BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//CERN//INDICO//EN
BEGIN:VEVENT
SUMMARY:Local Adaptive Refinement on Lattice Gauge Fields
DTSTART;VALUE=DATE-TIME:20170629T104500Z
DTEND;VALUE=DATE-TIME:20170629T111500Z
DTSTAMP;VALUE=DATE-TIME:20210121T153628Z
UID:indico-contribution-625043-2617559@indico.cern.ch
DESCRIPTION:Speakers: Edward White\, Jr. ()\nAdaptive Mesh Reﬁnement (AM
R) has been widely used in computational ﬂuid dynamics\, shock hydrodyna
mics\, astrophysics\, turbulence modeling and combustion to improve the pe
rformance of algorithms running large\, complex problems. The results are
often impressive. However\, for various reasons which will be discussed\,
traditional AMR is not directly applicable to lattice QCD. This talk propo
ses a new numerical method\, Local Adaptive Reﬁnement (LAR)\, for lattic
e QCD. \nLAR is strongly motivated by AMR\, and is designed to work with t
he aggregation-based algebraic multigrid framework of lattice QCD. The pre
mise behind all adaptive reﬁnement methods is that certain numerical pro
blems have solutions that exhibit local variations\, which can be examined
more fully at higher resolution (more closely spaced grid or lattice poin
ts) than the base level of \nresolution used for the simulation. As a resu
lt\, problems that would normally require very high resolution over a larg
e domain (which is very computationally expensive) could instead be solved
at lower resolution over the large domain\, with automated local reﬁnem
ent to higher resolution as needed. The scale and nature of the problems c
ommonly found at the leading edge of today’s lattice QCD work suggest th
at the field could be a beneﬁt from adaptive reﬁnement. Among the topi
cs that will be discussed are the need for an algebraic variable-based app
roach to reﬁnement\, as opposed to a geometric grid-based approach\, met
hods of determining the regions in which reﬁnement should occur\, and wa
ys to implement LAR in a parallel machine environment.\n\nhttps://indico.c
ern.ch/event/625043/contributions/2617559/
LOCATION:
URL:https://indico.cern.ch/event/625043/contributions/2617559/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Algebraic Multigrid: Theory and Practice
DTSTART;VALUE=DATE-TIME:20170628T104500Z
DTEND;VALUE=DATE-TIME:20170628T113000Z
DTSTAMP;VALUE=DATE-TIME:20210121T153628Z
UID:indico-contribution-625043-2631505@indico.cern.ch
DESCRIPTION:Speakers: James Brannick ()\nThis talk gives an overview of r
ecent progress made in the design and analysis of algebraic multigrid meth
ods. The focus is on the setup algorithm that automatically constructs the
multilevel hierarchy used in the solve phase. A sharp two-grid theory is
introduced and then used to derive various quality measures of the coarse
spaces constructed by the setup algorithm\, based on the ideas of compatib
le relaxation\, a related identity that assumes the use of the so-called i
deal interpolation operator\, and an optimal form of classical algebraic m
ultigrid interpolation that gives the best possible two-grid convergence r
ate. Various numerical results are presented to illustrate these theoretic
al results. As a test problem\, we focus on a finite volume discretization
of a scalar diffusion problem with highly varying (discontinuous) diffusi
on coefficient.\n\nhttps://indico.cern.ch/event/625043/contributions/26315
05/
LOCATION:
URL:https://indico.cern.ch/event/625043/contributions/2631505/
END:VEVENT
BEGIN:VEVENT
SUMMARY:HMC: Geometry and Applications
DTSTART;VALUE=DATE-TIME:20170627T083000Z
DTEND;VALUE=DATE-TIME:20170627T091500Z
DTSTAMP;VALUE=DATE-TIME:20210121T153628Z
UID:indico-contribution-625043-2631580@indico.cern.ch
DESCRIPTION:Speakers: Alessandro Barp ()\nhttps://indico.cern.ch/event/625
043/contributions/2631580/
LOCATION:
URL:https://indico.cern.ch/event/625043/contributions/2631580/
END:VEVENT
BEGIN:VEVENT
SUMMARY:An Eigensolver for the Hermitian Dirac Operator with Multigrid Acc
eleration
DTSTART;VALUE=DATE-TIME:20170629T101500Z
DTEND;VALUE=DATE-TIME:20170629T104500Z
