Tridiagonalization of a dense symmetric matrix on multiple GPUs and its application to symmetric eigenvalue problems

Concurrency and Computation: Practice and Experience

Volumn 26, Issue 16, 2014, Pages 2652-2666

  Yamazaki, Ichitaro ^a   Dong, Tingxing ^a   Solcà, Raffaele ^b   Tomov, Stanimire ^a   Dongarra, Jack ^a   Schulthess, Thomas ^b  

^a University of Tennessee   (United States)

^b ETH ZURICH   (Switzerland)

Author keywords

Dense linear algebra; GPU accelerators; Parallel eigensolver; Symmetric matrixvector multiplication; Symmetric tridiagonal reduction

Indexed keywords

COMPUTER GRAPHICS; COMPUTER HARDWARE; EIGENVALUES AND EIGENFUNCTIONS; GRAPHICS PROCESSING UNIT; MEMORY ARCHITECTURE; PROGRAM PROCESSORS; SCHEDULING;

DENSE LINEAR ALGEBRA; EIGENSOLVERS; GPU ACCELERATORS; MATRIX VECTOR MULTIPLICATION; TRIDIAGONAL;

MATRIX ALGEBRA;

EID: 84908118037     PISSN: 15320626     EISSN: 15320634     Source Type: Journal    
DOI: 10.1002/cpe.3152     Document Type: Article

References (35)

