> It is a PC running Linux with two processors. As I repeated the 1 ns
> simulation a few times, the average time is about 10-11 days for dimer and
> about 2 days for monomer. There was no other job running at the same time. I
> assume it is probable due to the way GB model treats the solvent effects on
> the molecule's surface but I coulndn't give any theoretical explanations.
It could be that there is an order of N^2 cost of the GB
algorithm, or possibly the size of the problem has reached
a threshold whereby swapping is occurring. Do you detect
more disk activity with the larger system?
Are you using MPI?
Bill
-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber.scripps.edu
To unsubscribe, send "unsubscribe amber" to majordomo.scripps.edu
Received on Wed Jan 28 2004 - 18:53:01 PST
