摘要
算力服务是数字中国建设的重要基础和支撑,是提升国家数字化能力和核心竞争力的关键因素。在数字中国背景下,算力服务体系需要适应不同领域、层次和场景的算力需求,实现算力资源的合理配置和高效利用,引领数字经济的发展和创新。本文基于全球视角和我国现状,厘清了算力服务的内涵,剖析了我国算力服务发展中存在的供需矛盾突出、算力基础资源分布不均匀、资源流通途径尚未建立、技术服务标准不统一等痛点问题,从算力服务形态基础、演进模型、顶层设计三个方面提出了我国算力服务体系建设的总体架构,并全面阐述了我国算力服务体系重点战略和发展路径。研究建议:加强算力顶层设计,推进算网融合发展;优化算力资源布局,降低算力使用门槛;搭建算力共享平台,盘活社会算力价值;健全算力利用机制,建立算力租赁制度;激发科技攻关潜力,发挥算力人才优势。
Computing power services are a key factor for strengthening the digital capabilities and core competitiveness of China.In the context of the Digital China initiative,the computing power service system of China needs to adapt to the computing power demands of different fields,levels,and scenarios,and the reasonable allocation and efficient utilization of computing resources should be realized,thereby guiding the development and innovation of the digital economy.Based on a global perspective and China’s current situation,this study clarifies the implications of computing power services and analyzes the problems regarding the development of China’s computing power services,including prominent contradictions between supply and demand,uneven distribution of basic computing resources,lack of circulation channels for resources,non-unified technical service standards.An overall architecture for the construction of China’s computing power service system is proposed from the aspects of computing power service form basis,evolution model,and top-level design,and key strategies and development paths are explored.Furthermore,the following suggestions are proposed:(1)strengthening top-level design of computing power to promote the integrated development of computing networks;(2)optimizing computing power resource layout while lowering the threshold for using computing power;(3)building computing power sharing platforms to revitalize social computing values;(4)improving computing power utilization mechanisms and establishing computing power leasing systems;and(5)stimulating potentials in scientific and technological research by exploiting computing power talent advantages.
作者
陈晓红
许冠英
徐雪松
田志平
霍杨杰
易国栋
Chen Xiaohong;Xu Guanying;Xu Xuesong;Tian Zhiping;Huo Yangjie;Yi Guodong(Xiangjiang Laboratory,Changsha 410205,China;Business School of Central South University,Changsha 410083,China;School of Advanced Interdisciplinary Studies,Hunan University of Technology and Business,Changsha 410205,China)
出处
《中国工程科学》
CSCD
北大核心
2023年第6期49-60,共12页
Strategic Study of CAE
基金
中国工程院咨询项目“产业链供应链安全创新发展基础理论与方法”(2022-JB-01),“数据安全与数字监管的战略研究”(2023-33-08)。
关键词
算力服务
算网融合
算力服务演进模型
体系构建
computing power service
integration of computing and networking
system construction
computility service evolution model
