Parameter reduction for variability analysis by Slice Inverse Regression (SIR) method

Alexander Mitev, Michael Mahmoud Marefat, Dongsheng Ma, Meiling Wang

Research output: Chapter in Book/Report/Conference proceedingConference contribution

6 Citations (Scopus)

Abstract

With semiconductor fabrication technologies scaled below 100 nm, the design-manufacturing interface becomes more and more complicated. The resultant process variability causes a number of issues in the new generation IC design. One of the biggest challenges is the enormous number of process variation related parameters. These parameters represent numerous local and global variations, and pose a heavy burden in today's chip verification and design. This paper proposes a new way of reducing the statistical variations (which include both process parameters and design variables) according to their impacts on the overall circuit performance. The new approach creates an effective reduction subspace (ERS) and provides a transformation matrix by using the mean and variance of the response surface. With the generated transformation matrix, the proposed method maps the original statistical variations to a smaller set of variables with which we process variability analysis. Thus, the computational cost due to the number of variations is greatly reduced. Experimental results show that by using new method we can achieve 20% to 50% parameter reduction with only less than 8% error on average.

Original languageEnglish (US)
Title of host publicationProceedings of the Asia and South Pacific Design Automation Conference, ASP-DAC
Pages468-473
Number of pages6
DOIs
StatePublished - 2007
EventASP-DAC 2007 - Asia and South Pacific Design Automation Conference 2007 - Yokohama, Japan
Duration: Jan 23 2007Jan 27 2007

Other

OtherASP-DAC 2007 - Asia and South Pacific Design Automation Conference 2007
CountryJapan
CityYokohama
Period1/23/071/27/07

Fingerprint

Semiconductor materials
Fabrication
Networks (circuits)
Costs
Integrated circuit design

ASJC Scopus subject areas

  • Engineering(all)

Cite this

Mitev, A., Marefat, M. M., Ma, D., & Wang, M. (2007). Parameter reduction for variability analysis by Slice Inverse Regression (SIR) method. In Proceedings of the Asia and South Pacific Design Automation Conference, ASP-DAC (pp. 468-473). [4196076] https://doi.org/10.1109/ASPDAC.2007.358030

Parameter reduction for variability analysis by Slice Inverse Regression (SIR) method. / Mitev, Alexander; Marefat, Michael Mahmoud; Ma, Dongsheng; Wang, Meiling.

Proceedings of the Asia and South Pacific Design Automation Conference, ASP-DAC. 2007. p. 468-473 4196076.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Mitev, A, Marefat, MM, Ma, D & Wang, M 2007, Parameter reduction for variability analysis by Slice Inverse Regression (SIR) method. in Proceedings of the Asia and South Pacific Design Automation Conference, ASP-DAC., 4196076, pp. 468-473, ASP-DAC 2007 - Asia and South Pacific Design Automation Conference 2007, Yokohama, Japan, 1/23/07. https://doi.org/10.1109/ASPDAC.2007.358030
Mitev A, Marefat MM, Ma D, Wang M. Parameter reduction for variability analysis by Slice Inverse Regression (SIR) method. In Proceedings of the Asia and South Pacific Design Automation Conference, ASP-DAC. 2007. p. 468-473. 4196076 https://doi.org/10.1109/ASPDAC.2007.358030
Mitev, Alexander ; Marefat, Michael Mahmoud ; Ma, Dongsheng ; Wang, Meiling. / Parameter reduction for variability analysis by Slice Inverse Regression (SIR) method. Proceedings of the Asia and South Pacific Design Automation Conference, ASP-DAC. 2007. pp. 468-473
@inproceedings{5f0e27a2287b40eca772ae2ae1656d9a,
title = "Parameter reduction for variability analysis by Slice Inverse Regression (SIR) method",
abstract = "With semiconductor fabrication technologies scaled below 100 nm, the design-manufacturing interface becomes more and more complicated. The resultant process variability causes a number of issues in the new generation IC design. One of the biggest challenges is the enormous number of process variation related parameters. These parameters represent numerous local and global variations, and pose a heavy burden in today's chip verification and design. This paper proposes a new way of reducing the statistical variations (which include both process parameters and design variables) according to their impacts on the overall circuit performance. The new approach creates an effective reduction subspace (ERS) and provides a transformation matrix by using the mean and variance of the response surface. With the generated transformation matrix, the proposed method maps the original statistical variations to a smaller set of variables with which we process variability analysis. Thus, the computational cost due to the number of variations is greatly reduced. Experimental results show that by using new method we can achieve 20{\%} to 50{\%} parameter reduction with only less than 8{\%} error on average.",
author = "Alexander Mitev and Marefat, {Michael Mahmoud} and Dongsheng Ma and Meiling Wang",
year = "2007",
doi = "10.1109/ASPDAC.2007.358030",
language = "English (US)",
isbn = "1424406293",
pages = "468--473",
booktitle = "Proceedings of the Asia and South Pacific Design Automation Conference, ASP-DAC",

}

TY - GEN

T1 - Parameter reduction for variability analysis by Slice Inverse Regression (SIR) method

AU - Mitev, Alexander

AU - Marefat, Michael Mahmoud

AU - Ma, Dongsheng

AU - Wang, Meiling

PY - 2007

Y1 - 2007

N2 - With semiconductor fabrication technologies scaled below 100 nm, the design-manufacturing interface becomes more and more complicated. The resultant process variability causes a number of issues in the new generation IC design. One of the biggest challenges is the enormous number of process variation related parameters. These parameters represent numerous local and global variations, and pose a heavy burden in today's chip verification and design. This paper proposes a new way of reducing the statistical variations (which include both process parameters and design variables) according to their impacts on the overall circuit performance. The new approach creates an effective reduction subspace (ERS) and provides a transformation matrix by using the mean and variance of the response surface. With the generated transformation matrix, the proposed method maps the original statistical variations to a smaller set of variables with which we process variability analysis. Thus, the computational cost due to the number of variations is greatly reduced. Experimental results show that by using new method we can achieve 20% to 50% parameter reduction with only less than 8% error on average.

AB - With semiconductor fabrication technologies scaled below 100 nm, the design-manufacturing interface becomes more and more complicated. The resultant process variability causes a number of issues in the new generation IC design. One of the biggest challenges is the enormous number of process variation related parameters. These parameters represent numerous local and global variations, and pose a heavy burden in today's chip verification and design. This paper proposes a new way of reducing the statistical variations (which include both process parameters and design variables) according to their impacts on the overall circuit performance. The new approach creates an effective reduction subspace (ERS) and provides a transformation matrix by using the mean and variance of the response surface. With the generated transformation matrix, the proposed method maps the original statistical variations to a smaller set of variables with which we process variability analysis. Thus, the computational cost due to the number of variations is greatly reduced. Experimental results show that by using new method we can achieve 20% to 50% parameter reduction with only less than 8% error on average.

UR - http://www.scopus.com/inward/record.url?scp=46649088771&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=46649088771&partnerID=8YFLogxK

U2 - 10.1109/ASPDAC.2007.358030

DO - 10.1109/ASPDAC.2007.358030

M3 - Conference contribution

AN - SCOPUS:46649088771

SN - 1424406293

SN - 9781424406296

SP - 468

EP - 473

BT - Proceedings of the Asia and South Pacific Design Automation Conference, ASP-DAC

ER -