算力资源调度优化技术综述OA
A survey of techniques for optimizing computational resource scheduling
随着云计算、边缘计算、人工智能及高性能计算的快速发展,算力资源调度优化成为提升资源利用率、降低延迟与能耗的关键技术.异构资源管理、动态负载响应及多目标权衡对调度策略提出更高要求.智能调度方法逐渐取代传统算法,Kubernetes、Mesos等平台促进了应用落地,但实时性和扩展性仍有待提升.本文系统综述了算力调度的理论基础、关键技术及典型应用,聚焦云-边-HPC领域最新进展,探讨智能化、协同化与绿色化趋势,为相关研究与实践提供参考.
With the rapid advancement of cloud computing,edge computing,artificial intelligence,and high-performance comput-ing,computational resource scheduling has become a key enabler for improving resource utilization while reducing latency and energy consumption.The growing complexity of heterogeneous resources,dynamic workloads,and multi-objective trade-offs poses increasing challenges to scheduling strategies.Intelligent scheduling approaches are progressively replacing traditional algorithms,and platforms such as Kubernetes and Mesos have facilitated practical deployment,though issues of real-time performance and scalability persist.This paper provides a comprehensive survey of the theoretical foundations,core technologies,and representative applications of computational scheduling,with a focus on recent developments across cloud-edge-high-performance computing paradigms.Trends in intelligent,col-laborative,and energy-efficient scheduling are discussed to inform future research and practice.
刘娟
中国电信股份有限公司四川分公司,四川 成都 611756
信息技术与安全科学
算力资源调度高性能计算智能调度
Computational resource schedulingHigh performance computingIntelligent scheduling
《通信与信息技术》 2026 (1)
49-52,4
评论