Acceleration of conjugate gradient method for circuit simulation using CUDA

Author

Maringanti, Anirudh ; Athavale, Viraj ; Patkar, Sachin B.

Author_Institution

Dept. of Electr. Eng., Indian Inst. of Technol., Mumbai, India

fYear

2009

fDate

16-19 Dec. 2009

Firstpage

438

Lastpage

444

Abstract

The Conjugate Gradient method is a popular iterative method to solve a system of linear equations and is used in a variety of applications. The DC Analyser is a circuit simulator built at IIT Bombay to solve large circuits containing resistances, voltage and current sources and which employs the conjugate gradient method. Current generation of graphics cards offer extremely high raw processing power and memory bandwidths compared to conventional CPUs. We have accelerated the conjugate gradient part of the DC Analyser using an Nvidia GTX 280 GPU and the new CUDA technology and successfully obtained a speedup of over 10Ã— for the CG method and more than 4Ã— for the entire application for very large circuits when compared to a single-threaded CPU implementation.

Keywords

circuit simulation; computer graphics; coprocessors; gradient methods; CUDA technology; DC analyser; Nvidia GTX 280 GPU; circuit simulation; compute unified device architecture; conjugate gradient method; graphics cards; graphics processing unit; Acceleration; Analytical models; Circuit analysis; Circuit simulation; Equations; Gradient methods; Graphics; Iterative methods; Power generation; Voltage;

fLanguage

English

Publisher

ieee

Conference_Titel

High Performance Computing (HiPC), 2009 International Conference on

Conference_Location

Kochi

Print_ISBN

978-1-4244-4922-4

Electronic_ISBN

978-1-4244-4921-7

Type

conf

DOI

10.1109/HIPC.2009.5433184

Filename

5433184