聚焦"AI计算开放架构" 光合组织2025人工智能创新大会即将举办 编辑：王永乐 广告等商务合作，请点击这里 光合组织秘书长任京暘向媒体介绍光合组织2025人工智能创新大会相关情况。中新网记者 孙自法 摄 光合组织秘书长任京暘介绍说，光合组织2025人工智能创新大会将首次大规模集中展示国产人工智能 (AI)加速计算全栈技术，发布AI计算开放架构联合实验室行动计划，基于该架构打造的全球领先大规模 智算超集群系统也将亮相。该系统采用中国自主研发的高速互联网络技术，在算力密度与系统性能上实 现双重突破，验证了开放技术路线的可行性与竞争力。 光合组织2025人工智能创新大会将设立六大分论坛，邀请百余位专家分享前沿案例，促进产业全链协同 创新，为中国AI计算的可持续发展提供开放平台与实践样板。 据悉，光合组织自成立以来，始终以构建自主可控、开放协同的国产计算生态为使命，旨在汇聚产业全 链、共破发展瓶颈，目前已有超6000家合作伙伴，建成28个生态适配中心。(完) 来源：中国新闻网 中新网北京12月5日电 (记者 孙自法)中国国内首个以"AI计算开放架构"为核心的行业盛会——光合组织 2025人工智能创新大会(HAIC2 ...

Core Insights - The World Computing Conference held in Changsha showcased the ScaleX640 super node by Inspur, which is the world's first single-rack 640-card super node, achieving a 20-fold increase in computing density and supporting million-card-level cluster expansion, thus injecting strong domestic computing power into the development of new productivity [1][3]. Group 1: Product Features - The ScaleX640 super node consists of a single immersion liquid cooling heat exchange module and two immersion liquid cooling dedicated computer cabinets, representing a significant breakthrough as it surpasses traditional solutions that would require multiple cabinets to achieve similar computing power [3]. - The device utilizes immersion phase change liquid cooling technology, allowing it to maintain a stable operating temperature below 35 degrees Celsius, which is lower than typical gaming PCs, and achieves a Power Usage Effectiveness (PUE) of 1.04, meaning 96% of electricity is used for computing, resulting in over 30% energy savings compared to traditional air cooling solutions [4]. Group 2: Industry Impact - The ScaleX640 super node adheres to the "AI computing open architecture" concept, initiated by Inspur in collaboration with over 20 industry partners, aiming to build an open and collaborative innovation system to overcome challenges in domestic AI development, such as insufficient software-hardware synergy and fragmented ecosystems [5]. - The super node supports multi-brand acceleration cards and is compatible with mainstream computing ecosystems, enabling users to flexibly select domestic computing power for rapid deployment of trillion-parameter model training and high-throughput inference, effectively addressing the practical pain points of application adaptation and software-hardware disconnection in domestic intelligent computing [5].

Core Insights - The 2025 National College Student Computer System Capability Competition - Intelligent Computing Innovation Design Competition (Pilot Cup) successfully concluded in Hefei, emphasizing the importance of AI computing open architecture in the integration of industry, academia, and research [1][3] Group 1: Competition Overview - The Pilot Cup is the only intelligent computing track in the national competition, featuring a high prize pool and employment referral opportunities for winners [3] - This year, the competition introduced a "teaching and training competition" model, enhancing the relevance of competition topics, coverage of scenarios, and integration of education and industry [3] - The competition attracted nearly 10,000 students from over 1,200 universities, with 58 teams awarded [3] Group 2: Industry Impact - The competition addressed engineering challenges in AI implementation with three key topics: "MoE language model end-to-end efficiency optimization," "ONNX Runtime operator performance optimization," and "GMRES algorithm optimization" [3] - The Senior Vice President of Sugon, Li Bin, highlighted the launch of China's first AI computing open architecture, indicating a new phase for the Chinese AI industry and a growing demand for interdisciplinary AI talent [3] Group 3: Educational Initiatives - Sugon's computing platform is user-friendly for students, comparable to CUDA, and provides extensive free learning resources [4] - The Chief Scientist of Intelligent Computing at Sugon noted an explosive growth in AI talent demand, particularly in the era of large models, and emphasized the competition's role in expanding the audience for AI talent [4] - Sugon has consistently focused on talent education, launching various initiatives such as the Pilot Cup, developer communities, and joint laboratories to explore new models for AI talent cultivation [4]

