenue reaching 1.774 billion yuan, a 16.17% increase from the previous year [4]. - The revenue from self-developed software products accounted for 54.83% of total revenue in 2024, up from 51.24% in 2023, with industrial digitalization showing the most significant growth at 21.27% [4]. Group 4: Incentives and Growth Strategy - In April 2024, the company announced a restricted stock incentive plan, proposing to grant 41.98 million shares to 268 incentive targets, which represents 4.26% of the total share capital [5]. - The incentive plan includes ambitious revenue and net profit targets for 2025, which are expected to enhance the company's competitive capabilities and motivate employees [5]. Group 5: Investment Outlook - The company is projected to have EPS of 0.25, 0.31, and 0.38 yuan for 2025, 2026, and 2027 respectively, with corresponding PE ratios of 80.00, 64.29, and 52.94 times [6]. - Given its position as a leading digital service provider and the integration of advanced technologies with ecosystem partners, the company is expected to accelerate the deployment of AI agents in enterprises, leading to a "buy" rating [6].

Core Viewpoint - Xueersi has launched three series of learning machines (P, S, T) aimed at reshaping the learning experience through "good AI + good content" [1] Group 1: AI and Content Features - The "good AI" aspect relies on the self-developed Jiuzhang model and DeepSeek, achieving a comprehensive upgrade of the intelligent learning system [3] - The learning machines have added 4 million minutes of premium courses and 850,000 sets of real questions since their launch, promoting the idea that "asking questions is a better way to learn" [3] - The "Xiao Si AI 1-on-1" interactive feature breaks the traditional passive response model, actively guiding learning and engaging in continuous dialogue with students [3][4] Group 2: Learning Machine Series - The P series offers a complete and foundational learning solution covering all educational stages from enlightenment to senior high school, integrating classic practice modules and 50 AI learning tools [4] - The S series incorporates a twelve-level mathematics ability training system tailored to different cognitive characteristics of students, with 55 AI tools and a "Gold Medal Learning" feature for personalized learning planning [4] - The T series, as the flagship product, includes the Xiao Si Smart Island Pro, providing comprehensive upgrades and one-stop intelligent homework tutoring services [4][5] Group 3: Learning Efficiency and Engagement - The "Filtering Learning" and "Filtering Practice" functions in the Precision Learning 3.0 system help students focus on key topics and avoid blind practice, enhancing learning efficiency and reducing burden [3] - The homework mode in the T series supports detailed homework reports and allows parents to monitor their children's learning progress [5] - The exploration mode uses engaging tasks to stimulate children's interest in learning and develop their autonomous learning abilities [5] Group 4: Industry Challenges - The intelligent learning head, Zhao Puzheng, analyzed the "impossible triangle" dilemma in education, highlighting the difficulty of balancing high quality, large scale, and personalization [5]

Core Insights - The MCP protocol is fundamentally reshaping the AI agent ecosystem, emphasizing the symbiotic relationship between infrastructure and applications rather than a linear progression from one to the other [3][10][30] - The rapid growth of MCP services indicates a strong market demand for a well-recognized protocol to expand the capabilities of AI agents beyond dialogue [4][17] Infrastructure and Application Symbiosis - The relationship between infrastructure and applications is crucial for understanding the MCP ecosystem, as new applications drive the need for more complex infrastructure, which in turn fosters new application categories [10][11] - The case of Arcade.dev illustrates how application demands can lead to infrastructure development, showcasing the cyclical nature of this relationship [11][30] MCP's Problem-Solving Capabilities - MCP addresses the challenges faced by AI agents in high-demand application scenarios, enabling them to connect with external services for real-time data and operations, thus transforming AI assistants into practical tools [12][14] - Traditional APIs conflict with AI models' probabilistic outputs, but MCP standardizes interactions, allowing AI models to discover and use tools effectively while maintaining conversational context [14][15] Current State of MCP - The MCP ecosystem is experiencing explosive growth, driven by the supply of servers and development tools, with significant increases in server creation and SDK downloads [17][18] - The distribution of actual applications is highly concentrated, with a few servers achieving high installation numbers, indicating a power-law distribution [20][24] Emerging Patterns from Supply and Demand - The current demand is focused on desktop applications and single-user scenarios, with AI assistants like Cursor becoming primary sources of MCP demand [30] - The rise of reasoning tools and web/UI automation indicates that infrastructure development is aligning with the capabilities required for advanced AI applications [30] Opportunities for Entrepreneurs - Infrastructure and tool providers should embrace MCP while maintaining strategic flexibility, as the protocol is still evolving and may face adoption barriers [35] - Developers can focus on creating value by addressing real developer pain points and enhancing the capabilities of AI agents through the MCP framework [36][38] Future Outlook - The MCP ecosystem is accelerating the interaction between infrastructure and applications, with the potential for rapid evolution in AI capabilities [39] - Developers who can identify and solve past challenges will set the benchmarks for the next generation of applications in the AI-driven software landscape [39]

