Function gpu_matmul 

pub fn gpu_matmul(
    a_data: Vec<f64>,
    a_rows: usize,
    a_cols: usize,
    b_data: Vec<f64>,
    b_cols: usize,
) -> PyResult<Vec<f64>>

Multiply two row-major matrices: C = A (m×k) × B (k×n).

Returns the product as a flat Vec<f64> of length m * n in row-major order together with the output shape (m, n).

Arguments

• a_data – flat row-major elements of A, length must equal a_rows * a_cols
• a_rows – number of rows in A (= m)
• a_cols – number of columns in A (= k)
• b_data – flat row-major elements of B, length must equal a_cols * b_cols
• b_cols – number of columns in B (= n)

Errors

Returns PyValueError when any length constraint is violated.