工业和信息化部日前印发《算力互联互通行动计划》，旨在实现全国公共算力标准化互联，使算力像 水、电一样便捷使用，形成可查询、可对话、可调用的服务能力，进而构建数字经济发展新底座，有力 支撑制造强国、网络强国和数字中国建设。 实现算力互联互通，首要的是筑牢技术基础。一方面，加强基础软硬件技术攻关，包括减少算力互联时 产生的损失或错误，提高多样化算力感知能力，优化各类算力资源管理、协同调用等，从底层技术上为 算力互联互通奠定基础；另一方面，推动国家级互联网骨干直联点和新型互联网交换中心改造升级，鼓 励基础电信运营商扩大带宽容量，有效降低数据传输网络时延，提升网络通行能力。 当务之急，是为算力流通打造一套统一规范的"交通规则"。建立标准体系，是开启算力互联互通大门的 钥匙。比如，在标准中明确通信网络互通、算力资源互联、业务调度互通、数据传输流动等关键环节要 求，才能保障数据在算力网络中安全、快速流动，助力各类应用更快找到并用好不同类型的算力资源， 发挥最大效能。再如，业界关注的算力互联市场规则，考虑各方利益，形成合理的价格体系，并为算力 交易营造公平、公正、透明的市场环境，促进算力市场长远健康发展。 一个重要前提是让 ...

Group 1 - The computing power industry chain is experiencing a collective surge, with significant gains in sectors such as optical packaging, components, power IoT, MLOps, and AI PCs, highlighted by Shenghong Technology reaching a historical high [1] - The Chinese government is deploying 10 national data element comprehensive experimental zones in regions like Beijing, Zhejiang, and Anhui to enhance the integration of the real economy and digital economy [1] - The National Energy Administration has initiated pilot projects for a new power system, focusing on the collaboration between computing power and electricity, with an emphasis on regions with favorable energy resources [1] Group 2 - The Ministry of Industry and Information Technology is promoting the development of the artificial intelligence industry and its integration into new industrialization, emphasizing the need for a supportive ecosystem and innovation stimulation [2] - The ministry aims to strengthen computing power supply and enhance the deployment of general and industry-specific large models, accelerating the intelligent upgrade of manufacturing processes [2] Group 3 - The "Computing Power Interconnection Action Plan" aims to establish a comprehensive standard and rule system for computing power interconnectivity by 2026, with significant growth projected in China's intelligent computing power scale [3] - By 2025, China's intelligent computing power is expected to reach 1037.3 EFLOPS, with a 43% year-on-year growth, and is projected to double by 2026 [3] Group 4 - The construction of computing power networks is crucial for enhancing the utilization of domestic computing power cards and ensuring the sustainability of computing power leasing services for government and enterprises [4] - Policies at both central and local levels are setting clear goals for domestic AI computing power construction, with an increase in project bidding for intelligent computing centers since 2024 [4]

算力产业链走强 创业板指涨逾1% 智通财经6月5日电，午后指数走高，创业板指拉升涨逾1%，沪指涨0.09%，深成指涨0.45%。兵装集 团、体育、算力、虚拟电厂等方向涨幅居前，沪深京三市上涨个股近2400只。 ...

6月4日，从宁夏回族自治区政府新闻办举行的新闻发布会上获悉，宁夏数字经济呈现总量持续扩大、质 效持续提升、数实融合持续深化的良好发展态势，在算力基础设施建设、产业生态培育、数据要素开发 应用、数智赋能转型及跨域协同合作五个方面实现"新突破"。 算力基础设施建设实现"新突破"。截至目前，宁夏已建成大型、超大型数据中心9个，在建项目27个， 标准机架数达14.2万架，智算算力规模7.4万P，中卫数据中心集群智算规模位居全国第三。"总体看，项 目落地实现新突破，建设进度跑出加速度，形成多点发力、梯次推进的跨越式发展格局，截至2024年 底，智算项目投资、标准机架、智算算卡、算力规模同比实现'四个翻番'。"自治区发展改革委副主 任、自治区数据局局长麦欣甫表示，预计今年底，以上指标将再次实现同比翻番。宁夏算力综合指数进 入全国前十，算力质效分指数排名西部第一。 算力产业生态培育成效显著。银川、石嘴山等地集聚电子制造项目发展初具规模，为4000余家企业提供 云计算服务。20多个大模型在宁夏落地训练，凯添燃气AI、农业农村经济运行智能AI等人工智能行业 垂域模型陆续投入应用。6所学校入选全国中小学人工智能教育基地。 数据要 ...

