Efficient high-dimensional molecular dynamics for studies of the glass/jamming transition

Many of the most exciting recent advances in the theory of amorphous materials have come from studies of high-dimensional liquids.  Simulations of such liquids, however, have been limited by the lack of a publicly-available, open-source, efficiently-parallel molecular dynamics code that works for d > 3.  We have developed such a code (hdMD).  All its routines work in arbitrary d; the maximum simulated d is limited only by available computing resources.  These routines include several that are particularly useful for studies of the glass/jamming transition, such as SWAP Monte Carlo and FIRE energy minimization.  Scaling of simulation runtimes with the number of particles N and number of simulation threads n_threads is comparable to popular MD codes such as LAMMPS, and the latter scaling actually improves with increasing d.  This efficient parallel implementation allows simulation of systems that are much larger than those employed in previous high-dimensional studies.   As a demonstration of the code's capabilities, we show that dynamics in d = 6 supercooled liquids can be much more heterogeneous than previously reported.