This paper is concerned with the parallel implementation of two splitting methods for solving block tridiagonal linear systems on a transputer network. For both methods, we describe the data allocation, the workload distribution and the interprocessor communications and we evaluate the effectiveness of the two parallel algorithms by numerical experiments on a set of test problems.
Index Terms:
parallel algorithms; linear algebra; transputer systems; splitting methods implementation; block tridiagonal linear systems; transputers; data allocation; workload distribution; interprocessor communications; parallel algorithms
Citation:
E. Galligani, V. Ruggiero, "Implementation of splitting methods for solving block tridiagonal linear systems on transputers," pdp, pp.409, 3rd Euromicro Workshop on Parallel and Distributed Processing, 1995