Trait collenchyma_blas::plugin::Gemm [−] [src]

pub trait Gemm<F> {
    fn gemm(&self, alpha: &mut SharedTensor<F>, at: Transpose, a: &mut SharedTensor<F>, bt: Transpose, b: &mut SharedTensor<F>, beta: &mut SharedTensor<F>, c: &mut SharedTensor<F>) -> Result<(), Error>;
    fn gemm_plain(&self, alpha: &SharedTensor<F>, at: Transpose, a: &SharedTensor<F>, bt: Transpose, b: &SharedTensor<F>, beta: &SharedTensor<F>, c: &mut SharedTensor<F>) -> Result<(), Error>;
}

Provides the gemm operation.

Required Methods

fn gemm(&self, alpha: &mut SharedTensor<F>, at: Transpose, a: &mut SharedTensor<F>, bt: Transpose, b: &mut SharedTensor<F>, beta: &mut SharedTensor<F>, c: &mut SharedTensor<F>) -> Result<(), Error>

Computes a matrix-matrix product with general matrices.

Saves the result into c. This is a Level 3 BLAS operation.

For a no-memory managed version see gemm_plain.

fn gemm_plain(&self, alpha: &SharedTensor<F>, at: Transpose, a: &SharedTensor<F>, bt: Transpose, b: &SharedTensor<F>, beta: &SharedTensor<F>, c: &mut SharedTensor<F>) -> Result<(), Error>

Computes a matrix-matrix product with general matrices.

Saves the result into c. This is a Level 3 BLAS operation.

Attention:
For a correct computation result, you need to manage the memory allocation and synchronization yourself.
For a memory managed version see gemm.

Implementors