Group 1 - The company has approved a guarantee for its wholly-owned subsidiary, Suzhou Unisplendour Digital Group Co., Ltd., to provide a total guarantee amount of up to RMB 1.06 billion for six subsidiaries' bank credit applications within 12 months from the shareholders' meeting held on May 20, 2025 [1][2] - Recently, Suzhou Unisplendour Digital signed a maximum guarantee contract with CITIC Bank Tianjin Branch, providing a joint liability guarantee for RMB 80 million for Unisplendour E-commerce's credit application, increasing the total guarantee balance to RMB 160 million [2][3] - The total amount of guarantees provided by the company and its subsidiaries is RMB 18.38 billion and USD 25.12 million, accounting for 270.79% of the company's equity attributable to shareholders as of the end of 2024 [4] Group 2 - The company has no overdue guarantees or guarantees involved in litigation, and there are no losses due to guarantees resulting from adverse judgments [5] - The board of directors believes that Unisplendour E-commerce has a sound internal control system and the ability to continue operations and repay debts, which helps mitigate risks associated with the guarantee [4][5] - The company has approved the proposal to provide guarantees for bank credit applications at the sixth temporary shareholders' meeting held on December 19, 2025, with 99.16% of the voting rights in favor [17]

证券日报网讯 12月19日晚间，紫光股份发布公告称，公司2025年第六次临时股东会审议通过《关于 2026年度日常业务外汇套期保值额度的议案》等多项议案。 （文章来源：证券日报） ...

Group 1 - The core stock price of Unisplendour Corporation (000938) closed at 24.43 yuan on December 19, 2025, with a market capitalization of 69.872 billion yuan, reflecting a 0.45% increase from the previous trading day [1] - The stock opened at 24.45 yuan, reached a high of 24.72 yuan, and a low of 24.43 yuan, with a trading volume of 699 million yuan and a turnover rate of 0.99% [1] - The company held its sixth extraordinary general meeting of shareholders in 2025, where it approved the foreign exchange hedging limit for routine business in 2026 [1] Group 2 - The meeting was chaired by Chairman Yu Yingtao, with 1,756 shareholders and representatives present, representing 890,173,199 shares, which accounted for 31.1241% of the total voting shares [1] - Beijing Zhongguang Law Firm provided a legal opinion confirming the legality and validity of the meeting's convening, proceedings, and voting results [1]

Core Viewpoint - The company is actively developing liquid cooling technology and aims to provide a comprehensive range of advanced solutions for AI computing and other fields [1] Group 1: Liquid Cooling Technology - The company has fully laid out its liquid cooling technology, offering various technical routes including cold plate, immersion, and spray cooling [1] - The company is committed to creating advanced and high-performance full-stack liquid cooling solutions [1] - The company has launched a series of liquid cooling cabinet products specifically designed for AI computing and other applications [1] Group 2: Research and Integration Capabilities - The company possesses self-research and integration capabilities across the entire supply chain, from cold plates and quick connectors to CDU, cabinets, and monitoring systems [1] - The company has experience and case studies in delivering liquid cooling infrastructure for large internet companies [1]

股票简称：紫光股份 股票代码：000938 公告编号：2025－075 紫光股份有限公司 关于为子公司提供担保的进展公告 本公司及董事会全体成员保证信息披露的内容真实、准确和完整，没有虚假记载、 误导性陈述或重大遗漏。 一、担保情况概述 紫光股份有限公司（以下简称"公司"）2025 年 4 月 28 日召开的第九届董事会第 十二次会议及 2025 年 5 月 20 日召开的 2024 年度股东大会审议通过了《关于为子公 司申请银行综合授信额度提供担保的议案》，同意公司全资子公司紫光数码（苏州） 集团有限公司（以下简称"苏州紫光数码"）为紫光电子商务有限公司（以下简称"紫 光电子商务"）等六家子公司自公司 2024 年度股东大会审议通过之日起 12 个月内申 请的银行综合授信额度提供总额不超过人民币 10.6 亿元的担保（以下简称"本次额 度"）。具体内容详见公司于 2025 年 4 月 29 日在巨潮资讯网等披露的《第九届董事会 第十二次会议决议的公告》（公告编号：2025-018）、《关于为子公司申请银行综合授信 额度提供担保的公告》（公告编号：2025-022）及于 2025 年 5 月 21 日披露的《 ...