1. Vetter J, Glassbrook R, Dongarra J, Schwan K, Loftis B, McNally S, Meredith J, Rogers J, Roth P, Spafford K, Yalamanchili S. Keeneland: bringing heterogeneous GPU computing to the computational science community. IEEE Computing in Science and Engineering 2011; 13:90-5. available also at http://dx.doi.org/10.1109/MCSE.2011.83.
2. Anderson E, Bai Z, Bischof C, Blackford S, Demmel J, Dongarra J, Croz J, Greenbaum A, Hammarling S, McKenney A, Sorensen D. LAPACK Users' Guide, 3rd Edition. Society for Industrial and Applied Mathematics: Philadelphia, PA, 1999.
3. Tomov S, Nath R, Du P, Dongarra J. MAGMA users' guide, 2009. available at http://icl.eecs.utk.edu/magma.
4. Lawson C, Hanson R, Kincaid D, Krogh F. Basic linear algebra subprograms for FORTRAN usage. Acm Transactions on Mathematical Software 1979; 5:308-323.
5. Nath R, Tomov S, Dongarra J. An improved MAGMA GEMM for Fermi graphics processing units. International Journal of High Performance Computing Applications 2010; 24:511-515.
6. Tomov S, Dongarra J, Baboulin M. Towards dense linear algebra for hybrid GPU accelerated manycore systems. Parallel Computing 2010; 36:232-240.
7. Tomov S, Nath R, Dongarra J. Accelerating the reduction to upper Hessenberg, tridiagonal, and bidiagonal forms through hybrid GPU-based computing. Parallel Computing 2010; 36:645-654.
8. Yamazaki I, Tomov S, Dongarra J. One-sided dense matrix factorizations on a multicore with multiple GPU accelerators. International Conference on Computational Science (ICCS), Omaha, Nebraska, 2012; 37-46.
9. Auckenthaler T, Blum V, Bungartz H, Huckle T, Jahanni R, Krämer L, Lang B, Lederer H, Willems P. Parallel solution of partial symmetric eigenvalue problems from electronic structure calculations. Parallel Computing 2011; 37:783-794.
10. Kent P. Computational challenges of large-scale long-time first-principles molecular dynamics. Journal of Physics: Conference Series 2008; 125:012-058.
11. Martin R. Electronic Structure: Basic Theory and Practical Methods. Cambridge University Press, 2004.
12. Exciting. software (Available at http://exciting-code.org/) [Accessed date: September 12, 2013].
13. Dewhurst J, Sharma S, Nordström L, Cricchio F, Bultmark F, Gross E. The ELK code manual version 1.4.18. software (Available at http://elk.sourceforge.net/) [Accessed date: September 12, 2013].
14. Yamazaki I, Wu K, Simon H. nu-TRLan User Guide. Technical Report LBNL-1288E, Lawrence Berkeley National Lab, 2008. software available at https://codeforge.lbl.gov/projects/trlan/.
15. Wu K, Simon H. Thick-restart Lanczos method for large symmetric eigenvalue problems. SIAM Journal on Matrix Analysis and Applications 2000; 22:602-616.
16. Tomov S, Langou J, Dongarra J, Canning A, Wang LW. Conjugate-gradient eigenvalue solvers in computing electronic properties of nanostructure architectures. International Journal of Computational Science and Engineering 2006; 2:205-212.
17. Wang LW, Zunger A. Solving Schrodingers equation around a desired energy: application to silicon quantum dots. Journal of Chemical Physics 1994; 100:2394-2397.
18. Higham N. Functions of Matrices: Theory and Computation. Society for Industrial and Applied Mathematics: Philadelphia, PA, USA, 2008.
19. Auckenthaler T, Bungartz HJ, Huckle T, Kramer L, Lang B, Willems P. Developing algorithms and software for the parallel solution of the symmetric eigenvalue problem. Journal of Computational Science; 2:272-278. software available at http://elpa.rzg.mpg.de/.
20. QuantumEspresso. software (Available at http://www.quantum-espresso.org/) [Accessed date: September 12, 2013].
21. Gonze X, Amadon B, Anglade P-M, Beuken J-M, Bottin F, Boulanger P, Bruneval F, Caliste D, Caracas R, Côté M, Deutsch T, Genovese L, Ghosez P, Giantomassi M, Goedecker S, Hamann DR, Hermet P, Jollet F, Jomard G, Leroux S. ABINIT: First-principles approach of materials and nanosystem properties, 2009. software (Available at www.http://abinit.org/) [Accessed date: September 12, 2013].
22. Choi J, Dongarra J, Ostrouchov S, Petitet A, Walker D, Whaley R. A proposal for a set of parallel basic linear algebra subprograms. The Proceedings of the Second International Workshop on Applied Parallel Computing (PARA), Lyngby, Denmark, 1995; 107-114.
23. van de Geijn R. Using Plapack: Parallel Linear Algebra Package. MIT press: Cambridge, MA, 1997.
24. Abdelfattah A, Keyes D, Dongarra J, Ltaief H. Optimizing memory-bound numerical kernels on GPU hardware accelerators. The Proceedings of the 10th International Meeting on High-Performance Computing for Computational Science (VECPAR), Kobe, Japan, 2012; 72-79.
25. Imamura T, Yamada S, Machida M. A high performance SYMV kernel on a Fermi-core GPU. The Proceedings of the 10th International Meeting on High-Performance Computing for Computational Science (VECPAR), Kobe, Japan, 2012; 59-71.
26. Nath R, Tomov S, Dong T, Dongarra J. Optimizing symmetric dense matrix-vector multiplication on GPUs. Acm/IEEE Conference on Supercomputing (SC11), 2011.
27. Yamada S, Imamura T, Machida M. Dynamic variation of eigenvalue problems in density-matrix renormalization-group code, 2012. presented at SIAM conference on Parallel Processing for Scientific Computing.
28. NVIDIA CUDA Compute Unified Device Architecture-Programming Guide.
29. Kurzak J, Tomov S, Dongarra J. Autotuning GEMMs for Fermi. LAPACK working note 245: Technical Report UT-CS-11-671, Electrical Engineering and Computer Science Department, University of Tennessee, 2011.
30. Cuppen J. A divide and conquer method for the symmetric tridiagonal eigenproblem. Numerical Mathematics 1981; 36:177-195.
31. Vömel C, Tomov S, Dongarra J. Divide and conquer on hybrid GPU-accelerated multicore systems. SIAM Journal of Scientific Computing 2012; 34:C70-C82.
32. Yamazaki I, Wu K. A communication-avoiding thick-restart Lanczos method on a distributed-memory system. Workshop on Algorithms and Programming Tools for Next-Generation High-Performance Scientific and Software (HPCC), 2011.
33. Yamazaki I, Bai Z, Simon H, Wang LW, Wu K. Adaptive projection subspace dimension for the thick-restart Lanczos method. ACM Transactions on Mathematical Software 2010; 37:-1-1.
34. Dhillon I, Parlett B. Multiple representations to compute orthogonal eigenvectors of symmetric tridiagonal matrices. Linear Algebra and its Applications 2004; 387:1-28.
35. Yamazaki I, Dong T, Tomov S, Dongarra J. Tridiagonalization of a symmetric dense matrix on a GPU cluster. The Proceedings of the Third International Workshop on Accelerators and Hybrid Exascale Systems (AsHES), Boston, Massachusetts, 2013; 1070-1079.