An Overview of Symbolic Analysis Techniques Needed for the Effective Parallelization of the Perfect Benchmarks

William Blume, Rudolf Eigenmann
1994 1994 International Conference on Parallel Processing (ICPP'94)  
We have identi ed symbolic analysis techniques that will improve the e ectiveness o f p arallelizing Fortran compilers, with emphasis upon data dependence analysis. We have done this by comparing the automatically and manually parallelized versions of the Perfect Benchmarks R . T h e t e chniques include: symbolic data dependence tests for nonlinear expressions, c onstraint propagation, array summary information, and run time tests.
doi:10.1109/icpp.1994.59 dblp:conf/icpp/BlumeE94 fatcat:dbo4n2zydzhjne2isjfabhsv2u