Workflow
通算
icon
Search documents
国家数据基础设施建设取得积极进展 一定范围内实现数据“一点发布、全域可见”
Ren Min Ri Bao· 2025-09-06 22:07
国家数据局组织北京、上海等18个城市,开展可信数据空间、数场等6条技术路线先行先试。为推动各 地数据基础设施协同联动,国家数据局构建全域功能节点、区域功能节点、城市业务节点三层技术架 构,已布局建设7个区域功能节点,建成25个城市业务节点,初步构建了覆盖16个省(区、市)的数据 基础设施主体架构,在一定范围内实现了数据"一点发布、全域可见",应用"一点部署、跨区通用"。预 计到今年底,城市业务节点规模将进一步扩大至50个左右,覆盖80%的省(区、市)。 组织建设一体化算力网试验验证平台,以63项指标绘制全国算力资源"动态地图",初步实现对东数西算 八大枢纽十大集群、部分非枢纽省份、三大运营商及曙光超算互联网的监测调度,覆盖901个算力设 施,可调度4万PFLOPS跨主体跨地域多元算力,提供40余款通算、智算、超算等产品服务。 (文章来源:人民日报) 本报北京9月6日电 (记者王云杉)记者从国家数据局获悉:一年来,国家数据基础设施建设取得了积 极进展,正在逐步形成以释放数据要素价值为目标、以网络为支撑、以算力为底座、以数据流通利用设 施建设为重点的新型基础设施体系。 发布《国家数据基础设施建设指引》,明确提出国 ...
算力领域,密集迎大事件
财联社· 2025-08-30 11:09
Core Viewpoint - The article highlights significant developments in the computing power sector, emphasizing the government's push for high-quality development and the need to address the current supply-demand imbalance in computing resources [1][4][5]. Group 1: Government Initiatives - The State Council recently released a top-level document on "Artificial Intelligence+", focusing on strengthening the overall coordination of intelligent computing power [1][6]. - During the 2025 Digital Expo, the National Data Bureau launched a nationwide integrated computing power network monitoring and scheduling platform, aiming to enhance resource allocation and efficiency [6][7]. - A joint initiative was proposed to strengthen the collaborative development of computing power hubs, emphasizing the need for unified technical standards and regional cooperation [7]. Group 2: Industry Development - As of June 2023, China has 10.85 million operational computing centers, with intelligent computing power reaching 788 EFLOPS [4]. - The computing power industry is experiencing a structural mismatch in supply and demand, with industry insiders expressing concerns about the quality of current resources [4][8]. - The government is signaling the importance of improving the quality of computing power supply, with a focus on avoiding disorderly competition and ensuring sustainable development [5][6]. Group 3: Challenges in Computing Power - The industry faces challenges such as insufficient core supply, communication bottlenecks, and complex scheduling of computing resources [8]. - By 2025, the demand for intelligent computing in China is expected to reach 300E, but the supply distribution remains uneven [8][14]. - The mismatch in computing resources is particularly evident in the government cloud sector, where various factors contribute to the imbalance [9][12]. Group 4: Perspectives on Resource Allocation - Some industry experts argue that the current supply-demand imbalance is a normal market adjustment and not necessarily a true mismatch [14][15]. - The lack of effective commercial closure for many large model enterprises contributes to underutilization of computing resources [15]. - There is a consensus among industry players on the need for high-quality development and various proposed solutions to optimize resource utilization [16][17].
“连接+算力”双轮驱动,谢峻石畅谈中兴数智化转型之路
Huan Qiu Wang· 2025-04-30 02:12
Core Insights - The 2025 ZTE China Ecosystem Partner Conference focuses on the integration of digitalization and intelligence, aiming to accelerate customers' digital transformation [1] - ZTE's Executive Vice President and COO, Xie Junshi, discussed the company's strategies in AI and computing power during an interview, highlighting the importance of AI in driving technological and industrial revolutions [3] Group 1: AI Market and Investment Trends - The global AI market is projected to exceed $800 billion by 2028, with a compound annual growth rate (CAGR) of over 32% [3] - Domestic operators are expected to maintain over 20% growth in computing power investments by 2025, while leading internet companies will invest over 500 billion in AI over the next three years [3] Group 2: ZTE's Strategic Focus Areas - ZTE is focusing on three main areas in AI: enhancing connectivity through innovations in 5G-A, 6G, and optical networks; developing computing power solutions such as unified cloud management platforms; and deepening AI-native ICT infrastructure [3][4] - The company aims to strengthen AI applications on the consumer side with its "AI for All" strategy, which includes a full range of AI terminals and home upgrades [4] Group 3: Financial Performance and Future Plans - In Q1, ZTE reported revenue of 32.97 billion, a year-on-year increase of 7.8%, with over 35% of revenue coming from computing power and terminal businesses [4] - ZTE plans to expand its cooperation in the internet market and is well-prepared to meet the 2025 market demand with comprehensive computing power solutions [5] - The company will continue to invest in AI, aiming to integrate AI technology into industry applications and everyday life, fostering an open and collaborative AI ecosystem [5]