Automatic Co-scheduling Based on Main Memory Bandwidth Usage

Most applications running on supercomputers achieve only a fraction of a system’s peak performance. It has been demonstrated that co-scheduling applications improves overall system utilization. In this case, however, applications being co-scheduled need to fulfill certain criteria such that mutual slowdown is kept at a minimum. In this paper we present a set of libraries and a first HPC scheduler prototype that automatically detects an application’s main memory bandwidth utilization and prevents the co-scheduling of multiple main memory bandwidth limited applications. We demonstrate that our prototype achieves almost the same performance as we achieved with manually tuned co-schedules in previous work.