Abstract
This is a class B proposal, asking for continuation of time on the clusters to study the systematic errors of non-perturbative renormalization (NPR) with improved staggered fermions.
The project has two main focuses. First, the calculation of Z-factors for bilinear operators of all tastes and spins so as to allow a detailed study of the chiral and continuum extrapolations and the comparison with perturbation theory. In the 2010-11 allocation year, we calculated such Z-factors with valence asqtad and HYP-smeared fermions on the coarse MILC lattices, and during the present (2011-12) allocation year we have been extending these results to the fine lattices. Our allocation is sufficient for somewhat less than 8 configurations per mass, but we now think that 16 configurations per mass is preferable. Thus part of our request for the upcoming year is to allow this more than doubling of statistics.
The second focus is to implement state-of-the-art NPR techniques, in particular the use of non-exceptional momenta and twisted boundary conditions. Our main aim here is a calculation of $Z_m$. We have preliminary results using non-exceptional momenta, and find that these lead to the expected improvement in the chiral extrapolation and reduction in long-distance non-perturbative contributions. The second part of our request is for the continuation of this study.
We use the Chroma package, augmented by the addition of momentum sources and general bilinear operators. We are asking for time on clusters, and prefer to continue running at Fermilab.
We are requesting 2,500,000 Jpsi core-hours on clusters (165,000 of which are for storage).
Abstract
This is a class B proposal, asking for continuation of time on the clusters to calculate matching (``Z-'') factors using non-perturbative renormalization (NPR) for bilinear and four-fermion operators with staggered quarks.
By the end of the present (2010-11) allocation year, we will have completed the calculation of Z-factors for bilinear operators of all tastes and spins using valence asqtad and HYP-smeared fermions on the coarse MILC lattices, with some preliminary results on the fine lattices. This is the first such detailed study using improved staggered fermions. We have compared these to one-loop perturbative predictions, and find that ratios of Z-factors (which have no anomalous dimensions) for vector, axial-vector and tensor spins agree at the 1% level, while the agreement is 5-10% for the Z-factors themselves.
We propose here to extend this calculation in several ways. First, to complete the calculations on the fine MILC lattices. Second, to implement twisted boundary conditions for the valence quarks so as to allow continuous tuning of momenta. Third, aided by the use of twisted boundary conditions, to consider bilinears with non-exceptional momenta (thus reducing non-perturbative contamination). Our final goal is then to apply these improved methods to the calculation of four-fermion Z-factors with HYP-smeared staggered valence quarks, results which are needed to reduce and control the systematic error in the calculation of B_K and other kaon matrix elements.
We use the Chroma package, augmented by the addition of momentum sources and general bilinear operators. We are asking for time on clusters, and prefer to continue running at Fermilab.
We are requesting 2,500,000 Jpsi-equivalent node-hours on clusters (135,000 of which are for storage).
Abstract
This is a class B proposal, asking for continuation of time on the clusters to calculate matching (``Z'') factors using non-perturbative renormalization (NPR) for the quark mass, bilinear and four-fermion operators with staggered quarks.
During the present (2009-10) allocation year, we will have completed the calculation of the propagators and vertices necessary for determining the mass renormalization for asqtad quarks (as well as those for related bilinears) on coarse and fine MILC lattices. We propose here to extend this calculation to HYP-smeared staggered valence quarks, for both the bilinears and four-fermion operators. These matching factors are needed to reduce and control the systematic error in the calculation of B_K and other kaon matrix elements using HYP-fermions. Matching is the dominant source of error in this calculation.
We use the Chroma package, augmented by the addition of momentum sources. We are asking for time on clusters, and prefer to continue running at Fermilab.
We are requesting 1,640,000 Jpsi-equivalent node-hours on the clusters (140,000 of which are for storage).
Abstract
This is a class B proposal, asking for continuation of time on the clusters to calculate matching (``Z'') factors using non-perturbative renormalization (NPR) for the quark mass, bilinear and four-fermion operators with staggered quarks.
Due to a variety of factors, we have made less progress during the present (2008-9) allocation year than anticipated, and are likely only to have completed the calculation of the mass renormalization for asqtad quarks (as well as those for related bilinears) on coarse and fine MILC lattices. We propose here to extend this calculation to HYP-smeared staggered valence quarks, for both the bilinears and four-fermion operators. These matching factors are needed to reduce and control the systematic error in the calculation of B_K and other kaon matrix elements using HYP-fermions. Matching is, or is likely soon to be, the dominant source of error in this calculation. We propose to use the standard NPR methodology, adapted to staggered fermions, which our results to date indicate will be successful with relatively small ensembles. We use the Chroma package, augmented by the addition of momentum sources. We are asking for time on clusters, and prefer to continue running at Fermilab.
We are requesting 500,000 6n-equivalent node-hours on the clusters.
Abstract
This is a class B proposal, asking for time on the clusters to calculate matching (``Z'') factors using non-perturbative renormalization (NPR) for the quark mass, bilinear and four-fermion operators, using valence asqtad and HYP-smeared staggered quarks on a subset of the coarse and fine MILC lattices. These matching factors are needed to reduce and control the systematic error in the quark mass obtained by the MILC collaboration using asqtad fermions, and in the calculation of BK and other kaon matrix elements using HYP-fermions. For both calculations, matching is, or is likely soon to be, the dominant source of error. We propose to use the standard NPR methodology, adapted to staggered fermions. We use the Chroma package, augmented by the addition of momentum sources. We are asking for time on clusters, and have a slight preference for running at Fermilab (where we have done our test runs).
We are requesting 500,000 6n-equivalent node-hours on the clusters.