Core Insights - The European Centre for Medium-Range Weather Forecasts (ECMWF) has announced the opening of its real-time data catalog to global users, providing access to high-resolution weather forecast data [2] - Zhongke Tianji, a subsidiary of Zhongke Shuguang, has launched a "High-Resolution Meteorological Data Sharing Plan," offering global and regional weather data with higher resolution and frequency [2][3] - The trend of data sharing and openness in the meteorological service industry is gaining momentum, driven by technological advancements and policy support [3][4] Group 1: Data Sharing Initiatives - ECMWF provides daily forecasts covering 1-15 days with a maximum resolution of 0.1° grid point data, including various meteorological variables [2] - Zhongke Tianji offers over 160 meteorological elements with a resolution of 12 kilometers globally and 3 kilometers for China, outputting data hourly for 15 days [2][3] - The simultaneous data sharing initiatives by ECMWF and Zhongke Tianji highlight a significant shift in the meteorological data service industry [2] Group 2: Policy and Technological Support - The acceleration of data sharing is supported by systematic restructuring from top-level design, with various provinces in China implementing market-oriented data reform plans [3] - The Chinese government is the largest contributor to data sharing, with platforms providing millions of data downloads and fostering numerous AI application scenarios [3] - The World Meteorological Organization's initiatives have increased the demand for forecast products, further promoting data openness [4] Group 3: Competitive Advantages of Zhongke Tianji - Zhongke Tianji's data offers higher resolution for China at 3 kilometers compared to ECMWF's 12 kilometers, catering to specific industry needs [4][5] - The data output frequency from Zhongke Tianji is more granular, providing hourly updates, while ECMWF uses tiered output intervals [5] - Access to Zhongke Tianji's data is free for domestic users, contrasting with ECMWF's cost structure, which can reach approximately 1.9 million for comprehensive data access [5] Group 4: Technological Innovations - Zhongke Tianji's data sharing is supported by advanced technologies, including deep integration of dynamics and physics, enhancing the quality of meteorological applications [6] - The company aims to address challenges in high-resolution simulations and timeliness through seven foundational technologies [6] - The establishment of an AI open computing architecture by Zhongke Shuguang represents a significant advancement in the AI industry, promoting collaborative innovation [7][8]

Core Viewpoint - The article emphasizes the need for a unified and collaborative computing ecosystem in China's computing industry, particularly in the context of AI development, highlighting the gap between national strategic goals and the current fragmented industry reality [1][2][12][13]. Group 1: Industry Challenges - The Chinese computing industry faces a fundamental dilemma: whether to pursue a vertically integrated model like Apple or an open collaborative model like Android [5][6]. - Currently, the industry is characterized by strong competition and weak cooperation, leading to resource wastage and inefficiencies due to incompatible technologies and private protocols [11][12]. - The lack of unified technical standards has resulted in significant challenges for users, who often face compatibility issues when integrating products from different manufacturers [34]. Group 2: National Strategy - National strategic documents, such as the "High-Quality Development Action Plan for Computing Infrastructure," emphasize the need for optimizing computing resource allocation and promoting standardization [9][10]. - The goal is to build a strong, unified national computing market, which is now a core focus of industry development rather than just a vision of individual companies [9]. Group 3: Company Initiatives - Inspiringly, the leading company in the field, Zhongke Shuguang, has taken the initiative to establish an "AI Computing Open Architecture" in collaboration with over twenty industry partners [5][6]. - This open architecture aims to facilitate systemic innovation across various segments of the computing ecosystem, moving away from isolated advancements to a more integrated approach [37][40]. - Zhongke Shuguang's commitment to open standards and data sharing is seen as a significant step towards addressing industry bottlenecks and enhancing collaboration [35][54]. Group 4: Technological Advancements - Zhongke Shuguang has invested in liquid cooling technology since 2011, achieving significant energy savings and increasing power density in data centers [18][19]. - The company’s comprehensive business model spans high-end computing, storage, and cloud computing, demonstrating its capability to understand and address the entire computing ecosystem [21][22]. Group 5: Financial Performance - In the first half of 2025, Zhongke Shuguang reported a revenue of 5.85 billion yuan, a 2.41% increase year-on-year, with a notable 29.39% growth in net profit [26][27]. - This financial stability allows the company to focus on long-term strategic initiatives that benefit the entire industry ecosystem rather than just short-term profits [28]. Group 6: Data Sharing Initiatives - Zhongke Shuguang has launched a high-resolution meteorological data sharing plan, providing valuable data resources to various industries, which can enhance operational efficiency and innovation [42][44]. - The shared data includes over 160 meteorological elements, significantly lowering the barriers to access for industries that previously faced high costs and complexity in obtaining such data [44][49]. Group 7: Vision for the Future - The ultimate goal of Zhongke Shuguang's initiatives is to create an open and inclusive computing ecosystem that supports the digital transformation of various sectors [53][54]. - By establishing a closed-loop ecosystem from foundational computing to application and data, the company aims to drive continuous optimization and innovation within the industry [52].

