Skip to content

LAMMPS

LAMMPS is a large scale classical molecular dynamics code, and stands for Large-scale Atomic/Molecular Massively Parallel Simulator. LAMMPS has potentials for soft materials (biomolecules, polymers), solid-state materials (metals, semiconductors) and coarse-grained or mesoscopic systems. It can be used to model atoms or, more generically, as a parallel particle simulator at the atomic, meso, or continuum scale.

Availability

Software Version Dependent Toolchain Module Load Command
LAMMPS 29Aug2024_update2-kokkos foss/2024a module load foss/2024a LAMMPS/29Aug2024_update2-kokkos
LAMMPS 29Aug2024_update2-kokkos-CUDA-12.6.0 foss/2024a module load foss/2024a LAMMPS/29Aug2024_update2-kokkos-CUDA-12.6.0
LAMMPS 2Aug2023_update2-kokkos-CUDA-12.6.0 foss/2024a module load foss/2024a LAMMPS/2Aug2023_update2-kokkos-CUDA-12.6.0

Note

To know the deatils about dependent toolchain please go to Toolchains

Application Information, Documentation and Support

The official LAMMPS is available at LAMMPS Online Manual. LAMMPS has a large user base and a good user support. Question related to using LAMMPS can be posted to the LAMMPS User forum. Archived user mailing list are also useful to resolve some of the common user issues.

Tip

If after checking the above forum, if you believe that there is an issue with the module, please file a ticket with Service Now

Using LAMMPS

Sample Batch Script to Run LAMMPS 
 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
#!/bin/bash
#SBATCH -J test_lammps
#SBATCH --output=%x.%j.out # %x.%j expands to slurm JobName.JobID
#SBATCH --error=%x.%j.err # prints the error message
#SBATCH --partition=general 
#SBATCH --nodes=1
#SBATCH --ntasks-per-node=128
#SBATCH --mem-per-cpu=4000M # Maximum allowable memory per CPU 4G
#SBATCH --qos=standard
#SBATCH --account=PI_ucid # Replace PI_ucid which the NJIT UCID of PI
#SBATCH --time=71:59:59  # D-HH:MM:SS

###############################################
#
# Purge and load modules needed for run
#
################################################
module purge
module load wulver # Load slurm, easybuild
module load foss/2024a LAMMPS

srun lmp -in test.in

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
#!/bin/bash
#SBATCH -J gpu_lammps
#SBATCH --output=%x.%j.out # %x.%j expands to slurm JobName.JobID
#SBATCH --error=%x.%j.err # prints the error message
#SBATCH --partition=gpu 
#SBATCH --nodes=1
#SBATCH --ntasks-per-node=128
#SBATCH --mem-per-cpu=4000M # Maximum allowable memory per CPU 4G
#SBATCH --qos=standard
#SBATCH --gres=gpu:2
#SBATCH --account=PI_ucid # Replace PI_ucid which the NJIT UCID of PI
#SBATCH --time=71:59:59  # D-HH:MM:SS

###############################################
#
# Purge and load modules needed for run
#
################################################
module purge
module load wulver # Load slurm, easybuild
module load foss/2024a LAMMPS/29Aug2024_update2-kokkos-CUDA-12.6.0

export OMP_NUM_THREADS=$SLURM_CPUS_PER_TASK
export OMP_PROC_BIND=spread
export OMP_PLACES=threads


srun lmp -in in.lj -k on g 2 -sf kk -pk kokkos newton off neigh full comm device

Then submit the job script using the sbatch command, e.g., assuming the job script name is test_lammps.slurm:

1
sbatch test_lammps.slurm

Building LAMMPS from source

Some users may be interested in building LAMMPS from source to enable more specific LAMMPS packages. The source files for LAMMPS can be downloaded as either a tar file or from the LAMMPS Github repository.

Building on Cluster

The following procedure was used to build LAMMPS on Wulver. In the terminal:

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
module purge
module load wulver
module load foss
module load CMake

git clone https://github.com/lammps/lammps.git
cd lammps
mkdir build
cd build

cmake -DCMAKE_INSTALL_PREFIX=$PWD/../install_hsw -DCMAKE_CXX_COMPILER=mpicxx \
            -DCMAKE_BUILD_TYPE=Release -D BUILD_MPI=yes -DKokkos_ENABLE_OPENMP=ON \
            -DKokkos_ARCH_HSW=ON -DCMAKE_CXX_STANDARD=17 -D PKG_MANYBODY=ON \
            -D PKG_MOLECULE=ON -D PKG_KSPACE=ON -D PKG_REPLICA=ON -D PKG_ASPHERE=ON \
            -D PKG_RIGID=ON -D PKG_KOKKOS=ON -D DOWNLOAD_KOKKOS=ON \
            -D CMAKE_POSITION_INDEPENDENT_CODE=ON -D CMAKE_EXE_FLAGS="-dynamic" ../cmake
make -j16
make install

User Contributed Information

Please help us improve this page

Users are invited to contribute helpful information and corrections through our Github repository.