Mengjuei,
You have to focus on the numerical solver in the profiling analysis.
Initialization is indeed not significant. Also, DMP is probably too
small to be representative.
All the best,
Ray
==========================================
Ray Luo, Ph.D.
Associate Professor in Computational
Biochemistry and Molecular Biophysics
Dept of Molecular Biology and Biochemistry
Dept of Biomedical Engineering
University of California, Irvine, CA 92697-3900
Email: rluo.uci.edu Phones: (949) 824-9528, 9562
Web: http://rayl0.bio.uci.edu/rayl/
==========================================
Mengjuei Hsieh wrote:
> On Wed, Feb 10, 2010 at 08:55:48AM -0500, case wrote:
>
>> On Tue, Feb 09, 2010, Mengjuei Hsieh wrote:
>>
>>> Yet another issues for pbsa integration is that I am still doing
>>> single point calculation for every step of the nab minimization.
>>> I.e. redundant initialization and deallocation will definitely
>>> slow down the performance.
>>>
>> OK...something to have on the to-do list. But you might want to check about
>> how big the performance hit is.
>>
>
> Hi,
>
> I went to check the profile of the whole calculation, the initialization
> cost is not even significant enough to show up in the list.
> figure: http://farm5.static.flickr.com/4018/4346919309_8af9a4fc53_o.png
> So I'll just forget this, unless other people show me otherwise.
>
>
_______________________________________________
AMBER-Developers mailing list
AMBER-Developers.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber-developers
Received on Wed Feb 10 2010 - 19:00:03 PST