Statistical timing and power analysis of VLSI considering non-linear dependence

Lerong Cheng, Wenyao Xu, Fengbo Ren, Fang Gong, Puneet Gupta, Lei He

Research output: Contribution to journalArticle

2 Citations (Scopus)

Abstract

Majority of practical multivariate statistical analysis and optimizations model interdependence among random variables in terms of the linear correlation. Though linear correlation is simple to use and evaluate, in several cases non-linear dependence between random variables may be too strong to ignore. In this paper, we propose polynomial correlation coefficients as simple measure of multi-variable non-linear dependence and show that the need for modeling non-linear dependence strongly depends on the end function that is to be evaluated from the random variables. Then, we calculate the errors in estimation resulting from assuming independence of components generated by linear de-correlation techniques, such as PCA and ICA. The experimental results show that the error predicted by our method is within 1% error compared to the real simulation of statistical timing and leakage analysis. In order to deal with non-linear dependence, we further develop a target-function-driven component analysis algorithm (FCA) to minimize the error caused by ignoring high order dependence. We apply FCA to statistical leakage power analysis and SRAM cell noise margin variation analysis. Experimental results show that the proposed FCA method is more accurate compared to the traditional PCA or ICA.

Original languageEnglish (US)
Pages (from-to)487-498
Number of pages12
JournalIntegration, the VLSI Journal
Volume47
Issue number4
DOIs
StatePublished - 2014
Externally publishedYes

Fingerprint

Random variables
Independent component analysis
Static random access storage
Statistical methods
Polynomials

Keywords

  • Statistical modeling
  • VLSI
  • Yield analysis

ASJC Scopus subject areas

  • Hardware and Architecture
  • Software
  • Electrical and Electronic Engineering

Cite this

Statistical timing and power analysis of VLSI considering non-linear dependence. / Cheng, Lerong; Xu, Wenyao; Ren, Fengbo; Gong, Fang; Gupta, Puneet; He, Lei.

In: Integration, the VLSI Journal, Vol. 47, No. 4, 2014, p. 487-498.

Research output: Contribution to journalArticle

Cheng, Lerong ; Xu, Wenyao ; Ren, Fengbo ; Gong, Fang ; Gupta, Puneet ; He, Lei. / Statistical timing and power analysis of VLSI considering non-linear dependence. In: Integration, the VLSI Journal. 2014 ; Vol. 47, No. 4. pp. 487-498.
@article{b62991e4c3ae449499feb45cc92ff44b,
title = "Statistical timing and power analysis of VLSI considering non-linear dependence",
abstract = "Majority of practical multivariate statistical analysis and optimizations model interdependence among random variables in terms of the linear correlation. Though linear correlation is simple to use and evaluate, in several cases non-linear dependence between random variables may be too strong to ignore. In this paper, we propose polynomial correlation coefficients as simple measure of multi-variable non-linear dependence and show that the need for modeling non-linear dependence strongly depends on the end function that is to be evaluated from the random variables. Then, we calculate the errors in estimation resulting from assuming independence of components generated by linear de-correlation techniques, such as PCA and ICA. The experimental results show that the error predicted by our method is within 1{\%} error compared to the real simulation of statistical timing and leakage analysis. In order to deal with non-linear dependence, we further develop a target-function-driven component analysis algorithm (FCA) to minimize the error caused by ignoring high order dependence. We apply FCA to statistical leakage power analysis and SRAM cell noise margin variation analysis. Experimental results show that the proposed FCA method is more accurate compared to the traditional PCA or ICA.",
keywords = "Statistical modeling, VLSI, Yield analysis",
author = "Lerong Cheng and Wenyao Xu and Fengbo Ren and Fang Gong and Puneet Gupta and Lei He",
year = "2014",
doi = "10.1016/j.vlsi.2013.12.004",
language = "English (US)",
volume = "47",
pages = "487--498",
journal = "Integration, the VLSI Journal",
issn = "0167-9260",
publisher = "Elsevier",
number = "4",

}

TY - JOUR

T1 - Statistical timing and power analysis of VLSI considering non-linear dependence

AU - Cheng, Lerong

AU - Xu, Wenyao

AU - Ren, Fengbo

AU - Gong, Fang

AU - Gupta, Puneet

AU - He, Lei

PY - 2014

Y1 - 2014

N2 - Majority of practical multivariate statistical analysis and optimizations model interdependence among random variables in terms of the linear correlation. Though linear correlation is simple to use and evaluate, in several cases non-linear dependence between random variables may be too strong to ignore. In this paper, we propose polynomial correlation coefficients as simple measure of multi-variable non-linear dependence and show that the need for modeling non-linear dependence strongly depends on the end function that is to be evaluated from the random variables. Then, we calculate the errors in estimation resulting from assuming independence of components generated by linear de-correlation techniques, such as PCA and ICA. The experimental results show that the error predicted by our method is within 1% error compared to the real simulation of statistical timing and leakage analysis. In order to deal with non-linear dependence, we further develop a target-function-driven component analysis algorithm (FCA) to minimize the error caused by ignoring high order dependence. We apply FCA to statistical leakage power analysis and SRAM cell noise margin variation analysis. Experimental results show that the proposed FCA method is more accurate compared to the traditional PCA or ICA.

AB - Majority of practical multivariate statistical analysis and optimizations model interdependence among random variables in terms of the linear correlation. Though linear correlation is simple to use and evaluate, in several cases non-linear dependence between random variables may be too strong to ignore. In this paper, we propose polynomial correlation coefficients as simple measure of multi-variable non-linear dependence and show that the need for modeling non-linear dependence strongly depends on the end function that is to be evaluated from the random variables. Then, we calculate the errors in estimation resulting from assuming independence of components generated by linear de-correlation techniques, such as PCA and ICA. The experimental results show that the error predicted by our method is within 1% error compared to the real simulation of statistical timing and leakage analysis. In order to deal with non-linear dependence, we further develop a target-function-driven component analysis algorithm (FCA) to minimize the error caused by ignoring high order dependence. We apply FCA to statistical leakage power analysis and SRAM cell noise margin variation analysis. Experimental results show that the proposed FCA method is more accurate compared to the traditional PCA or ICA.

KW - Statistical modeling

KW - VLSI

KW - Yield analysis

UR - http://www.scopus.com/inward/record.url?scp=84903278014&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84903278014&partnerID=8YFLogxK

U2 - 10.1016/j.vlsi.2013.12.004

DO - 10.1016/j.vlsi.2013.12.004

M3 - Article

AN - SCOPUS:84903278014

VL - 47

SP - 487

EP - 498

JO - Integration, the VLSI Journal

JF - Integration, the VLSI Journal

SN - 0167-9260

IS - 4

ER -