{"title":"Preliminary estimates of parallelism in CONSUL programs","authors":"D. Baldwin","doi":"10.1109/HICSS.1989.48086","DOIUrl":null,"url":null,"abstract":"CONSUL is a prototype constraint-based programming language for multiprocessors. Parallelism in CONSUl programs is intended to be detected by compilers instead of being explicitly coded by people. Initial results from an experiment to study the parallelism available from real CONSUL programs are described. The experiment estimates the parallelism that is inherent in the language, independent of overheads introduced by specific implementations, machines, or run-time environments. The resulting numbers are thus an upper limit on the parallelism that real implementations can be expected to deliver. Measurements have been made on several simple programs, with overall speedups due to parallelism ranging from 2 to over 50.<<ETX>>","PeriodicalId":325958,"journal":{"name":"[1989] Proceedings of the Twenty-Second Annual Hawaii International Conference on System Sciences. Volume II: Software Track","volume":"259 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"1989-01-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"2","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"[1989] Proceedings of the Twenty-Second Annual Hawaii International Conference on System Sciences. Volume II: Software Track","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/HICSS.1989.48086","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 2
Abstract
CONSUL is a prototype constraint-based programming language for multiprocessors. Parallelism in CONSUl programs is intended to be detected by compilers instead of being explicitly coded by people. Initial results from an experiment to study the parallelism available from real CONSUL programs are described. The experiment estimates the parallelism that is inherent in the language, independent of overheads introduced by specific implementations, machines, or run-time environments. The resulting numbers are thus an upper limit on the parallelism that real implementations can be expected to deliver. Measurements have been made on several simple programs, with overall speedups due to parallelism ranging from 2 to over 50.<>