Efficient implementation of multi-dimensional array redistribution

Array redistribution is required very often in programs on distributed memo ry parallel computers. It is essential to use efficient algorithms for redi stribution, otherwise the performance of programs may degrade considerably. In this paper, we focus on automatic generation of communication routines for multi-dimensional redistribution. The principal advantage of this work is to gain the ability to handle redistribution between arbitrary source an d destination processor sets and between arbitrary source and destination d istribution schemes. We have implemented these algorithms using Parallelwar e communication library. Some experimental results show the efficiency and flexibility of our techniques compared to the other redistribution works.