{"title":"Scheduling moldable tasks on homogeneous multi-cluster platforms with GPUs","authors":"Fangfang Wu , Run Zhang , Xiandong Zhang","doi":"10.1016/j.cor.2025.107041","DOIUrl":null,"url":null,"abstract":"<div><div>This paper examines task scheduling in homogeneous multi-cluster platforms, equipped with Graphics Processing Units (GPUs), with the aim of minimizing the makespan. We assume that tasks can be parallelized across these platforms under the moldable model. Recognizing the NP-hard nature of the problem, our goal is to develop algorithms that provide approximation ratios. While existing research has established algorithms for single-cluster GPU environments, scaling these to multi-cluster platforms introduces new challenges, especially due to the restriction that tasks cannot use processors from different clusters. We propose an integer programming-based algorithm that achieves an approximation ratio of <span><math><mrow><mfrac><mrow><mn>3</mn></mrow><mrow><mn>2</mn></mrow></mfrac><mo>+</mo><mi>ϵ</mi></mrow></math></span>, trading off runtime for an improved approximation ratio. Additionally, leveraging recent theoretical advancements, we have created a polynomial-time algorithm with an approximation ratio of <span><math><mrow><mn>2</mn><mo>+</mo><mi>ϵ</mi></mrow></math></span>. Empirical computational experiments show that our algorithms surpass their counterparts in empirical approximation ratios.</div></div>","PeriodicalId":10542,"journal":{"name":"Computers & Operations Research","volume":"179 ","pages":"Article 107041"},"PeriodicalIF":4.1000,"publicationDate":"2025-03-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Computers & Operations Research","FirstCategoryId":"5","ListUrlMain":"https://www.sciencedirect.com/science/article/pii/S0305054825000693","RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q2","JCRName":"COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS","Score":null,"Total":0}
引用次数: 0
Abstract
This paper examines task scheduling in homogeneous multi-cluster platforms, equipped with Graphics Processing Units (GPUs), with the aim of minimizing the makespan. We assume that tasks can be parallelized across these platforms under the moldable model. Recognizing the NP-hard nature of the problem, our goal is to develop algorithms that provide approximation ratios. While existing research has established algorithms for single-cluster GPU environments, scaling these to multi-cluster platforms introduces new challenges, especially due to the restriction that tasks cannot use processors from different clusters. We propose an integer programming-based algorithm that achieves an approximation ratio of , trading off runtime for an improved approximation ratio. Additionally, leveraging recent theoretical advancements, we have created a polynomial-time algorithm with an approximation ratio of . Empirical computational experiments show that our algorithms surpass their counterparts in empirical approximation ratios.
期刊介绍:
Operations research and computers meet in a large number of scientific fields, many of which are of vital current concern to our troubled society. These include, among others, ecology, transportation, safety, reliability, urban planning, economics, inventory control, investment strategy and logistics (including reverse logistics). Computers & Operations Research provides an international forum for the application of computers and operations research techniques to problems in these and related fields.