LAMMPS is a classical molecular dynamics code. It can be applied to metals, semiconductors, biomolecules, polymers in all-atom or coarse-grained models. The source code is distributed under the terms of GPL by Sandia National Laboratories. See LAMMPS home page http://lammps.sandia.gov/ for additional information.
B. How to download LAMMPS
There are several different mechanisms to download LAMMPS. See complete instructions for downloading the code. You can download a recent stable version from the LAMMPS github repository https://github.com/lammps/lammps using wget. For example, the October, 2020 stable release can be obtained by running
C. How to build LAMMPS
On Blue Waters, LAMMPS can be built under the GNU environment. Building under other compiler environments may be possible. Please contact email@example.com if you need to use another environment.
C.1 Building under GNU environment
module swap PrgEnv-cray PrgEnv-gnu module add fftw module add cmake/3.17.3 mkdir ~/lammps cd ~/lammps wget https://github.com/lammps/lammps/archive/stable_29Oct2020.zip unzip stable_29Oct2020.zip cd lammps-stable_29Oct2020 mkdir build cd build cmake -D CMAKE_CXX_COMPILER=CC -D CMAKE_C_COMPILER=cc -D CMAKE_Fortran_COMPILER=ftn \ -D CMAKE_INSTALL_PREFIX=$HOME/lammps -D PKG_MOLECULE=yes ../cmake make make install
D. Sample test
LAMMPS comes with an extensive suit of tests. To make sure that the compiled code works, we wil run one of the supplied tests.
Create run.pbs file having the following content:
#!/bin/bash #PBS -l nodes=1:ppn=32:xe #PBS -l walltime=00:05:00 #PBS -q debug cd $PBS_O_WORKDIR export OMP_NUM_THREADS=8 aprun -n 4 -N 4 -d 8 $HOME/lammps/bin/lmp -in in.micelle -log job.out -screen none
Submit the job
Open the file job.out with text editor and make sure that line "Loop time of 0.121433 on 4 procs for 1000 steps with 1200 atoms" is present near the end of the file. It indicates that the job performed 1,000 dynamics steps on 4 processors. Timing of 0.121433 will be different on different machines and may somewhat vary from run to run, which is normal.