Are Nonblocking Networks Really Needed for High-End-Computing Workloads?
High-speed interconnects are frequently used to provide scalable communication on increasingly large high-end computing systems. Often, these networks are non-blocking, where there exist independent paths between all pairs of nodes in the system allowing for simultaneous communication with zero network contention. This performance, however, comes at a heavy cost as the number of components needed (and hence cost) increases super-linearly with the number of nodes in the system. In this paper, the authors study the behavior of real and synthetic supercomputer workloads to understand the impact of the network's non-blocking capability on overall performance.