{"title":"一种新的NoC上多dag在线核块分配框架和策略","authors":"U. Boregowda, V. R. Chakaravarthy","doi":"10.1109/PDGC.2012.6449813","DOIUrl":null,"url":null,"abstract":"Computer industry has widely accepted that future performance increases must largely come from increasing the number of processing cores on a die. This has led to NoC processors. Task scheduling is one of the most challenging problems facing parallel programmers today which is known to be NP-complete. A good principle is space-sharing of cores and to schedule multiple DAGs simultaneously on NoC processor. Hence the need to find optimal number of cores for a DAG and further which region of cores on NoC, to be allotted for a DAG . In this work, a method is proposed to find near-optimal minimal block of cores for a DAG on a NoC processor. Further, a time efficient framework and three on-line block allotment policies to submitted DAGs are experimented. The objectives of the policies, is to improve the total completion time for the submitted set of DAGs, hence the throughput. The policies are experimented on a simulator and found to deliver better performance than the policies found in literature.","PeriodicalId":166718,"journal":{"name":"2012 2nd IEEE International Conference on Parallel, Distributed and Grid Computing","volume":"30 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2012-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"A novel framework and policies for on-line block of cores allotment for multiple DAGs on NoC\",\"authors\":\"U. Boregowda, V. R. Chakaravarthy\",\"doi\":\"10.1109/PDGC.2012.6449813\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"Computer industry has widely accepted that future performance increases must largely come from increasing the number of processing cores on a die. This has led to NoC processors. Task scheduling is one of the most challenging problems facing parallel programmers today which is known to be NP-complete. A good principle is space-sharing of cores and to schedule multiple DAGs simultaneously on NoC processor. Hence the need to find optimal number of cores for a DAG and further which region of cores on NoC, to be allotted for a DAG . In this work, a method is proposed to find near-optimal minimal block of cores for a DAG on a NoC processor. Further, a time efficient framework and three on-line block allotment policies to submitted DAGs are experimented. The objectives of the policies, is to improve the total completion time for the submitted set of DAGs, hence the throughput. The policies are experimented on a simulator and found to deliver better performance than the policies found in literature.\",\"PeriodicalId\":166718,\"journal\":{\"name\":\"2012 2nd IEEE International Conference on Parallel, Distributed and Grid Computing\",\"volume\":\"30 1\",\"pages\":\"0\"},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2012-12-01\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"2012 2nd IEEE International Conference on Parallel, Distributed and Grid Computing\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/10.1109/PDGC.2012.6449813\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"2012 2nd IEEE International Conference on Parallel, Distributed and Grid Computing","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/PDGC.2012.6449813","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
A novel framework and policies for on-line block of cores allotment for multiple DAGs on NoC
Computer industry has widely accepted that future performance increases must largely come from increasing the number of processing cores on a die. This has led to NoC processors. Task scheduling is one of the most challenging problems facing parallel programmers today which is known to be NP-complete. A good principle is space-sharing of cores and to schedule multiple DAGs simultaneously on NoC processor. Hence the need to find optimal number of cores for a DAG and further which region of cores on NoC, to be allotted for a DAG . In this work, a method is proposed to find near-optimal minimal block of cores for a DAG on a NoC processor. Further, a time efficient framework and three on-line block allotment policies to submitted DAGs are experimented. The objectives of the policies, is to improve the total completion time for the submitted set of DAGs, hence the throughput. The policies are experimented on a simulator and found to deliver better performance than the policies found in literature.