Abstract
This paper reports the performance of Eigen-G, which is a GPU-based eigenvalue solver for real-symmetric matrices. We confirmed that Eigen-G outperforms state-of-the-art GPU-based eigensolvers such as magma_dsyevd and magma_dsyevd_2stage implemented in the MAGMA version 1.4.0. Applying the best-tuned CUDA BLAS libraries and the GPU-CPU hybrid DGEMM yields an even better performance improvement. We observe an approximately 2.3 times speedup over magma_ dsyevd on a Tesla K20c.
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., Demmel, J., et al.: Numerical linear algebra on emerging architectures: the PLASMA and MAGMA projects. J. Phys. Conf. Ser. 180(1), 012037 (2009)
Tomov, S., Nath, R., Dongarra, J.: Accelerating the reduction to upper Hessenberg, tridiagonal, and bidiagonal forms through hybrid GPU-based computing. Parallel Comput. 36(12), 645–654 (2010)
Vömel, C., Tomov, S., Dongarra, J.: Divide and conquer on hybrid GPU-accelerated multicore systems. SIAM J. Sci. Comput. 34, 70–82 (2012)
Yamada, S., Imamura, T., Kano, T., Machida, M.: High-performance computing for exact numerical approaches to quantum many body problems on the earth simulator. In: Proceedings of the 2006 ACM/IEEE Conference on Supercomputing SC06 (2006)
Imamura, T., Yamada, S., Machida, M.: Development of a high performance eigensolver on the peta-scale next generation supercomputer system. Prog. Nucl. Sci. Technol. AESJ 2, 643–650 (2011)
Imamura, T.: ASPEN-K2: automatic-tuning and stabilization for the performance of CUDA BLAS level 2 kernels. In: 15th SIAM Conference on Parallel Processing for Scientific Computing (2012)
Yamazaki, I., Dong, T., et al.: Tridiagonalization of a dense symmetric matrix on multiple GPUs and its application to symmetric eigenvalue problems. Concurrency Comput. Pract. Exper. (2013). doi:10.1002/cpe.3152
Acknowledgment
This research was supported in part by the Ministry of Education, Scientific Research on Priority Areas, 21013014.
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2014 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Imamura, T., Yamada, S., Machida, M. (2014). Eigen-G: GPU-Based Eigenvalue Solver for Real-Symmetric Dense Matrices. In: Wyrzykowski, R., Dongarra, J., Karczewski, K., Waśniewski, J. (eds) Parallel Processing and Applied Mathematics. PPAM 2013. Lecture Notes in Computer Science(), vol 8384. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-642-55224-3_63
Download citation
DOI: https://doi.org/10.1007/978-3-642-55224-3_63
Published:
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-642-55223-6
Online ISBN: 978-3-642-55224-3
eBook Packages: Computer ScienceComputer Science (R0)