=================================================
More fine-tuning of the Unrolled Assembler Kernel                        [TOC]
=================================================

In the last implementation the pointer increments were done direclty after each
other:

---- CODE(type=c) --------------------------------------------------------------
    "addq      $4*32,     %%rax    \n\t"  // A += 16;
    "addq      $4*32,     %%rbx    \n\t"  // B += 16;
--------------------------------------------------------------------------------

These instructions compete for the same execution unit of the CPU.  Pulling
these apart futher improves performance.


Select the demo-sse-asm-unrolled-v3 Branch
==========================================
Again, we do a `make clean` before switching a branch:

    *--[SHELL(height=6)]------------------------------------------------*
    |                                                                   |
    |  cd ulmBLAS                                                       |
    |  make clean                                                       |
    |                                                                   |
    *-------------------------------------------------------------------*

Then we are checking out the `demo-sse-asm-unrolled-v3` branch:

    *--[SHELL(path=ulmBLAS)]--------------------------------------------*
    |                                                                   |
    |  git branch -a                                                    |
    |  git checkout -B demo-sse-asm-unrolled-v3                      +++|
    |                  remotes/origin/demo-sse-asm-unrolled-v3          |
    |                                                                   |
    *-------------------------------------------------------------------*

Then we compile the project

    *--[SHELL(path=ulmBLAS,height=15)]----------------------------------*
    |                                                                   |
    |  make                                                             |
    |                                                                   |
    *-------------------------------------------------------------------*


The dgemm_nn Code
=================
:import: ulmBLAS/src/level3/dgemm_nn.c [linenumbers]


Benchmark Results
=================
We run the benchmarks

    *--[SHELL(path=ulmBLAS)]--------------------------------------------*
    |                                                                   |
    |  cd bench                                                         |
    |  ./xdl3blastst > report                                           |
    |                                                                   |
    *-------------------------------------------------------------------*

and filter out the results for the `demo-sse-asm-unrolled-v3`
branch:

    *--[SHELL(path=ulmBLAS/bench)]--------------------------------------*
    |                                                                   |
    |  grep PASS report > demo-sse-asm-unrolled-v3                      |
    |                                                                   |
    *-------------------------------------------------------------------*

With the gnuplot script

:import: ulmBLAS/bench/bench12.gps

we feed gnuplot

    *--[SHELL(path=ulmBLAS/bench)]--------------------------------------*
    |                                                                   |
    |  gnuplot bench12.gps                                              |
    |                                                                   |
    *-------------------------------------------------------------------*

and get

    ---- IMAGE --------------
    ulmBLAS/bench/bench12.svg
    -------------------------

:navigate: __up__    -> doc:index
           __back__  -> doc:page10/index
           __next__  -> doc:page12/index