Skip to content

Optimization_4x4_6

Jump to bottom
Jianyu Huang edited this page Aug 11, 2016 · 4 revisions

Copy the contents of file MMult_4x4_5.c into a file named MMult_4x4_6.c and change the contents:

Change the first lines in the makefile to

OLD  := MMult_4x4_5
NEW  := MMult_4x4_6
  • make run
octave:3> PlotAll        % this will create the plot

This time the performance graph will look something like

The performance benefits from the fact that we use (regular) registers for the 4x4 block of C and the elements of the current 4x1 column of A. Notice that we are using more regular registers than there exist, so it is anyone's guess what the compiler does with that.

Clone this wiki locally