首页> 外文期刊>Microprocessors and microsystems >A CAD approach for pre-layout optimal PDN design and its post-layout verification
【24h】

A CAD approach for pre-layout optimal PDN design and its post-layout verification

机译:布局前最佳PDN设计的CAD方法及其布局后验证

获取原文
获取原文并翻译 | 示例
           

摘要

Design of power distribution network (PDN) draws great attention to integrated circuit (IC) designers as it directly affects the circuit performance. The challenge is to immune the circuit from the noise arising due to PDN by means of effective placement of decoupling capacitance (decap) with optimal power, delay, energy and area utilization. Our work involves extraction of circuit parameters in the pre-layout stage and estimating decoupling capacitance as an early prediction activity based on dynamic power dissipation. We propose a new modular pre-layout PDN (MPLPDN) design approach, which modularizes the circuit and involves a computer aided design (CAD) methodology to investigate the changes in noise, power and delay parameter based on predictive decap allocation across the modules. A pruning methodology has been adopted to choose the best combination of the modules. The modularization algorithm is applied in a bottom up approach i.e. the lower level modules are combined at first and the top level module is created at the last stage. The experimental results based on pre and post layout simulation of application circuits show that MPLPDN achieves a considerable PDN noise suppression with a negligible increase in power, delay, energy and area. The comparison between pre-layout and post-layout simulation results also establishes the correctness of our early prediction. (C) 2019 Elsevier B.V. All rights reserved.
机译:配电网络(PDN)的设计引起了集成电路(IC)设计人员的极大关注,因为它直接影响电路性能。挑战在于如何通过有效放置去耦电容(decap)以及最佳功率,延迟,能量和面积利用率,使电路免受PDN引起的噪声的影响。我们的工作包括在预布局阶段提取电路参数,并根据动态功耗估算去耦电容,作为早期的预测活动。我们提出了一种新的模块化预布局PDN(MPLPDN)设计方法,该方法对电路进行了模块化,并涉及一种计算机辅助设计(CAD)方法,以基于模块之间的预测开端分配来研究噪声,功率和延迟参数的变化。已采用修剪方法来选择模块的最佳组合。模块化算法以自下而上的方式应用,即,较低级别的模块首先合并,而顶层模块则在最后阶段创建。基于应用电路的前后布局仿真的实验结果表明,MPLPDN可实现相当大的PDN噪声抑制,而功率,延迟,能量和面积的增加可忽略不计。布局前和布局后仿真结果之间的比较也确定了我们早期预测的正确性。 (C)2019 Elsevier B.V.保留所有权利。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号