新消费概念的外延不断扩大。 昨天，新品（茶啤）、新渠道（山姆超市）带火了啤酒、休闲食品板块。今天上午，"苏超"概念大涨。 有观点认为，"苏超"的火爆并不局限于足球比赛，而是通过足球折射了江苏省13个设区市的地域特色和 地域文化，让观众在观赛时带有一定的情绪归属感，本质上是情绪消费的体现。因此"苏超"概念也属于 新消费范畴。 不过，今天上午，新消费行情回调，美容护理、宠物经济等板块下跌；港股新消费"三姐妹"泡泡玛特、 蜜雪集团、老铺黄金下跌。 算力产业链上午持续拉升 ，光模块龙头股中际旭创（300308）上涨5.14%，成交额为41.03亿元，居A股 第一；PCB龙头股胜宏科技（300476）上涨9.23%，股价创历史新高，成交额为31.61亿元，居A股第二 位。此外，博创科技、仕佳光子、太辰光等个股也大涨。 分析人士表示，近两日算力产业链连续走强。一方面，市场对AI算力资本开支放缓的担忧缓解，海外 AI龙头股持续走强；另一方面，经历前期的连续调整后，当前A股TMT板块拥挤度回落至低位。 上午收盘，上证指数上涨0.08%，深证成指上涨0.16%，创业板指上涨0.32%。 | 足球概念大涨 | 今天上午，足球概念 ...

Core Insights - Ningxia aims to quadruple investment, standard racks, intelligent computing cards, and computing power scale by the end of 2024, with a comprehensive computing power index entering the top 10 nationally [1] - The region is recognized for its favorable climate for large data centers and is the first national new energy comprehensive demonstration zone, providing abundant green electricity resources [1] - By 2027, Ningxia's digital economy is projected to reach 270 billion RMB, accounting for over 40% of the total economy, with core digital economy industries expected to double their value added compared to 2021 [2] Investment and Infrastructure - As of now, Ningxia has established 9 large and super-large data centers, with 27 more under construction, and a total of 142,000 standard racks [1] - The region has built 116,000 intelligent computing cards and achieved a computing power scale of 74,000 P, ranking third among the top ten data center clusters in the country [1] Industry Ecosystem Development - Ningxia is fostering a new industrial ecosystem by attracting projects in electronic materials, photovoltaic equipment, high-performance servers, and smart wearable manufacturing [2] - Over 20 large models are being trained in Ningxia, with applications in AI healthcare and over 40 typical application scenarios developed [2] - The region has established more than 1,200 smart classrooms and laboratories, 186 smart factories, and 2,600 small and medium-sized enterprises utilizing cloud computing and data intelligence [2]

记者4日从工业和信息化部获悉，工业和信息化部近日印发《算力互联互通行动计划》（以下简称《计 划》），旨在加快构建算力互联互通体系，实现不同主体、不同架构的公共算力资源标准化互联，提高 公共算力资源使用效率和服务水平，促进算力高质量发展。 同时，《计划》将推动算力互联互通赋能典型应用场景：推动算力互联在算力资源服务、任务调度、市 场交易、开源社区运营等新业态场景应用；推动算力互联在人工智能、科学计算、智能制造、远程医 疗、视联网等企业级场景，以及智能驾驶、云渲染、云电脑、云游戏等消费级场景应用；推动算力互联 与能源互联网、工业互联网、移动互联网等融合创新应用。 当前人工智能产业快速发展，全球算力资源使用需求不断攀升，算力互联互通已成为国内外关注焦点。 工业和信息化部信息通信管理局有关负责人介绍，算力互联互通是在现有互联网体系架构基础上，通过 构建统一算力标识、增强异构计算和弹性网络能力等方式，将不同主体、不同架构的公共算力资源标准 化互联，形成可查询、可对话、可调用的服务能力，实现数据和应用在算力间高效供需匹配、流动互 通、迁移计算，使算力像水、电一样便捷使用。 （文章来源：科技日报） 《计划》提出，到2026 ...

