LAMMPS setup with MPI

LAMMPS (Large-scale Atomic/Molecular Massively Parallel Simulator) is a classical molecular dynamics (MD) simulator designed for modeling atomic, polymeric, and coarse-grained systems. It is open-source and highly parallelized, making it suitable for running on high-performance computing clusters. Its versatility allows researchers to simulate a wide range of materials and phenomena, from simple liquids to complex biological molecules, under various physical conditions.

This guide shows how to build and run LAMMPS on Arch Linux using OpenMPI. It includes checking prerequisites, building from source, verifying MPI support, and running a sample simulation.

1. Install Required Tools

LAMMPS requires openmpi and cmake. First, verify if they are installed:

which openmpi
which mpicxx
which cmake

If any are missing, install via pacman:

sudo pacman -S openmpi cmake git

2. Clone LAMMPS Repository

Get the latest LAMMPS source code:

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

3. Build LAMMPS with MPI Support

Create a build directory and configure with CMake:

mkdir build && cd build
cmake ../cmake -D BUILD_MPI=YES 
make -j$(nproc)

This builds LAMMPS using all available CPU cores. And you will get a lmp named executable file which will be used in all LAMMPS calculations.

4. Verify MPI Support

Check if the built LAMMPS binary has MPI enabled:

cd ~/lammps/build/
./lmp -h | grep MPI

If MPI is listed, your LAMMPS build is ready for parallel execution.

5. Run a Sample Simulation

Test LAMMPS on a simple Lennard-Jones system using 2 MPI processes:

mpirun -np 2 ./lmp -in ../examples/UNITS/in.ar.lj

This command will start the simulation and output the trajectory, which can later be visualized using OVITO.