Accelerating Volkov’s hybrid implementation of Cholesky factorization on a Fermi GPU

機構典藏 > College of Business and Management > Graduate Institute & Department of Information Management > Proceeding > Item 987654321/96252

Please use this identifier to cite or link to this item: https://tkuir.lib.tku.edu.tw/dspace/handle/987654321/96252

Title:	Accelerating Volkov’s hybrid implementation of Cholesky factorization on a Fermi GPU
Authors:	Wei, Shih-Chieh;Huang , Bormin
Contributors:	淡江大學資訊管理學系
Keywords:	Cholesky factorization;general purpose graphics processing unit;parallel computing
Date:	2012-12
Issue Date:	2014-03-06 17:08:15 (UTC+8)
Abstract:	In linear algebra, Cholesky factorization is useful in solving a system of equations with a symmetric positive definite coefficient matrix. Cholesky factorization is roughly twice as fast relative to LU factorization which applies to general matrices. In recent years, with advances in technology, a Fermi GPU card can accommodate hundreds of cores compared to the small number of 8 or 16 cores on CPU. Therefore a trend is seen to use the graphics card as a general purpose graphics processing unit (GPGPU) for parallel computation. In this work, Volkov's hybrid implementation of Cholesky factorization is evaluated on the new Fermi GPU with others and then some improvement strategies were proposed. After experiments, compared to the CPU version using Intel Math Kernel Library (MKL), our proposed GPU improvement strategy can achieve a speedup of 3.85x on Cholesky factorization of a square matrix of dimension 10,000.
Relation:	Parallel and Distributed Systems (ICPADS), 2012 IEEE 18th International Conference on, pp.896-900
DOI:	10.1109/ICPADS.2012.147
Appears in Collections:	[Graduate Institute & Department of Information Management] Proceeding

Files in This Item:

File	Description	Size	Format
Accelerating Volkov’s Hybrid Implementation of Cholesky Factorization.pdf	全文檔	377Kb	Adobe PDF	371	View/Open
index.html		0Kb	HTML	332	View/Open