Group 1 - Zhongke Shuguang's subsidiary, Zhongke Tianji, has launched a high-resolution meteorological data sharing initiative, marking a significant milestone in China's AI industry for open sharing and industrial integration [1][2] - The shared meteorological data includes global 12 km and regional 3 km resolution data, covering over 160 meteorological elements such as temperature, wind field, and precipitation, with a time scale of up to 15 days and hourly updates [2][3] - The initiative addresses the data bottleneck, a key factor in AI development, as China has abundant data resources but lacks effective and high-quality domain-specific data [3][5] Group 2 - The AI computing open architecture, announced by Zhongke Shuguang in collaboration with over 20 industry partners, aims to foster collaborative innovation across the AI ecosystem, moving away from isolated technological advancements [4][5] - China's AI industry is projected to exceed 900 billion yuan in 2024, with a year-on-year growth of 24%, and over 5,300 companies operating in the sector [5] - The establishment of standards and frameworks for collaborative innovation, along with the development of legal regulations and ethical guidelines, positions China's AI industry for high-quality growth and a more significant role on the global stage [5][6]

Core Viewpoint - The Chinese computing industry faces a fundamental choice between a vertically integrated model, akin to Apple's approach, and an open collaborative model, similar to Android's, as it prepares for the AI-driven transformation by 2025 [1][2][5] Industry Context - The Chinese computing industry has made significant progress, but it remains in a state of "strong competition, weak cooperation," with many companies using proprietary interfaces and protocols, leading to resource waste and inefficiencies [4][5] - National strategic documents emphasize the need for a unified and strong national computing market, highlighting the importance of optimizing computing resource layout and promoting standard recognition [3][5] Company Actions - Inspiring collective action, Inspiring company, Zhongke Shuguang, has initiated the first "AI Computing Open Architecture" in collaboration with over twenty industry partners, marking a shift towards an open collaborative approach [1][2][15] - The open architecture aims to address industry bottlenecks by providing a system for collaborative innovation across various computing components, thus reducing the need for redundant R&D efforts among different manufacturers [14][15] Technical Innovations - Zhongke Shuguang has invested in liquid cooling technology since 2011, achieving over 30% energy savings compared to traditional air cooling, and significantly increasing power density in data centers [6][10] - The company’s comprehensive business model covers high-end computing, storage, security, data centers, and cloud computing, demonstrating its capability to understand and integrate the entire computing ecosystem [7][8] Economic Performance - In the first half of 2025, Zhongke Shuguang reported revenues of 5.85 billion yuan, a 2.41% increase year-on-year, with net profit rising by 29.39% to 729 million yuan, indicating a shift from hardware expansion to value enhancement through technology and solutions [10] Data Sharing Initiatives - Following the launch of the open architecture, Zhongke Shuguang announced the sharing of high-resolution meteorological data, which aims to support various industries by providing critical data resources that were previously costly and difficult to obtain [17][18] - This initiative has already shown practical applications in sectors such as energy, agriculture, aviation, and insurance, demonstrating the transformative potential of shared data in enhancing operational efficiency and decision-making [18][20] Strategic Vision - The actions taken by Zhongke Shuguang reflect a broader vision of creating an open and inclusive computing ecosystem in China, positioning itself not just as a technology leader but as a builder of the entire industry ecosystem [21][22]

