Skip to main content

matmul_par

oxiblas_ndarray::parallel

Function matmul_par 

Source 
pub fn matmul_par<T>(a: &Array2<T>, b: &Array2<T>) -> Array2<T>
where
    T: Zeroable + Clone + Field + GemmKernel,
Expand description

Parallel matrix multiplication: C = A * B

Simplified parallel version that allocates a new output matrix. Uses all available Rayon threads for computation.

§Arguments

  • a - Left matrix (m x k)
  • b - Right matrix (k x n)

§Returns

New matrix C = A * B in column-major order

§Panics

Panics if inner dimensions do not match.