Mohd Saharul Nizam Abdul Rahim, L. Mazalan, S. F. S. Adnan
{"title":"集群计算中CPU与GPGPU并行性分析","authors":"Mohd Saharul Nizam Abdul Rahim, L. Mazalan, S. F. S. Adnan","doi":"10.1109/ISCAIE.2014.7010238","DOIUrl":null,"url":null,"abstract":"Computer cluster is a collection of multiple computers connected together so that it can be viewed and perform as a single computer using an interconnected device. Graphical Processing Unit is a specialized hardware used to accelerate the creation of an image by manipulating and altering the memory to help CPU to produce an image. Previously, only CPU is used to perform the task execution in a cluster, however since the advancement of technology in cluster computing, GPU's can now be used as alternative processor to perform execution of a task or job. Due to its relatively cheap in Flops/price ratio, GPGPU cluster system has gained a lot of popularity, however, the real world performance for both CPU cluster system and GPGPU cluster system has not been widely tested. In this paper, the performance between CPU cluster system and GPGPU cluster system simulating proteins using GROMACS 4.6.5 will be measured and analyzed, and how the parallelism between the two will be evaluate by comparing two set of tasks. The cluster's protein folding performance is measured in ns/day, average CPU utilization, average memory usage, and average network bandwidth will be taken into account in this simulation.","PeriodicalId":385258,"journal":{"name":"2014 IEEE Symposium on Computer Applications and Industrial Electronics (ISCAIE)","volume":"138 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2014-04-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"1","resultStr":"{\"title\":\"Analysis on parallelism between CPU and GPGPU processing on cluster computing\",\"authors\":\"Mohd Saharul Nizam Abdul Rahim, L. Mazalan, S. F. S. Adnan\",\"doi\":\"10.1109/ISCAIE.2014.7010238\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"Computer cluster is a collection of multiple computers connected together so that it can be viewed and perform as a single computer using an interconnected device. Graphical Processing Unit is a specialized hardware used to accelerate the creation of an image by manipulating and altering the memory to help CPU to produce an image. Previously, only CPU is used to perform the task execution in a cluster, however since the advancement of technology in cluster computing, GPU's can now be used as alternative processor to perform execution of a task or job. Due to its relatively cheap in Flops/price ratio, GPGPU cluster system has gained a lot of popularity, however, the real world performance for both CPU cluster system and GPGPU cluster system has not been widely tested. In this paper, the performance between CPU cluster system and GPGPU cluster system simulating proteins using GROMACS 4.6.5 will be measured and analyzed, and how the parallelism between the two will be evaluate by comparing two set of tasks. The cluster's protein folding performance is measured in ns/day, average CPU utilization, average memory usage, and average network bandwidth will be taken into account in this simulation.\",\"PeriodicalId\":385258,\"journal\":{\"name\":\"2014 IEEE Symposium on Computer Applications and Industrial Electronics (ISCAIE)\",\"volume\":\"138 1\",\"pages\":\"0\"},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2014-04-07\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"1\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"2014 IEEE Symposium on Computer Applications and Industrial Electronics (ISCAIE)\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/10.1109/ISCAIE.2014.7010238\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"2014 IEEE Symposium on Computer Applications and Industrial Electronics (ISCAIE)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/ISCAIE.2014.7010238","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
Analysis on parallelism between CPU and GPGPU processing on cluster computing
Computer cluster is a collection of multiple computers connected together so that it can be viewed and perform as a single computer using an interconnected device. Graphical Processing Unit is a specialized hardware used to accelerate the creation of an image by manipulating and altering the memory to help CPU to produce an image. Previously, only CPU is used to perform the task execution in a cluster, however since the advancement of technology in cluster computing, GPU's can now be used as alternative processor to perform execution of a task or job. Due to its relatively cheap in Flops/price ratio, GPGPU cluster system has gained a lot of popularity, however, the real world performance for both CPU cluster system and GPGPU cluster system has not been widely tested. In this paper, the performance between CPU cluster system and GPGPU cluster system simulating proteins using GROMACS 4.6.5 will be measured and analyzed, and how the parallelism between the two will be evaluate by comparing two set of tasks. The cluster's protein folding performance is measured in ns/day, average CPU utilization, average memory usage, and average network bandwidth will be taken into account in this simulation.