An Offloading Architecture of Lossless Compression Based on Smart NIC

Canbang Zhang, Yaming Xu, Hongliang Wang, Wei Liu, Qi Mu, Wei Guo
{"title":"An Offloading Architecture of Lossless Compression Based on Smart NIC","authors":"Canbang Zhang, Yaming Xu, Hongliang Wang, Wei Liu, Qi Mu, Wei Guo","doi":"10.1109/icceic51584.2020.00049","DOIUrl":null,"url":null,"abstract":"With the explosive growth of information, data transmission and compression are becoming more and more important. Most of the existing compression algorithms are implemented by software architecture, which results in the waste of CPU resources. By unloading the algorithm to hardware, it is beneficial to reasonably allocate resources and improve the processing efficiency of CPU. The smart Network Interface Cards (NIC) based on FPGA can not only complete the task of data unloading, but also have the function of sending the compressed data to the cloud storage device directly. It provides a new operating environment for the compression algorithm with efficient network transmission speed, and is able to reduce the system power consumption within a certain range.In this paper, we design an FPGA-based smart NIC that has the ability to unload CPU’s LZ4 compression algorithm to FPGA to improve the utilization rate of CPU. Experiments show that the architecture is able to successfully open up the data path, and the network speed is nearly 46 Gb/s. At the same time, it can be seen from the experiments that the time required to run the LZ4 compression algorithm on the CPU is proportional to the size of files. It shows that for large files, using a smart NIC for data offloading has strong practical significance.","PeriodicalId":135840,"journal":{"name":"2020 International Conference on Computer Engineering and Intelligent Control (ICCEIC)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2020-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"2020 International Conference on Computer Engineering and Intelligent Control (ICCEIC)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/icceic51584.2020.00049","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 0

Abstract

With the explosive growth of information, data transmission and compression are becoming more and more important. Most of the existing compression algorithms are implemented by software architecture, which results in the waste of CPU resources. By unloading the algorithm to hardware, it is beneficial to reasonably allocate resources and improve the processing efficiency of CPU. The smart Network Interface Cards (NIC) based on FPGA can not only complete the task of data unloading, but also have the function of sending the compressed data to the cloud storage device directly. It provides a new operating environment for the compression algorithm with efficient network transmission speed, and is able to reduce the system power consumption within a certain range.In this paper, we design an FPGA-based smart NIC that has the ability to unload CPU’s LZ4 compression algorithm to FPGA to improve the utilization rate of CPU. Experiments show that the architecture is able to successfully open up the data path, and the network speed is nearly 46 Gb/s. At the same time, it can be seen from the experiments that the time required to run the LZ4 compression algorithm on the CPU is proportional to the size of files. It shows that for large files, using a smart NIC for data offloading has strong practical significance.
一种基于智能网卡的无损压缩卸载架构
随着信息的爆炸式增长,数据的传输和压缩变得越来越重要。现有的压缩算法大多采用软件架构实现,造成了CPU资源的浪费。将算法卸载到硬件上,有利于合理分配资源,提高CPU的处理效率。基于FPGA的智能网卡不仅可以完成数据卸载任务,还具有将压缩后的数据直接发送到云存储设备的功能。它以高效的网络传输速度为压缩算法提供了新的运行环境,并能在一定范围内降低系统功耗。本文设计了一种基于FPGA的智能网卡,能够将CPU的LZ4压缩算法卸载到FPGA上,从而提高CPU的利用率。实验表明,该架构能够成功打通数据通路,网络速度接近46 Gb/s。同时,从实验中可以看出,在CPU上运行LZ4压缩算法所需的时间与文件大小成正比。这表明,对于大文件,使用智能网卡进行数据卸载具有很强的现实意义。
本文章由计算机程序翻译,如有差异,请以英文原文为准。
求助全文
约1分钟内获得全文 求助全文
来源期刊
自引率
0.00%
发文量
0
×
引用
GB/T 7714-2015
复制
MLA
复制
APA
复制
导出至
BibTeX EndNote RefMan NoteFirst NoteExpress
×
提示
您的信息不完整,为了账户安全,请先补充。
现在去补充
×
提示
您因"违规操作"
具体请查看互助需知
我知道了
×
提示
确定
请完成安全验证×
copy
已复制链接
快去分享给好友吧!
我知道了
右上角分享
点击右上角分享
0
联系我们:info@booksci.cn Book学术提供免费学术资源搜索服务,方便国内外学者检索中英文文献。致力于提供最便捷和优质的服务体验。 Copyright © 2023 布克学术 All rights reserved.
京ICP备2023020795号-1
ghs 京公网安备 11010802042870号
Book学术文献互助
Book学术文献互助群
群 号:604180095
Book学术官方微信