Hello LAMMPS community,
I am trying to calculate SAED from a system of 2,992,400 atoms (orthogonal box).
I am consistently getting either ‘unable enough to allocate enough memory’ error or an empty output .vtk file. Has anyone experienced a similar problem?
Thank you,
My input script:
units metal
atom_style atomic
read_data lmp_data_2.in
neighbor 0.5 bin
neigh_modify delay 5 every 1
compute 1 all saed 0.0001 C Kmax 8.0 Zone 0 0 1 dR_Ewald 0.01 c 5.0 5.0 5.0
fix 1 all saed/vtk 1 1 1 c_1 file out.saed
When running a calculation with a large number of atoms, it is quite possible that you run out of memory unless you are using a sufficient number of MPI processes, so the data can be distributed.
Compute saed is a special case, since the memory consumption is in addition also determined by the settings for reciprocal space. Because of the relation of real space to reciprocal space, you may not get a lot of useful information, if you simulation cell is large.
Please note that you are using a fix, but do have any command like run that will actually trigger its execution. So the fact that you get an empty output file is not a surprise.
1 Like