{"title":"Effective SIMD code generation for the high-level declarative data-parallel language 8 1/2","authors":"D. D. Vito, Olivier Michel","doi":"10.1109/EURMIC.1996.546372","DOIUrl":null,"url":null,"abstract":"We describe in this paper an effective code generation scheme for a declarative data-parallel language targeted towards sequential, vector or SIMD architectures. The code generation is produced from the declarative data-parallel language 8 1/2, a language that relies on the notions of stream and collection in a high-level declarative framework. We first describe the language and its specificities focusing on data-parallelism. Then we present two optimizations for the generated code: the sharing of common control expressions and the optimization of delay copies. Next, we give some elements for the evaluation of the generated code. As a conclusion, we recall the overall effectiveness of the execution scheme and draw some plans for the future.","PeriodicalId":311520,"journal":{"name":"Proceedings of EUROMICRO 96. 22nd Euromicro Conference. Beyond 2000: Hardware and Software Design Strategies","volume":"445 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"1996-09-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"4","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Proceedings of EUROMICRO 96. 22nd Euromicro Conference. Beyond 2000: Hardware and Software Design Strategies","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/EURMIC.1996.546372","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 4
Abstract
We describe in this paper an effective code generation scheme for a declarative data-parallel language targeted towards sequential, vector or SIMD architectures. The code generation is produced from the declarative data-parallel language 8 1/2, a language that relies on the notions of stream and collection in a high-level declarative framework. We first describe the language and its specificities focusing on data-parallelism. Then we present two optimizations for the generated code: the sharing of common control expressions and the optimization of delay copies. Next, we give some elements for the evaluation of the generated code. As a conclusion, we recall the overall effectiveness of the execution scheme and draw some plans for the future.