AMBER Archive (2004)

Subject: RE: AMBER: question about parallel simulation

From: Ross Walker (ross_at_rosswalker.co.uk)
Date: Wed Dec 01 2004 - 14:40:19 CST

Dear Qiang,
 
> In order to run go model in AMBER8, I changed some parameters
> in sander.

What do you mean by you changed some parameters? What parameters? Do you
mean you actually change stuff in the code?

> However the results from single CPU and parallel versions are quite
> different.

If you have changed the source code you need to be very careful since it is
very very easy to break the parallel code unless you check where every
variable you change is allocated, filled and broadcast.

> Could anyone tell me what's wrong.

You need to provide a lot more information about what exactly you changed.
Did all the standard simulation tests cases work before you made changes?
Both in serial and in parallel?

All the best
Ross

/\
\/
|\oss Walker

| Department of Molecular Biology TPC15 |
| The Scripps Research Institute |
| Tel:- +1 858 784 8889 | EMail:- ross_at_rosswalker.co.uk |
| http://www.rosswalker.co.uk/ | PGP Key available on request |

-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber_at_scripps.edu
To unsubscribe, send "unsubscribe amber" to majordomo_at_scripps.edu