转自：新华财经 "仅靠硬件难以触达客户核心业务价值。"一位接近恒为科技的行业人士表示，数珩科技的加入，恰好补上了这一短板。数珩科技是国内少有的AI RaaS模 式实践者，已为快消、汽车、金融等行业数十家品牌提供服务，近三年营收、利润均实现快速增长。 公开资料显示，数珩科技成立于2017年，是一家领先的企业级场景化AI解决方案服务商。数珩信息融合业务策略与可本地化部署的S-GPTAI引擎，将大模 型技术真正应用至企业业务场景中，通过RaaS模式，颠覆性地赋能用户的业务场景，贴身满足行业个性化需求，能够真正做到大模型的行业化和本地 化。 当前，数珩科技核心产品线已形成完整矩阵。在内容创作领域，AIGC多模态模型为日化、家装行业生成广告级图片与电影级短片；在智能交互领域，实 时响应集成服务机器人IRISBOT已接入数千C端用户并周级迭代，可在汽车4S店自然解答顾客疑问；在企业培训领域，AI陪练系统通过沉浸式场景模拟与 全维度评价体系，提升员工销售等能力；知识管理领域，企业定制智库整合全量信息，方便金融等行业员工智能查询等。 恒为科技收购数珩科技，是其AI战略从基础设施向应用场景的延伸。一方面，恒为科技基于昇腾、昆仑 ...

Core Viewpoint - Tianrongxin is actively engaging with investors regarding its collaboration with Huawei and the development of its intelligent computing cloud business, highlighting significant growth and future potential in these areas [1][3]. Group 1: Investor Relations Activity - The investor activity was part of the "2025 Guangdong District Investor Online Reception Day" held on September 19, 2025, via the "Panorama Roadshow" website [2]. - Key personnel from the company included Chairman and General Manager Li Xueying, and other executives [2]. Group 2: Collaboration with Huawei and Business Growth - Since launching its cloud computing products in 2019, Tianrongxin's cloud computing business has experienced a compound annual growth rate (CAGR) of 44.76%, with this segment accounting for 9.06% of total revenue in 2024 [3]. - The company launched the intelligent computing integrated machine in February 2025, upgrading its cloud computing business to intelligent computing cloud, with nearly 100 million yuan in business opportunities [3]. - Recent collaboration with Huawei has led to the release of intelligent computing integrated machine models based on the Kunpeng + Ascend architecture, currently generating revenue in the million-level range [3]. Group 3: Market Trends and Future Outlook - The intelligent computing cloud accounted for 12.40% of overall business in the first half of 2025, a year-on-year increase of 3% [3]. - The AI and security market is projected to grow significantly, with IDC forecasting the Chinese security intelligent body market to reach $1.6 billion by 2028, with a CAGR exceeding 230% [3]. - The network security industry is expected to see a market size of $17.1 billion by 2028, with a CAGR of 9.2% [3]. Group 4: International Expansion and Industry Positioning - Tianrongxin is expanding its international business, participating in digital infrastructure security construction along the "Belt and Road" initiative [3]. - The company is focusing on the automotive internet security sector, collaborating with major enterprises like BAIC and GAC to implement relevant solutions [3]. - The company has seen improvements in gross margin and net margin, with a year-on-year increase of 4.1 percentage points and 15.73 percentage points, respectively, in the first half of the year [3].

Core Viewpoint - Tianrongxin (002212.SZ) is actively expanding its offerings in the intelligent computing sector, focusing on the development and deployment of integrated computing machines and cloud platforms to meet the growing domestic demand for intelligent cloud solutions [1] Group 1: Product Development - In February 2025, the company plans to launch its integrated intelligent computing machine, having already released the intelligent computing cloud platform and 17 models of integrated machines catering to various performance needs [1] - The integrated machines are designed to provide full-stack cloud capabilities for clients in government, finance, and telecommunications sectors [1] Group 2: Market Application - Tianrongxin's intelligent computing cloud has been successfully implemented in various fields, including government services, education, and healthcare, with current business opportunities nearing 100 million [1] Group 3: Strategic Partnerships - Recently, the company partnered with Huawei to launch a new model of integrated intelligent computing machine based on the Kunpeng and Ascend architecture, addressing the domestic demand for intelligent cloud construction [1] - The revenue from this collaboration is currently at a million-level, representing a small proportion of the overall revenue [1]

Group 1 - The core viewpoint of the article highlights the risks associated with AI servers being vulnerable to cyber attacks, emphasizing the need for a robust cybersecurity defense system [2] - The company has made investments and established a presence in the AI + security and intelligent computing cloud sectors, aiming to support the intelligent transformation of various industries [2] - In the AI + security domain, the company has launched innovative products and services such as AI version firewalls, large model security gateways, large model security assessments, and network data leakage prevention [2] Group 2 - In the intelligent computing cloud sector, the company has introduced products like the intelligent computing cloud platform and integrated intelligent computing machines, achieving a deep integration of "security + computing power" [2] - The intelligent computing cloud products include capabilities for full lifecycle management of intelligent agents, panoramic views of container assets, image vulnerability scanning, model service API control, and sensitive word detection [2] - These offerings are designed to meet the needs of intelligent computing centers, private deployment and training inference of large models, and security protection for large models [2]

