BLAS-like Extensions

oneMKL DPC++ provides additional routines to extend the functionality of the BLAS routines. These include routines to compute many independent matrix-matrix products.

Table “BLAS-like Extensions” lists these routines.

Routine

Data Types

Description

axpy_batch

float, double, std::complex<float>, std::complex<double>

Computes groups of vector-scalar product added to a vector.

gemm_batch

float, double, std::complex<float>, std::complex<double>

Computes groups of matrix-matrix product with general matrices.

trsm_batch

float, double, std::complex<float>, std::complex<double>

Solves a triangular matrix equation for a group of matrices.

gemm_bias

mixed int8_t, uint8_t, and int32_t

Computes a matrix-matrix product with general matrices and mixed precision.

gemmt

float, double, std::complex<float>, std::complex<double>

Computes a matrix-matrix product with general matrices, but updates only the upper/lower triangular part of the output matrix.
trsm (USM Version) axpy_batch