Speaker
Description
The low-mode contribution to the hadronic vacuum polarization dominates the total statistical error, making its efficient computation a critical bottleneck in lattice QCD determinations of the muon anomalous magnetic moment. We present and validate a strategy that leverages already available Fermilab Lattice/HPQCD/MILC HISQ eigenvectors to improve the statistics of our existing fat link calculation without generating new eigenvectors (except a small number for bias correction). Using the decomposition $D_{HISQ} = D_{Fat} + D_{Naik}$ for the HISQ Dirac operator, fat-link correlators are reconstructed from the known HISQ eigensystem at negligible, or no additional cost. We use an AMA-style bias correction computed on a small set of configurations to eliminate the systematic difference between the two actions. We demonstrate the method on a $48^3$, a=0.12 fm ensemble and then turn to the application of the method to a larger $144^3$, a=0.042 fm ensemble. Both ensembles are from the MILC collaboration.