[ptx] x64 benchmarks

[Daniel M. German]

Hi Pablo,


Any chance that you can run the profiler on your tests? Also, what
format did you use as output? I assume it was TIFF_m with no colour
correction.

daniel



 Pablo d'Angelo twisted the bytes to say:

 Pablo> Marek Januszewski schrieb:
 >> Hello,
 >> 
 >> So how about we set up some big benchmarking project and we all post
 >> results here?

 Pablo> I always had the feeling that the AMD64 bit version run a bit faster.

 Pablo> While I haven't set up a complete benchmark project, but I have remapped a
 Pablo> single image from a project (with lots of black, empty space). No crop
 Pablo> (nona) or acceleration (PTmender) was used.

 Pablo> This are the user times:
             x86 32 bit (s)    AMD64 (s)      speedup
 Pablo> PTStitcher	35.495
 Pablo> PTmender	35.675		28.995		23%
 Pablo> nona            27.588		22.179		24%

 Pablo> So at least for panoramic stitching, actually using the 64 bit instructions
 Pablo> on that the AMD64 provides seems to improve remapping speed by ~ 20%

 Pablo> Note: if fast transform or cropped tiff output would have been used, the
 Pablo> times would differ a lot. For smaller panos, the fast transform code used by
 Pablo> PTStitcher and PTmender would blow nona away, while for larger panos, nona
 Pablo> would be faster

 Pablo> If there is interest, I can provide the example. However, for very fast
 Pablo> machine, the IO bandwith will probably the limiting factor

 Pablo> Details:

 Pablo> I'm running ubuntu breezy on and AMD64 (almost 2 years old) machine with 1.5
 Pablo> GB of memory. All programs have been run during the same session. cpu
 Pablo> frequency was fixed to 2000 MHz during the test.

 >> gcc -v
 Pablo> Target: x86_64-linux-gnu
 Pablo> gcc-Version 4.0.2 20050808 (prerelease) (Ubuntu 4.0.1-4ubuntu9)

 >> gcc -v
 Pablo> Target: i486-linux-gnu
 Pablo> gcc version 4.0.2 20050808 (prerelease) (Ubuntu 4.0.1-4ubuntu9)


 Pablo> This is what /proc/cpu said about my processor:

 Pablo> vendor_id       : AuthenticAMD
 Pablo> cpu family      : 15
 Pablo> model           : 4
 Pablo> model name      : AMD Athlon(tm) 64 Processor 3000+
 Pablo> stepping        : 8
 Pablo> cpu MHz         : 2002.608
 Pablo> cache size      : 512 KB

 Pablo> ciao
  Pablo

--
Daniel M. German                  "And the world, to each individual,
                                   means the part of it with which
   John Stuart Mill ->             he comes in contact."
http://turingmachine.org/
http://silvernegative.com/
dmg (at) uvic (dot) ca
replace (at) with @ and (dot) with .