in these cases it is better to do a short simulation
under both sets of conditions and look more closely
at relative performance beyond how often it prints
coordinates. You can look at the information at the
end of the output file where it will tell you what part
of the simulations took a given fraction of the time.
If I understand you correctly, the simulation with moving
protein was slower than the simulation with frozen protein.
This may be reasonable since the protein-protein forces
do not need to be calculated with frozen protein. Check
the output file to see if the nonbond calculation is the
slower one. I am not sure if I understood the specifics,
though.
carlos
----- Original Message -----
From: nie beining
To: amber.scripps.edu
Sent: Monday, September 22, 2003 11:43 AM
Subject: AMBER: slower simulation
I ran 200ps MD simulation with protein fixed to equilibrate water molecules using belly command. The system contains 16,202 atoms. I am using sander of Amber 6 running simulations on single CPU of red hat linux cluster. I found that coordinates were printed out 3ps/hr after I used the restart coordinates to run simulation allowing everything to
move. The simulation without equilibrating water was printing coordinates at 7ps/hr. I am a pretty new user. Is there any reason for slower printing out? Or anyone had the same experience before? Many thanks!
------------------------------------------------------------------------------
Do you Yahoo!?
Yahoo! SiteBuilder - Free, easy-to-use web site design software
-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber.scripps.edu
To unsubscribe, send "unsubscribe amber" to majordomo.scripps.edu
Received on Mon Sep 22 2003 - 17:53:01 PDT
