Confused about “Matrix Multiplication“ in tutorial

[HonestDeng]

In the tutorial, the mattrix multiplication is described as:

# Do in parallel
for m in range(0, M, BLOCK_SIZE_M):
  # Do in parallel
  for n in range(0, N, BLOCK_SIZE_N):
    acc = zeros((BLOCK_SIZE_M, BLOCK_SIZE_N), dtype=float32)
    for k in range(0, K, BLOCK_SIZE_K):
      a = A[m : m+BLOCK_SIZE_M, k : k+BLOCK_SIZE_K]
      b = B[k : k+BLOCK_SIZE_K, n : n+BLOCK_SIZE_N]
      acc += dot(a, b) # why is "dot"? 
    C[m : m+BLOCK_SIZE_M, n : n+BLOCK_SIZE_N] = acc

In the second last line code, sub-matrix a is BLOCK_SIZE_M by BLOCK_SIZE_K, and sub-matrix b is BLOCK_SIZE_K by BLOCK_SIZE_N. Their shapes is not compatible to perform "dot" application. Why?

[HonestDeng]

Ok. I find out that "dot" of triton is not just "dot product" of 2 vectors or element-wise matrix multiplication. When the the inputs of dot is 2D, "dot" actually performs standard matrix mulplication.