基于CUDA的混合精度可变预处理GCR在GPU上的实现

Digests of the 2010 14th Biennial IEEE Conference on Electromagnetic Field Computation Pub Date : 2010-05-09 DOI:10.1109/CEFC.2010.5481534

S. Ikuno, N. Fujita, Susumu Yamamoto, S. Nakata

{"title":"基于CUDA的混合精度可变预处理GCR在GPU上的实现","authors":"S. Ikuno, N. Fujita, Susumu Yamamoto, S. Nakata","doi":"10.1109/CEFC.2010.5481534","DOIUrl":null,"url":null,"abstract":"The Variable Preconditioned GVR (VPGCR) with mixed precision on Graphics Processing Unit (GPU) using Compute Unified Device Architecture (CUDA) is numerically investigated. The convergence theorem of VPGCR is guaranteed that the residual equation for the preconditioned procedure can be solved in the range of single precision operation. The results of computations show that VPGCR with mixed precision operation on GPU demonstrated significant achievement than that of CPU. Especially, VPGCR on GPU with mixed precision operation is 22.53 times faster than that of Central Processing Unit (CPU).","PeriodicalId":148739,"journal":{"name":"Digests of the 2010 14th Biennial IEEE Conference on Electromagnetic Field Computation","volume":"16 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2010-05-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"Implementation of Variable Preconditioned GCR with mixed precision on GPU using CUDA\",\"authors\":\"S. Ikuno, N. Fujita, Susumu Yamamoto, S. Nakata\",\"doi\":\"10.1109/CEFC.2010.5481534\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"The Variable Preconditioned GVR (VPGCR) with mixed precision on Graphics Processing Unit (GPU) using Compute Unified Device Architecture (CUDA) is numerically investigated. The convergence theorem of VPGCR is guaranteed that the residual equation for the preconditioned procedure can be solved in the range of single precision operation. The results of computations show that VPGCR with mixed precision operation on GPU demonstrated significant achievement than that of CPU. Especially, VPGCR on GPU with mixed precision operation is 22.53 times faster than that of Central Processing Unit (CPU).\",\"PeriodicalId\":148739,\"journal\":{\"name\":\"Digests of the 2010 14th Biennial IEEE Conference on Electromagnetic Field Computation\",\"volume\":\"16 1\",\"pages\":\"0\"},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2010-05-09\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Digests of the 2010 14th Biennial IEEE Conference on Electromagnetic Field Computation\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/10.1109/CEFC.2010.5481534\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Digests of the 2010 14th Biennial IEEE Conference on Electromagnetic Field Computation","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/CEFC.2010.5481534","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}

引用次数: 0

摘要

采用计算统一设备架构(CUDA)对混合精度的可变预置GVR (VPGCR)进行了数值研究。利用VPGCR的收敛性定理，保证了在单精度运算范围内，预条件过程的残差方程可以得到解。计算结果表明，在GPU上进行混合精度运算的VPGCR比在CPU上取得了显著的效果。特别是在混合精度运算的GPU上，VPGCR比CPU (Central Processing Unit)快22.53倍。

本文章由计算机程序翻译，如有差异，请以英文原文为准。

查看原文本刊更多论文

Implementation of Variable Preconditioned GCR with mixed precision on GPU using CUDA

The Variable Preconditioned GVR (VPGCR) with mixed precision on Graphics Processing Unit (GPU) using Compute Unified Device Architecture (CUDA) is numerically investigated. The convergence theorem of VPGCR is guaranteed that the residual equation for the preconditioned procedure can be solved in the range of single precision operation. The results of computations show that VPGCR with mixed precision operation on GPU demonstrated significant achievement than that of CPU. Especially, VPGCR on GPU with mixed precision operation is 22.53 times faster than that of Central Processing Unit (CPU).

求助全文

通过发布文献求助，成功后即可免费获取论文全文。去求助

来源期刊

Digests of the 2010 14th Biennial IEEE Conference on Electromagnetic Field Computation

自引率

0.00%

发文量