An accumulative parallel skeleton for all

Abstract

Parallel skeletons intend to encourage programmers to builda parallel program from ready-made components for which efficient implementation sare known to exist, making the parallelization process simpler.However, it is neither easy to develop efficient parallel program susing skeletons nor to use skeletons to manipulate irregular data, and moreover there lacks a systematic way to optimize skeletal parallel programs.To remedy this situation, we propose a novel parallel skeleton,called accumulate, which not only efficiently describes data dependencyin computation but also exhibits nice algebraic properties for manipulation.We show that this skeleton significantly eases skeletal parallel programming in practice, efficiently manipulating both regular and irregulardata, and systematically optimizing skeletal parallel programs.