证券时报·数据宝统计显示，海天瑞声今日收盘价为129.29元，上涨1.74%，日换手率为6.22%，成交额 为4.91亿元，全天主力资金净流入2656.90万元，近5日该股累计上涨0.93%，近5日资金合计净流入 529.95万元。 两融数据显示，该股最新融资余额为3.52亿元，近5日增加1488.46万元，增幅为4.42%。（数据宝） 8月18日海天瑞声大宗交易一览 海天瑞声8月18日大宗交易平台出现一笔成交，成交量8.93万股，成交金额1000.16万元，大宗交易成交 价为112.00元，相对今日收盘价折价13.37%。该笔交易的买方营业部为华泰证券股份有限公司总部，卖 方营业部为华泰证券股份有限公司南京中山东路华泰证券大厦证券营业部。 进一步统计，近3个月内该股累计发生15笔大宗交易，合计成交金额为6560.16万元。 | 成交量 | 成交金额 | 成交价 | 相对当日收盘 | | | | --- | --- | --- | --- | --- | --- | | （万 | （万元） | 格 | 折溢价（%） | 买方营业部 | 卖方营业部 | | 股） | | （元） | | | | | 8.93 | ...

| 交易日期 | 证券简称 | 证券代码 | 成交价(元) 成交金额(万元) 成交量( * ) 买入营业部 | | | 卖出营业部 | | --- | --- | --- | --- | --- | --- | --- | | 2025-08-18 | 海天瑞声 | 688787 | 112 1000.16 | 8.93 | 华泰证券股份有限 公司总部 | 华泰证券股份有限 公司南京中山东路 | | | | | | | | 华泰证券大厦证券 | 8月18日，海天瑞声大宗交易成交8.93万股，成交额1000.16万元，占当日总成交额的2%，成交价112元，较市场收盘价129.29元折价13.37%。 ...

Group 1 - The first humanoid robot data training center in Beijing was inaugurated, showcasing innovations and applications in various fields such as new retail, home care, and special operations [1][2] - Current challenges in the robotics industry include heavy and expensive traditional robotic arms, single-function service robots, and a lack of adaptability to complex environments [1] - A reliable operational capability and acceptable costs are essential for the large-scale application of robots across various industries, creating a positive feedback loop of "application-data-intelligence-application" [1] Group 2 - The humanoid robot data training center covers an area of 3,000 square meters and features a "nine-year consistent" training area and a "robot university" scenario area [2] - The center is equipped with 108 diverse robots and has established ten real-world application scenarios, focusing on data collection and model training [2] - It is expected that the center will produce over one million high-quality multimodal data entries annually [2]

Company Dynamics - The Shijingshan Humanoid Robot Data Training Center Open Day was held on August 15, with participation from nearly 40 representatives from various academic and research institutions [2] - Beijing Haitian Ruisheng Technology Co., Ltd. (Haitian Ruisheng) unveiled its embodied intelligence data training ground, marking its official establishment [2] - Both parties aim to leverage their respective advantages in artificial intelligence data resources and robot R&D technology for future collaboration in embodied intelligence development and training ground construction [2][6] Industry Development - The humanoid robot data training center was established on March 29, showcasing rapid development from "bare shell" to "108 robots in operation" in just 22 days, highlighting the "Shijingshan efficiency" [4] - The newly unveiled embodied intelligence data training ground integrates advanced technologies and products from over ten companies in the embodied intelligence industry chain, creating a comprehensive industrial ecosystem [4] - The training ground aims to establish a closed-loop capability covering scene data collection, model training, and application validation, accelerating the transition of robotic technology from laboratory to application [6]

Core Viewpoint - The stock price of Airong Software reached 60.10 yuan as of August 15, 2025, reflecting a rise of 2.33 yuan from the previous trading day, indicating positive market sentiment towards the company and its sector [1]. Company Summary - Airong Software operates in the software development industry, focusing on financial technology solutions for banks and financial institutions. The company's services include financial IT system development, big data applications, and artificial intelligence technology research and development [1]. Industry Summary - The digital currency sector has shown active performance recently, with multiple stocks in this category experiencing collective gains. Market information indicates a strong overall trend in the digital currency sector, with several stocks rising over 5% [1].

Core Viewpoint - As of August 15, 2025, Haitan Ruisheng (688787) reported a closing price of 127.08 yuan, reflecting a 0.47% increase, with a trading volume of 21,000 lots and a transaction amount of 266 million yuan [1] Financial Performance - For the first quarter of 2025, the company reported total operating revenue of 69.81 million yuan, representing a year-on-year growth of 71.75% [1] - The net profit attributable to shareholders was 371,600 yuan, showing a year-on-year increase of 158.60% [1] - The non-deductible net profit was 684,800 yuan, with a year-on-year growth of 74.12% [1] - The current ratio stood at 7.034, the quick ratio at 6.659, and the debt-to-asset ratio at 8.22% [1] Investment and Intellectual Property - Haitan Ruisheng has invested in 14 companies and participated in 93 bidding projects [2] - The company holds 36 trademark registrations and 46 patents, along with 13 administrative licenses [2]

