Floorplan Driven High Level Synthesis for Crosstalk Noise Minimization in Macro-cell Based Designs

Hariharan Sankaran, S. Katkoori
{"title":"Floorplan Driven High Level Synthesis for Crosstalk Noise Minimization in Macro-cell Based Designs","authors":"Hariharan Sankaran, S. Katkoori","doi":"10.1109/ISVLSI.2009.59","DOIUrl":null,"url":null,"abstract":"In DSM regime, due to higher interconnect densities, the coupling noise between adjacent signals is aggravated and can lead to many timing violations. In traditional high-level synthesis (HLS), due to lack detailed physical details, it is difficult to accurately estimate crosstalk. Crosstalk minimization is typically done during routing, which makes it computationally expensive to be used within an iterative design flow. In this paper, we propose a floorplan driven highlevel synthesis framework for minimizing crosstalk in a bus-based architecture. The proposed framework employs a Simulated Annealing engine to simultaneously explore HLS (scheduling, allocation, and binding) and floorplan (module swap, module move, and module rotate) subspaces. The effect of a high-level decision is evaluated by updating the floorplan and identifying crosstalk prone buses (i.e., those buses exceeding Lcrit). The primary goal is to minimize the number of crosstalk violations with minimum area and latency overheads. We have validated the approach by synthesizing netlists down to layout-level using Cadence-SOC encounter followed by detailed crosstalk noise analysis using Cadence Celtic. Experimental results for three DSP benchmarks (DCT, EWF, and FFT) demonstrate that the proposed approach can reduce crosstalk violations by as much as 96% (in 180 nm technology node) with an average reduction of 75% over the designs synthesized with traditional sequential flow.","PeriodicalId":137508,"journal":{"name":"2009 IEEE Computer Society Annual Symposium on VLSI","volume":"4 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2009-05-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"7","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"2009 IEEE Computer Society Annual Symposium on VLSI","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/ISVLSI.2009.59","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 7

Abstract

In DSM regime, due to higher interconnect densities, the coupling noise between adjacent signals is aggravated and can lead to many timing violations. In traditional high-level synthesis (HLS), due to lack detailed physical details, it is difficult to accurately estimate crosstalk. Crosstalk minimization is typically done during routing, which makes it computationally expensive to be used within an iterative design flow. In this paper, we propose a floorplan driven highlevel synthesis framework for minimizing crosstalk in a bus-based architecture. The proposed framework employs a Simulated Annealing engine to simultaneously explore HLS (scheduling, allocation, and binding) and floorplan (module swap, module move, and module rotate) subspaces. The effect of a high-level decision is evaluated by updating the floorplan and identifying crosstalk prone buses (i.e., those buses exceeding Lcrit). The primary goal is to minimize the number of crosstalk violations with minimum area and latency overheads. We have validated the approach by synthesizing netlists down to layout-level using Cadence-SOC encounter followed by detailed crosstalk noise analysis using Cadence Celtic. Experimental results for three DSP benchmarks (DCT, EWF, and FFT) demonstrate that the proposed approach can reduce crosstalk violations by as much as 96% (in 180 nm technology node) with an average reduction of 75% over the designs synthesized with traditional sequential flow.
基于大单元设计的串扰噪声最小化的平面驱动高电平合成
在DSM体制下,由于较高的互连密度,相邻信号之间的耦合噪声会加剧,并可能导致许多时序违规。在传统的高阶合成(HLS)中,由于缺乏详细的物理细节,难以准确地估计串扰。串扰最小化通常在路由过程中完成,这使得在迭代设计流程中使用它的计算成本很高。在本文中,我们提出了一个平面图驱动的高级综合框架,以减少基于总线的体系结构中的串扰。该框架采用模拟退火引擎同时探索HLS(调度、分配和绑定)和floorplan(模块交换、模块移动和模块旋转)子空间。通过更新平面图和识别容易串扰的总线(即那些超过Lcrit的总线)来评估高层决策的效果。主要目标是以最小的面积和延迟开销最小化串扰违规的数量。我们通过使用Cadence- soc遭遇合成网络列表到布局级别,然后使用Cadence Celtic进行详细的串扰噪声分析,验证了该方法。三个DSP基准(DCT, EWF和FFT)的实验结果表明,该方法可以减少多达96%的串扰违反(在180 nm技术节点上),平均减少75%的设计与传统的顺序流合成。
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