近日，记者来到位于南京市江宁开发区的英飞源华东研发生产基地项目现场，坐落在正方大道北侧的主 体大楼雏形初现，今年8月就要竣工交付。深圳英飞源技术有限公司董事长朱春辉已经习惯了深圳、南 京两头跑的节奏。"年底全部搬进来，未来这里将是我们的中国区总部，预计3年内产值达到30亿元。" 项目建设如火如荼，企业研发生产一刻也没闲着。位于江宁区秦淮路、九龙湖的两处临时过渡性厂房 里，充电模块、全液冷超充桩等自研产品整齐装箱，搭乘新能源汽车产业爆发的"快车"不断出货，去年 就实现5亿元产值，今年一季度又跑出30%的增速。 建设、生产"两手抓"，英飞源在江宁不是个例。总投资40亿元的省重大产业项目美埃高端环保装备生产 基地，今年底竣工、明年6月投产，工业除尘器、自带风机过滤机组、医药用净化设备等产品销往半导 体工厂、商场医院等，投资方美埃(中国)三年前在科创板挂牌上市。"一周前我刚从匈牙利回来，在那 里新谈成一笔订单！"美埃中国区常务总经理颜文礼一脸兴奋。 持续释放猛攻项目的强烈信号，江宁区近日召开一季度经济运行推进点评会，会前由党政"一把手"带 队，先去看了以英飞源、美埃为代表的4个项目现场。"要扭住不放增量项目，用好用足 ...

Group 1 - The article highlights the importance of weekly announcements from Sunday to Thursday, focusing on significant stock market events such as suspensions, investments, acquisitions, and performance reports [1] - A company plans to invest over 16 million yuan in building a humanoid robot data collection factory, utilizing a hyper-converged computing scheduling platform for multiple intelligent computing centers [1] - Another company is involved in humanoid robots and cloud computing, with products widely used in robotics and strategic partnerships with firms like UBTECH [1] Group 2 - A company is set to participate in establishing a joint venture focused on the localization of high-end packaging and testing equipment, supported by national funds in the semiconductor sector [1]

Group 1 - The core concept of "AI Agent" is gaining significant attention from major tech companies globally, including Microsoft, Google, Amazon, OpenAI, Alibaba, Tencent, ByteDance, and Baidu, as they view it as a key business direction [1][2] - Market research firms like Forrester and Gartner predict that AI Agents will be among the critical emerging technologies by 2025, with Gartner ranking it as the top technology trend [1] - According to Gartner, only about 1% of enterprise software will have AI Agent capabilities by 2024, but this is expected to rise to 33% by 2028, with AI expected to automate 15% of daily business decisions [1] Group 2 - AI Agents are defined as systems capable of autonomous planning and task execution, differing from traditional AI systems that require continuous human interaction [4] - AI Agents can be either virtual or embodied, with the former existing in digital environments and the latter having physical forms like self-driving cars and humanoid robots [5] - The development of open standard communication protocols for AI Agents, such as MCP, ANP, and A2A, enables them to utilize external tools and collaborate with one another, enhancing their capabilities [7][9] Group 3 - The rise of AI Agents is expected to disrupt the existing platform-centric business ecosystem, leading to new business forms, organizational structures, and models [2][10] - AI Agents will change the decision-making landscape in business, as they will operate with a focus on optimal solutions, contrasting with human decision-making, which often seeks satisfactory outcomes [12] - The traditional "data is king" paradigm may shift, as AI Agents will not rely on human behavior data for decision-making, altering the competitive landscape [18] Group 4 - The emergence of AI Agents could significantly impact platform-based business models, as they can efficiently match transactions without the need for intermediaries, reducing the value of platforms [13][14] - Current business strategies that rely on capturing human attention, such as auction-based advertising and recommendation algorithms, may become less effective as AI Agents take over information retrieval tasks [16][17] - The nature of collaboration in business may evolve, with AI Agents facilitating deeper and broader cooperation without the constraints of traditional organizational structures [19][20] Group 5 - The traditional frameworks for analyzing business competition, such as Porter's Five Forces and Resource-Based View, may become less applicable in the context of AI Agents [23][26] - A shift in focus is necessary to understand the new dynamics introduced by AI Agents, emphasizing their network properties and collaborative capabilities [27] - The competitive landscape will require a reevaluation of metrics and strategies, moving from human-centric models to those that prioritize AI Agents and their interactions [27]

