Section: New Results

Workload-aware Loop Scheduling of Irregular Loops

Participants : Pedro Henrique de Mello Morado Penna, Marcio Castro [UFSC, Universidade Federal de Santa Catarina, Brazil] , Henrique Cota de Freitas [Pontifical Catholic University of Minas Gerais, Brazil] , Francois Broquedis, Jean-Francois Mehaut.

The input workload of an irregular application must be evenly distributed among its threads to enable cutting-edge performance. To address this need in OpenMP, several loop scheduling strategies were proposed. While having this ever-increasing number of strategies at disposal is helpful, it has become a non-trivial task to select the best one for a particular application. Nevertheless, this challenge becomes easier to be tackled when existing scheduling strategies are extensively evaluated. Therefore, we present a performance and scalability evaluation of the recently proposed loop scheduling strategy named Smart Round-Robin (SRR). To deliver a comprehensive analysis, we coupled a synthetic kernel benchmarking technique with several rigorous statistical tools, and considered OpenMP's Static and Dynamic loop schedulers as our baselines. Our results unveiled that SRR performs better on irregular applications with symmetric workloads and coarse-grained parallelization, achieving up to 1.9x and 1.5x speedup over OpenMP's Static and Dynamic schedulers, respectively.

This work laid the foundations of a collaboration between CORSE, the UFSC, and PUC Minas, that led to the publication of two conference papers (ICCS'17 [20], WSCAD'17 [27]) and one international journal paper (CCPE'17 [5]). The WSCAD'17 paper has also been selected to be extended for a special issue of the CCPE journal. This extension will be based on recent works with the MHM (Multiscale Hybrid-Mixed Methods) simulator developed at LNCC for the H2020 HPC4e project 8.3.1.3 between Europe and Brazil.