A comparison between parallelization approaches in molecular dynamics simulations on GPUs

@article{Rovigatti2015ACB,
  title={A comparison between parallelization approaches in molecular dynamics simulations on GPUs},
  author={Lorenzo Rovigatti and Petr {\vS}ulc and Istv{\'a}n Z. Reguly and Flavio Romano},
  journal={Journal of Computational Chemistry},
  year={2015},
  volume={36},
  pages={1 - 8}
}
We test the relative performances of two different approaches to the computation of forces for molecular dynamics simulations on graphics processing units. A “vertex‐based” approach, where a computing thread is started per particle, is compared to an “edge‐based” approach, where a thread is started per each potentially non‐zero interaction. We find that the former is more efficient for systems with many simple interactions per particle while the latter is more efficient if the system has more… 
Classical molecular dynamics on graphics processing unit architectures
TLDR
Focusing on published works in the field of classical MD, the chosen implementation methods and algorithmic techniques used for porting to GPU are described, as well as how recent advances of GPU architectures will provide even more optimization possibilities in the future.
Molecular Dynamics of Heterogeneous Systems on GPUs and Their Application to Nucleation in Gas Expanding to a Vacuum.
TLDR
This work presents an implementation of molecular dynamics on graphics processing units based on the Verlet list and applies it to several systems for which experimental data are available, reproducing sufficiently sized clusters but not the experimentally observed clusters of irregular shape.
Graphics processing units in bioinformatics, computational biology and systems biology
TLDR
A collection of GPU tools recently developed to perform computational analyses in life science disciplines, emphasizing the advantages and the drawbacks in the use of these parallel architectures.
RUMD: A general purpose molecular dynamics package optimized to utilize GPU hardware down to a few thousand particles
TLDR
RUMD is a general purpose, high-performance molecular dynamics simulation package running on graphical processing units (GPU's) that has a performance that is comparable to other GPU-MD codes at large system sizes and substantially better at smaller sizes.
Accelerating the Finite-Element Method for Reaction-Diffusion Simulations on GPUs with CUDA
TLDR
Reaction-diffusion equations are studied (a DNA-based predator-prey system) in a tortuous geometry (a maze), which was shown experimentally to generate subtle geometric effects and solve the partial differential equations on a GPU, demonstrating a speedup of ∼100 over the same resolution on a 20 cores CPU.
Molecular Dynamics Simulations of DNA-Functionalized Nanoparticle Building Blocks on GPUs
TLDR
This thesis implements the nBLOCK computational model as an extension to the coarse grain molecular simulator oxDNA to explore greater simulation time scales and provides descriptions of the scripting utilities which support nBLock assembly generation, simulation, and analysis.
Hierarchical bounding structures for efficient virial computations: Towards a realistic molecular description of cholesterics.
TLDR
The application of bounding volume hierarchies to accelerate second-virial evaluations for arbitrary complex particles interacting through hard and soft finite-range potentials and highlights the complex relationship between microscopic structure and helical twisting power in such model systems.
Brownian Dynamics Simulation on the GPU: Virtual Colloidal Suspensions
TLDR
This paper focuses on the problem of neighbourhood search to accelerate computations on a single GPU forBrownian Dynamics simulations and outperforms existing implementations by introducing a novel dynamic test for one kind of particle.
Connectivity, dynamics, and structure in a tetrahedral network liquid.
TLDR
A detailed computational study by Brownian dynamics simulations of the structure and dynamics of a liquid of patchy particles which forms an amorphous tetrahedral network upon decreasing the temperature, understood in terms of an evolving network of long time metastable cooperative domains of particles.
...
1
2
3
4
5
...

References

SHOWING 1-10 OF 111 REFERENCES
Accelerating molecular modeling applications with graphics processors
TLDR
An overview of recent advances in programmable GPUs is presented, with an emphasis on their application to molecular mechanics simulations and the programming techniques required to obtain optimal performance in these cases.
GPU-accelerated molecular modeling coming of age.
Harvesting graphics power for MD simulations
TLDR
An implementation of molecular dynamics simulations on a graphic processing unit (GPU) in the NVIDIA CUDA language and results for two MD algorithms suitable for short-ranged and long-ranged interactions, and a congruential shift random number generator are presented.
Sop‐GPU: Accelerating biomolecular simulations in the centisecond timescale using graphics processors
TLDR
It is found that the mechanical molecular response critically depends on the conditions of force application and that the kinetics and pathways for unfolding change drastically even upon a modest 10‐fold increase in vf, implying that, to resolve accurately the free energy landscape and to relate the results of single‐molecule experiments in vitro and in silico, molecular simulations should be carried out under the experimentally relevant force loads.
Scalable molecular dynamics with NAMD
NAMD is a parallel molecular dynamics code designed for high‐performance simulation of large biomolecular systems. NAMD scales to hundreds of processors on high‐end parallel platforms, as well as
...
1
2
3
4
5
...