AMBER Archive (2008)

Subject: Re: AMBER: sgi4700 the time of MD simulation is 30 minute for 1ps about 2500 atom using 1cpu

From: steinbrt_at_rci.rutgers.edu
Date: Sat Dec 13 2008 - 12:58:18 CST

Hi,

you are running a GB calculation with esentially unlimited cutoff. Due to
the large number of pair-interactions, GB can be fairly slow for large
systems. Check the timings, 100% of your calculation is spend doing the GB
part. Did you compare your results to a simulation of the same system in a
waterbox with a reasonable cutoff?

If your wondering if your installation is ok, check if the Amber
benchmarks give you good timings (the amber webpage gives some for
comparison I think).

On Sat, December 13, 2008 9:36 am, llduan_at_itcc.nju.edu.cn wrote:
> Dear amber committee:
> The MD simulation is proformed using AMBER10 at sgi4700, but the time
> is too long for 1ps about 30 miniute using 1cpu and the atom number is
> 2500. attached is the input and output files. Thank you very much.
>

Dr. Thomas Steinbrecher
BioMaps Institute
Rutgers University
610 Taylor Rd.
Piscataway, NJ 08854
-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber_at_scripps.edu
To unsubscribe, send "unsubscribe amber" (in the *body* of the email)
      to majordomo_at_scripps.edu