Group 1: Overall Performance - The overall performance for Q4 2024 and Q1 2025 met expectations, with resilience in both domestic and foreign demand[1] - The retail sales of passenger vehicles in Q4 2024 increased by 17% year-on-year, supported by trade-in policies and government subsidies[26] - In Q1 2025, retail, export, and wholesale figures all showed positive year-on-year growth of 3%, 6%, and 13% respectively[26] Group 2: Automotive Sector Insights - The automotive sector is expected to benefit from a potential easing of the US-China trade war, which may alleviate previous concerns regarding external demand[2] - The AI and robotics sectors are prioritized for investment, with companies like Xpeng Motors and Horizon Robotics highlighted as key players[2] - The penetration rate of new energy vehicles in retail sales showed a slight decline, indicating a need for strategic adjustments[27] Group 3: Company-Specific Performance - Xpeng Motors reported a 23% increase in revenue for Q4 2024, with monthly deliveries exceeding 30,000 units despite seasonal disruptions[5] - BYD's revenue grew by 53% in Q4 2024, with a profit increase of 73%, driven by strong export performance[5] - The gross profit margin for the automotive sector showed mixed results, with some companies experiencing margin pressure due to increased competition and pricing strategies[3] Group 4: Risks and Challenges - Risks include the potential escalation of the trade war, lower-than-expected global economic recovery, and uncertainties in geopolitical conditions[2] - The automotive industry faces challenges from rising raw material costs and the need for continuous innovation in L3-L4 autonomous driving technologies[2]

Group 1 - The core viewpoint of the articles revolves around the rapid development and commercialization of AI agents, particularly following the success of Manus, which has sparked significant interest and investment in this sector [2][3][4]. - Major tech companies are intensifying their efforts in the AI agent space, with ByteDance reportedly forming at least five teams to develop various AI agent products, and Baidu launching the "Xinxiang" app, which aims to compete with Manus [4][5]. - The investment landscape is also shifting, as evidenced by the $75 million funding round for Manus's parent company, Butterfly Effect, which has raised its valuation to nearly $500 million [2]. Group 2 - The emergence of AI agents is seen as a solution to the unmet business needs and technological gaps left by previous enterprise digital transformation efforts [3]. - Companies are adopting the MCP (Multi-Cloud Platform) mechanism to enhance the ecosystem of AI agents, with major players like Alibaba, Tencent, and Baidu integrating MCP protocols into their AI products [6]. - There is a growing concern regarding the safety and risk management of AI agents, as many companies lack a comprehensive understanding of the associated risks, with a significant portion of clients unaware of what AI agents entail [7][8]. Group 3 - The concept of AI agents is evolving, with new terminologies such as Agentic AI and Agentic Workflow gaining traction, indicating a shift towards more specialized and collaborative AI systems [10][11]. - The industry is focused on making AI agents adaptable to complex application scenarios, requiring advancements in perception, understanding, planning, and execution [11][12]. - There is a call for a more cautious approach to the deployment of AI agents, emphasizing the need for improved governance and risk assessment capabilities before widespread implementation [12].