基于CGRA的高效FFT实现

Proceedings of the 2020 5th International Conference on Mathematics and Artificial Intelligence Pub Date : 2020-04-10 DOI:10.1145/3395260.3395279

Li Yu, Yang Jinjiang, Liu Leibo

{"title":"基于CGRA的高效FFT实现","authors":"Li Yu, Yang Jinjiang, Liu Leibo","doi":"10.1145/3395260.3395279","DOIUrl":null,"url":null,"abstract":"In this paper, we present an efficient implementation of FFT algorithm on a CGRA-based reconfigurable architecture. Radix-4 method is used in this paper according to the advantages of proposed CGRA. The performance of the radix-4 FFT implementation is optimized by the parallelism of the Processing Elements (PEs) and the multi-access scheme of the shared memory (SM). Compared with other similar reconfigurable architectures, the proposed FFT implementation on the CGRA has performance advantages. Taking 1024-point FFT as an example, we achieve 1.93X to 6.22X advantages.","PeriodicalId":103490,"journal":{"name":"Proceedings of the 2020 5th International Conference on Mathematics and Artificial Intelligence","volume":"54 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2020-04-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"2","resultStr":"{\"title\":\"Efficient FFT Implementation on a CGRA\",\"authors\":\"Li Yu, Yang Jinjiang, Liu Leibo\",\"doi\":\"10.1145/3395260.3395279\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"In this paper, we present an efficient implementation of FFT algorithm on a CGRA-based reconfigurable architecture. Radix-4 method is used in this paper according to the advantages of proposed CGRA. The performance of the radix-4 FFT implementation is optimized by the parallelism of the Processing Elements (PEs) and the multi-access scheme of the shared memory (SM). Compared with other similar reconfigurable architectures, the proposed FFT implementation on the CGRA has performance advantages. Taking 1024-point FFT as an example, we achieve 1.93X to 6.22X advantages.\",\"PeriodicalId\":103490,\"journal\":{\"name\":\"Proceedings of the 2020 5th International Conference on Mathematics and Artificial Intelligence\",\"volume\":\"54 1\",\"pages\":\"0\"},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2020-04-10\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"2\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Proceedings of the 2020 5th International Conference on Mathematics and Artificial Intelligence\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/10.1145/3395260.3395279\",\"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 2020 5th International Conference on Mathematics and Artificial Intelligence","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1145/3395260.3395279","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}

引用次数: 2

摘要

在本文中，我们提出了一种在基于cgra的可重构架构上有效实现FFT算法的方法。根据所提出的CGRA的优点，本文采用了Radix-4方法。利用处理单元(pe)的并行性和共享内存(SM)的多访问机制优化了基数4 FFT实现的性能。与其他类似的可重构架构相比，本文提出的FFT在CGRA上的实现具有性能优势。以1024点FFT为例，我们实现了1.93倍到6.22倍的优势。

本文章由计算机程序翻译，如有差异，请以英文原文为准。

查看原文本刊更多论文

Efficient FFT Implementation on a CGRA

In this paper, we present an efficient implementation of FFT algorithm on a CGRA-based reconfigurable architecture. Radix-4 method is used in this paper according to the advantages of proposed CGRA. The performance of the radix-4 FFT implementation is optimized by the parallelism of the Processing Elements (PEs) and the multi-access scheme of the shared memory (SM). Compared with other similar reconfigurable architectures, the proposed FFT implementation on the CGRA has performance advantages. Taking 1024-point FFT as an example, we achieve 1.93X to 6.22X advantages.

求助全文

通过发布文献求助，成功后即可免费获取论文全文。去求助

来源期刊

Proceedings of the 2020 5th International Conference on Mathematics and Artificial Intelligence

自引率

0.00%

发文量