Group 1: Software ETF Performance - The Software ETF has a turnover rate of 6.78% during trading, with a transaction volume of 327 million yuan. Over the past month, the average daily transaction volume reached 567 million yuan, ranking first among comparable funds [3] - The Software ETF has seen a significant growth in scale, increasing by 1.803 billion yuan over the past three months, also ranking first among comparable funds [3] - As of September 10, 2025, the net value of the Software ETF has increased by 17.05% over the past three years. The highest monthly return since inception was 39.35%, with the longest consecutive monthly gains being three months and a maximum increase of 69.40%, averaging a monthly return of 10.06% [3] Group 2: Industry Insights - According to CITIC Securities, the computer industry is experiencing a noticeable acceleration in revenue and significant improvement in net profit in the first half of 2025. The computing power sector is driven by increased CAPEX, technological upgrades, and supply improvements, leading to high revenue growth [4] - The growth of the "Xinchuang" sector continues to be robust and is expected to accelerate in the second half of the year, with potential expansion into industrial software [4] - The contribution of AI applications to overall performance is significantly increasing, with notable improvements in profitability due to cost reduction and efficiency enhancement, indicating an upcoming performance inflection point [4] Group 3: Key Stocks in Software Sector - As of August 29, 2025, the top ten weighted stocks in the CSI Software Service Index include iFlytek, Tonghuashun, Kingsoft Office, and others, collectively accounting for 62.05% of the index [4] - The individual stock performances include iFlytek with a 2.35% increase and a weight of 13.93%, Kingsoft Office with a 3.81% increase and a weight of 8.15%, and Tonghuashun with a 4.65% increase and a weight of 7.39% [6] - Investors can also access AI software investment opportunities through the Software ETF linked fund (012620) [6]

Group 1 - Xinba, the founder of XinXuan Group, was reported to be taken away for investigation, but the company denied the claims [2] - The film "Nanjing Photo Studio" was released in the UK, providing an opportunity for audiences to understand the history of the Nanjing Massacre [3] - The first AI computing open architecture in China was launched at the World Intelligent Industry Expo, supporting significant AI computing capabilities [4] Group 2 - Yi Huiman, during his tenure as the chairman of the China Securities Regulatory Commission, saw the A-share market breach the 3000-point mark 20 times [5][6] - OpenAI is expected to generate nearly $10 billion in revenue through ChatGPT this year, with total revenue projected to reach $13 billion [10] - A Brazilian billionaire has named football star Neymar as the sole heir to his fortune, estimated to exceed $1 billion [12]

Core Viewpoint - The launch of China's first AI computing open architecture and AI supercluster system by Zhongke Shuguang and over 20 industry chain enterprises aims to enhance collaborative innovation in AI computing, addressing computing power bottlenecks and promoting accessible computing power [1] Group 1: AI Computing Open Architecture - The AI computing open architecture is designed for large-scale intelligent computing scenarios, focusing on efficient tightly-coupled system design centered around GPUs [1] - The initiative aims to connect AI industry chain enterprises, transitioning from single-point breakthroughs in computing, storage, networking, power, cooling, management, and software to cluster innovation [1] - Zhongke Shuguang has built over 20 large-scale computing clusters in the past decade, deploying more than 500,000 heterogeneous accelerator cards [1] Group 2: AI Supercluster System - The newly launched Shuguang AI supercluster system features four main characteristics: ultra-high performance, ultra-high efficiency, ultra-high reliability, and comprehensive openness [1] - This system is designed to provide a robust computing foundation for trillion-parameter large model training and inference, industry large model fine-tuning, high-throughput inference, and multi-modal large model development [1]