Searcharxiv⌕ Search

arXiv subjects

A. D. Kennedy

Publications and source records attributed to A. D. Kennedy.

At least 37 records · Page 2Linked to original sources

Speeding up HMC with better integrators

We discuss how dynamical fermion computations may be made yet cheaper by using symplectic integrators that conserve energy much more accurately without decreasing the integration step size. We first explain why symplectic integrators exactly conserve a ``shadow'' Hamiltonian close to the desired one, and how this Hamiltonian may be computed in terms of Poisson brackets. We then discuss how classical mechanics may be implemented on Lie groups and derive the form of the Poisson brackets and force terms for some interesting integrators such as those making use of second derivatives of the action (Hessian or force gradient integrators). We hope that these will be seen to greatly improve energy conservation for only a small additional cost and that their use will significantly reduce the cost of dynamical fermion computations.

hep-lat↗

2+1 flavor domain wall QCD on a (2 fm)^3 lattice: light meson spectroscopy with Ls = 16

We present results for light meson masses and pseudoscalar decay constants from the first of a series of lattice calculations with 2+1 dynamical flavors of domain wall fermions and the Iwasaki gauge action. The work reported here was done at a fixed lattice spacing of about 0.12 fm on a 16^3\times32 lattice, which amounts to a spatial volume of (2 fm)^3 in physical units. The number of sites in the fifth dimension is 16, which gives m_{res} = 0.00308(4) in these simulations. Three values of input light sea quark masses, m_l^{sea} \approx 0.85 m_s, 0.59 m_s and 0.33 m_s were used to allow for extrapolations to the physical light quark limit, whilst the heavier sea quark mass was fixed to approximately the physical strange quark mass m_s. The exact rational hybrid Monte Carlo algorithm was used to evaluate the fractional powers of the fermion determinants in the ensemble generation. We have found that f_π= 127(4) MeV, f_K = 157(5) MeV and f_K/f_π= 1.24(2), where the errors are statistical only, which are in good agreement with the experimental values.

hep-lat↗

Accelerating Staggered Fermion Dynamics with the Rational Hybrid Monte Carlo (RHMC) Algorithm

Improved staggered fermion formulations are a popular choice for lattice QCD calculations. Historically, the algorithm used for such calculations has been the inexact R algorithm, which has systematic errors that only vanish as the square of the integration step-size. We describe how the exact Rational Hybrid Monte Carlo (RHMC) algorithm may be used in this context, and show that for parameters corresponding to current state-of-the-art computations it leads to a factor of approximately seven decrease in cost as well as having no step-size errors.

hep-lat↗

Asymptotics of Fixed Point Distributions for Inexact Monte Carlo Algorithms

We introduce a simple general method for finding the equilibrium distribution for a class of widely used inexact Markov Chain Monte Carlo algorithms. The explicit error due to the non-commutivity of the updating operators when numerically integrating Hamilton's equations can be derived using the Baker-Campbell-Hausdorff formula. This error is manifest in the conservation of a ``shadow'' Hamiltonian that lies close to the desired Hamiltonian. The fixed point distribution of inexact Hybrid algorithms may then be derived taking into account that the fixed point of the momentum heatbath and that of the molecular dynamics do not coincide exactly. We perform this derivation for various inexact algorithms used for lattice QCD calculations.

hep-lat↗

First results from 2+1-Flavor Domain Wall QCD: Mass Spectrum, Topology Change and Chiral Symmetry with $L_s=8$

We present results for the static interquark potential, light meson and baryon masses, and light pseudoscalar meson decay constants obtained from simulations of domain wall QCD with one dynamical flavour approximating the $s$ quark, and two degenerate dynamical flavours with input bare masses ranging from $m_s$ to $m_s/4$ approximating the $u$ and $d$ quarks. We compare these quantities obtained using the Iwasaki and DBW2 improved gauge actions, and actions with larger rectangle coefficients, on $16^3\times32$ lattices. We seek parameter values at which both the chiral symmetry breaking residual mass due to the finite lattice extent in the fifth dimension and the Monte Carlo time history for topological charge are acceptable for this set of quark masses at lattice spacings above 0.1 fm. We find that the Iwasaki gauge action is best, demonstrating the feasibility of using QCDOC to generate ensembles which are good representations of the QCD path integral on lattices of up to 3 fm in spatial extent with lattice spacings in the range 0.09-0.13 fm. Despite large residual masses and a limited number of sea quark mass values with which to perform chiral extrapolations, our results for light hadronic physics scale and agree with experimental measurements within our statistical uncertainties.

hep-lat↗

Accelerating Dynamical Fermion Computations using the Rational Hybrid Monte Carlo (RHMC) Algorithm with Multiple Pseudofermion Fields

There has been much recent progress in the understanding and reduction of the computational cost of the Hybrid Monte Carlo algorithm for Lattice QCD as the quark mass parameter is reduced. In this letter we present a new solution to this problem, where we represent the fermionic determinant using (n) pseudofermion fields, each with an (\nth) root kernel. We implement this within the framework of the Rational Hybrid Monte Carlo (RHMC) algorithm. We compare this algorithm with other recent methods in this area and find it is competitive with them.

