GEMM

General Matrix-Matrix-Multiplication. More...

Collaboration diagram for GEMM:

Functions
template<std::floating_point Float>
auto	tatooine::blas::gemm (op TRANSA, op TRANSB, int M, int N, int K, Float ALPHA, Float const *A, int LDA, Float const *B, int LDB, Float BETA, Float *C, int LDC) -> void
template<std::floating_point Float, std::size_t M, std::size_t N, std::size_t K>
auto	tatooine::blas::gemm (Float const alpha, tensor< Float, M, K > const &A, tensor< Float, K, N > const &B, Float const beta, tensor< Float, M, N > &C)
	See GEMM. More...
template<std::floating_point Float>
auto	tatooine::blas::gemm (blas::op trans_A, blas::op trans_B, Float const alpha, tensor< Float > const &A, tensor< Float > const &B, Float const beta, tensor< Float > &C)
	See GEMM. More...
template<std::floating_point Float>
auto	tatooine::blas::gemm (Float const alpha, tensor< Float > const &A, tensor< Float > const &B, Float const beta, tensor< Float > &C)
	See GEMM. More...

Detailed Description

General Matrix-Matrix-Multiplication.

GEMM performs one of the matrix-matrix operations

\[\mC=\alpha\cdot\text{op}(\mA)\cdot \text{op}(\mB) + \beta\cdot\mC\]

where \(\text{op}(\mX)\) is one of

\[\text{op}(\mX)=\mX\ \ \text{ or }\ \ \text{op}(\mX)=\mX^\top\]

.

\(\alpha\) and \(\beta\) are scalars, and \(\mA\), \(\mB\) and \(\mC\) are matrices, with \(\text{op}(\mA)\) an \(m\times k\) matrix, \(\text{op}(\mB)\) a \(k\times n\) matrix and \(\mC\) an \(m\times n\) matrix.

• BLAS documentation

Function Documentation

gemm() [1/4]

template<std::floating_point Float>

auto tatooine::blas::gemm	(	blas::op	trans_A,
		blas::op	trans_B,
		Float const	alpha,
		tensor< Float > const &	A,
		tensor< Float > const &	B,
		Float const	beta,
		tensor< Float > &	C
	)

See GEMM.

gemm() [2/4]

template<std::floating_point Float>

auto tatooine::blas::gemm	(	Float const	alpha,
		tensor< Float > const &	A,
		tensor< Float > const &	B,
		Float const	beta,
		tensor< Float > &	C
	)

See GEMM.

gemm() [3/4]

template<std::floating_point Float, std::size_t M, std::size_t N, std::size_t K>

auto tatooine::blas::gemm	(	Float const	alpha,
		tensor< Float, M, K > const &	A,
		tensor< Float, K, N > const &	B,
		Float const	beta,
		tensor< Float, M, N > &	C
	)

See GEMM.

gemm() [4/4]

template<std::floating_point Float>

auto tatooine::blas::gemm	(	op	TRANSA,
		op	TRANSB,
		int	M,
		int	N,
		int	K,
		Float	ALPHA,
		Float const *	A,
		int	LDA,
		Float const *	B,
		int	LDB,
		Float	BETA,
		Float *	C,
		int	LDC
	)		-> void