Paper Info
Title
A novel hybrid CPU-GPU generalized eigensolver for electronic structure calculations based on fine-grained memory aware tasks
Abstract
The adoption of hybrid CPU-GPU nodes in traditional supercomputing platforms such as the Cray-XK6 opens acceleration opportunities for electronic structure calculations in materials science and chemistry applications, where medium-sized generalized eigenvalue problems must be solved many times. These eigenvalue problems are too small to effectively solve on distributed systems, but can benefit from the massive computing power concentrated on a single-node, hybrid CPU-GPU system. However, hybrid systems call for the development of new algorithms that efficiently exploit heterogeneity and massive parallelism of not just GPUs, but of multicore/manycore CPUs as well. Addressing these demands, we developed a generalized eigensolver featuring novel algorithms of increased computational intensity compared with the standard algorithms, decomposition of the computation into fine-grained memory aware tasks, and their hybrid execution. The resulting eigensolvers are state-of-the-art in high-performance computing, significantly outperforming existing libraries. We describe the algorithm and analyze its performance impact on applications of interest when different fractions of eigenvectors are needed by the host electronic structure code.
Year
DOI
Venue
2014
10.1177/1094342013502097
International Journal of High Performance Computing Applications
Keywords
DocType
Volume
Eigensolver, generalized eigensolver, two-stage, multicore, GPU, hybrid, electronic structure calculations, high performance
Journal
28
Issue
ISSN
Citations 
2
1094-3420
7
PageRank 
References 
Authors
0.48
14
5
Name
Order
Citations
PageRank
Azzam Haidar140935.39
Stanimire Tomov21214102.02
Jack J. Dongarra3176252615.79
Raffaele Solcà4353.74
Thomas C. Schulthess510615.16