[gmx-users] Scaling Problems

[K.A. Feenstra]

Dallas Warren wrote:
> 
[...]
>         42,000 atoms
>         500 steps
[...]
> Results (time to perform the mdrun calculation):
>         1 CPU / 1 box - 480 seconds
>         2 CPU / 1 box - 340 seconds
>         2 CPU / 2 box - 400 seconds
>         4 CPU / 2 box - 420 seconds

Your system seems large enough to expect better scaling. I have
also seen very different results on different clusters (of similar
setup). Did you include or exclude mdrun initialization time in 
the times reported? This might be considerable compared to 400s,
especially for multi-cpu runs. You can get an estimate of this by
running 5000 steps (in stead of 500). Note that runtimes reported 
in the logfile by mdrun are not always reliable (IIRC).


-- 
Groetjes,

Anton
 ________ ___________________________________________________________
|        | Anton Feenstra                                            |
| .      | Dept. of Pharmacochemistry - Vrije Universiteit Amsterdam |
| |----  | De Boelelaan 1083 - 1081 HV Amsterdam - The Netherlands   |
| |----  | Tel: +31 20 44 47608 - Fax: +31 20 44 47610               |
| ' __   | Feenstra at chem.vu.nl - http://www.chem.vu.nl/afdelingen/FAR|
|  /  \  |-----------------------------------------------------------|
| (    ) | Dept. of Biophysical Chemistry - University of Groningen  |
|  \__/  | Nijenborgh 4 - 9747 AG Groningen - The Netherlands        |
|   __   | Tel +31 50 363 4327 - Fax +31 50 363 4800                 |
|  /  \  | K.A.Feenstra at chem.rug.nl - http://md.chem.rug.nl/~anton   |
| (    ) |-----------------------------------------------------------|
|  \__/  | "If You See Me Getting High, Knock Me Down"               |
|        | (Red Hot Chili Peppers)                                   |
|________|___________________________________________________________|