An investigation of the effects of hard and soft errors on graphics processing unit‐accelerated molecular dynamics simulations

Abstract

SUMMARYMolecular dynamics (MD) simulations rely on the accurate evaluation and integration of Newton's equations of motion to propagate the positions of atoms in proteins during a simulation. As such, one can expect them to be sensitive to any form of numerical error that may occur during a simulation. Increasingly graphics processing units (GPUs) are being used to accelerate MD simulations. Current GPU architectures designed for high performance computing applications support error‐correcting codes (ECC) that detect and correct single bit‐flip soft error events in GPU memory; however, this error checking carries a penalty in terms of simulation speed. ECC is also a major distinguishing feature between high performance computing NVIDIA Tesla cards and the considerably more cost‐effective NVIDIA GeForce gaming cards. An argument often put forward for not using GeForce cards is that the results are unreliable because of the lack of ECC. In an initial attempt to quantify these concerns, an investigation of the reproducibility of GPU‐accelerated MD simulations using the AMBER software was conducted on the XSEDE supercomputer Keeneland, a cluster at Los Alamos National Laboratory, and a cluster at the San Diego Supercomputer Center. While the data collected are insufficient to make solid conclusions and more extensive testing is needed to provide quantitative statistics, the absence of ECC events and lack of any silent errors in all the simulations conducted to date suggest that these errors are exceedingly rare and as such the time and memory penalty of ECC may outweigh the utility of error checking functionality. However, a considerable amount of error originating from defective hardware was observed, which suggests that rigorous acceptance testing should be performed on new GPU‐based systems by repeatedly running reproducible yet realistic calculations.

GPU-enabled particle-particle particle-tree scheme for simulating dense stellar cluster system

“…The bottom panel of figure 7 shows that all curves eventually approach to constant values for both of large and small r cut . For large r cut , the timesteps of the nonisolated particles are determined by N , not by r cut [see equation (25)], whereas for small values of r cut the non-isolated particles have a timesteps ∆t max . This is because most neighbouring particles are in the buffer shell and not in the neighbour sphere.…”

Classical molecular dynamics on graphics processing unit architectures

“…Betz et al . investigate the effect of various kinds of hardware errors that can arise in GPUs to MD simulations.…”

“…Betz et al 116 investigate the effect of various kinds of hardware errors that can arise in GPUs to MD simulations. The main focus is on error-correcting codes (ECC) that can detect and fix single bit flips, however, using them introduces a noticeable time and memory penalty.…”

Molecular simulations and visualization: introduction and overview

“…Calculations using Reax-FF, which is a reactive force field, have been accelerated using GPUs [ 79 ]. Aspects of reliability and reproducibility have been studied in the context of error correcting code [ 80 ]. 95 ] is a relatively recent approach to molecular simulation that builds on distributed computing approaches like FightAIDS@home, SETI@home, and folding@home.…”