DocumentCode :
2927684
Title :
Parallel programming in multi-paradigm clusters
Author :
Leichtl, Johann ; Crandal, Phyllis E. ; Clement, Mark J.
Author_Institution :
Dept. of Comput. Sci. & Eng., Connecticut Univ., Storrs, CT, USA
fYear :
1997
fDate :
5-8 Aug 1997
Firstpage :
326
Lastpage :
335
Abstract :
An important development in cluster computing is the availability of multiprocessor workstations. These are able to provide additional computational power to the cluster without increasing network overhead and allow multiparadigm parallelism, which we define to be the simultaneous application of both distributed and shared memory parallel processing techniques to a single problem. In this paper we compare execution times and speedup of parallel programs written in a pure message-passing paradigm with those that combine message passing and shared-memory primitives in the same application. We consider three basic applications that are common building blocks for many scientific and engineering problems: numerical integration, matrix multiplication and Jacobi iteration. Our results indicate that the added complexity of combining shared- and distributed-memory programming methods in the same program does not contribute sufficiently to performance to justify the added programming complexity
Keywords :
computational complexity; matrix multiplication; message passing; parallel processing; parallel programming; Jacobi iteration; cluster computing; distributed memory parallel processing; matrix multiplication; message passing; message-passing paradigm; multi-paradigm clusters; multiparadigm parallelism; multiprocessor workstations; network overhead; numerical integration; parallel programming; programming complexity; shared memory parallel processing; shared-memory primitives; Computer networks; Computer science; Concurrent computing; Distributed computing; Jacobian matrices; Libraries; Parallel processing; Parallel programming; Power engineering computing; Workstations;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
High Performance Distributed Computing, 1997. Proceedings. The Sixth IEEE International Symposium on
Conference_Location :
Portland, OR
ISSN :
1082-8907
Print_ISBN :
0-8186-8117-9
Type :
conf
DOI :
10.1109/HPDC.1997.626438
Filename :
626438
Link To Document :
بازگشت