DTSTAMP;VALUE=DATE-TIME:20210121T153628Z
UID:indico-contribution-625043-2614503@indico.cern.ch
DESCRIPTION:Speakers: Artur Strebel (University of Wuppertal)\nIn this tal
k we present a Davidson type eigensolver combined with the DD-$\\alpha$AMG
multigrid solver library. The basic Davidson method is adjusted to our mu
ltigrid method and the structure of the hermitian Dirac operator in a way
that both methods benefit from each other.\n\nWe compare the resulting eig
ensolver with a Chebychev filtered Arnoldi method (PARPACK) and the multi
purpose eigensolver library PRIMME based on a variety of scaling and perfo
rmance studies.\n\nhttps://indico.cern.ch/event/625043/contributions/26145
03/
LOCATION:
URL:https://indico.cern.ch/event/625043/contributions/2614503/
END:VEVENT
BEGIN:VEVENT
SUMMARY:MILC code performance on high end CPU and GPU supercomputer cluste
rs
DTSTART;VALUE=DATE-TIME:20170627T163000Z
DTEND;VALUE=DATE-TIME:20170627T170000Z
DTSTAMP;VALUE=DATE-TIME:20210121T153628Z
UID:indico-contribution-625043-2614504@indico.cern.ch
DESCRIPTION:Speakers: Ruizi Li ()\nWith recent developments in parallel su
percomputing architecture\, \nmany core\, multi-core\, and GPU processors\
nare now commonplace resulting in more levels of parallelism\, memory\nhie
rarchy\, and programming complexity. It has been necessary to adapt\nthe
MILC code to these new processors starting with NVIDIA\nGPUs and more rece
ntly the Intel Xeon Phi processors. We report on \nour efforts to port an
d optimize our code for the Intel Knights Landing\narchitecture. We consi
der performance of the MILC code with MPI and OpenMP\, \nand optimizations
with QOPQDP and QPhiX. For the latter approach we \nconcentrate on the s
taggered conjugate gradient and gauge force. We also\nconsider performanc
e on recent NVIDIA GPUs using the QUDA library.\n\nhttps://indico.cern.ch/
event/625043/contributions/2614504/
LOCATION:
URL:https://indico.cern.ch/event/625043/contributions/2614504/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Algorithmic advances in NSPT
DTSTART;VALUE=DATE-TIME:20170628T083000Z
DTEND;VALUE=DATE-TIME:20170628T091500Z
DTSTAMP;VALUE=DATE-TIME:20210121T153628Z
UID:indico-contribution-625043-2614511@indico.cern.ch
DESCRIPTION:Speakers: Mattia Dalla Brida (Universita' di Milano-Bicocca &
INFN)\nNumerical stochastic perturbation theory (NSPT) is a powerful tool
that allows perturbation expansions in QCD and other interesting theories
to be estimated to high order in the interactions. The standard algorithm
s on which NSPT is based on\, however\, suffer from several limitations wh
ich in practice restrict the potential of these techniques. In this talk I
will review the recent algorithmic advances in this field and show how th
ese significantly reduce the computational effort for precise and accurate
determinations. This opens up the way to tackle challenging and interesti
ng new problems\, as will be illustrated by a highly non-trivial computati
on.\n\nhttps://indico.cern.ch/event/625043/contributions/2614511/
LOCATION:
URL:https://indico.cern.ch/event/625043/contributions/2614511/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Deflation for Monte-Carlo estimation of the trace of a matrix inve
rse
DTSTART;VALUE=DATE-TIME:20170627T091500Z
DTEND;VALUE=DATE-TIME:20170627T094500Z
DTSTAMP;VALUE=DATE-TIME:20210121T153628Z
UID:indico-contribution-625043-2614514@indico.cern.ch
DESCRIPTION:Speakers: Eloy Romero Alcalde (College of William and Mary )\n
In the context of computing disconnected diagrams\, we investigate the\nef
ficient estimation of the trace of large-scale matrix inverses. Our appro
ach\nis based on the Hutchinson method (Monte-Carlo averaging over matrix\
nquadratures). Previous work showed that combining deflation against the l
owest\npart of the spectrum with Hierarchical Probing can accelerate the c
onvergence\nsignificantly. As the size of the matrix grows\, however\, the
computation of\nenough singular vectors to achieve a reduction in the var
iance can be a\nbottleneck.\n\nIn this work we take advantage of the fact
