网络处理器——在NIC上传输层卸载的一种简化方法

G. Gadre, S. Badhe, Kedar Kulkarni
{"title":"网络处理器——在NIC上传输层卸载的一种简化方法","authors":"G. Gadre, S. Badhe, Kedar Kulkarni","doi":"10.1109/ICACCI.2016.7732440","DOIUrl":null,"url":null,"abstract":"High performance Network Interconnect is the most important component in High Performance Computing systems. The Network Interconnect mainly consists of three components: Network Interface Card, Switch fabric, and Light weight protocol stack. In order to achieve high bandwidth and low latency, transport protocol is offloaded to Network Interface Card. In transport offload model, the Network Interface Card is not only involved in data transfer but also involved in protocol processing. Protocol processing is a very complex task that involves multiple variables and intricate functionality. Therefore, it would be very beneficial if the Network Interface Card can provide enough flexibility to support multiple protocols. Enhancing support for multiple functionalities will empower the Network Interface Card to support multiple applications. Traditionally, the fixed custom logic in FPGA or ASIC was used for protocol processing. But, it is very difficult to achieve flexibility in the fixed custom logic. To overcome this issue, Network Processor, which provides a lot of flexibility, has emerged as an alternative for conventional hardwired logic. In this paper, we present a novel architecture that makes use of a RISC processor as the Network processor for high-speed network interfaces. We focus on the use of Network Processor in protocol processing. We also share how useful Network Processor is for supporting additional features, even after the complete NIC architecture was finalized. We also explain the debugging strategy, which is very helpful for debugging complex protocol processing.","PeriodicalId":371328,"journal":{"name":"2016 International Conference on Advances in Computing, Communications and Informatics (ICACCI)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2016-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"Network processor — A simplified approach for transport layer offloading on NIC\",\"authors\":\"G. Gadre, S. Badhe, Kedar Kulkarni\",\"doi\":\"10.1109/ICACCI.2016.7732440\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"High performance Network Interconnect is the most important component in High Performance Computing systems. The Network Interconnect mainly consists of three components: Network Interface Card, Switch fabric, and Light weight protocol stack. In order to achieve high bandwidth and low latency, transport protocol is offloaded to Network Interface Card. In transport offload model, the Network Interface Card is not only involved in data transfer but also involved in protocol processing. Protocol processing is a very complex task that involves multiple variables and intricate functionality. Therefore, it would be very beneficial if the Network Interface Card can provide enough flexibility to support multiple protocols. Enhancing support for multiple functionalities will empower the Network Interface Card to support multiple applications. Traditionally, the fixed custom logic in FPGA or ASIC was used for protocol processing. But, it is very difficult to achieve flexibility in the fixed custom logic. To overcome this issue, Network Processor, which provides a lot of flexibility, has emerged as an alternative for conventional hardwired logic. In this paper, we present a novel architecture that makes use of a RISC processor as the Network processor for high-speed network interfaces. We focus on the use of Network Processor in protocol processing. We also share how useful Network Processor is for supporting additional features, even after the complete NIC architecture was finalized. We also explain the debugging strategy, which is very helpful for debugging complex protocol processing.\",\"PeriodicalId\":371328,\"journal\":{\"name\":\"2016 International Conference on Advances in Computing, Communications and Informatics (ICACCI)\",\"volume\":\"1 1\",\"pages\":\"0\"},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2016-09-01\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"2016 International Conference on Advances in Computing, Communications and Informatics (ICACCI)\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/10.1109/ICACCI.2016.7732440\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"2016 International Conference on Advances in Computing, Communications and Informatics (ICACCI)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/ICACCI.2016.7732440","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 0

摘要

高性能网络互连是高性能计算系统中最重要的组成部分。网络互连主要由三个部分组成:网络接口卡、交换结构和轻量级协议栈。为了实现高带宽和低延迟,传输协议被卸载到网络接口卡上。在传输卸载模型中,网卡不仅参与数据传输,还参与协议处理。协议处理是一项非常复杂的任务,涉及多个变量和复杂的功能。因此,如果网络接口卡能够提供足够的灵活性来支持多种协议,将是非常有益的。增强对多种功能的支持将使网络接口卡能够支持多种应用程序。传统上,协议处理采用FPGA或ASIC中的固定自定义逻辑。但是,在固定的自定义逻辑中很难实现灵活性。为了克服这个问题,网络处理器作为传统硬连接逻辑的替代方案出现了,它提供了很大的灵活性。在本文中,我们提出了一种利用RISC处理器作为高速网络接口的网络处理器的新架构。重点讨论了网络处理器在协议处理中的应用。我们还分享了网络处理器在支持附加功能方面是多么有用,甚至在完整的NIC架构最终确定之后也是如此。我们还解释了调试策略,这对调试复杂的协议处理非常有帮助。
本文章由计算机程序翻译,如有差异,请以英文原文为准。
Network processor — A simplified approach for transport layer offloading on NIC
High performance Network Interconnect is the most important component in High Performance Computing systems. The Network Interconnect mainly consists of three components: Network Interface Card, Switch fabric, and Light weight protocol stack. In order to achieve high bandwidth and low latency, transport protocol is offloaded to Network Interface Card. In transport offload model, the Network Interface Card is not only involved in data transfer but also involved in protocol processing. Protocol processing is a very complex task that involves multiple variables and intricate functionality. Therefore, it would be very beneficial if the Network Interface Card can provide enough flexibility to support multiple protocols. Enhancing support for multiple functionalities will empower the Network Interface Card to support multiple applications. Traditionally, the fixed custom logic in FPGA or ASIC was used for protocol processing. But, it is very difficult to achieve flexibility in the fixed custom logic. To overcome this issue, Network Processor, which provides a lot of flexibility, has emerged as an alternative for conventional hardwired logic. In this paper, we present a novel architecture that makes use of a RISC processor as the Network processor for high-speed network interfaces. We focus on the use of Network Processor in protocol processing. We also share how useful Network Processor is for supporting additional features, even after the complete NIC architecture was finalized. We also explain the debugging strategy, which is very helpful for debugging complex protocol processing.
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
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