Lammps 2020-5-5 version restart

LAMMPS LAMMPS General Discussion
,
1

Hi,
I ran a long lammps job using Lammps 2020-5-5 version with these packages;
KSPACE MANYBODY MOLECULE RIGID GPU OMP USER-OMP

I am trying to restart the runs after maintenance in the cluster I am working in.
So I used the restart.file to build the data file.
When I am trying to restart I get the below error:
ERROR: Unrecognized pair style ‘lj/cut/gpu’ is part of the GPU package which is not enabled in this LAMMPS binary. (src/force.cpp:283)

I tried running older data file as well , but it gives me the same error.
I need to know whether my files are corrupted and if so is there any way to fix it .
If not is there any other app or software to create a restart.data file to continue the runs.

Below is the input script:
log lammps.9e1_2
units lj
newton off
package gpu 2
neighbor 0.4 bin
neigh_modify delay 5 every 1 one 5000
#neigh_modify delay 0 every 1 check yes
boundary p p p
#processors 4 2 2
atom_style bond
bond_style fene
special_bonds lj/coul 0 1 1
pair_style lj/cut/gpu 2.5
#kspace_style pppm 1.0e-4
read_restart restart.42ae2_2
#change_box all triclinic remap
#kspace_style pppm 1.0e-4
bond_coeff * 30.0 1.5 1.0 1.0
#bond_coeff 2 50.0 3.0 1.0 1.0
pair_coeff * * 1.0 1.0 2.5
#pair_coeff 3 3 1.0 1.0 2.5
#pair_coeff 1 2 1.0 1.0 2.5
#dielectric 30
#reset_timestep 0
#replicate 2 1 3
#write_data linear-5%sticker-e1.data
#fix 1 all recenter 0.0 0.0 0.0 units box
#velocity all create 1.0 330083
fix 1 all npt temp 1.0 1.0 10 iso 0.1 0.1 10.0
#fix 1 all nve
#fix 1 all nve/limit 0.01
#fix 2 all langevin 1.0 1.0 100.0 85675
#fix 3 all balance 100 xyz 10 1.05 out tmp.balance
#fix 4 all deform 1 x final -28.11 28.11 y final -28.11 28.11 z final -28.11 28.11 units box
#fix 3 all recenter 0.0 0.0 0.0 units box
fix 10 all momentum 1 linear 1 1 1
thermo 1000
thermo_style custom step temp press vol density cpu
#thermo_modify flush yes
timestep 0.01
#reset_timestep 0
restart 100000 restart.13e1 restart.14e1
dump 1 polymer custom 1000000 atom.1a id type xs ys zs
dump 1 all custom 100000 atom.9e1 id type xs ys zs ix iy iz
dump_modify 1 append yes
run 100000
exit

Regards,
Priya

2

Please spend some time to read the forum guidelines: Please Read This First: Guidelines and Suggestions for posting LAMMPS questions

In particular, you posted to the wrong forum category (Materials Project) and didn’t pay attention to the what you need to do to keep the forum software from incorrectly formatting your quoted input file.

That said, the error message is very clear. You are now using a LAMMPS binary that has no GPU package support. You should be able to confirm this by looking at the output of lmp -h. As to how this can happen, that is a question you need to resolve with your local administrators and user support folks.
To give advice from remote requires the ability to read people’s minds (and of their computers) and that is a very rare trait.

3

Thanks for the reply.Sorry for the confusion.
My concern is that I ran the earlier jobs with GPU .
As far as I know the same lammp version is recompiled with the GPU by IT team.
So want to know whether the restart file (binary) has got currupted .If so,is there a way to fix them.

4

There is still not enough information here to make any assessment. As I already mentioned, nobody here can read minds.
There is no indication that the restart file is corrupted.
BTW: there is no fixing of a corrupted file. Once the data is gone, it cannot be recovered.

That said, you can try to convert the restart into a data file with lmp -r2data restart.42ae2_2 data.42ae2_2 and then use read_data instead of read_restart.

Please also note that your LAMMPS version is quite old and proper restarting where you can drop the GPU acceleration is only supported since 2023. Until then restarting has to use the exact same LAMMPS version compiled the same way and the same command line.