Ming Yang, Yuan Yao, Shuai Wei, Yuanyuan Zhang, Lei Huang
{"title":"基于SIMD矢量化优化中矢量寄存器重用的效益分析技术","authors":"Ming Yang, Yuan Yao, Shuai Wei, Yuanyuan Zhang, Lei Huang","doi":"10.1109/ISISE.2010.18","DOIUrl":null,"url":null,"abstract":"The reuse-rate of vector register is one of the most important aspects that influence the SIMD performance. However, the reuse of vector register probably can lead ton on-continuous memory access and non-hit of cache. Based on register reuse analysis, we establish a cost analysis that guide the multiple code generation. Then we perform NPB test based on different problem-scale and the result shows that program can reap an optimal performance as a result of this optimization, which can avoid low cache hit rate arose by vector register reuse for SIMD code generation.","PeriodicalId":206833,"journal":{"name":"2010 Third International Symposium on Information Science and Engineering","volume":"6 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2010-12-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"1","resultStr":"{\"title\":\"A Technology Based Benefit Analysis on Reuse of Vector Register for SIMD Vectorization Optimization\",\"authors\":\"Ming Yang, Yuan Yao, Shuai Wei, Yuanyuan Zhang, Lei Huang\",\"doi\":\"10.1109/ISISE.2010.18\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"The reuse-rate of vector register is one of the most important aspects that influence the SIMD performance. However, the reuse of vector register probably can lead ton on-continuous memory access and non-hit of cache. Based on register reuse analysis, we establish a cost analysis that guide the multiple code generation. Then we perform NPB test based on different problem-scale and the result shows that program can reap an optimal performance as a result of this optimization, which can avoid low cache hit rate arose by vector register reuse for SIMD code generation.\",\"PeriodicalId\":206833,\"journal\":{\"name\":\"2010 Third International Symposium on Information Science and Engineering\",\"volume\":\"6 1\",\"pages\":\"0\"},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2010-12-24\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"1\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"2010 Third International Symposium on Information Science and Engineering\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/10.1109/ISISE.2010.18\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"2010 Third International Symposium on Information Science and Engineering","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/ISISE.2010.18","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
A Technology Based Benefit Analysis on Reuse of Vector Register for SIMD Vectorization Optimization
The reuse-rate of vector register is one of the most important aspects that influence the SIMD performance. However, the reuse of vector register probably can lead ton on-continuous memory access and non-hit of cache. Based on register reuse analysis, we establish a cost analysis that guide the multiple code generation. Then we perform NPB test based on different problem-scale and the result shows that program can reap an optimal performance as a result of this optimization, which can avoid low cache hit rate arose by vector register reuse for SIMD code generation.