Forum Discussion

Altera_Forum's avatar
Altera_Forum
Icon for Honored Contributor rankHonored Contributor
16 years ago

Floating Point Matrix Multiplication

I want to multiply a 56x56 matrix with a 56x1 matrix in floating point. There is a altfp_matrix_mult megafunction which does not compute this quickly enough. I am looking for ideas to implement ...
Altera_Forum's avatar
Altera_Forum
Icon for Honored Contributor rankHonored Contributor
16 years ago

You could also try rearranging the matrix and vector and check if the implementation works on rows faster then columns. Try using BB transpose for AA and AA transpose for BB. Mathmatically, the result = AA * BB = (BB' * AA')'. See http://en.wikipedia.org/wiki/matrix_multiplication#common_properties. Depending on how your matrices are stored, transposes can be "free" in FPGAs.