Advanced code generation for high performance Fortran

• Authors: Vikram Adve;John Mellor-Crummey
• Venue: Compiler optimizations for scalable parallel systems
• Year: 2001

For data-parallel languages such as High Performance Fortran to achieve wide acceptance, parallelizing compilers must be able to provide consistently high performance for a broads pectrum of scientific applications. Although compilation of regular data-parallel applications for message-passing systems have been widely studied, current state-of-the-art compilers implement only a small number of key optimizations, and the implementations generally focus on optimizing programs using a "case-based" approach. For these reasons, current compilers are unable to provide consistently high levels of performance. In this paper, we describe techniques developed in the Rice dHPF compiler to address key code generation challenges that arise in achieving high performance for regular applications on message-passing systems. We focus on techniques requiredto implement advanced optimizations and to achieve consistently high performance with existing optimizations. Many of the core communication analysis andco de generation algorithms in dHPF are expressed in terms of abstract equations manipulating integer sets. This approach enables general andy et simple implementations of sophisticated optimizations, making it more practical to include a comprehensive set of optimizations in data-parallel compilers. It also enables the compiler to support much more aggressive computation partitioning algorithms than in previous compilers. We therefore believe this approach can provide higher and more consistent levels of performance than are available today.