enblend starts 4 threads on my 4 core MacPro and does very briefly
reach a CPU utilisation of 350%. The whole process is so fast that it
is difficult get a good average result from the activity monitor.
Blending 16 x 72MB (2800x4300 16bit) images takes a little more than
1.5 real minutes and 4 cpu minutes. This corresponds to 2.5 effective
cores for the whole process. And best off all: no crashes.

Martin

On Jan 22, 6:56 pm, Harry van der Wolf <[email protected]> wrote:
> 2012/1/22 Martin Middelhoek <[email protected]>
>
>
>
>
>
>
>
>
>
> > It says:
>
> > Extra feature: OpenMP: yes
> >  - version 2008-5
> >  - support for nested parallelism;
> >    nested parallelism disabled by default
> >  - support for dynamic adjustment of the number of threads;
> >    dynamic adjustment disabled by default
> >  - using 4 processors and up to 4 threads
>
> > It is pretty fast, so maybe 4 cores is just overkill. I will increase
> > the sampling rate of the activity monitor and try again.
>
> > I also notice that the GPU acceleration is enabled by default. Is that
> > OpenCL? In that case it may be that my 4870 gets part of the workload.
>
> GPU accelaration is only used for a very small part of the calculations and
> it's very experimental. You can enable it by adding --gpu to the enblend
> options.
> There's a great possibility that ti crashes, but try for yourself.
>
> Harry

-- 
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