Group 1: Core Insights - The year 2025 is widely regarded as the "Year of AI Agents," marking the mainstream adoption of agent-based AI technology [2][12] - Agent products are evolving from simple tools to complex, collaborative ecosystems that will significantly change how businesses automate complex tasks and make decisions [2][3] - There is a growing concern regarding user interest and renewal rates for previously popular agent products, attributed to issues such as exaggerated capabilities and unclear pricing structures [3][4] Group 2: Industry Trends - The integration of AI agents into enterprise software is expected to rise dramatically, with projections indicating that by 2028, approximately 33% of enterprise software applications will embed agent-based AI [12] - The financial sector is seen as a promising area for AI application, but it faces challenges such as budget constraints, talent shortages, and planning issues [8][9] - By 2026, it is anticipated that at least 30% of generative AI projects in the banking sector will be abandoned due to poor data quality and budget limitations [8][9] Group 3: Technological Developments - The emergence of integrated computing solutions, such as all-in-one generative AI stacks, is expected to enhance the capabilities of AI agents, allowing for faster iterations and improved decision-making [6][7] - The development of specialized models tailored to specific industries or tasks is anticipated to increase performance and reduce issues like hallucinations in AI outputs [7][11] - The financial industry is moving towards a new phase of AI application, transitioning from basic automation to more autonomous decision-making capabilities [11][12]

Core Viewpoint - ZTE Corporation is focusing on the development of liquid-cooled servers, AI solutions, and high-performance computing infrastructure, positioning itself as a leader in the intelligent computing industry [1] Group 1: Liquid-Cooled Servers - The company’s liquid-cooled servers are primarily aimed at the high-performance computing market, with performance ranking among the best in the industry, catering to diverse computing scenarios [1] - ZTE has successfully delivered liquid-cooled servers to domestic operators and the internet sector [1] Group 2: Robotics - ZTE is collaborating with ecosystem partners on humanoid robots, focusing on computing power, networking, and intelligent hardware and software research and development [1] Group 3: Artificial Intelligence - The company is committed to investing in artificial intelligence, positioning intelligent computing as a long-term strategic focus, enhancing product development and industry ecosystem collaboration [1] - ZTE is developing a full-stack computing infrastructure, including self-developed chips, intelligent computing servers, and large-scale clusters, to provide end-to-end AI solutions [1] - The self-developed high-performance custom processor chips aim to reduce total cost of ownership (TCO) for applications like cloud computing and big data processing [1] - The company’s AI Booster software platform offers a one-stop AI solution, simplifying AI innovation and reducing model migration costs for clients [1] - ZTE's self-developed Nebula large models have achieved top rankings in various AI inference benchmarks, indicating strong capabilities in algorithm innovation and data engineering [1] Group 4: Computing Infrastructure - ZTE provides a range of high-performance training servers and cost-effective inference servers, meeting diverse intelligent computing needs from central to edge [1] - The company’s intelligent computing super-node system, based on its self-developed AI large-capacity switching chip "Lingyun," supports large-scale interconnection of domestic GPU cards [1] - ZTE has launched high-density switches with leading performance in the market, catering to various sectors including telecommunications and finance [1] Group 5: Military Sector - The company has not engaged in any business within the military sector [1]

Core Insights - The article highlights the rapid development and increasing importance of computing power in China's digital economy, emphasizing its role as a fundamental resource akin to water and electricity [1][2][6]. Group 1: Current Developments in Computing Power - Huawei showcased its Ascend 384 super node with a total computing power of 300 PFLOPS at the 2025 World Artificial Intelligence Conference [1]. - By mid-2023, China had 10.85 million standard racks in use, ranking second globally in total computing power, with a storage capacity exceeding 1680 EB [2]. - The "East Data West Computing" initiative has led to significant infrastructure developments, including the expansion of data centers and the construction of over 250 optical cables [2][3]. Group 2: Industry Applications and Innovations - Computing power is transforming various industries, including automotive and healthcare, with significant improvements in efficiency and productivity [4][6]. - In the automotive sector, computing power has reduced manufacturing cycles from 60 months to 24 months, showcasing its impact on production efficiency [3]. - In healthcare, the demand for computing power is surging, with a 36% annual growth rate in health data, leading to innovative applications that enhance diagnostic capabilities [6]. Group 3: Future Directions and Challenges - The computing power industry in China is moving towards high-quality, large-scale development, facing challenges such as supply-demand mismatches and technological gaps [7]. - Predictions indicate that by 2035, artificial intelligence could contribute over 11 trillion yuan to China's GDP, significantly increasing the demand for computing power [7]. - Experts suggest that the construction of intelligent computing centers is crucial for meeting future computing power needs, emphasizing the importance of infrastructure and technology integration [7][8].

