Optimizing Fortran 90 Shift Operations on Distributed-Memory Multicomputers

When executing Fortran 90 style data-parallel array operations on distributed-memory multiprocessors, intraprocessor data movement due to shift operations can account for a signiicant fraction of the execution time. This paper describes a strategy for minimizing data movement caused by Fortran 90 CSHIFT operations and presents a compiler technique that exploits this strategy automatically. The compiler technique is global in scope and can reduce data movement even when a deenition of an array and its uses are separated by control ow. This technique supersedes those whose scope is restricted to a single statement. We focus on the application of this strategy on distributed-memory architectures, although it is more broadly applicable.