Power-aware clock routing in 7nm designs

Mohamed Chentouf, Lekbir Cherif, Zine El Abidine Alaoui Ismaili
{"title":"Power-aware clock routing in 7nm designs","authors":"Mohamed Chentouf, Lekbir Cherif, Zine El Abidine Alaoui Ismaili","doi":"10.1109/ICOA.2018.8370505","DOIUrl":null,"url":null,"abstract":"The latest advancement in nanotechnology across a different range of industries, and the increased microelectronics market demand for high-performance, high complexity, and low power System on Chips (SoCs) have pushed the Electronic Design Automation (EDA) vendors to explore and to innovate in all stages and aspects of the design development cycle. Nowadays, some IC foundries have enabled the 7nm node for mass production and have seduced many industries to target this technology for their future devices. This market trend has brought many challenges for EDA vendors and physical design and verification specialists since they need to account for many new physical constraints and design rules in order to meet the foundry requirements. On the other side, 7nm node came with new opportunities and advantages that did not exist in earlier design nodes. In this paper, we will use one of these benefits while exploring the low power clock routing problem to achieve better clock power reduction without impacting the circuit timing or area. Our approach takes advantage of the big resistivity differences between SADP and Non-SADP layer to reduce the overall clock parasitic load to be balanced by the clock driver, which will reduce the number of inverters and buffers needed to drive all clock leaf cells, and by consequence, results in an important clock power reduction. By applying our new clock routing approach on a real 7nm testcase using Nitro-SoC place and route tool of Mentor Graphics, we achieved an improvement of 14.7% in clock nets power, 4% in clock cells power, a timing improvement of 1.7% in TNS, and 9.8% in WNS, with better utilization and less total wirelength.","PeriodicalId":433166,"journal":{"name":"2018 4th International Conference on Optimization and Applications (ICOA)","volume":"647 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2018-04-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"3","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"2018 4th International Conference on Optimization and Applications (ICOA)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/ICOA.2018.8370505","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 3

Abstract

The latest advancement in nanotechnology across a different range of industries, and the increased microelectronics market demand for high-performance, high complexity, and low power System on Chips (SoCs) have pushed the Electronic Design Automation (EDA) vendors to explore and to innovate in all stages and aspects of the design development cycle. Nowadays, some IC foundries have enabled the 7nm node for mass production and have seduced many industries to target this technology for their future devices. This market trend has brought many challenges for EDA vendors and physical design and verification specialists since they need to account for many new physical constraints and design rules in order to meet the foundry requirements. On the other side, 7nm node came with new opportunities and advantages that did not exist in earlier design nodes. In this paper, we will use one of these benefits while exploring the low power clock routing problem to achieve better clock power reduction without impacting the circuit timing or area. Our approach takes advantage of the big resistivity differences between SADP and Non-SADP layer to reduce the overall clock parasitic load to be balanced by the clock driver, which will reduce the number of inverters and buffers needed to drive all clock leaf cells, and by consequence, results in an important clock power reduction. By applying our new clock routing approach on a real 7nm testcase using Nitro-SoC place and route tool of Mentor Graphics, we achieved an improvement of 14.7% in clock nets power, 4% in clock cells power, a timing improvement of 1.7% in TNS, and 9.8% in WNS, with better utilization and less total wirelength.
7nm设计中的功耗感知时钟路由
纳米技术在不同行业的最新进展,以及微电子市场对高性能、高复杂性和低功耗芯片系统(soc)的需求增加,推动了电子设计自动化(EDA)供应商在设计开发周期的各个阶段和方面进行探索和创新。如今,一些集成电路代工厂已经实现了7nm节点的大规模生产,并吸引了许多行业将该技术用于其未来的设备。这种市场趋势给EDA供应商和物理设计和验证专家带来了许多挑战,因为他们需要考虑许多新的物理约束和设计规则,以满足代工要求。另一方面,7nm节点带来了早期设计节点不存在的新机遇和优势。在本文中,我们将利用这些优势之一,同时探索低功耗时钟路由问题,在不影响电路时序或面积的情况下实现更好的时钟功耗降低。我们的方法利用SADP和非SADP层之间的大电阻率差异来减少由时钟驱动器平衡的总体时钟寄生负载,这将减少驱动所有时钟叶单元所需的逆变器和缓冲器的数量,从而导致重要的时钟功耗降低。通过使用Mentor Graphics的硝基soc器件和路由工具将我们的新时钟路由方法应用于一个真实的7nm测试用例,我们实现了时钟网功率提高14.7%,时钟单元功率提高4%,TNS时序提高1.7%,WNS时序提高9.8%,并且具有更好的利用率和更少的总无线长度。
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:481959085
Book学术官方微信