{"title":"一个快速,简单的路由器,用于数据密集型架构(DIVA)系统","authors":"C. Kang, J. Draper","doi":"10.1109/MWSCAS.2000.951617","DOIUrl":null,"url":null,"abstract":"This paper presents a fast, simple router design for implementing the Red Rover algorithm for a bidirectional ring. This design is very suitable for the Data-Intensive Architecture (DIVA) system, a system which demonstrates the benefits of embedded DRAM technology, because of its high performance as well as simple architecture and low cost. The key attributes of this router are one clock node-to-node latency, high channel throughput, and simple hardware implementation. The router architecture employs short-cut FIFO data paths, which makes the router speed independent of the channel buffer size (in terms of flits). A prototype implementation of the router achieves a maximum channel bandwidth of 5.12 Gb/s and runs at 80 MHz using 3.3 V CMOS signaling in 0.5 /spl mu/m technology. This high throughput and low latency were achieved without resorting to the use of complex high-speed signaling technologies.","PeriodicalId":437349,"journal":{"name":"Proceedings of the 43rd IEEE Midwest Symposium on Circuits and Systems (Cat.No.CH37144)","volume":"8 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2000-08-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"6","resultStr":"{\"title\":\"A fast, simple router for the Data-Intensive Architecture (DIVA) system\",\"authors\":\"C. Kang, J. Draper\",\"doi\":\"10.1109/MWSCAS.2000.951617\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"This paper presents a fast, simple router design for implementing the Red Rover algorithm for a bidirectional ring. This design is very suitable for the Data-Intensive Architecture (DIVA) system, a system which demonstrates the benefits of embedded DRAM technology, because of its high performance as well as simple architecture and low cost. The key attributes of this router are one clock node-to-node latency, high channel throughput, and simple hardware implementation. The router architecture employs short-cut FIFO data paths, which makes the router speed independent of the channel buffer size (in terms of flits). A prototype implementation of the router achieves a maximum channel bandwidth of 5.12 Gb/s and runs at 80 MHz using 3.3 V CMOS signaling in 0.5 /spl mu/m technology. This high throughput and low latency were achieved without resorting to the use of complex high-speed signaling technologies.\",\"PeriodicalId\":437349,\"journal\":{\"name\":\"Proceedings of the 43rd IEEE Midwest Symposium on Circuits and Systems (Cat.No.CH37144)\",\"volume\":\"8 1\",\"pages\":\"0\"},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2000-08-08\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"6\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Proceedings of the 43rd IEEE Midwest Symposium on Circuits and Systems (Cat.No.CH37144)\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/10.1109/MWSCAS.2000.951617\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Proceedings of the 43rd IEEE Midwest Symposium on Circuits and Systems (Cat.No.CH37144)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/MWSCAS.2000.951617","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 6
摘要
本文提出了一种快速、简单的路由器设计,用于实现双向环的Red Rover算法。本设计非常适用于数据密集型架构(DIVA)系统,这是一个体现嵌入式DRAM技术优势的系统,因为它具有高性能、简单的架构和低成本。该路由器的主要特性是一个时钟节点到节点延迟、高通道吞吐量和简单的硬件实现。路由器架构采用了先进先出(FIFO)数据路径的捷径,这使得路由器的速度与通道缓冲区大小无关(就flits而言)。该路由器的原型实现实现了5.12 Gb/s的最大信道带宽,并使用3.3 V CMOS信号在0.5 /spl mu/m技术下运行在80 MHz。这种高吞吐量和低延迟是在不使用复杂的高速信令技术的情况下实现的。
A fast, simple router for the Data-Intensive Architecture (DIVA) system
This paper presents a fast, simple router design for implementing the Red Rover algorithm for a bidirectional ring. This design is very suitable for the Data-Intensive Architecture (DIVA) system, a system which demonstrates the benefits of embedded DRAM technology, because of its high performance as well as simple architecture and low cost. The key attributes of this router are one clock node-to-node latency, high channel throughput, and simple hardware implementation. The router architecture employs short-cut FIFO data paths, which makes the router speed independent of the channel buffer size (in terms of flits). A prototype implementation of the router achieves a maximum channel bandwidth of 5.12 Gb/s and runs at 80 MHz using 3.3 V CMOS signaling in 0.5 /spl mu/m technology. This high throughput and low latency were achieved without resorting to the use of complex high-speed signaling technologies.