Core Insights - The demand for computing power in China is rapidly increasing, with significant advancements in infrastructure and technology being made to support this growth [1][2][5] - The Chinese government has issued policies to enhance the development of computing power, recognizing it as a critical resource in the digital economy [1][5] Infrastructure Development - As of mid-2023, China has 10.85 million standard racks in use for computing centers, ranking second globally in total computing power [2] - The "East Data West Computing" initiative has led to the construction of over 250 optical cable lines, enhancing the country's data transmission capabilities [2] - Major expansions in computing centers have been reported, such as the China Telecom Central Intelligent Computing Center in Wuhan, which added 300 intelligent computing servers, reaching over 2000 PFLOPS [1][2] Technological Innovations - The intelligent computing capacity in China has reached 788 EFLOPS, driven by advancements in artificial intelligence technologies [2] - ZTE has launched an intelligent computing supernode system that integrates self-developed AI chips, enabling flexible assembly of domestic GPUs for high-speed operations [2] - The development of a national integrated computing network is underway, with a tiered layout of hub nodes, regional centers, and local edge computing [3] Industry Applications - Computing power is transforming various industries, including automotive and healthcare, with significant improvements in operational efficiency and productivity [4][5] - In the automotive sector, companies like Geely are leveraging intelligent computing centers to reduce manufacturing cycles significantly [3][4] - The healthcare sector is experiencing explosive growth in data, with a 36% annual compound growth rate, leading to increased demand for computing power to support innovative applications [4] Future Directions - The computing power industry in China is transitioning towards high-quality, large-scale development, with a focus on optimizing resource allocation and enhancing technological collaboration [5][6] - By 2035, it is projected that artificial intelligence will contribute over 11 trillion yuan to China's GDP, driving a tenfold increase in computing power demand [5] - The government plans to implement a standardized interconnectivity framework for computing resources by 2028, aiming to improve efficiency and accessibility [6]

Investment Rating - The report maintains a "Buy" rating for the company, indicating an expected stock price increase of over 15% compared to the benchmark index [4][6]. Core Insights - The company reported a slight decline in revenue but significantly reduced its losses, with a 68.56% year-on-year decrease in net profit loss for the first half of 2025 [1]. - The company is focusing on enhancing revenue quality and operational efficiency, leading to a 4.1 percentage point increase in gross margin and a 14.04% reduction in total sales, R&D, and management expenses [1]. - The company's intelligent computing cloud business is growing, contributing 12.40% to total revenue, with a 3.00% year-on-year increase [2]. - Key industries such as finance, telecommunications, energy, and transportation have shown significant growth, with transportation growing by 60.78% year-on-year [3]. - The company aims to become a leading provider of cybersecurity and intelligent computing cloud solutions in China, with substantial growth potential in the future [4]. Financial Performance Summary - For the first half of 2025, the company achieved revenue of 826 million yuan, a 5.38% decrease year-on-year, while net profit loss was 64.7 million yuan, a 68.56% reduction in loss compared to the previous year [1]. - The company adjusted its revenue forecasts for 2025-2027 to 2.902 billion, 3.111 billion, and 3.314 billion yuan, respectively, with net profit forecasts of 161 million, 224 million, and 304 million yuan [4][7]. - The earnings per share (EPS) are projected to be 0.14, 0.19, and 0.26 yuan for 2025, 2026, and 2027, respectively, with corresponding price-to-earnings (P/E) ratios of 81.98, 58.79, and 43.40 [4][7].

Investment Rating - The report maintains a "Buy" rating for the company [4][6] Core Views - The company achieved a turnaround in Q2 with a revenue of 4.91 billion yuan, a year-on-year increase of 8.72%, and a net profit of 3.7 million yuan, marking a significant recovery [1] - The company is actively advancing its "AI + Security" strategy, with a strong demand for integrated AI computing machines and a reported business opportunity of nearly 100 million yuan [3] - The company has improved its revenue quality, with a gross margin of 67.41%, up 4.11 percentage points year-on-year, due to a focus on core business and control of low-margin projects [2] Financial Performance - In the first half of 2025, the company reported a revenue of 8.26 billion yuan, a decrease of 5.38% year-on-year, and a net loss of 65 million yuan, which is a 68.56% reduction in losses compared to the previous year [1] - The company expects revenues for 2025-2027 to be 30.46 billion, 33.18 billion, and 36.44 billion yuan respectively, with net profits projected at 1.79 billion, 2.33 billion, and 3.07 billion yuan [4] - The company’s R&D expense ratio was 36.26%, down 4.33 percentage points year-on-year, while management and sales expense ratios also saw declines [2] Market Opportunities - The company has received orders for integrated AI computing machines exceeding 8 million yuan, with over 5 million yuan in revenue already recognized [3] - The application of AI technology is expected to significantly enhance production and operational efficiency, with 61% of clients planning to purchase AI-based security products and services in the next 1-3 years [3]