DaSH Benchmark Suite


DaSH is a comprehensive benchmark suite for hybrid dataflow and shared memory programming models. It features 11 benchmarks, each representing a single Berkeley dwarf that capture patterns of communication and computation common to a wide range of emerging applications. Each benchmark includes sequential and shared-memory implementations based on OpenMP and Intel TBB to facilitate easy comparison between hybrid dataflow implementations and traditional shared memory implementations based on work-sharing and/or tasks. Currently, we also provide two dataflow implementations of each benchmark realized using Intel TBB flow graph framework and OmpSs programming model. The dataflow implementations using the Atomic Dataflow model will be included once the model has been realesed.