这些不仅是市场关注的问题，也是政策研究部门关注的问题。 2026年两会前夕，在国家信息中心的支持下，《财经》将2026年初的一次算力政策讨论与意见征集，梳理出了七个关键问题。这次国家信息中心组织的讨 论主要聚焦"算力调度"对产业发展的价值，参与者以电信运营商、科研院所、咨询机构的专家为主。 国家信息中心是国家发展改革委直属事业单位，长期进行全国一体化大数据中心、"东数西算"工程、全国一体化算力网等政策研究。多年来，国家信息中 心在不断组织大型科技公司、创业公司、咨询机构、科研院所等进行算力领域的专题讨论和意见征集。 在历经多年的长期研讨中，各方对"算力调度"的理解不断深化，并逐渐达成了一些共识： 算力调度，是充分利用现有算力，减少闲置资源，实现合理配置的重要方法。算力调度的复杂度，要远超水电的调度。水电是同质的物理资源，但算力是 高度异构和非标准化的。全国算力调度"一张网"的建设，不能只靠行政指令推进，而是通过政府搭台、标准引领、市场运作的综合模式，让全国一体化算 力网真正高效运转起来。 中国算力产业布局中，产业政策一直发挥着重要作用。其中有两大关键节点——2022年2月，"东数西算"工程启动。2023年12 ...

Core Viewpoint - The A-share market saw a significant rise in computing power concept stocks, with notable gains from companies like UCloud and others, driven by an announcement regarding price adjustments due to rising costs in the global supply chain [1] Group 1: Stock Performance - UCloud (688158) experienced a rise of 14.09%, with a total market capitalization of 19.5 billion and a year-to-date increase of 52.47% [2] - Dazhi Technology (600589) increased by 10%, with a market cap of 20.6 billion and a year-to-date rise of 93.58% [2] - Jinkai New Energy (600821) saw a gain of 9.97%, with a market cap of 13 billion and a year-to-date increase of 27.65% [2] - Dongfang Guoxin (300166) rose by 7.48%, with a market cap of 17 billion and a year-to-date increase of 46.18% [2] - Taicheng Light (300570) increased by 7.06%, with a market cap of 32.4 billion and a year-to-date rise of 23.63% [2] - Other notable performers include Xiechuang Data (300857) up 6.86%, Shoudu Online (300846) up 6.22%, and Tianfu Communication (300394) up 5.96% [2] Group 2: Price Adjustment Announcement - UCloud announced a price increase for all products and services starting March 1, 2026, due to significant and structural increases in core hardware procurement and infrastructure costs [1]

A股开盘，三大指数小幅高开，沪指高开0.12%报4136.99点，深证成指高开0.12%，创业板指高开 0.3%。盘面上，算力概念股高开，大位科技、特发信息涨停；影视院线板块继续调整。 ...

Group 1 - The AI and computing industry continues to receive significant catalysts, with major tech stocks in the US experiencing a collective surge, including Nvidia up 7.87% and TSMC and ASML up nearly 5% [1] - Amazon's projected capital expenditure for 2026 is expected to reach $200 billion, significantly exceeding market expectations of $146.11 billion, representing a 50% increase from 2025 [1] - The overall financial reports from tech stocks indicate a high level of confidence in AI, with major companies significantly raising their capital expenditure guidance for 2026 [1] Group 2 - The Ministry of Industry and Information Technology in China has issued a notice to develop a national computing power interconnection node construction plan, aimed at creating a standardized and efficient resource utilization system [2] - This initiative is significant for the domestic computing power industry, emphasizing resource integration and utilization during a period of rapid expansion [2] Group 3 - Key companies in the optical module sector include LianTe Technology and HuiLv Ecology [3] - Companies involved in optical interconnection include ZhiShang Technology and JuGuang Technology [3] Group 4 - Domestic computing power companies include ZhongKe Shuguang, Dongyangguang, and XieChuang Data [4] Group 5 - AI application companies include iFlytek, Decai Co., and Kunlun Wanwei [5]

首先从流通性来说，很多智算中心之间确实存在技术壁垒，尤其对于不同的用户场景很难平替使用，即便付出巨大的迁移成本完成 异构部署，往往也无法发挥出绝对性能。 scaleX的AI计算开放架构很好的解决了这一点，由于集群底层架构本身就支持不同类别的AI加速卡，而且兼容主流的AI计算软件生 态，所以对绝大多数开放者用户来说，基本不存在太多被动调优环节，很多情况下都能拿来就用。 这两天，国产算力行业热度爆棚。 据说国家超算互联网核心节点已经进入上线试运行阶段，底层重磅部署了3套scaleX万卡超集群，成功打造出全国首个实现3万卡部 署、且实际投入运营的最大国产AI算力池。 业内媒体认为，这标志着国产万卡超集群技术正完成从"示范验证"到"规模化应用"的关键跨越。尤其在AI大模型参数爆发期，大规模 AI计算场景的算力需求仍极为迫切，率先"走进现实"的万卡超集群，有望迎来巨大市场利好。 前段时间，业内还在探讨算力闲置的问题，很多算力资源确实流通不出来，更用不起来。但这显然不代表国内用户都不缺算力了。 事实上，闲置下来的大多都是同质化困于异构壁垒的算力，而真正具备流通价值，且满足高并发场景需求的算力仍然极为抢手。 scaleX万 ...

