Oct 22

Print this Post

10th CCoE Seminar: “Application of GPUs in statistical physics”

odor02On Thursday, Dec. 11, at 3 p.m. (at TU Dresden, Willersbau Room A317) Géza Ódor from the Institute for Technical Physics and Materials Science at the Hungarian Academy of Sciences will present his work on using GPUs in statistical physics.


Application of GPUs for long standing open problems in nonequilibrium statistical physics, with an emphasis on the Kardar-Parisi-Zhang model and its mapping onto binary driven lattice gases is overviewed [1-4]. I point out how pattern formation [5] and disorder effects [3,6] can be studied by these cellular automaton algorithms.

I present another application area, which is related to network science. I show how CUDA sparse matrix libraries perform for the spectral analysis of scale-free models with spreading models running on top of them [7]. I present localization studies obtained for graphs up to one million nodes. Finally, I show how Acceleware software is used in simulations of hierarchical modular network models, describing brain dynaics.

[1] Géza Ódor, Bartosz Liedke and Karl-Heinz Heinig, Mapping of 2+1-dimensional Kardar-Parisi-Zhang growth onto a driven lattice gas model of dimers, Phys. Rev. E 79 (2009) 021125
[2]  Géza Ódor, Bartosz Liedke and Karl-Heinz Heinig Directed d-mer diffusion describing the Kardar-Parisi-Zhang-type surface growth, Phys. Rev. E 81 (2010) 031112
[3]  Henrik Schulz, Géza Ódor, Gergely Ódor, Máte Ferenc Nagy Simulation of 1+1 dimensional surface growth and lattices gases using GPUs Comp. Phys. Comm. 182 (2011) 1467
[4]  Jeffrey Kelling and Géza Ódor, Extremely large-scale simulation of a Kardar-Parisi-Zhang model using graphics cards, Phys. Rev. E 84 (2011) 061150
[5]  Géza Ódor, Bartosz Liedke and Karl-Heinz Heinig, Surface pattern formation and scaling described by conserved lattice gases, Phys. Rev. E81 (2010) 051114
[6]  R. Juhász, G. Ódor, Anomalous coarsening in disordered exclusion processes, J. Stat. Mech. (2012) P08004
[7]  Géza Ódor, Localization transition, Lifschitz tails, and rare-region effects in network models, Phys. Rev. E 90, 032110 (2014)

Permanent link to this article: https://gcoe-dresden.de/10th-ccoe-seminar-application-of-gpus-in-statistical-physics/