On Wed, Mar 25, 2009 at 5:39 PM, Edward d'Auvergne
<edward.dauvergne@xxxxxxxxx> wrote:
Done! Cheers.
Edward
Hi Ed
I have sent some changes to you for the poster If they are to late it doesn't matter as they are only minor
sorry for the late reply it all looks very good! A couple of things
1. here are some better graphs
They show scaling in the same way but the bottom one shows the total
expected speedup achieved which doesn't give the rather odd wiggles as
you get to smaller numbers of processors and the accuracy in the timing
gets worse.
2. My address is Astbury Centre for Structural Molecular biology
University of Leeds
3. through the MPI messaging interface and other parallel libraries via
a simple plug-in architecture...
4. some possibly better wordings
a. para 1 Is controlled by special user functions either by the built in
python prompt or through the flexibility of pythons fully featured
scripting language.
b. execute them in-line execute them as a sub-process and then read the
output. This allows relax to use the following programs as replacements
for the built-in optimization engine
c. The 1.3 relax versions
At three years old...
5. I don't think I will do better than in preparation
regards
gary
Re: ENC 2009 poster.