hep-lat↗

Light meson masses and decay constants in 2+1 flavour domain wall QCD

We present results for light meson masses and psedoscalar meson decay constants in 2+1 flavour domain wall QCD with the DBW2 and Iwasaki gauge actions, using lattices with linear sizes in the range 1.6 to 2.2fm and $u$ and $d$ quark masses as low as one quarter of the strange quark mass. All data were generated on the QCDOC machines at the University of Edinburgh and Brookhaven National Laboratory. Despite large residual masses and a limited number of sea quark mass values with which to perform chiral extrapolations, our results agree with experiment and scale within errors.

hep-lat↗

Baryons in 2+1 flavour domain wall QCD

We present results for some of the light baryon masses and their excited states in 2+1 flavour domain wall QCD. We considered several lattice spacings, with the DBW2 and Iwasaki gauge actions and different sea quark masses on a volume of $16^3\times32$ and a fifth dimension of size 8. All data were generated on the QCDOC machines. Despite large residual massses and a limited number of sea quark mass values with which to perform chiral extrapolations, our results are in reasonable agreement with experiment and scale within errors. Finite size effects on most ensembles appear to be small.

hep-lat↗

Algorithm Shootout: R versus RHMC

We present initial results comparing the RHMC and R algorithms on large lattices with small quark masses using chiral fermions. We also present results concerning staggered fermions near the deconfinement/chiral phase transition. We find that the RHMC algorithm not only eliminates the step-size error of the R algorithm, but is also considerably more efficient. We discuss several possibilities for further improvement to the RHMC algorithm.

hep-lat↗

Fast Evaluation of Zolotarev Coefficients

We review the theory of elliptic functions leading to Zolotarev's formula for the sign function over the range (ε\leq|x| \leq1). We show how Gauss' arithmetico-geometric mean allows us to evaluate elliptic functions cheaply, and thus to compute Zolotarev coefficients ``on the fly'' as a function of (ε). This in turn allows us to calculate the matrix functions (\sgn H), (\sqrt H), and (1/\sqrt H) both quickly and accurately for any Hermitian matrix (H) whose spectrum lies in the specified range.

hep-lat↗

Algorithms for Lattice QCD with Dynamical Fermions

We consider recent progress in algorithms for generating gauge field configurations that include the dynamical effects of light fermions. We survey what has been achieved in recent state-of-the-art computations, and examine the trade-offs between performance and control of systematic errors. We briefly review the use of polynomial and rational approximations in Hybrid Monte Carlo algorithms, and some of the theory of on-shell chiral fermions on the lattice. This provides a theoretical framework within which we compare algorithmic alternatives for their implementation; and again we examine the trade-offs between speed and error control.

hep-lat↗

Exact 2+1 flavour RHMC simulations

We consider the Rational Hybrid Monte Carlo algorithm for performing exact 2+1 flavour fermion simulations. The specific cases of ASQTAD and domain wall fermions are considered. We find that in both cases the naive performance is similar to conventional hybrid algorithms.

hep-lat↗

Accelerating Fermionic Molecular Dynamics

We consider how to accelerate fermionic molecular dynamics algorithms by introducing n pseudofermion fields coupled with the nth root of the fermionic kernel. This reduces the maximum pseudofermionic force, and thus allows a larger molecular dynamics integration step size without hitting an instability in the integrator.

hep-lat↗

Structure functions and form factors close to the chiral limit from lattice QCD

Results for nucleon matrix elements (arising from moments of structure functions) and form factors from a mixture of runs using Wilson, clover and overlap fermions (both quenched and unquenched) are presented and compared in an effort to explore the size of the chiral `regime', lattice spacing errors and quenching artefacts. While no run covers this whole range of effects the partial results indicate a picture of small lattice spacing errors, small quenching effects and only reaching the chiral regime at rather light quark masses.

hep-lat↗

Approximation Theory for Matrices

We review the theory of optimal polynomial and rational Chebyshev approximations, and Zolotarev's formula for the sign function over the range (ε\leq |z| \leq1). We explain how rational approximations can be applied to large sparse matrices efficiently by making use of partial fraction expansions and multi-shift Krylov space solvers.

hep-lat↗

Quark spectra and light hadron phenomenology from overlap fermions with improved gauge field action

We present first results from a simulation of quenched overlap fermions with improved gauge field action. Among the quantities we study are the spectral properties of the overlap operator, the chiral condensate and topological charge, quark and hadron masses, and selected nucleon matrix elements. To make contact with continuum physics, we compute the renormalization constants of quark bilinear operators in perturbation theory and beyond.

hep-lat↗

The RHMC Algorithm for 2 Flavours of Dynamical Staggered Fermions

We describe an implementation of the Rational Hybrid Monte Carlo (RHMC) algorithm for dynamical computations with two flavours of staggered quarks. We discuss several variants of the method, the performance and possible sources of error for each of them, and we compare the performance and results to the inexact R algorithm.

hep-lat↗

The Paraldor Project

Paraldor is an experiment in bringing the power of categorical languages to lattice QCD computations. Our target language is Aldor, which allows the capture of the mathematical structure of physics directly in the structure of the code using the concepts of categories, domains and their inter-relationships in a way which is not otherwise possible with current popular languages such as Fortran, C, C++ or Java. By writing high level physics code portably in Aldor, and implementing switchable machine dependent high performance back-ends in C or assembler, we gain all the power of categorical languages such as modularity, portability, readability and efficiency.

hep-lat↗