TY - JOUR
T1 - Practical Considerations of Parallel Simulations and Architecture Independent Parallel Algorithm Design
AU - Gerbessiotis, Alexandros V.
N1 - Funding Information:
This research was primarily conducted while the author was with the Aiken Computation Laboratory, Harvard University, Cambridge, MA 02138, and supported in part by the National Science Foundation under Grants NSF-CCR-89-02500 and NSF-CCR-9200884. The experimental work presented in this paper was carried out at the Computing Laboratory, The University of Oxford, Oxford, UK, while the author was supported in part by the Engineering and Physical Sciences Research Council (EPSRC) under Grant GR K16999. The support of the Edinburgh Parallel Computing Centre in granting access to a Cray T3D is gratefully acknowledged.
PY - 1998/8/25
Y1 - 1998/8/25
N2 - We examine combinatorial properties of a class of hash functions and its application to the simulations of classical models of parallel computation on other models, such as theBSPand theS*PRAM, optimally in communication to within additive lower order terms. The BSP model can serve as a programming paradigm as well; we also examine the implications of architecture independent parallel algorithm design in the context of the BSP model and show how it can lead to portable and scalable implementations of algorithms that can work on a multiplicity of hardware platforms with only recompilation of the source program code. Toward this end, dense Cholesky factorization algorithms are presented and their performance on three parallel hardware platforms, an SGI Power Challenge, IBM SP2, and Cray T3D, is examined and analyzed.
AB - We examine combinatorial properties of a class of hash functions and its application to the simulations of classical models of parallel computation on other models, such as theBSPand theS*PRAM, optimally in communication to within additive lower order terms. The BSP model can serve as a programming paradigm as well; we also examine the implications of architecture independent parallel algorithm design in the context of the BSP model and show how it can lead to portable and scalable implementations of algorithms that can work on a multiplicity of hardware platforms with only recompilation of the source program code. Toward this end, dense Cholesky factorization algorithms are presented and their performance on three parallel hardware platforms, an SGI Power Challenge, IBM SP2, and Cray T3D, is examined and analyzed.
UR - http://www.scopus.com/inward/record.url?scp=0006412244&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=0006412244&partnerID=8YFLogxK
U2 - 10.1006/jpdc.1998.1477
DO - 10.1006/jpdc.1998.1477
M3 - Article
AN - SCOPUS:0006412244
SN - 0743-7315
VL - 53
SP - 1
EP - 25
JO - Journal of Parallel and Distributed Computing
JF - Journal of Parallel and Distributed Computing
IS - 1
ER -