Peter A. Boyle

The IBM Blue Gene/Q Compute Chip

Blue Gene/Q aims to build a massively parallel high-performance computing system out of power-efficient processor chips, resulting in power-efficient, cost-efficient, and floor-space- efficient systems. Focusing on reliability during design helps with scaling to large systems and lowers the total cost of ownership. This article examines the architecture and design of the Compute chip, which combines processors, memory, and communication functions on a single chip.

Physical results from 2+1 flavor domain wall QCD and SU(2) chiral perturbation theory

We have simulated QCD using 2+1 flavors of domain wall quarks on a (2.74fm)3 volume with an inverse lattice scale of a?1=1.729(28) GeV. The up and down (light) quarks are degenerate in our calculations and we have used four values for the ratio of light quark masses to the strange (heavy) quark mass in our simulations: 0.217, 0.350, 0.617 and 0.884. We have measured pseudoscalar meson masses and decay constants, the kaon bag parameter BK and vector meson couplings. We have used SU(2) chiral perturbation theory, which assumes only the up and down quark masses are small, and SU(3) chiral perturbation theory to extrapolate to the physical values for the light quark masses. While next-to-leading order formulae from both approaches fit our data for light quarks, we find the higher order corrections for SU(3) very large, making such fits unreliable. We also find that SU(3) does not fit our data when the quark masses are near the physical strange quark mass. Thus, we rely on SU(2) chiral perturbation theory for accurate results. We use the masses of the ? baryon, and the ? and K mesons to set the lattice scale and determine the quark masses. We then find f?=124.1(3.6)stat(6.9)systMeV, fK=149.6(3.6)stat(6.3)systMeV and fK/f?=1.205(0.018)stat(0.062)syst. Using non-perturbative renormalization to relate lattice regularized quark masses to RI-MOM masses, and perturbation theory to relate these to MS¯ we find mMS¯ud(2GeV)=3.72(0.16)stat(0.33)ren(0.18)systMeV and mMS¯s(2GeV)=107.3(4.4)stat(9.7)ren(4.9)systMeV.

Continuum limit physics from 2+1 flavor domain wall QCD

We present physical results obtained from simulations usin g 2+1 flavors of domain wall quarks and the Iwasaki gauge action at two values of the lattice spac ing a, (a−1= 1.73 (3) GeV and a−1= 2.28 (3) GeV). On the coarser lattice, with 24 3×64×16 points (where the 16 corresponds to Ls, the extent of the 5 th dimension inherent in the domain wall fermion (DWF) formula tion

Domain wall QCD with physical quark masses

We present results for several light hadronic quantities ( f π , f K , B K , m ud , m s , t 0 ½, w 0 ) obtained from simulations of 2+1 flavor domain wall lattice QCD with large physical volumes and nearly physical pion masses at two lattice spacings. We perform a short, O (3) %, extrapolation in pion mass to the physical values by combining our new data in a simultaneous chiral/continuum “global fit” with a number of other ensembles with heavier pion masses. We use the physical values of m π , m K and m Ω to determine the two quark masses and the scale—all other quantities are outputs from our simulations. We obtain results with subpercent statistical errors and negligible chiral and finite-volume systematics for these light hadronic quantities, including f π = 130.2(9) MeV; f K = 155.5(8) MeV; the average up/down quark mass and strange quark mass in the ‾MS scheme at 3 GeV, 2.997(49) and 81.64(1.17) MeV respectively; and the neutral kaon mixing parameter, B K , in the renormalization group invariant scheme, 0.750(15) and the ‾MS scheme at 3 GeV, 0.530(11).

Lattice Determination of the Hadronic Contribution to the Muon

We present a calculation of the leading order hadronic contribution to the anomalous magnetic moment of the muon for a dynamical simulation of 2+1 avour QCD using domain wall fermions. The electromagnetic 2-point function is evaluated on the RBC-UKQCD lattice gauge congurations and this is tted to a continuous form motivated by models of vector dominance. We determine a robust and reliable technique for performing this t, allowing us to extract the most accurate results possible from our ensembles. This combined with data at very light quark masses produces the result a (2)had = 641(33)(32) 10 10 at the physical point, where the rst uncertainty is statistical, and the second is an estimate of systematics, which is in agreement with previous results. We outline various methods by which this calculation can and will be improved in order to compete with the accuracy of alternative techniques of deducing this quantity from experimental scattering data.

g-2

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 × 32 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 sea l ≈ 0.85m s , 0.59m s and 0.33m s were used to allow for extrapolations to the physical light quark limit, while 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.

using Dynamical Domain Wall Fermions

We present the first results for the K13 form factor from simulations with 2+1 flavors of dynamical domain wall quarks. Combining our result, namely, f+(0)=0.964(5) with the latest experimental results for Kl3 decays leads to |V us|=0.2249(14), reducing the uncertaintity in this important parameter. For the O(p6) term in the chiral expansion we obtain Delta f=-0.013(5).

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

We discuss in general the construction of gauge-invariant nonlocal meson operators on the lattice. We use such operators to study the {ital P}- and {ital D}-wave mesons as well as hybrid mesons in quenched QCD, with quark masses near the strange quark mass. The resulting spectra are compared with experiment for the orbital excitations. For the states produced by gluonic excitations (hybrid mesons) we find evidence of mixing for nonexotic quantum numbers. We give predictions for masses of the spin-exotic hybrid mesons with {ital J}{sup {ital PC}}=1{sup {minus}+}, 0{sup +{minus}}, and 2{sup +{minus}}. {copyright} {ital 1996 The American Physical Society.}

Kl3 semileptonic form factor from 2 + 1 flavor lattice QCD

The QCDSP and QCDOC computers are two generations of multithousand-node multidimensional mesh-based computers designed to study quantum chromodynamics (QCD), the theory of the strong nuclear force. QCDSP (QCD on digital signal processors), a four-dimensional mesh machine, was completed in 1998; in that year, it won the Gordon Bell Prize in the price/performance category. Two large installations--of 8,192 and 12,288 nodes, with a combined peak speed of one teraflops--have been in operation since. QCD-on-a-chip (QCDOC) utilizes a sixdimensional mesh and compute nodes fabricated with IBM systemon-a-chip technology. It offers a tenfold improvement in price/ performance. Currently, 100-node versions are operating, and there are plans to build three 12,288-node, 10-teraflops machines. In this paper, we describe the architecture of both the QCDSP and QCDOC machines, the operating systems employed, the user software environment, and the performance of our application-- lattice QCD.

Orbitally excited and hybrid mesons from the lattice.

We present the first results for neutral-kaon mixing using (2+1)-flavors of domain-wall fermions. A new approach is used to extrapolate to the physical up and down quark masses from our numerical studies with pion masses in the range 240-420 MeV; only SU(2)_{L}xSU(2)_{R} chiral symmetry is assumed and the kaon is not assumed to be light. Our main result is B_{K};{MS[over ]}(2 GeV)=0.524(10)(28) where the first error is statistical and the second incorporates estimates for all systematic errors.