Comment by inglor

Can you explain why you did the naive algorithm here and not any of the fast matrix multiplication ones that trade multiplications for more additions? Just for educational purposes or is there a performance benefit in the technique?