G. K. Egan
ABSTRACT
The SDEM code models systems of interacting blocks of rock using the distinct element (DE) method, which represents these systems as discontinuums with each block acting under Newton's laws of motion. The data structures associated with the DE method make the task of obtaining performance gains through vectorisation difficult. Typical systems, however, contain thousands of blocks and there is the potential to perform calculations associated with groups of blocks in parallel.
This paper details the analysis and program refinement steps used in implementing a parallel version of SDEM on the Kendall Square Research KSR-1 distributed memory multiprocessor. Performance gains from a simple translation of the original Cray FORTRAN code are poor, but satisfactory performance is obtained, with minimum changes to the code, by addressing specific sources of overhead. The refinement steps focus on reducing lock costs, ensuring data locality and improving load balance.
- 1.Cray Research, CF77 Compiling System Volume 4,: Parallel Processing Guide, Cray Research, Incorporated, SG-3074 5.0, 1991.Google Scholar
- 2.Cundall, P.A., A computer model .for simulating progressive large-scale movements in blocky rock systems, Proceedings ISRM Symposium on Rock Fracture, Nancy, Vol. 1, Paper II-8. 1971.Google Scholar
- 3.Cundall, P.A. et al., Computer modelling of jointed rock masses, Technical Report N-78-4, U.S. Army Engineer Waterways Experiment Station, Vicksburg, Mississippi, 1978.Google Scholar
- 4.Egan G.K. and M.A. Coulthard, Parallel processing for the Distinct Element Method of Stress Analysis, 3rd Australian Supercomputing Conference, Melbourne, December, 1990.Google Scholar
- 5.Egan, G.K., Parallelisation of the SDEM Distinct Element Stress Analysis Code, Applications of Supercomputing in Engineering III, Computational Mechanics Publications, Elsevier Applied Science, London New York, pp 297-312, 1993.Google Scholar
- 6.Itasca, UDEC- Universal distinct element code, Version ICG1.6; User's manual. Itasca Consulting Group, Incorporated, Minneapolis, 1990.Google Scholar
- 7.Itasca 3DEC- 3-D distinct element code, Version 1.2; User's Manual, Itasca Consulting Group, Incorporated, Minneapolis, 1990.Google Scholar
- 8.K.S.R., KSR Fortran Programming, Kendall Square Research, 170 Tracer Lane, Waltham, MA, 15 Feb. 1992.Google Scholar
- 9.Lemos, J.V. A hybrid distinct element- boundary element computational model for the half-plane, M.S. Thesis, Department of Civil and Mineral Engineering, University of Minnesota, Minneapolis, 1983.Google Scholar
Index Terms
- Parallelisation of the SDEM distinct element stress analysis code on the KSR-1
Recommendations
Using an interactive parallelisation toolkit to parallelise an ocean modelling code
AbstractThis paper describes an interactive parallelisation toolkit that can be used to generate parallel code suitable for either a distributed memory system (using message passing) or a shared memory system (using OpenMP). This study focuses ...
Performance evaluation of a seismic data analysis kernel on the KSR multiprocessors
IPPS '95: Proceedings of the 9th International Symposium on Parallel ProcessingThe paper presents the results of performance analyses of a seismic analysis kernel code on the KSR multiprocessors. The purpose of such analysis is to understand the performance behaviors of a class of applications on shared memory parallel machines. ...
Comments