北京市重光律师事务所 关于紫光股份有限公司 2025 年第六次临时股东会 法律意见书 致：紫光股份有限公司 北京市重光律师事务所（以下简称"本所"）接受公司的委托，指派李静律师和刘 伟东律师（以下简称"本所律师"）出席会议，并依据《中华人民共和国证券法》（以下 简称"《证券法》"）、《中华人民共和国公司法》（以下简称"《公司法》"）、中国证券监 督管理委员会《上市公司股东会规则》（以下简称"《股东会规则》"）和《紫光股份有 限公司章程》（以下简称"《公司章程》"）出具本法律意见书。 本所律师依法对公司 2025 年第六次临时股东会召集、召开程序是否合法及是否符 合《公司章程》、召集人资格及出席会议人员资格的合法有效性和股东会表决程序、表 决结果的合法有效性予以见证并发表法律意见，本所律师同意将本法律意见书作为公司 2025 年第六次临时股东会的必备文件公告，本所律师确认法律意见书中不存在虚假、严 重误导性陈述及重大遗漏，否则愿意承担相应的法律责任。 按照律师行业公认的业务标准、道德规范和勤勉尽责精神，本所律师对公司本次股 东会的相关文件和有关事实进行了核查和验证，现就本次股东会的如下问题发表法律意 见： 一、 ...

股票简称：紫光股份 股票代码：000938 公告编号：2025－074 紫光股份有限公司 2025 年第六次临时股东会决议公告 本公司及董事会全体成员保证信息披露的内容真实、准确和完整，没有虚假记 载、误导性陈述或重大遗漏。 特别提示： 现场会议召开日期和时间：2025 年 12 月 19 日（星期五）下午 14 时 30 分 网络投票时间：2025 年 12 月 19 日。其中，通过深圳证券交易所交易系统进行投 票的具体时间为：2025 年 12 月 19 日上午 9:15-9:25，9:30-11:30 和下午 13:00-15:00 期间的任意时间；通过互联网投票系统进行网络投票的起止时间为 2025 年 12 月 19 日 9:15-15:00 期间的任意时间。 7、 会议出席情况：出席会议的股东及股东代表 1,756 名，代表股份数 890,173,199 股，占公司有表决权股份总数的 31.1241%。其中，出席现场会议的股东及股东代表 3 名，代表股份数 801,343,818 股，占公司有表决权股份总数的 28.0182%；参加网络 投票的股东 1,753 名，代表股份数 88,829,381 ...

Core Insights - The article discusses the launch of the H3C UniPoD S80000 super node product by Unisplendour Corporation's subsidiary, H3C, aimed at addressing the challenges of communication walls and computing power utilization in large model training and inference [1][5]. Group 1: Product Features and Innovations - The H3C UniPoD S80000 super node utilizes a "computing power × connectivity" technology concept, achieving full interconnection of GPUs through a Scale-up architecture, resulting in an 8-fold increase in inter-card bandwidth compared to traditional 8-card servers and an 80% improvement in single-card inference efficiency [1][5]. - The super node supports liquid cooling for high-density deployment and is compatible with multiple brands of GPUs, addressing the long-term stability requirements for large model training through software and hardware collaborative optimization [1][5][7]. Group 2: Market Context and Demand - As the market for high-performance computing power surges, driven by the increasing prevalence of high-parameter MoE large models like DeepSeek, the ability to efficiently train and infer large models becomes critical for gaining a competitive edge in the rapidly evolving AI landscape [2][3]. - The article highlights the importance of building robust and efficient AI computing infrastructure, with super node products emerging as a key focus in the current computing power sector [2][3]. Group 3: Technical Advantages - The article emphasizes that traditional cross-node communication methods lead to significant communication overhead, reducing computing power utilization. The Scale-up technology enables direct high-speed communication between GPUs, significantly enhancing GPU utilization and reducing idle time [3][4]. - In the inference phase, the super node's support for independent scaling of computing and storage resources allows for efficient resource allocation, particularly in scenarios requiring frequent access to KV Cache, thus minimizing resource waste and ensuring low latency [4][5]. Group 4: Stability and Reliability - The H3C UniPoD S80000 is designed with a focus on stability and maintainability, crucial for preventing training interruptions that could lead to resource waste and model performance degradation. The product incorporates collaborative optimization of software and hardware to ensure uninterrupted long-term training [7][8]. - The company is actively investing in optical interconnection technology to leverage the benefits of high speed, low latency, and low energy consumption while addressing the reliability issues associated with optical components [7][9]. Group 5: Future Outlook - H3C aims to continue developing super node products that support large-scale deployments of 1024 cards and above, enhancing the scale and efficiency of intelligent computing clusters [7][8]. - The company is committed to building a strong, diverse, and continuously evolving computing infrastructure to support the AI industry's growth and transformation [8][9].

