code not match the problem description in matrix-core blog ?

[ZJLi2013]

hi, AMDer, I am recently read the blog matrix core, looks the problem size describe here is :

Input:
   A :  16X4 floats 
   B :  4x16 floats 
Output:
    D: 16x16 floats 

while in the 2nd example, matrix A [16, 4], B[4, 16], looks the code is directly copied from matrix-core/samples, which has a different matrix shape :

Input:
  A : 16 x 16 x 4 floats (four 16x16 matrices)
  B : 16 x 16 x 4 floats (four 16x16 matrices)

Output:
  D : 16 x 16 x 4 floats (four 16x16 matrices)

thanks for confirmation

[saadrahim]

@jychang48 can your team look into this?