Core Insights - The Ministry of Industry and Information Technology has issued the "Computing Power Interconnection Action Plan" aiming to establish a comprehensive system of standards, identifiers, and rules for computing power interconnection by 2026 [1] Group 1: Infrastructure Interconnection - The plan promotes new high-performance transmission protocols to enhance the interconnection level among computing power nodes [1] - It aims to build national, regional, and industry computing power interconnection platforms to unify and aggregate public computing power identifiers, facilitating interconnection of resources among leading computing power enterprises nationwide [1] Group 2: Resource Utilization - The initiative focuses on achieving interconnection of computing power resources, enabling cross-entity, cross-architecture, and cross-regional scheduling of computing power supply and demand [1] Group 3: Application Scenarios - The plan includes the establishment of pilot projects for computing power internet experimental networks to empower industries with accessible computing power [1] - By 2028, the goal is to achieve standardized interconnection of public computing power across the country, gradually forming a computing power internet capable of intelligent perception, real-time discovery, and on-demand access [1]

Core Viewpoint - The merger between Haiguang Information and Zhongke Shuguang aims to enhance the integration of resources in China's AI industry, facilitating the development of comprehensive AI solutions and addressing industry gaps [1][2]. Group 1: Merger and Integration - The merger is seen as a beneficial attempt to "fill gaps and strengthen capabilities" in China's computing industry, leveraging technology complementarity and market resource reuse to achieve scale effects [2]. - The merger will allow both companies to pool their core strengths in high-end computing, storage, and cloud computing, enhancing their competitive edge in the AI sector [2][3]. - The merger is a strategic response to global industry trends and national strategic needs, promoting collaborative development and cost efficiency [2][3]. Group 2: Financial Performance - Haiguang Information reported a net profit of 1.931 billion yuan last year, a year-on-year increase of approximately 50%, with a 75% growth in the first quarter of this year [5]. - Zhongke Shuguang achieved a net profit of 1.911 billion yuan last year, with a 30% increase in the first quarter of this year [5]. Group 3: Strategic Focus and R&D - The companies plan to focus on high-end chip and solution development to enhance customer satisfaction and promote the large-scale application of domestic chips in key industries such as government, finance, and energy [3][6]. - Zhongke Shuguang aims to increase R&D spending and strengthen self-research capabilities in response to external challenges posed by U.S. sanctions, while also expanding its overseas market presence [6]. Group 4: Industry Positioning - Zhongke Shuguang has established a comprehensive industry chain layout covering chips, servers, cloud computing, and data services, positioning itself as a competitive player in the domestic computing industry [4]. - The company has successfully integrated liquid cooling technology into its IT equipment, achieving significant energy efficiency improvements in data center operations [7].

Group 1 - The core viewpoint of the news is the strategic merger between Zhongke Shuguang and Haiguang Information, which aims to enhance the development of the information industry and significantly impact the industry landscape [1][2] - The merger will involve Haiguang Information issuing A-shares to absorb and merge Zhongke Shuguang, along with raising supporting funds through the issuance of A-shares [1] - Both companies will suspend trading of their stocks starting May 26, with the suspension expected to last no more than 10 trading days [1] Group 2 - Zhongke Shuguang holds a 27.96% stake in Haiguang Information, and the merger will optimize the industrial layout from chips to software and systems, enhancing the supply chain [1] - The merger aligns with the global trend of extending industrial chains and aims to promote the large-scale application of domestic chips in key sectors such as government, finance, communication, and energy [1][2] - The combined entity will enhance its technical strength and market competitiveness, leveraging scale effects to increase profits and improve the quality and efficiency of the listed companies [2]