Long-Time Dynamics through Parallel Trajectory Splicing

Abstract

Simulating the atomistic evolution of materials over long time scales is a longstanding challenge, especially for complex systems where the distribution of barrier heights is very heterogeneous. Such systems are difficult to investigate using conventional long-time scale techniques, and the fact that they tend to remain trapped in small regions of configuration space for extended periods of time strongly limits the physical insights gained from short simulations. We introduce a novel simulation technique, Parallel Trajectory Splicing (ParSplice), that aims at addressing this problem through the timewise parallelization of long trajectories. The computational efficiency of ParSplice stems from a speculation strategy whereby predictions of the future evolution of the system are leveraged to increase the amount of work that can be concurrently performed at any one time, hence improving the scalability of the method. ParSplice is also able to accurately account for, and potentially reuse, a substantial fraction of the computational work invested in the simulation. We validate the method on a simple Ag surface system and demonstrate substantial increases in efficiency compared to previous methods. We then demonstrate the power of ParSplice through the study of topology changes in Ag42Cu13 core-shell nanoparticles.

The mobility of small vacancy/helium complexes in tungsten and its impact on retention in fusion-relevant conditions

“…In order to access timescales that are sufficient to enable the motion of V N He M complexes in conditions typical of the first wall in fusion reactors ( T  ~ 1000  K ), we use Parallel Trajectory Splicing (ParSplice) 45 , a recently introduced AMD method that improves upon its predecessor, Parallel Replica Dynamics 32, 46 . AMD techniques are designed to extend the simulation timescales of systems evolving through rare, activated, events between persistent states.…”

“…A key innovation of ParSplice is that segments can be concurrently generated in multiple states, according to the expected likelihood that they will be used in the construction of the trajectory. Details can be found in the original publication 45 .…”

Evidence for percolation diffusion of cations and reordering in disordered pyrochlore from accelerated molecular dynamics

“…In this regime, we instead rely on the parallel trajectory splicing 32 method. ParSplice is an extension of parallel replica dynamics (PRD 54, 55 ).…”

“…In the present paper, we used molecular dynamics (MD) and parallel trajectory splicing (ParSplice) 32 , one of the accelerated molecular dynamics methods 33 , to simulate cation vacancy diffusion in disordered Gd 2 Ti 2 O 7 (GTO) pyrochlore on the microsecond timescale. The disordering processes, which can be caused by, for example, temperature or irradiation, are shown in Fig.…”

Reaction pathways in atomistic models of thin film growth

“…3 To access experimental time scales, many different approaches to the atomistic modelling thin film growth have been considered, including the methods developed at Los Alamos such as temperature accelerated dynamics (TAD), 3,4 hyperdynamics, [5][6][7] and more recently parallel replica or parallel splice. 8 In these approaches, Newton's equations of motion are still integrated but the time at which the system evolves is advanced.…”