Contents

namespace

SushiBLAS::Kernels::SYCL

Declared in
include/SushiBLAS/kernels/sycl/blas/level3/gemm.hpp

Functions

template <typename T, typename Epilogue = GemmIdentityEpilogue>
sycl::event gemm(sycl::queue &queue, Core::Layout layout, bool transA, bool transB, int64_t m, int64_t n, int64_t k, T alpha, const T *a, int64_t lda, int64_t str_a, const T *b, int64_t ldb, int64_t str_b, T beta, T *c, int64_t ldc, int64_t str_c, int64_t batch_size, const std::vector< sycl::event > &deps, Epilogue epilogue=Epilogue{})

Computes C = Epilogue(alpha * op(A) * op(B) + beta * C) with the kernel the profile selects.

Template parameters

Epilogue

Functor T operator()(T value, int64_t row, int64_t col) const, applied before a store.

Parameters

beta

Exactly zero skips reading C.

c

Must not overlap A or B.

See also

include/SushiBLAS/kernels/README.md