A combination technique enabling parallelization of nested loops with non-uniform dependences

Although several methods exist in order to parallelize loops with non-uniform dependences, most of them perform poorly due to irregular and complex dependence constraints. This paper proposes efficient methods such as the improved tiling method for flow dependence and the improved region partitioning method for maximizing parallelization of nested loops with non-uniform dependences. A combination of existing methods and our proposed methods exposes more parallelism.