Implementing Hybrid Monte Carlo with stout-smeared chirally improved Dirac operators

TL;DR

This paper presents the development and testing of a Hybrid Monte Carlo algorithm for dynamical Ginsparg-Wilson fermions using a stout-smeared chirally improved Dirac operator, analyzing its chiral properties and performance.

Contribution

It introduces a new implementation of HMC for this specific Dirac operator and evaluates its effectiveness in dynamical simulations.

Findings

The operator's chiral properties are preserved in dynamical setups.

The algorithm shows favorable autocorrelation and performance metrics.

Spectral comparisons between dynamical and quenched cases are provided.

Abstract

We discuss our implementation of dynamical Ginsparg-Wilson type fermions using a stout-smeared chirally improved Dirac operator. Such operators have been studied extensively in quenched calculations within the Bern-Graz-Regensburg (BGR) collaboration. Here we discuss the development and testing of the Hybrid Monte Carlo algorithm with this Dirac operator. We study the chiral properties of this operator in a dynamical setup, comparing, e.g., the spectra of the operator for the dynamical and quenched cases. We then discuss quantitative features of the algorithm like autocorrelation and performance.