radical-cybertools / radical.repex.at

This is the github location for RepEx developed by the RADICAL team in conjunction with the York Lab.
Other
4 stars 3 forks source link

RepEx: Replica Exchange Molecular Dynamics simulations Package

RepEx is a novel Replica Exchange Molecular Dynamics (REMD) simulations package for HPC clusters. The primary aim for the design of RepEx is to decouple the implementation of the Replica Exchange (RE) algorithm from the Molecular Dynamics (MD) simulation engine. Current version of RepEx supports Amber [1] and NAMD [2] simulation engines, and can be easily modified to support any other MD engine.

RepEx supports multi-dimensional REMD simulations with arbitrary number and ordering of dimensions. Currently RepEx supports three exchange types: salt concentration, temperature and umbrella exchange. RepEx explicitly decouples the number of replicas, from the computational resources (CPUs, GPUs, etc.). RepEx can be used as a testing platform for new or unexplored REMD methods.

To execute its workloads, RepEx relies on the concept of task-level parallelism, which is enabled by the use of the RADICAL-Pilot API Radical Pilot.

RepEx supports synchronous and asynchronous RE patterns. Asynchronous RE Pattern does not have a global synchronization barrier between simulation and exchange phase. While some replicas are in the simulation phase, others might be in the exchange phase. Based on some criterion, a subset of replicas transition into the exchange phase, while other replicas continue in the simulation phase. Selection of replicas that will transition may be based on a FIFO principle, e.g. first N replicas transition into an exchange phase. Alternatively, only replicas which have finished a predefined number of simulation time-steps (2 ps) during some real time interval (1 minute) transition into exchange phase.

[1] - Amber

[2] - NAMD

User guide

RepEx user guide can be found at: repex.readthedocs.org

Website

RepEx website can be found at: repex.io

ICPP-2016 paper

Paper in pdf format: paper.pdf

Presentation in pdf format: slides-icpp16.pdf

Presentations

A. Treikalis thesis: at-thesis-slides.pdf

Performance optimization slides: repex-performance-optimization.pdf