Effects of job and task placement on parallel scientific applications performance

Abstract—this paper studies the influence that task placement may have on the performance of applications, mainly due to the relationship between communication locality and overhead. This impact is studied for torus and fat-tree topologies. A simulation-based performance study is carried out, using traces of applications and application kernels, to measure the time taken to complete one or several concurrent instances of a given workload. As the purpose of the paper is not to offer a miraculous task placement strategy, but to measure the impact that placement have on performance, we selected simple strategies, including random placement. The quantitative re-sults of these experiments show that different workloads pre-sent different degrees of responsiveness to placement. Fur-thermore, both the number of concurrent parallel jobs sharing a machine and the size of its network has a clear impact on the time to complete a given workload. We conclude that the effi-cient exploitation of a parallel computer requires the utiliza-tion of scheduling policies aware of application behavior and network topology. Keywords-interconnection networks; parallel job scheduling; performance characterization; resource allocation; trace-driven simulation. I