On-the-fly memory compression for multibody algorithms

Eckhardt, W.; Glas, R.; Korzh, D.; Wallner, S.; Weinzierl, T.

doi:10.3233/978-1-61499-621-7-421

On-the-fly memory compression for multibody algorithms

Eckhardt, W.; Glas, R.; Korzh, D.; Wallner, S.; Weinzierl, T.

Authors

W. Eckhardt

R. Glas

D. Korzh

S. Wallner

Professor Tobias Weinzierl tobias.weinzierl@durham.ac.uk
Professor

Contributors

G. R. Joubert
Editor

H. Leather
Editor

M. Parsons
Editor

F. Peters
Editor

M. Sawyer
Editor

Abstract

Memory and bandwidth demands challenge developers of particle-based codes that have to scale on new architectures, as the growth of concurrency outperforms improvements in memory access facilities, as the memory per core tends to stagnate, and as communication networks cannot increase bandwidth arbitrary. We propose to analyse each particle of such a code to find out whether a hierarchical data representation storing data with reduced precision caps the memory demands without exceeding given error bounds. For admissible candidates, we perform this compression and thus reduce the pressure on the memory subsystem, lower the total memory footprint and reduce the data to be exchanged via MPI. Notably, our analysis and transformation changes the data compression dynamically, i.e. the choice of data format follows the solution characteristics, and it does not require us to alter the core simulation code.

Citation

Eckhardt, W., Glas, R., Korzh, D., Wallner, S., & Weinzierl, T. (2016). On-the-fly memory compression for multibody algorithms. In G. R. Joubert, H. Leather, M. Parsons, F. Peters, & M. Sawyer (Eds.), Parallel computing : on the road to exascale (421-430). https://doi.org/10.3233/978-1-61499-621-7-421

Conference Name	International Conference on Parallel Computing (ParCo) 2015
Conference Location	Edinburgh, Scotland
Start Date	Sep 1, 2015
End Date	Sep 4, 2015
Acceptance Date	Sep 29, 2015
Publication Date	Apr 1, 2016
Deposit Date	Oct 1, 2015
Publicly Available Date	Nov 9, 2015
Pages	421-430
Series Title	Advances in parallel computing
Series Number	27
Series ISSN	0927-5452
Book Title	Parallel computing : on the road to exascale.
ISBN	9781614996200
DOI	https://doi.org/10.3233/978-1-61499-621-7-421
Public URL	https://durham-repository.worktribe.com/output/1152898
Publisher URL	http://ebooks.iospress.nl/volumearticle/42680
Additional Information	Conference dates: 1-4 September 2015