memory

Dear Lammps users,
my simulation on new PC used up all the memory increasingly to 8GB. I did not have the problem on my lap top, but on the pc that I am running without calculating task bar shows lmp_serial uses ram increasingly.
what could be the problem?
thanks

Dear Lammps users,
my simulation on new PC used up all the memory increasingly to 8GB. I did
not have the problem on my lap top, but on the pc that I am running without
calculating task bar shows lmp_serial uses ram increasingly.
what could be the problem?

​impossible to say in such generality.

what LAMMPS version are you using? and on what platform/OS?

do you see this effect with any of the examples​ bundled with LAMMPS?
does you LAMMPS executable include custom code?

axel.

I am using lammps-64bit-20170127, and I tried before that the latest version, but both of them have this issue. I downloaded them from rpm website and there is no custom code.
I am using windows 8.
the input and data file that I am using would work on my lap top( I cannot remember the version of lammps on my lap top, but it would work).
Actually, I can see on my task manager that from the start lmp_serial is using 1 GB ram and it increases to 7GB and maybe more.
Thanks

I am using lammps-64bit-20170127, and I tried before that the latest
version, but both of them have this issue. I downloaded them from rpm
website and there is no custom code.
I am using windows 8.
the input and data file that I am using would work on my lap top( I cannot
remember the version of lammps on my lap top, but it would work).
Actually, I can see on my task manager that from the start lmp_serial is
using 1 GB ram and it increases to 7GB and maybe more.

​memory usage can increase over run time depending on the LAMMPS features
that you are using and how you are writing your input.
thus the first step to identify what the issue i would be to pick an input
from the LAMMPS examples that is somewhat similar to your production run,
increase the number of time steps, so that it is running long enough so you
would detect the increased memory use reliably, and then report back.

axel.