Hessam Mirsadeghi, M. S. Talebi, A. Khonsari, M. Ould-Khaoua
{"title":"光片多处理器中比例公平的缓冲区分配","authors":"Hessam Mirsadeghi, M. S. Talebi, A. Khonsari, M. Ould-Khaoua","doi":"10.1109/CADS.2010.5623523","DOIUrl":null,"url":null,"abstract":"Advances in CMOS-compatible photonic elements have made it plausible to exploit nanophotonic communication to overcome the limitations of traditional NoCs. In this paper, we consider buffer management as a resource allocation which can be cast as a convex optimization problem. Using this framework enables us to devise a token-based optimal buffer management algorithm in optical on-chip architecture that can address fairness issues, as well. The optimal buffer allocation algorithm can be implemented by a centralized controller that is supposed to send processor elements via a simple auxiliary embedded core. Simulation results confirm that the buffer allocation algorithm manages buffer space in a fair fashion as compared to simple naive schemes.","PeriodicalId":145317,"journal":{"name":"2010 15th CSI International Symposium on Computer Architecture and Digital Systems","volume":"23 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2010-11-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"1","resultStr":"{\"title\":\"Proportionally fair buffer allocation in optical chip multiprocessors\",\"authors\":\"Hessam Mirsadeghi, M. S. Talebi, A. Khonsari, M. Ould-Khaoua\",\"doi\":\"10.1109/CADS.2010.5623523\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"Advances in CMOS-compatible photonic elements have made it plausible to exploit nanophotonic communication to overcome the limitations of traditional NoCs. In this paper, we consider buffer management as a resource allocation which can be cast as a convex optimization problem. Using this framework enables us to devise a token-based optimal buffer management algorithm in optical on-chip architecture that can address fairness issues, as well. The optimal buffer allocation algorithm can be implemented by a centralized controller that is supposed to send processor elements via a simple auxiliary embedded core. Simulation results confirm that the buffer allocation algorithm manages buffer space in a fair fashion as compared to simple naive schemes.\",\"PeriodicalId\":145317,\"journal\":{\"name\":\"2010 15th CSI International Symposium on Computer Architecture and Digital Systems\",\"volume\":\"23 1\",\"pages\":\"0\"},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2010-11-09\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"1\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"2010 15th CSI International Symposium on Computer Architecture and Digital Systems\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/10.1109/CADS.2010.5623523\",\"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 15th CSI International Symposium on Computer Architecture and Digital Systems","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/CADS.2010.5623523","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
Proportionally fair buffer allocation in optical chip multiprocessors
Advances in CMOS-compatible photonic elements have made it plausible to exploit nanophotonic communication to overcome the limitations of traditional NoCs. In this paper, we consider buffer management as a resource allocation which can be cast as a convex optimization problem. Using this framework enables us to devise a token-based optimal buffer management algorithm in optical on-chip architecture that can address fairness issues, as well. The optimal buffer allocation algorithm can be implemented by a centralized controller that is supposed to send processor elements via a simple auxiliary embedded core. Simulation results confirm that the buffer allocation algorithm manages buffer space in a fair fashion as compared to simple naive schemes.