Investment Rating - The report maintains an "Overweight" rating for the computer industry [1] Core Insights - The report emphasizes the importance of "Artificial Intelligence +" as a key investment opportunity, highlighting the need to focus on specific applications rather than just AI technology itself [5][55] - It identifies significant differences in AI industry logic between China and the US, suggesting that China should leverage its comparative advantages in data, industrial chain, market size, and application scenarios [5][25] Summary by Sections 1. AI Application Stage - AI applications are on the verge of rapid growth, characterized by cost reductions and increased penetration rates [10] - The performance of large models has significantly improved, with costs dropping dramatically [13][17] 2. Comparative Advantages in AI - China has a unique advantage in data, with over 80% of data still undeveloped, presenting substantial economic potential [29][30] - The manufacturing sector in China is robust, contributing to 28.9% of global manufacturing value added in 2024 [11][38] - The domestic market is vast, with over 1.4 billion people and a growing middle-income group, providing rich application scenarios [12][41] 3. Industry Catalysts and Policy Trends - The report highlights the importance of industry catalysts, such as the release of new large model versions (e.g., GPT-5) and supportive government policies [55][56] - The recent policy initiatives emphasize the integration of AI into various sectors, aiming to enhance traditional industries [58][59] 4. Stock Selection Logic - Six key stock selection strategies are outlined: 1. Changes brought by new large model versions 2. Top-down policy relevance 3. Bottom-up event-driven or strong company fundamentals 4. Large-cap institutional stocks 5. Mapping to US stocks 6. Low valuation stocks [65]

Group 1 - The digital watermark concept index declined by 0.58%, ranking among the top declines in the concept sector, with notable declines in stocks such as Hengxin Oriental, ST Yilianzhong, and Haitai Ruisheng [1][2] - Among the digital watermark concept stocks, 19 experienced net outflows from main funds, with a total outflow of 173 million yuan, led by Hengxin Oriental with a net outflow of 120 million yuan [1][2] - The top gainers in the digital watermark concept included Beixin Source, Hongbo Co., and Digital Vision, with increases of 1.99%, 1.66%, and 1.54% respectively [1][2] Group 2 - The concept sectors with the highest gains included the military equipment restructuring concept at 5.58% and liquid-cooled servers at 3.24%, while the digital watermark concept was among the sectors with the largest losses [1] - The main fund inflows were observed in stocks such as Hongbo Co., Beixin Source, and Dongfang Guoxin, with inflows of 44 million yuan, 32 million yuan, and 29 million yuan respectively [1][2] - The digital watermark concept stocks with the largest net outflows included Ruike Laser and Haitai Ruisheng, with outflows of 62.9 million yuan and 42.1 million yuan respectively [1][2]

Core Insights - The demand for high-quality AI training data has surged due to the rise of large models, leading to increased costs for data service providers [1][2] - The market for AI pre-training data services is projected to reach 16 billion yuan by 2027, with a compound annual growth rate of 28.9% over five years [2] - Companies are facing pressure on their financial performance as they invest heavily in large model development, raising concerns about the return on investment [7][8] Group 1: Opportunities - The explosion of large models has created a significant demand for high-quality data across various industries, prompting AI data service companies to secure partnerships with large model developers and research institutions [2][3] - Major AI data service companies in China have announced collaborations with large model firms, indicating a robust market for high-quality data sets [3] - The need for diverse and complex data requirements has increased, as clients seek advanced capabilities from large models [3] Group 2: Costs - The costs associated with data services have risen significantly due to the need for enhanced computational power and skilled labor [4][5] - Data service providers are now required to invest in more powerful hardware and hire highly educated personnel, which has led to increased operational costs [5][6] - The shift from low-cost labor to a more skilled workforce for data annotation has further escalated costs, with companies now seeking university graduates or higher [5][6] Group 3: Challenges - Despite the enthusiasm for large models, AI data service companies are experiencing financial strain, as reflected in their quarterly reports [7][8] - Regulatory scrutiny has increased, with companies receiving inquiries about the necessity of their fundraising efforts for large model projects [7][8] - The current market for large models is still in its infancy, and the full potential for data demand has yet to be realized, leading to uncertainty about future revenue [8][9] Group 4: Industry Outlook - The data industry is viewed as a long-term investment, with companies encouraged to be patient as they build capabilities and market presence [9][10] - The emergence of large models is seen as a positive development for the data industry, with expectations for rapid growth in pre-training data demand as applications become more widespread [10]

天眼查商业履历信息显示，北京海天瑞声科技股份有限公司，成立于2005年，位于北京市，是一家以从 事互联网和相关服务为主的企业。企业注册资本6032.518万人民币，实缴资本3210万人民币。公司法定 代表人为贺琳。 资金流向方面，今日主力资金净流出3210.45万元，占比成交额11.43%。其中，超大单净流出1628.47万 元、占成交额5.8%，大单净流出1581.98万元、占成交额5.63%，中单净流出流入155.87万元、占成交额 0.55%，小单净流入3054.58万元、占成交额10.87%。 海天瑞声最新一期业绩显示，截至2025一季报，公司营业总收入6980.95万元、同比增长71.75%，归属 净利润37.16万元，同比增长158.60%，扣非净利润68.48万元，同比增长74.12%，流动比率7.034、速动 比率6.659、资产负债率8.22%。 金融界消息 截至2025年8月11日收盘，海天瑞声（688787）报收于128.1元，上涨0.16%，换手率 3.63%，成交量2.19万手，成交金额2.81亿元。 通过天眼查大数据分析，北京海天瑞声科技股份有限公司共对外投资了14家企业，参与招投 ...