Class LibMatrixCuMatMult

    • Constructor Detail

      • LibMatrixCuMatMult

        public LibMatrixCuMatMult()
    • Method Detail

      • matmult

        public static MatrixObject matmult​(ExecutionContext ec,
                                           GPUContext gCtx,
                                           String instName,
                                           MatrixObject left,
                                           MatrixObject right,
                                           String outputName,
                                           boolean isLeftTransposed,
                                           boolean isRightTransposed)
        Matrix multiply on GPU Examines sparsity and shapes and routes call to appropriate method from cuBLAS or cuSparse C = op(A) x op(B) The user is expected to call ec.releaseMatrixOutputForGPUInstruction(outputName);
        Parameters:
        ec - Current ExecutionContext instance
        gCtx - a valid GPUContext
        instName - name of the invoking instruction to recordStatistics.
        left - Matrix A
        right - Matrix B
        outputName - Name of the output matrix C (in code generated after LOP layer)
        isLeftTransposed - op for A, transposed or not
        isRightTransposed - op for B, tranposed or not
        Returns:
        output of matrix multiply