A Novel Hybrid CPU-GPU Generalized Eigensolver for Electronic Structure Calculations Based on Fine Grained Memory Aware Tasks

TitleA Novel Hybrid CPU-GPU Generalized Eigensolver for Electronic Structure Calculations Based on Fine Grained Memory Aware Tasks
Publication TypeJournal Article
Year of Publication2014
AuthorsHaidar, A., R. Solcà, M. Gates, S. Tomov, T. C. Schulthess, and J. Dongarra
JournalInternational Journal of High Performance Computing Applications
Volume28
Start Page196
Issue2
Pagination196-209
Date Published2014-05
KeywordsEigensolver, electronic structure calculations, generalized eigensolver, gpu, high performance, hybrid, Multicore, two-stage
Abstract

The adoption of hybrid CPUGPU 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 CPUGPU 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.

DOI10.1177/1094342013502097
Project Tags: