NIM for ALCHEMI Batched Molecular Dynamics

Type: Microservice Tags: NVIDIA, NIM, ALCHEMI, batched molecular dynamics, BMD, molecular simulation, materials science, MLIP, MACE, AIMNet2, TensorNet Related: NVIDIA-NIM, NIM-for-ALCHEMI-Batched-Geometry-Relaxation, NVIDIA-BioNeMo, NVIDIA-Clara, NVIDIA-AI-Enterprise, cuEquivariance, TensorRT, Triton-Inference-Server, NVIDIA-CUDA, NGC Sources: https://docs.nvidia.com/nim/alchemi/alchemi-bmd/latest/overview.html Last Updated: 2026-04-29

Summary

NVIDIA NIM for ALCHEMI Batched Molecular Dynamics (BMD) is a NIM microservice for running batched molecular dynamics simulations using machine learning interatomic potentials. Current NVIDIA docs describe it as enabling near-DFT-accuracy molecular dynamics at a fraction of traditional computational cost with MACE, AIMNet2, and TensorNet models optimized for NVIDIA GPUs.

Detail

Purpose

Batched molecular dynamics lets researchers study the time evolution of atomic systems across many molecules or materials candidates. It is relevant to thermodynamic properties, phase transitions, transport, ligand binding, conformational sampling, free-energy workflows, and reaction dynamics.

Current scope

  • Supports NVE, NVT, and NPT simulation ensembles.
  • Uses dynamic batching to improve GPU utilization across concurrent MD requests.
  • Returns trajectory snapshots, positions, velocities, energies, and optional stress tensors.
  • Supports restart from saved states for chunked long-running simulations.
  • Supports external electric fields for AIMNet2 models.
  • Allows per-request simulation parameters such as temperature, timestep, ensemble, and barostat settings.
  • Supports MLIP model types selected through ALCHEMI_NIM_MODEL_TYPE, including MACE, TensorNet, and AIMNet2.
  • Includes MACE-MPA-0 as a bundled/default model path when the container runs with an NGC API key.

NVIDIA context

BMD is the simulation counterpart to NIM-for-ALCHEMI-Batched-Geometry-Relaxation. Together they form an ALCHEMI NIM path for atomistic modeling pipelines that need GPU-accelerated relaxation and molecular dynamics in a consistent NVIDIA-NIM deployment model.

Connections

Source Excerpts

  • NVIDIA docs describe BMD as a high-performance engine for batched molecular dynamics simulations using MLIPs.
  • The current docs list NVE, NVT, and NPT ensembles plus trajectory output and restart capability.

Resources