lammps/lammps

Health v2 · maintenance only

Adoption not part of the score

3029 stars · 2030 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-30, confidence not recorded

LAMMPS (Large-scale Atomic/Molecular Massively Parallel Simulator) is an open-source classical molecular dynamics simulation code focused on materials modeling, developed originally at Sandia National Laboratories. It runs efficiently on parallel computers via MPI, OpenMP, and GPU acceleration (CUDA, OpenCL, HIP, SYCL, Kokkos), and can be used as a standalone executable or embedded as a library with C++, Fortran, and Python interfaces.

Use cases

• simulate atoms and molecules with molecular dynamics
• model materials like metals, semiconductors, and polymers
• run large-scale parallel MD simulations on HPC clusters
• simulate biomolecules, polymers, and coarse-grained systems
• run molecular dynamics on GPUs
• use machine-learned interatomic potentials for atomistic simulation
• model granular and mesoscopic particle systems
• couple MD simulation with other codes via a library interface

When to choose

• you need a mature, widely-cited classical MD code with hundreds of interatomic potentials
• you need massive parallelism across CPUs and GPUs on HPC clusters
• you need an extensible, scriptable simulator that can be embedded as a library
• you work in materials science, chemistry, or condensed matter physics

When to avoid

• you need quantum-mechanical (ab initio) simulation rather than classical MD
• you need a simple GUI-driven tool for small educational simulations
• you need a general-purpose game or graphics particle engine rather than scientific simulation

Facets

application · maturity stable

simulation gpu-computing concurrency sdk benchmarking simulation chemistry machine-learning windows cpp python cross-platform molecular-dynamics hpc mpi kokkos materials-modeling particle-simulation force-fields scientific-computing physics materials-science linux macos gpu

Member repositories

For agents

markdown · JSON · MCP: product_card(name="lammps/lammps")