rocblas_gemm_ext2 Interface Reference#
hipfort_rocblas::rocblas_gemm_ext2 Interface Reference
! More...
Public Member Functions | |
| integer(kind(rocblas_status_success)) function | rocblas_gemm_ext2_ (handle, m, n, k, alpha, a, a_type, row_stride_a, col_stride_a, b, b_type, row_stride_b, col_stride_b, beta, c, c_type, row_stride_c, col_stride_c, d, d_type, row_stride_d, col_stride_d, compute_type, algo, solution_index, flags) |
Detailed Description
!
BLAS EX API
\details
GEMM_EXT2 performs the matrix-matrix operations
D = alpha A B + beta C,
alpha and beta are scalars, and A, B, C, and D are matrices, with A a m by k
matrtix, B a k by n matrix, and C and D are m by n matrices. Each matrix A, B, C, D
has independent row and column strides.
This is a beta feature.
@param[in]
handle [rocblas_handle]
handle to the rocblas library context queue.
@param[in]
m [rocblas_int]
matrix dimension m.
@param[in]
n [rocblas_int]
matrix dimension n.
@param[in]
k [rocblas_int]
matrix dimension k.
@param[in]
alpha [ void ]
device pointer or host pointer specifying the scalar alpha. Same datatype as compute_type.
@param[in]
a [void ]
device pointer storing matrix A.
@param[in]
a_type [rocblas_datatype]
specifies the datatype of matrix A.
@param[in]
row_stride_a [rocblas_int]
specifies the row stride of A.
@param[in]
col_stride_a [rocblas_int]
specifies the column stride of A.
@param[in]
b [void ]
device pointer storing matrix B.
@param[in]
b_type [rocblas_datatype]
specifies the datatype of matrix B.
@param[in]
row_stride_b [rocblas_int]
specifies the row stride of B.
@param[in]
col_stride_a [rocblas_int]
specifies the column stride of B.
@param[in]
beta [ void ]
device pointer or host pointer specifying the scalar beta. Same datatype as compute_type.
@param[in]
c [void ]
device pointer storing matrix C.
@param[in]
c_type [rocblas_datatype]
specifies the datatype of matrix C.
@param[in]
row_stride_c [rocblas_int]
specifies the row stride of C.
@param[in]
col_stride_c [rocblas_int]
specifies the column stride of C.
@param[out]
d [void ]
device pointer storing matrix D.
@param[in]
d_type [rocblas_datatype]
specifies the datatype of matrix D.
@param[in]
row_stride_d [rocblas_int]
specifies the row stride of D.
@param[in]
col_stride_d [rocblas_int]
specifies the column stride of D.
@param[in]
compute_type
[rocblas_datatype]
specifies the datatype of computation.
@param[in]
algo [rocblas_gemm_algo]
enumerant specifying the algorithm type.
@param[in]
solution_index
[int32_t]
reserved for future use.
@param[in]
flags [uint32_t]
optional gemm flags.
Member Function/Subroutine Documentation
◆ rocblas_gemm_ext2_()
| integer(kind(rocblas_status_success)) function hipfort_rocblas::rocblas_gemm_ext2::rocblas_gemm_ext2_ | ( | type(c_ptr), value | handle, |
| integer(c_int), value | m, | ||
| integer(c_int), value | n, | ||
| integer(c_int), value | k, | ||
| type(c_ptr), value | alpha, | ||
| type(c_ptr), value | a, | ||
| integer(kind(rocblas_datatype_f16_r)), value | a_type, | ||
| integer(c_int64_t), value | row_stride_a, | ||
| integer(c_int64_t), value | col_stride_a, | ||
| type(c_ptr), value | b, | ||
| integer(kind(rocblas_datatype_f16_r)), value | b_type, | ||
| integer(c_int64_t), value | row_stride_b, | ||
| integer(c_int64_t), value | col_stride_b, | ||
| type(c_ptr), value | beta, | ||
| type(c_ptr), value | c, | ||
| integer(kind(rocblas_datatype_f16_r)), value | c_type, | ||
| integer(c_int64_t), value | row_stride_c, | ||
| integer(c_int64_t), value | col_stride_c, | ||
| type(c_ptr), value | d, | ||
| integer(kind(rocblas_datatype_f16_r)), value | d_type, | ||
| integer(c_int64_t), value | row_stride_d, | ||
| integer(c_int64_t), value | col_stride_d, | ||
| integer(kind(rocblas_datatype_f16_r)), value | compute_type, | ||
| integer(kind(rocblas_gemm_algo_standard)), value | algo, | ||
| integer(c_int32_t), value | solution_index, | ||
| integer(kind=4), value | flags | ||
| ) |
The documentation for this interface was generated from the following file: