Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Log inSign up
Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Blog
Job Board
Q3AS

© 2026 Aqora Quantum S.A.S.

TermsPrivacyLegal Notice

Physical review. D. Particles, fields, gravitation, and cosmology/Physical review. D. Particles and fields

Possible barrier at z≈1 for local algorithms

Georgios Bathas, Herbert Neuberger B

Published May 15, 1992• 4 citations
Publisher

Abstract

It is shown that a certain class of generalizations of overrelaxation algorithms is incapable of further reducing the dynamical exponent $z$ below its standard overrelaxed value of $z\ensuremath{\approx}1$. The mean-field value is unity and is obtained in a theory that is free in the static limit, while the effect of interactions and dimensionality could be estimated with dynamical renormalization-group methods. The generalizations are obtained by viewing overrelaxation as a slightly deformed deterministic algorithm and should, therefore, hold for hybrid Monte Carlo algorithms as well.

DOI
10.1103/physrevd.45.3880

Related papers

The Local Hybrid Monte Carlo algorithm for free field theory: Reexamining overrelaxation

Ivan Horváth A, Anthony D. Kennedy

arXiv:hep-lat/9708024•Nuclear Physics B•Jan 1, 1998•15 citations

Dynamic Monte Carlo Renormalization Group

Jan Tobochnik, Sanjoy Kumar Sarker, Robert Cordery

A novel dynamic Monte Carlo renormalization-group method is devised. Application to the kinetic Ising model in one dimension yields a value of $z$, the dynamic critical exponent, which is in excellent agreement with the exact result. In two dimensions we find $z=2.22\ifmmode\pm\else\textpm\fi{}0.13$ with the most probable value $z=2.17$.

Physical Review Letters•May 25, 1981•72 citations

Monte Carlo renormalization group study of the two dimensional kinetic Ising model

John P. Bartel

The critical dynamics of the two dimensional Kinetic Ising Model have been studied using Monte Carlo Real Space Renormalization Group techniques. Time dependent correlation functions were calculated for square lattices of 16 x 16 and 32 x 32 spins, using periodic boundary conditions wth 60,000 and 15,000 Monte Carlo steps per spin, respectively;The critical exponent z was estimated using a matching process for a variety of renormalization group operators. The estimates for z were surprisingly sensitive to small changes in the value of the renormalization group operator, with physically reasonable operators giving estimates for z less than the rigorous lower bound for z of z(,o) = 1.75. The estimates for z from a given renormalization group operator showed poor convergence properties. Attempts to estimate z from the scaling form of the correlation function were inconclusive, as were attempts to estimate the static exponent (eta);Possible explanations of these unexpected results are discussed. The most plausible source of the complex behavior seen in this work is the discrete nature of time in the Monte Carlo computer simulation, which generates phenomena not present in standard theoretical treatments of the Kinetic Ising Model.

•Jan 1, 1984•0 citations

Renormalization Group

Giuseppe Mussardo

Abstract At a critical point, the correlation length diverges: the statistical influctuations extend on all scales of the system and any attempt to solve the dynamics by taking into account only a finite number of degrees of freedom fails. In the absence of an exact solution of the model under consideration, the computation of the critical exponents is often obtained only by numerical methods and Monte Carlo simulations.

Oxford University Press eBooks•Aug 27, 2009•0 citations

Nonuniversal critical dynamics in Monte Carlo simulations

Robert H. Swendsen, Jian‐Sheng Wang

A new approach to Monte Carlo simulations is presented, giving a highly efficient method of simulation for large systems near criticality. The algorithm violates dynamic universality at second-order phase transitions, producing unusually small values of the dynamical critical exponent.

Physical Review Letters•Jan 12, 1987•2,820 citations

Monte Carlo Algorithms for the Fully Frustrated XY Model

S. Große Pawig, K. Pinn

We investigate local update algorithms for the fully frustrated XY model on a square lattice. In addition to the standard updating procedures like the Metropolis or heat bath algorithm we include overrelaxation sweeps, implemented through single spin updates that preserve the energy of the configuration. The dynamical critical exponent (of order two) stays more or less unchanged. However, the integrated autocorrelation times of the algorithm can be significantly reduced.

arXiv:cond-mat/9807137•International Journal of Modern Physics C•Jul 1, 1998•17 citations
It is shown that a certain class of generalizations of overrelaxation algorithms is incapable of further reducing the dynamical exponent zzz below its standard overrelaxed value of z\ensuremath≈1z\ensuremath{\approx}1z\ensuremath≈1. The mean-field value is unity and is obtained in a theory that is free in the static limit, while the effect of interactions and dimensionality could be estimated with dynamical renormalization-group methods. The generalizations are obtained by viewing overrelaxation as a slightly deformed deterministic algorithm and should, therefore, hold for hybrid Monte Carlo algorithms as well.
A novel dynamic Monte Carlo renormalization-group method is devised. Application to the kinetic Ising model in one dimension yields a value of zzz, the dynamic critical exponent, which is in excellent agreement with the exact result. In two dimensions we find $z=2.22\ifmmode\pm\else\textpm\fi{}0.13$ with the most probable value z=2.17z=2.17z=2.17.
The critical dynamics of the two dimensional Kinetic Ising Model have been studied using Monte Carlo Real Space Renormalization Group techniques. Time dependent correlation functions were calculated for square lattices of 16 x 16 and 32 x 32 spins, using periodic boundary conditions wth 60,000 and 15,000 Monte Carlo steps per spin, respectively;The critical exponent z was estimated using a matching process for a variety of renormalization group operators. The estimates for z were surprisingly sensitive to small changes in the value of the renormalization group operator, with physically reasonable operators giving estimates for z less than the rigorous lower bound for z of z(,o) = 1.75. The estimates for z from a given renormalization group operator showed poor convergence properties. Attempts to estimate z from the scaling form of the correlation function were inconclusive, as were attempts to estimate the static exponent (eta);Possible explanations of these unexpected results are discussed. The most plausible source of the complex behavior seen in this work is the discrete nature of time in the Monte Carlo computer simulation, which generates phenomena not present in standard theoretical treatments of the Kinetic Ising Model.