Core Viewpoint - The company is making significant investments in high-performance computing servers, totaling 21.2 billion yuan, to prepare for the anticipated growth in the computing power market, which is likened to the real estate boom of the 1990s in China [1][2][9]. Group 1: Investment Strategy - The company plans to procure high-performance servers worth 21.2 billion yuan, with a focus on meeting the growing demand for computing power in the AI sector [1][2]. - The company has raised its credit limit to 28 billion yuan within a year to support its aggressive investment strategy [1]. - The company expects a net profit of 1.05 to 1.25 billion yuan for 2025, representing a year-on-year growth of 51.78% to 80.69% [2]. Group 2: Market Position and Demand - The computing power rental market is showing signs of consolidation, with the company positioned as a leading player due to its long-term contracts and established customer relationships [6][12]. - The demand for computing power is expected to outstrip supply, driven by the increasing needs of AI applications, which will lead to sustained growth in the computing power industry [10][12]. - The company has already secured five-year long-term contracts, ensuring a stable revenue stream to support its server procurement [7][8]. Group 3: Future Outlook - The ultimate goal of the company is to develop "robot brains" through its investments in computing power and AI platforms, with expectations for significant advancements in the robotics sector between 2030 and 2035 [3][16][17]. - The company is also focusing on building a comprehensive ecosystem that integrates computing power, cloud services, and intelligent terminals to support its long-term vision [16]. - The company anticipates that the AI-driven demand for computing power will continue to expand, making it a critical component of the digital economy [10][14].

21世纪经济报道记者 吴佳楠 深圳报道 "上午还在长三角见客户，下午已飞往北京洽谈合作。"这是协创数据（300857）董事长耿康铭近期的工 作常态。 在他身后，这家2005年成立于深圳，如今市值超七百亿的企业，正以大规模的投资，将战略重心转向一 场席卷全球的算力浪潮与具身智能的未来之战。 2025年，协创数据（300857.SZ）接连发布五次采购公告，累计拟用于采购高算力服务器的额度达212亿 元。 然而，从经营情况看，截至2025年三季度末，公司实现营业收入83.31亿元，净利润6.98亿元，资产负债 率高达78.33%。 年内服务器采购高达212亿元，这似乎是一场小马拉大车的资本豪赌。与此同时，协创数据也铺开融资 大网，在2025年一年内三度提升授信额度，总额已达280亿元。 面对在外界看来"激进"决策的不解，这位在行业摸爬滚打30多年的业内老兵却显得十分冷静与克制，又 透露出在时代风口前的果敢。 "这笔投入其实非常保守，客户的需求远远大于我们的投入。"在接受21世纪经济报道记者专访时，耿康 铭将当下算力产业发展比作"房地产的90年代"，认为未来五到十年，算力领域的投资规模将超越房地产 巅峰时期，而协创数 ...

Group 1 - The core viewpoint of the news highlights the strategic collaboration between Qichuang Data and Tashan Technology, aiming to enhance AI hardware capabilities and accelerate the application of AI technologies in robotics and other sectors [1][2]. - Qichuang Data's Fcloud platform showcases its comprehensive service capabilities in embodied intelligence, integrating NVIDIA's advanced AI robotics technology to streamline the robot development process and reduce costs [1][3]. - The partnership is expected to leverage Tashan Technology's expertise in tactile sensors and AI chips, combined with Qichuang Data's strengths in computing platform construction and large-scale deployment, creating significant synergies [2]. Group 2 - Qichuang Data's global industrial layout focuses on commercializing physical AI and embodied intelligence technologies through a three-pronged approach: computing foundation, cloud services, and intelligent terminals [3]. - The Fcloud platform provides an end-to-end development environment for embodied robot developers, including simulation, data, training, and model services, enhancing the efficiency of AI infrastructure management [3]. - In the first half of 2025, Qichuang Data reported a revenue of 4.944 billion yuan, a year-on-year increase of 38.18%, and a net profit of 457 million yuan, reflecting a growth of 27.68% [3].

Core Viewpoint - The company is focusing on building a global computing power network centered around embodied intelligence, aiming to accelerate the commercialization of physical AI and embodied intelligence technologies through a three-pronged approach: computing power foundation, cloud services, and intelligent terminals [1]. Group 1: AI Strategy and Development - The company plans to construct large-scale computing clusters, including GPU and edge computing infrastructure, to optimize large model scenarios and provide efficient AI computing solutions for enterprises [1]. - The company has launched the Omnibot embodied intelligence service platform, utilizing NVIDIA's full-stack technology, to support developers in creating embodied robots with end-to-end simulation and AI training capabilities [2]. - The FCloud intelligent training and推 platform will be officially launched in 2025, providing training and推 services across various sectors, including robotics, biomedicine, chip manufacturing, and finance [2]. Group 2: Financial Investments and Growth - The company has invested significantly in computing power, purchasing servers with self-funds amounting to 3 billion and 4 billion yuan, and has increased capital in its wholly-owned subsidiary, further enhancing its competitive edge in intelligent computing services [2]. - The company is experiencing a continuous growth trend in traditional business orders, with revenue from intelligent computing and cloud services also increasing, expected to be reflected in the third and fourth quarters of 2025 [3]. - Future plans include increasing R&D investments and deepening industry chain collaborations to explore new directions in AI technology innovation and application [3].