Proceedings. 15th Symposium on Computer Architecture and High Performance Computing
DOI: 10.1109/cahpc.2003.1250330
|View full text |Cite
|
Sign up to set email alerts
|

Parallel implementation of a lattice-gauge-theory code: studying quark confinement on PC clusters

Abstract: We consider the implementation of a parallel Monte Carlo code for high-performance simulations on PC clusters with MPI. We carry out tests of speedup and efficiency. The code is used for numerical simulations of pure SU (2) lattice gauge theory at very large lattice volumes, in order to study the infrared behavior of gluon and ghost propagators. This problem is directly related to the confinement of quarks and gluons in the physics of strong interactions.

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
3
1
1

Citation Types

1
12
0

Publication Types

Select...
6

Relationship

1
5

Authors

Journals

citations
Cited by 7 publications
(13 citation statements)
references
References 39 publications
(64 reference statements)
1
12
0
Order By: Relevance
“…We used the hybrid overrelaxed algorithm (HOR) [42,43] for thermalization and the stochastic overrelaxation algorithm or the so-called Cornell method for gauge-fixing thermalized configurations to the lattice Landau gauge [44,45,46]. The numerical code is parallelized using MPI [47]. For the random number generator we use a double-precision implementation of RANLUX (version 2.1) with luxury level set to 2.…”
Section: Simulationsmentioning
confidence: 99%
“…We used the hybrid overrelaxed algorithm (HOR) [42,43] for thermalization and the stochastic overrelaxation algorithm or the so-called Cornell method for gauge-fixing thermalized configurations to the lattice Landau gauge [44,45,46]. The numerical code is parallelized using MPI [47]. For the random number generator we use a double-precision implementation of RANLUX (version 2.1) with luxury level set to 2.…”
Section: Simulationsmentioning
confidence: 99%
“…At the same time, the vector u -with components u j normalized to 1 − u 2 0 -points along a uniformly chosen random direction in three-dimensional space [2].…”
Section: The 4-vector Spin Model (On a 1-d Lattice) Is Defined By Thementioning
confidence: 99%
“…In this formulation the theory becomes equivalent to a statistical mechanical model, which can be studied numerically using Monte Carlo simulations (see for example [1,2] and references therein). As a consequence, the system considered evolves according to a Markov process in the so-called Monte Carlo time and the actionweighted configuration-space average of the observables is substituted by a time average over successive (independent) field configurations of the system.…”
Section: Introductionmentioning
confidence: 99%
See 2 more Smart Citations