Hi Chris,

>         First of all CPU load is a silly
> measure.[...] For
> a user only wall-clock time matters.
>   
sure. I did not yet take times, but the new version is definitely faster 
than the old one. Of course 3 busy-waiting cpus and only one doing 
something useful will stress a 4-core cpu the same as 4 useful threads - 
the same in 'top', different wall clock time.

> Second, read about Amdahl's law
>         http://en.wikipedia.org/wiki/Amdahl's_law
> It matters.
>   
I know Amdahl's law, (good old parallel programming lectures) but I 
thought enblends work was somewhat simpler to split up than it seems.
> The current implementation does _not_ scale
> well beyond two processors.  The reasons for
> this are yet unknown.  You are welcome to run
> Enblend with your favorite profiler, identify
> the bottleneck(s), and send me a bunch of
> patches that rectify the congestion.
>   
Can you propose a (free) multi-thread profiler for linux? (Looking out 
for one for some time now...) If so, I'll have a look at it, enough of 
exploiting hugin/panotools now, time to give something back ;)

> IMO, you did nothing wrong.  My results are
> similar.
Fine :) I will check later how much faster enblend has become. I'm 
missing the output which image is processed (or, is finished being 
processed) a little...

Thanks for reply :)
Benjamin



--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups 
"hugin and other free panoramic software" group.
A list of frequently asked questions is available at: 
http://wiki.panotools.org/Hugin_FAQ
To post to this group, send email to [email protected]
To unsubscribe from this group, send email to 
[email protected]
For more options, visit this group at http://groups.google.com/group/hugin-ptx
-~----------~----~----~----~------~----~------~--~---

Reply via email to