Polynomial filtered HMC–an algorithm for lattice QCD with dynamical quarks Original Research Article

Polynomial approximations to the inverse of the fermion matrix are used to filter the dynamics of the upper energy scales in Hybrid Monte Carlo (HMC) simulations. The use of a multiple time-scale integration scheme then allows the filtered pseudofermions to be evolved using a coarse step size. The algorithm is applied to image flavour simulations at two different quark masses. Polynomial filtering yields a reduction in the computational expense of the molecular dynamics integration of between three- and fivefold when compared to standard HMC. Significantly, the observed speedup is better at the lower quark mass.