Computational Science Technical Note CSTN-093

CSTN Home

Regular Lattice and Small-World Spin Model Simulations using CUDA and GPUs

K. A. Hawick, A. Leist, D. P. Playne

Archived June 2009

Abstract

Data-parallel accelerator devices such as Graphical Processing Units (GPUs) are providing dramatic performance improvements over even multicore CPUs for lattice-oriented applications in computational physics. Models such as the Ising and Potts models continue to play a role in investigating phase transitions on small-world and scale-free graph structures. These models are particularly well-suited to the performance gains possible using GPUs and relatively high-level device programming languages such as NVIDIA's Compute Unified Device Architecture (CUDA). We report on algorithms and CUDA data-parallel programming techniques for implementing Metropolis Monte Carlo updates for the Ising using bit-packing storage, and adjacency neighbour lists for various graph structures in addition to regular hypercubic lattices. We report on parallel performance gains and also memory and performance tradeoffs using GPU/CPU and algorithmic combinations.

Keywords: Ising model; GPU; CUDA; data-parallel; cluster update.

Full Document Text: PDF version.

Citation Information: BiBTeX database for CSTN Notes.

BiBTeX reference:

@ARTICLE{CSTN-093,
  author = {K. A. Hawick and A. Leist and D. P. Playne},
  title = {{Regular Lattice and Small-World Spin Model Simulations using CUDA
	and GPUs}},
  journal = {Int. J. Parallel Prog.},
  year = {2011},
  volume = {39},
  pages = {183-201},
  number = {CSTN-093},
  doi = {10.1007/s10766-010-143-4},
  institution = {Computer Science, Massey University},
  keywords = {GPGPU, Simulation, Monte Carlo},
  timestamp = {2009.09.06}
}


[ CSTN Index | CSTN BiBTeX ]