Combined Iterative and Model-driven Optimization in an Automatic Parallelization Framework

Louis-Noël Pouchet (1) , Uday Bondhugula (2) , Cédric Bastoul (3) , Albert Cohen (3) , Jagannathan Ramanujam (4) , Ponnuswamy Sadayappan (1)

1 OSU - Ohio State University [Columbus]
2 IBM Watson Research Center
3 ALCHEMY - Architectures, Languages and Compilers to Harness the End of Moore Years
4 ECE - Department of Electrical and Computer Engineering - Louisiana State University

Today's multi-core era places significant demands on an optimizing compiler, which must parallelize programs, exploit memory hierarchy, and leverage the ever-increasing SIMD capabilities of modern processors. Existing model-based heuristics for performance optimization used in compilers are limited in their ability to identify profitable parallelism/locality trade-offs and usually lead to sub-optimal performance. To address this problem, we distinguish optimizations for which effective model-based heuristics and profitability estimates exist, from optimizations that require empirical search to achieve good performance in a portable fashion. We have developed a completely automatic framework in which we focus the empirical search on the set of valid possibilities to perform fusion/code motion, and rely on model-based mechanisms to perform tiling, vectorization and parallelization on the transformed program. We demonstrate the effectiveness of this approach in terms of strong performance improvements on a single target as well as performance portability across different target architectures.

Domaines: Calcul parallèle, distribué et partagé [cs.DC]; Performance et fiabilité [cs.PF]

Louis-Noël Pouchet, Uday Bondhugula, Cédric Bastoul, Albert Cohen, Jagannathan Ramanujam, et al.. Combined Iterative and Model-driven Optimization in an Automatic Parallelization Framework. Conference on Supercomputing (SC'10), Nov 2010, New Orleans, LA, United States. ⟨inria-00551067⟩