On Wed, 28 Mar 2012 17:31:47 +0200
Albert wrote:
> I am using gromacs for membrane simulation (under CHARMM36 FF) which
> contains around 80,000 atoms. I've submitted over 200 CPU in the
> cluster for such system with 2 fs time step. And what really
> astonished is that the efficiency for such
Hi Albert,
You are doing neighboursearching every step! So every step all 200
CPUs need to know the how and what of all the other 199. Imagine the
communication overhead. Furthermore, you have 400 atoms per CPU
(neglecting the PME dedication). That will also make communication a
bottle neck. Which
Albert wrote:
Dear:
I am using gromacs for membrane simulation (under CHARMM36 FF) which
contains around 80,000 atoms. I've submitted over 200 CPU in the cluster
for such system with 2 fs time step. And what really astonished is that
the efficiency for such simulation is only 3ns/day.
Dear:
I am using gromacs for membrane simulation (under CHARMM36 FF) which
contains around 80,000 atoms. I've submitted over 200 CPU in the cluster
for such system with 2 fs time step. And what really astonished is that the
efficiency for such simulation is only 3ns/day. I am wondering what
4 matches
Mail list logo