C. Bobda, H. Ishebabi, Philipp Mahr, Joel Mandebi Mbongue, S. Saha
{"title":"MeXT: A Flow for Multiprocessor Exploration","authors":"C. Bobda, H. Ishebabi, Philipp Mahr, Joel Mandebi Mbongue, S. Saha","doi":"10.1109/HPEC.2019.8916428","DOIUrl":null,"url":null,"abstract":"This paper presents an extended design approach for heterogeneous multiprocessor systems. The goal in this particular design exploration approach is to ease the implementation of an adaptive multiprocessor system by creating components such as processing nodes or memories from an application. A program is profiled and analysed to gather information about task precedence, communication cost or computational patterns for hardware accelerator generation. This information is then used to solve an optimization problem using Integer Linear Programming or Answer Set Programming with the goal of 1) creating suitable multiprocessor hardware architecture and 2) mapping of tasks onto the processors. A lightweight message-passing library for on-chip communication of parallel programs is provided. The resulting abstract architecture is further processed using the vendor tool-chain to generate the target platform’s configuration. Two real-world case studies are used to show the feasibility of our design-space exploration approach.","PeriodicalId":184253,"journal":{"name":"2019 IEEE High Performance Extreme Computing Conference (HPEC)","volume":"41 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2019-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"2","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"2019 IEEE High Performance Extreme Computing Conference (HPEC)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/HPEC.2019.8916428","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
This paper presents an extended design approach for heterogeneous multiprocessor systems. The goal in this particular design exploration approach is to ease the implementation of an adaptive multiprocessor system by creating components such as processing nodes or memories from an application. A program is profiled and analysed to gather information about task precedence, communication cost or computational patterns for hardware accelerator generation. This information is then used to solve an optimization problem using Integer Linear Programming or Answer Set Programming with the goal of 1) creating suitable multiprocessor hardware architecture and 2) mapping of tasks onto the processors. A lightweight message-passing library for on-chip communication of parallel programs is provided. The resulting abstract architecture is further processed using the vendor tool-chain to generate the target platform’s configuration. Two real-world case studies are used to show the feasibility of our design-space exploration approach.