A technique used to distribute the workload uniformly across all processing nodes in a parallel computing system.