Lokendra S. Panwar, Ashwin M. Aji, Jiayuan Meng, P. Balaji, Wu-chun Feng
{"title":"异构gpu集群的在线性能投影","authors":"Lokendra S. Panwar, Ashwin M. Aji, Jiayuan Meng, P. Balaji, Wu-chun Feng","doi":"10.1109/ICPADS.2013.48","DOIUrl":null,"url":null,"abstract":"We present a fully automated approach to project the relative performance of an OpenCL program over different GPUs. Performance projections can be made within a small amount of time, and the projection overhead stays relatively constant with the input data size. As a result, the technique can help runtime tools make dynamic decisions about which GPU would run faster for a given kernel. Usage cases of this technique include scheduling or migrating GPU workloads over a heterogeneous cluster with different types of GPUs.","PeriodicalId":160979,"journal":{"name":"2013 International Conference on Parallel and Distributed Systems","volume":"25 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"1900-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"2","resultStr":"{\"title\":\"Online Performance Projection for Clusters with Heterogeneous GPUs\",\"authors\":\"Lokendra S. Panwar, Ashwin M. Aji, Jiayuan Meng, P. Balaji, Wu-chun Feng\",\"doi\":\"10.1109/ICPADS.2013.48\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"We present a fully automated approach to project the relative performance of an OpenCL program over different GPUs. Performance projections can be made within a small amount of time, and the projection overhead stays relatively constant with the input data size. As a result, the technique can help runtime tools make dynamic decisions about which GPU would run faster for a given kernel. Usage cases of this technique include scheduling or migrating GPU workloads over a heterogeneous cluster with different types of GPUs.\",\"PeriodicalId\":160979,\"journal\":{\"name\":\"2013 International Conference on Parallel and Distributed Systems\",\"volume\":\"25 1\",\"pages\":\"0\"},\"PeriodicalIF\":0.0000,\"publicationDate\":\"1900-01-01\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"2\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"2013 International Conference on Parallel and Distributed Systems\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/10.1109/ICPADS.2013.48\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"2013 International Conference on Parallel and Distributed Systems","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/ICPADS.2013.48","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
Online Performance Projection for Clusters with Heterogeneous GPUs
We present a fully automated approach to project the relative performance of an OpenCL program over different GPUs. Performance projections can be made within a small amount of time, and the projection overhead stays relatively constant with the input data size. As a result, the technique can help runtime tools make dynamic decisions about which GPU would run faster for a given kernel. Usage cases of this technique include scheduling or migrating GPU workloads over a heterogeneous cluster with different types of GPUs.
求助内容：
应助结果提醒方式：
京公网安备 11010802042870号
