Presentation 2012-11-27
An Acceleration Method by GPGPU for Analytical Placement using Quasi-Newton Method
Yukihide KOHIRA, Yasuhiro TAKASHIMA,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) In this paper, we propose an acceleration method by GPGPU for an analytical placement method using a quasi-Newton method. In recent, analytical placement methods are promising to obtain a placement for circuits with many blocks. In the existing analytical placement methods using quasi-Newton methods, it takes huge execution time to calculate the evaluation and the gradient of an objective function iteratively. In our proposed method, we accelerate the analytical placement method by the parallelization of the calculation of the gradient. Experiments show that our proposed method is effective by the comparison on the quality of obtained placement and execution time.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) analytical placement / quasi-Newton method / GPGPU / acceralation
Paper # VLD2012-74,DC2012-40
Date of Issue

Conference Information
Committee VLD
Conference Date 2012/11/19(1days)
Place (in Japanese) (See Japanese page)
Place (in English)
Topics (in Japanese) (See Japanese page)
Topics (in English)
Chair
Vice Chair
Secretary
Assistant

Paper Information
Registration To VLSI Design Technologies (VLD)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) An Acceleration Method by GPGPU for Analytical Placement using Quasi-Newton Method
Sub Title (in English)
Keyword(1) analytical placement
Keyword(2) quasi-Newton method
Keyword(3) GPGPU
Keyword(4) acceralation
1st Author's Name Yukihide KOHIRA
1st Author's Affiliation School of Computer Science and Engineering, the University of Aizu()
2nd Author's Name Yasuhiro TAKASHIMA
2nd Author's Affiliation Faculty of Environmental Engineering Department of Information and Media Engineering, the University of Kitakyushu
Date 2012-11-27
Paper # VLD2012-74,DC2012-40
Volume (vol) vol.112
Number (no) 320
Page pp.pp.-
#Pages 6
Date of Issue