Solving Newton's Equations of Motion with Large Timesteps using Recurrent Neural Networks based Operators
This provides a faster method for molecular dynamics simulations, but it is incremental as it builds on existing neural network approaches for dynamics.
The paper tackled the problem of slow molecular dynamics simulations by introducing recurrent neural network-based operators that solve Newton's equations with timesteps up to 4000 times larger than Verlet, achieving significant speedup in systems of up to 16 particles.
Classical molecular dynamics simulations are based on solving Newton's equations of motion. Using a small timestep, numerical integrators such as Verlet generate trajectories of particles as solutions to Newton's equations. We introduce operators derived using recurrent neural networks that accurately solve Newton's equations utilizing sequences of past trajectory data, and produce energy-conserving dynamics of particles using timesteps up to 4000 times larger compared to the Verlet timestep. We demonstrate significant speedup in many example problems including 3D systems of up to 16 particles.