Dice Question Streamline Icon: https://streamlinehq.com

Effectiveness of open boundary conditions at fine lattice spacings

Determine the effectiveness of open boundary conditions in the time direction for lattice QCD ensemble generation at lattice spacings 1/a approximately 3–5 GeV, specifically whether such open boundaries achieve decorrelation of the topological charge across the bulk of the lattice beyond the small number of time slices near the boundaries, as assessed using Wilson flow and the 5Li definition of topological charge per time slice Q(t, τ).

Information Square Streamline Icon: https://streamlinehq.com

Background

Reducing autocorrelation, particularly of the topological charge, is a major challenge in lattice QCD as lattice spacing decreases. Open boundary conditions in the time direction have been proposed to mitigate critical slowing down. The authors present preliminary observations of Q(t, τ) after Wilson flow (t_f=32) that suggest the impact of open boundaries may be confined to regions near the temporal boundaries.

In this context, the authors explicitly state that the overall effectiveness of open boundary conditions for ensembles at fine lattice spacings (1/a ∼ 3–5 GeV) remains unclear, motivating a focused determination of their efficacy in decorrelating long-distance modes across the lattice volume.

References

However, our preliminary study of topological charge per time slice $Q(t,\tau)$ measured by 5Li definition after Wilson flow with flow time $t_f=32$, as shown in Figure \ref{fig:top_t}, suggests the effect of the open boundary on the decorrelation the topological charge may be limited to the small number of time slices near the boundary, and the effectiveness of open boundaries in the ensemble generation for the lattice spacings in the range of $1/a \sim $ 3-5Gev is not clear.

Riemannian Manifold HMC with fermions (2401.13226 - Jung et al., 24 Jan 2024) in Section 1 (Introduction), paragraph discussing Figure 1 (Topological charge per timeslice)