Core Insights - Shanghai's computing power scale has surpassed 120,000 PFLOPS, with nearly 20,000 PFLOPS managed by the municipal computing power scheduling platform, indicating significant resource aggregation effects in the city [1] - The city is actively implementing the "Artificial Intelligence +" initiative to lead the digital transformation of the manufacturing industry, with a focus on enhancing computing power applications in key sectors [1] - The Shanghai Computing Power Industry High-Quality Development Initiative has been launched to promote the construction of next-generation computing centers and a high-level cloud service system [2] Group 1 - Shanghai is advancing the construction of computing infrastructure and accelerating the large-scale deployment of domestically produced chips [1][2] - The city aims to create a self-controllable computing industry and establish a "national chip, model, and cloud" system to enhance the Shanghai brand in computing power [1] - The computing power industry is transitioning from "facility scale" to "comprehensive efficiency," with efforts to upgrade infrastructure and develop a domestically controllable ecosystem [2] Group 2 - The "Computing Power Economy Empowerment Co-Creation System" has been launched to address the challenge of converting computing resources into economic value [2] - A "Scientific Intelligent Open Community Computing Support Plan" has been signed, focusing on building computing infrastructure and cloud services for global scientists [2]

Core Viewpoint - The report from CITIC Securities anticipates continued growth in the computing power sector driven by sustained capital expenditures (Capex) from major CSPs, increasing demand for tokens, and continuous product capability enhancements [1] Group 1: Performance Outlook - The computing power sector's performance is expected to maintain a favorable growth trajectory through 2025 [1] - AI applications are projected to show structural highlights, particularly in areas such as AI fintech, AI healthcare, and AI data, with general agent applications also demonstrating resilient growth [1] - Overall profitability in the sector is likely to improve significantly [1] Group 2: Future Projections - Looking ahead to 2026, the computing power sector is expected to continue its favorable conditions, with AI applications poised to reach pivotal opportunities [1]

Core Insights - The article highlights the successful implementation of a mixed-rate network trial featuring 400G and 800G speeds in key computing power regions of China, particularly in the Guizhou-Guangzhou corridor, enhancing efficient interconnection and collaborative development of computing resources [1] - The "East Data West Computing" initiative has positioned Guizhou as a national hub for integrated computing networks, transitioning from a "data warehouse" to a "computing factory" [1] - The establishment of the "Xirang" computing scheduling platform in Guizhou aims to improve the efficiency of computing power allocation across various sectors [3] Group 1 - Guizhou has successfully completed a trial for a mixed-rate network, enhancing computing resource interconnectivity [1] - The "East Data West Computing" project has accelerated Guizhou's transformation into a computing power hub, with a focus on building a green data center that integrates general, super, and intelligent computing [1] - The computing capacity in Guizhou has increased from an initial 2P to over 10,000P, significantly expanding its computational capabilities [1] Group 2 - The network infrastructure has been upgraded to allow direct connections to 24 cities, reducing data transmission time by 33% and costs by over 30% for businesses in the Guangdong-Hong Kong-Macau Greater Bay Area [2] - The "Xirang" platform facilitates nationwide computing power scheduling, making it as accessible as utilities like water and electricity [3] - Collaborations with local meteorological agencies have improved weather prediction accuracy, enhancing disaster prevention efforts [3] Group 3 - Guizhou's digital economy strategy has led to a projected revenue of over 100 billion yuan in software and information technology services by 2024, with continuous growth rates among the highest in the country [4] - The region has established a comprehensive digital economy ecosystem, attracting multiple enterprises and creating a full industry chain from cloud services to data applications [4] - Guizhou currently operates 50 data centers with a computing capacity of 150 billion billion calculations per second, making it one of the strongest regions in terms of intelligent computing resources [4]

Core Insights - CIMG Inc. has executed initial commercial contracts related to computing power, with a total contract value of approximately $124 million in Q4 2025, indicating a strategic shift towards technology-driven growth [1][2][3] Group 1: Strategic Deployment - The company has elevated its computing power business to a core strategy, recognizing its importance in the digital economy and its role in driving AI and big data analytics [2] - Mr. Wenlong Tong has been appointed as President to lead the expansion of the computing power business, aiming to transform technological capabilities into sustainable market competitiveness [2][3] Group 2: Initial Contracts and Commercial Progress - CIMG's subsidiary, Xinmiao Shidai, was shortlisted for a server procurement tender by Guangzhou Bank with a benchmark bid of approximately $15.95 million, although this does not guarantee a contract [4] - A sales contract valued at approximately $106.5 million was signed by Zhongyan Shangyue Technology Co., Ltd., subject to inspection and acceptance [5] - Two separate sales contracts for computing power products were signed, totaling approximately $1.78 million, covering high-performance computing servers and GPUs [6] Group 3: Synergy with Core Business - The computing power deployment is synergistic with the company's traditional digital health business, exemplified by the development of the "LifeNode" product in collaboration with FLock Technology Holdings [7] - The CEO emphasized the goal of establishing computing power as a foundational technology connecting various business segments, aiming to create an innovative commercial ecosystem [8][9]

Core Insights - The National Bureau of Statistics has released implementation opinions on strengthening data technology innovation, emphasizing the construction of a unified national computing power network to support data technology research and verification [1] Group 1: Industry Trends - AI computing power is becoming a major growth engine in the telecommunications industry, with investment opportunities in the computing power supply chain, particularly in optical chips, OCS, hollow-core fibers, liquid cooling, and power supply [1] - The computing power sector has outperformed the market since 2025, with significant growth in both models and applications, indicating potential for a long-term bullish market similar to the US stock market in 2023 [1] Group 2: Domestic Developments - Domestic computing power is gaining momentum, with performance elasticity and investment certainty, driven by advancements in semiconductor equipment and AI chips under overseas restrictions, making domestic substitution a prevailing trend [1] - The collaboration between AI large models and domestic chip manufacturers is forming a complete technological ecosystem, with continuous capacity expansion expected to accelerate the integration into the domestic computing power infrastructure supply chain [2]