Stream-K
Muhammad Osama, Duane Merrill, Cris Cecka, Michael Garland, John D. Owens
Abstract
We introduce Stream-K, a work-centric parallelization of matrix multiplication (GEMM) and related computations in dense linear algebra. Whereas contemporary decompositions are primarily tile-based, our method operates by partitioning an even share of the aggregate inner loop iterations among physical processing elements. This provides a near-perfect utilization of computing resources, regardless of how efficiently the output tiling for any given problem quantizes across the underlying processing elements.
Topics & Concepts
Computer scienceMatrix multiplicationComputationParallel computingStream processingLinear algebraMultiplication (music)Concurrent computingNumerical linear algebraMatrix (chemical analysis)Matrix algebraAlgorithmTheoretical computer scienceComputational scienceMathematicsLinear systemCombinatoricsComposite materialMaterials scienceQuantum mechanicsGeometryEigenvalues and eigenvectorsQuantumPhysicsMathematical analysisParallel Computing and Optimization TechniquesMatrix Theory and AlgorithmsInterconnection Networks and Systems