LAMMPS with KOKKOS/GPU Build Guide (A40, CUDA 12.2)

This guide walks through building LAMMPS from source with GPU acceleration via the KOKKOS package. It is tested on an HPC system with NVIDIA A40 GPUs and CUDA 12.2.

1. Load Required Modules

module purge
module load gcc/10.3.0
module load cuda/12.2
module load cmake
module load openmpi

2. Clone LAMMPS Repository

git clone -b stable https://github.com/lammps/lammps.git
cd lammps

3. Create Build Directory and Configure with CMake

mkdir build-kokkos
cd build-kokkos

Clear the cache if re-running:

rm -rf CMakeCache.txt CMakeFiles

Then configure:

cmake ../cmake \
  -D CMAKE_BUILD_TYPE=Release \
  -D BUILD_MPI=on \
  -D BUILD_OMP=on \
  -D PKG_KOKKOS=on \
  -D Kokkos_ENABLE_CUDA=ON \
  -D Kokkos_ENABLE_MPI=ON \
  -D Kokkos_ENABLE_OPENMP=ON \
  -D Kokkos_ARCH_AMPERE86=ON \
  -D LAMMPS_MACHINE=kokkos \
  -D BUILD_SHARED_LIBS=off

4. Compile LAMMPS

make -j 8

5. Create Test Input Script (in.lj_kokkos)

units       lj
atom_style  atomic
dimension   3
boundary    p p p
lattice     fcc 0.8442
region      box block 0 10 0 10 0 10
create_box  1 box
create_atoms 1 box
mass        1 1.0

velocity    all create 1.44 87287 loop geom

pair_style  lj/cut/kk 2.5
pair_coeff  * * 1.0 1.0 2.5

neighbor    0.3 bin
neigh_modify every 20 delay 0 check no

fix         1 all nve
thermo      100
run         1000

6. Run LAMMPS with GPU and KOKKOS

OMP_PROC_BIND=spread OMP_PLACES=threads mpirun -np 1 ./lmp_kokkos -k on g 1 -sf kk -in in.lj_kokkos

7. Optional: Monitor GPU Usage

watch -n 1 nvidia-smi

Notes

• Use gcc ≤ 10 for CUDA compatibility.
• Always include -sf kk to enable KOKKOS-accelerated styles.
• Kokkos_ENABLE_OPENMP ensures host execution space is defined.