{"title":"动态可重构noc中的任务分散度量","authors":"Mohammad Hosseinabady, J. Núñez-Yáñez, A. Coppola","doi":"10.1109/ISVLSI.2010.91","DOIUrl":null,"url":null,"abstract":"Large Network-on-Chips (NoCs) with thousands of tiles benefit from dynamic task mapping techniques in which the assignment of tasks to tiles is decided at run-time. Dynamic task mapping technique can suffer from fragmentation with free nodes being dispersed as application are mapped and released, continuously. Using these dispersed free tiles results in interference among the task traffics of different applications. The aim of this paper is to distinguish and measure the traffic contention among the tasks of an application (internal contention) and the tasks of different applications (external contention). External contention is one of the factors which decrease the application efficiency and increase the power consumption and latency. This paper also proposes a method to decrease the external contention during task mapping scenario. The proposed technique shows up to 66.34% reduction in external communication in comparison with state-of-the-art task mapping techniques.","PeriodicalId":187530,"journal":{"name":"2010 IEEE Computer Society Annual Symposium on VLSI","volume":"60 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2010-07-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"1","resultStr":"{\"title\":\"Task Dispersal Measurement in Dynamic Reconfigurable NoCs\",\"authors\":\"Mohammad Hosseinabady, J. Núñez-Yáñez, A. Coppola\",\"doi\":\"10.1109/ISVLSI.2010.91\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"Large Network-on-Chips (NoCs) with thousands of tiles benefit from dynamic task mapping techniques in which the assignment of tasks to tiles is decided at run-time. Dynamic task mapping technique can suffer from fragmentation with free nodes being dispersed as application are mapped and released, continuously. Using these dispersed free tiles results in interference among the task traffics of different applications. The aim of this paper is to distinguish and measure the traffic contention among the tasks of an application (internal contention) and the tasks of different applications (external contention). External contention is one of the factors which decrease the application efficiency and increase the power consumption and latency. This paper also proposes a method to decrease the external contention during task mapping scenario. The proposed technique shows up to 66.34% reduction in external communication in comparison with state-of-the-art task mapping techniques.\",\"PeriodicalId\":187530,\"journal\":{\"name\":\"2010 IEEE Computer Society Annual Symposium on VLSI\",\"volume\":\"60 1\",\"pages\":\"0\"},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2010-07-05\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"1\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"2010 IEEE Computer Society Annual Symposium on VLSI\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/10.1109/ISVLSI.2010.91\",\"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 IEEE Computer Society Annual Symposium on VLSI","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/ISVLSI.2010.91","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
Task Dispersal Measurement in Dynamic Reconfigurable NoCs
Large Network-on-Chips (NoCs) with thousands of tiles benefit from dynamic task mapping techniques in which the assignment of tasks to tiles is decided at run-time. Dynamic task mapping technique can suffer from fragmentation with free nodes being dispersed as application are mapped and released, continuously. Using these dispersed free tiles results in interference among the task traffics of different applications. The aim of this paper is to distinguish and measure the traffic contention among the tasks of an application (internal contention) and the tasks of different applications (external contention). External contention is one of the factors which decrease the application efficiency and increase the power consumption and latency. This paper also proposes a method to decrease the external contention during task mapping scenario. The proposed technique shows up to 66.34% reduction in external communication in comparison with state-of-the-art task mapping techniques.