Abstract
To solve a generalized eigensystem problem, we firstly need to transform the generalized eigenproblem to a standard eigenproblem, and then reduce a matrix to tridiagonal form. These are based on both blocked Cholesky decomposition and blocked Householder tridiagonalization method. We present parallel implementations of standard transformation which combines the Cholesky into the transformation from generalized to standard form, and reduction of a dense matrix to tridiagonal form on GPU accelerator using CUBLAS. Experimental results clearly demonstrate the potential of data-parallel coprocessors for scientific computations. When comparing against the CPU implementation, the GPU implementations achieve above 16-fold and 20-fold speedups in double precision respectively.
This is a preview of subscription content, log in via an institution.
Buying options
Tax calculation will be finalised at checkout
Purchases are for personal use only
Learn about institutional subscriptionsReferences
Agullo E, Augonnet C, Dongarra J, Faverge M, Ltaief H, Thibault S, Tomov S (2010) QR factorization on a multicore node enhanced with multiple GPU accelerators, University of Tennessee Computer Science, Tech. Rep. ICL-UT-10-04.
Anderson E, Bai Z, Bischof C, Blackford S, Demmel J, Dongarra J, Du Croz J, Greenbaum A, Hammarling S, McKenney A et al. (1999), LAPAck Users’ Guide SIAM.
Barrachina S, Castillo M, Igual FD, Mayo R, Quintana-Ort‘ı ES, Quintana-Ort‘ı G, Exploiting the capabilities of modern GPUs for dense matrix computations, Concurrency and Computation: Practice and Experience, 21(18):2457–2477, 2009. (Online). Available: http://dx.doi.org/10.1002/cpe.1472
Bientinesi P, Dhillon IS, van de Geijn RA (2005) A parallel eigensolver for dense symmetric matrices based on multiple relatively robust representations. SIAM J Sci Comput 27(1):43–66
Bischof C, Van Loan C (1987) The WY representation for products of Householder matrices, SIAM J. Sci. Stat. Comp. 8, no. 1, S2–S13, Parallel processing for scientific computing (Norfolk, Va., 1985). MR 88f:65070.
Blackford L, Cleary A, Choi J, d’Azevedo d’Azevedo E, Demmel J, Dhillon I, Dongarra J, Hammarling S, Henry G, Petitet A et al. (1997), ScaLAPACK users’ guide. Society for Industrial Mathematics.
Cao X, Chi X, Gu N (2002) Parallel solving symmetric eigenproblems. 5th international conference on algorithms and architectures for parallel processing. IEEE, Beijing, China.
Dhillon Inderjit S, Parlett Beresford N (2004) Multiple representations to compute orthogonal eigenvectors of symmetric tridiagonal matrices. Linear Algebra Appl 387:1–28
Igual FD, Quintana-Ort’ı G, van de Geijn R (2009) Level-3 BLAS on a GPU: Picking the low hanging fruit. FLAME Working Note #37. DICC 2009–04-01, Universitat Jaume I. Dept. ICC.
Lessig C, Bientinesi P (2009) On parallelizing the MRRR algorithm for data-parallel coprocessors, In: Wyrzykowski R, Dongarra J, Karczewski K, Wasniewski J (eds) in PPAM (1), ser. Lecture Notes in Computer Science, vol 6067. Springer, Heidelberg, pp 396–402 (Online)
Ltaief H, Tomov S, Nath R, Du P, Dongarra J (2009) A Scalable High Performant Cholesky Factorization for Multicore with GPU Accelerators, Tech. report, LAPACK Working Note 223, Tech. Rep.
Nvidia C (2008) CUBLAS library. NVIDIA Corporation, Santa Clara, California
Tomov S, Nath R, Ltaief H, Dongarra J, (2010) Dense linear algebra solvers for multicore with GPU accelerators, IEEE international symposium on parallel and distributed processing, workshops and Ph.D. forum (IPDPSW), pp 1–8.
Tomov S, Nath R, Du P, Dongarra J (2009) MAGMA version 0.2 User Guide
Tomov S, Nath R, Dongarra J (2010) Accelerating the reduction to upper Hessenberg, tridiagonal, and bidiagonal forms through hybrid GPU-based computing. Parallel Comput 36(12):645–654
Volkov V, Demmel J, LU, QR and Cholesky factorizations using vector capabilities of GPUs, EECS Department, University of California, Berkeley, Tech. Rep. UCB/EECS-2008-49, pp 2008–49
Volkov V, Demmel JW (2008) Using GPUs to accelerate the bisection algorithm for finding eigenvalues of symmetric tridiagonal matrices, Department of Computer Science, University of Tennessee, Knoxville, inst-UT-CS:adr, LAPACK Working Note 197, (Online).
Acknowledgments
This work is supported by the National Science Foundation of China (Grant No. 60873113) and 863 Program (2009AA01A134, 2010AA012301).
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2013 Springer-Verlag Berlin Heidelberg
About this chapter
Cite this chapter
Zhao, Y., Liu, F., Wang, Y., Chi, X. (2013). Implementations of Main Algorithms for Generalized Symmetric Eigenproblem on GPU Accelerator. In: Yuen, D., Wang, L., Chi, X., Johnsson, L., Ge, W., Shi, Y. (eds) GPU Solutions to Multi-scale Problems in Science and Engineering. Lecture Notes in Earth System Sciences. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-642-16405-7_33
Download citation
DOI: https://doi.org/10.1007/978-3-642-16405-7_33
Published:
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-642-16404-0
Online ISBN: 978-3-642-16405-7
eBook Packages: Earth and Environmental ScienceEarth and Environmental Science (R0)