{"title":"在日立SR2201超级计算机上并行计算小波变换的进程间通信","authors":"A. Smyk, M. Tudruj","doi":"10.1109/PCEE.2000.873638","DOIUrl":null,"url":null,"abstract":"The paper presents results of experiments which studied the impact of communication control preprocessing on efficiency of parallel programs which compute two-dimensional wavelet transforms on multiprocessor computer Hitachi SR2201. The transforms are computed using the fast Mallat algorithm. Two communication models were applied in the experiments: the MPI persistent communication requests and the Remote-DMA communication based on multiple buffering areas with the reuse of the transmission control word. It has been shown that using the Remote-DMA communication with the preprocessed communication control infrastructure important speed-up of program execution can be obtained.","PeriodicalId":369394,"journal":{"name":"Proceedings International Conference on Parallel Computing in Electrical Engineering. PARELEC 2000","volume":"4 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2000-08-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"Inter-process communication for parallel computations of wavelet transforms on Hitachi SR2201 supercomputer\",\"authors\":\"A. Smyk, M. Tudruj\",\"doi\":\"10.1109/PCEE.2000.873638\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"The paper presents results of experiments which studied the impact of communication control preprocessing on efficiency of parallel programs which compute two-dimensional wavelet transforms on multiprocessor computer Hitachi SR2201. The transforms are computed using the fast Mallat algorithm. Two communication models were applied in the experiments: the MPI persistent communication requests and the Remote-DMA communication based on multiple buffering areas with the reuse of the transmission control word. It has been shown that using the Remote-DMA communication with the preprocessed communication control infrastructure important speed-up of program execution can be obtained.\",\"PeriodicalId\":369394,\"journal\":{\"name\":\"Proceedings International Conference on Parallel Computing in Electrical Engineering. PARELEC 2000\",\"volume\":\"4 1\",\"pages\":\"0\"},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2000-08-27\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Proceedings International Conference on Parallel Computing in Electrical Engineering. PARELEC 2000\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/10.1109/PCEE.2000.873638\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Proceedings International Conference on Parallel Computing in Electrical Engineering. PARELEC 2000","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/PCEE.2000.873638","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
Inter-process communication for parallel computations of wavelet transforms on Hitachi SR2201 supercomputer
The paper presents results of experiments which studied the impact of communication control preprocessing on efficiency of parallel programs which compute two-dimensional wavelet transforms on multiprocessor computer Hitachi SR2201. The transforms are computed using the fast Mallat algorithm. Two communication models were applied in the experiments: the MPI persistent communication requests and the Remote-DMA communication based on multiple buffering areas with the reuse of the transmission control word. It has been shown that using the Remote-DMA communication with the preprocessed communication control infrastructure important speed-up of program execution can be obtained.