Optimization_4x4_8

Copy the contents of file MMult_4x4_7.c into a file named MMult_4x4_8.c and change the contents:

|| from || to || ||<^> -Include(HowToOptimizeGemm/Details/MMult_4x4_7)- ||<^> -Include(HowToOptimizeGemm/Details/MMult_4x4_8)- ||

Change the first lines in the makefile to {{{ OLD := MMult_4x4_7 NEW := MMult_4x4_8 }}}

• make run {{{ octave:3> PlotAll % this will create the plot }}}

This time the performance graph will look something like

ImageLink(http://www.cs.utexas.edu/users/rvdg/HowToOptimizeGemm/Graphs/compare_MMult-4x4-7_MMult-4x4-8.png,http://www.cs.utexas.edu/users/rvdg/HowToOptimizeGemm/Graphs/compare_MMult-4x4-7_MMult-4x4-8.png,width=40%)

We now use registers to store the elements of the current row of B. (Notice that we did not do this for the case where we computed C four elements at a time.) The performance actually drops slightly. But this step enables further optimizations.