Core Insights - The company is enhancing its product and solution capabilities through an "AI in ALL" strategy, focusing on building an integrated AI empowerment platform that includes a comprehensive digital infrastructure, cloud and intelligent platforms, and an AI training and application platform [2] Group 1: AI Empowerment Strategy - The company has adopted a "1+N" large model strategy to accelerate "AI for ALL" and is deeply engaging in industry-specific scenarios [2] - The intelligent platform has been adapted to over 20 types of GPUs and more than 60 large models, optimizing computing units to improve training and inference efficiency while lowering the barriers for enterprises to apply AI [2] Group 2: Industry-Specific Solutions - The company has launched the LinSeer ICT intelligent agent, which enhances the intelligence level of equipment inspection and operation and maintenance based on extensive ICT knowledge training [2] - A series of vertical industry AI solutions, including specialized models for medical, rail transportation, and education sectors, have been released and are already being applied in various scenarios [2] Group 3: Business Model and Services - The company provides one-stop computing power services through the "Turing Town" model, achieving a commercial closed loop from "construction, ecology, operation, empowerment" to "sales, service" [2] - The company will continue to strengthen its full-stack AI capabilities in both software and hardware to empower industry clients in their digital and intelligent transformation [2]

Investment Rating - The industry investment rating is "Buy" [3] Core Insights - The report highlights the potential of HBF (High Bandwidth Flash) technology to meet the memory capacity requirements of AI large models, which have been increasing due to the growth in model parameters and context lengths [13][20] - Companies with a database technology foundation are well-positioned to develop data infrastructure software based on HBF, with major players including Huawei and Alibaba, as well as independent firms like StarRing Technology and PingCAP [29][32] - The maturity of HBF technology is expected to drive the application of related data infrastructure software, particularly in AI inference tasks [33] Summary by Sections HBF Technology Development - HBF is anticipated to become the optimal solution for meeting the memory capacity demands of AI large models, as current HBM (High Bandwidth Memory) solutions are insufficient [20][23] - The first batch of HBF samples is expected to be delivered in the second half of 2026, with integrated AI inference hardware projected for early 2027 [23] Data Infrastructure Software Opportunities - The report identifies a growing need for data processing software that can handle the high memory requirements of various AI models, particularly large language models and video generation models [28][30] - Companies with existing database technology capabilities are likely to develop software optimized for HBF, enhancing their competitive edge in the market [29][32] Key Company Valuations and Financial Analysis - StarRing Technology (688031.SH) has a current price of CNY 97.99 with a rating of "Increase" [8] - Cambricon (688256.SH) is rated "Buy" with a target value of CNY 1480.92, reflecting strong growth potential [8] - Inspur Information (000977.SZ) is rated "Increase" with a current price of CNY 62.85 [8] - Unisplendour (000938.SZ) is rated "Buy" with a target value of CNY 39.18 [8]