Choice of Integrator in the Hybrid Monte Carlo Algorithm

Abstract

We study efficiency of higher order integrator schemes for the hybrid Monte Carlo (HMC) algorithm. Numerical tests are performed for Quantum Chromo Dynamics (QCD) with two flavors of Wilson fermions. We compare 2nd, 4th and 6th order integrators at various quark masses. The performance depends on both volume and quark mass. On currently accessible large lattices ($V\sim 24^4$), higher order integrators can be more efficient than the 2nd order one only in heavy quark region, $m_{q}a>0.3$. Thus we conclude that for most full QCD simulations, except for heavy quark case, the usual 2nd order integrator is the best choice.

20 pages, 14 figures

References in corpus (1)

• Full QCD light hadron spectrum from the CP-PACS

Cited by in corpus (24)

• 2+1 Flavor Lattice QCD toward the Physical Point
• Higher representations on the lattice: numerical simulations. SU(2) with adjoint fermions
• Testing and tuning symplectic integrators for Hybrid Monte Carlo algorithm in lattice QCD
• Speeding up Lattice QCD simulations with clover-improved Wilson Fermions
• Sp(4) gauge theory on the lattice: towards SU(4)/Sp(4) composite Higgs (and beyond)
• Polynomial Hybrid Monte Carlo algorithm for lattice QCD with an odd number of flavors
• Shadow Hamiltonians, Poisson Brackets, and Gauge Theories
• Simulating the Schroedinger functional with two pseudo-fermions
• Higher Order Hybrid Monte Carlo at Finite Temperature
• Full QCD Algorithms towards the Chiral Limit
• Bayesian Inference of Stochastic Volatility Model by Hybrid Monte Carlo
• Bayesian estimation of GARCH model by hybrid Monte Carlo
• Applying polynomial filtering to mass preconditioned Hybrid Monte Carlo
• Financial Time Series Analysis of SV Model by Hybrid Monte Carlo
• Simulation of Wilson fermion at finite isospin density
• Hessian-free force-gradient integrators
• Empirical Analysis of Stochastic Volatility Model by Hybrid Monte Carlo Algorithm
• Bayesian estimation of realized stochastic volatility model by Hybrid Monte Carlo algorithm
• GPU Computing in Bayesian Inference of Realized Stochastic Volatility Model
• Recent algorithm and machine developments for lattice QCD
• A modified Cayley transform for SU(3) molecular dynamics simulations
• Extended framework for the hybrid Monte Carlo in lattice gauge theory
• Numerical stability of force-gradient integrators and their Hessian-free variants in lattice QCD simulations
• Nf=1 QCD simulation with improved gauge action at finite temperature