# jax-md/jax-md

Differentiable, Hardware Accelerated, Molecular Dynamics

Repository: https://github.com/jax-md/jax-md
Canonical: https://ross.abutalabs.com/products/jax-md
Language: Jupyter Notebook
License: Apache-2.0
License Family: permissive
Last push: 2026-08-18T12:17:06+00:00

## Health v2 (maintenance only)
Score: 86/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 98, release rhythm 63, longevity 100
- inputs: {"age_days": 2669, "days_push": 15, "days_rel": 84, "gap_med": 107.0, "n_releases_24m": 5}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1458, forks 248 (observed 2026-08-28T04:04:47.149377+00:00)

## What it is
JAX MD is a Python library for molecular dynamics simulations built on JAX, making them hardware accelerated on CPU, GPU, and TPU and end-to-end differentiable. It is a functional, data-driven research library for computational condensed matter physics and materials simulation.

## Use cases
- simulate molecular dynamics of materials on GPU or TPU
- run differentiable physics simulations end-to-end in Python
- compute forces and energies with automatic differentiation
- study how small-scale interactions produce large-scale material behavior
- prototype molecular simulation code without writing specialized CUDA
- train machine learning models coupled to physics simulations

## When to choose
- you need differentiable molecular dynamics integrated with JAX/ML workflows
- you want hardware-accelerated simulations without maintaining separate CPU/GPU code
- you are doing research in computational physics or materials science in Python

## When to avoid
- you need a mature, feature-complete production MD package like LAMMPS or HOOMD Blue
- you require API stability - the project is a research effort with possible breaking changes
- you work outside the JAX/Python ecosystem

## Facets
- artifact type: library
- maturity: active
- function: simulation, machine-learning, gpu-computing, math
- domain: simulation, chemistry, machine-learning
- platform: python, cross-platform
- tags: molecular-dynamics, jax, differentiable-programming, computational-physics, materials-science, physics, gpu

## Member repositories
- jax-md/jax-md (main) score 86

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:04:47.149377+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T04:35:30.556227+00:00, confidence not recorded.
  - readme: https://github.com/jax-md/jax-md (fetched 2026-08-28T04:04:47.149377+00:00, sha 0c2d34460d71)
- Data as of 2026-08-30T08:39:29.467469+00:00.