that the singular vectors\ncorresponding to the lowest modes can be repres
ented well in a sparse basis.\nThis allows us to compute and store efficie
ntly the first 500~1000 lowest modes\nof the matrix spectrum\, which is en
ough to obtain a good reduction in the\nvariance. Moreover we discuss diff
erent projectors for deflating and the\nperformance impact when the matrix
is close to singular.\n\nhttps://indico.cern.ch/event/625043/contribution
s/2614514/
LOCATION:
URL:https://indico.cern.ch/event/625043/contributions/2614514/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Progress and Challenge of Lattice Quantum Finite Elements (QFE) on
Spheres
DTSTART;VALUE=DATE-TIME:20170627T101500Z
DTEND;VALUE=DATE-TIME:20170627T111500Z
DTSTAMP;VALUE=DATE-TIME:20210121T153628Z
UID:indico-contribution-625043-2614512@indico.cern.ch
DESCRIPTION:Speakers: Richard Brower (Boston University)\nExtending latti
ce field to ultraviolet complete quantum field theory on any smooth Rieman
n manifolds is a challenging problem. By adapting element methods (FEM) a
nd Regge geometry one recovers classical (IR) solution in the continuum.
However to correctly handle UV divergences requires new counter terms to c
onstruct a what we call a "Quantum Finite Elements" (QFE) discrete Lagran
gian on the simplicial complex. These UV counters for 2d phi 4th theory a
nd free fermions on the two sphere (S2) have been tested numerically to hi
gh precession against the exact Ising solution. Methods to generalize the
QFE construction to radial quantized 3d super renormalizable theories on
R x S2 and challenges for asymptotical free 4d gauge theories on R x S3 w
ill be presented.\n\nhttps://indico.cern.ch/event/625043/contributions/261
4512/
LOCATION:
URL:https://indico.cern.ch/event/625043/contributions/2614512/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Novel Approaches for Staggered Multigrid Algorithms
DTSTART;VALUE=DATE-TIME:20170629T091500Z
DTEND;VALUE=DATE-TIME:20170629T094500Z
DTSTAMP;VALUE=DATE-TIME:20210121T153628Z
UID:indico-contribution-625043-2614517@indico.cern.ch
DESCRIPTION:Speakers: Evan Weinberg ()\nCritical slowing down in the fermi
on sector is a leading obstacle facing the approach to the continuum in la
ttice gauge theory simulations. Adaptive multigrid ($\\alpha$-MG) methods
offer a solution to the resulting superlinear growth in the cost of iterat
ive Krylov solves. Ongoing research has suggested that previously develope
d $\\alpha$-MG methods\, such as the successful formulation for Wilson-clo
ver fermions\, cannot be applied in a black-box fashion to other discretiz
ations. In this talk I will expand on subtle issues offered by the stagger
ed discretization. I will also discuss novel approaches that may sidestep
these issues.\n\nhttps://indico.cern.ch/event/625043/contributions/2614517
/
LOCATION:
URL:https://indico.cern.ch/event/625043/contributions/2614517/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Scaling Multigrid to the Exascale
DTSTART;VALUE=DATE-TIME:20170629T083000Z
DTEND;VALUE=DATE-TIME:20170629T091500Z
DTSTAMP;VALUE=DATE-TIME:20210121T153628Z
UID:indico-contribution-625043-2614518@indico.cern.ch
DESCRIPTION:Speakers: Kate Clark (NVIDIA)\nOwing to its success in removin
g the critical slowing down of Dirac linear systems\, adaptive multigrid i
s now a standard solver in the arsenal of tools that the lattice field the
orist expects. In this work we report on the latest progress in improving
the strong scaling of adaptive multigrid algorithms when running on GPU-a
ccelerated architectures using the QUDA library. Techniques include Schwa
rz preconditioning\, pipelined solvers\, precision truncation and RDMA-ena
bled MPI. Furthermore\, we report on progress on optimizing the adaptive
setup process in order to increase its applicability to Hybrid Monte Carlo
. Finally we discuss the challenges in scaling multigrid to the Exascale-
generation of supercomputers.\n\nhttps://indico.cern.ch/event/625043/contr
ibutions/2614518/
LOCATION:
URL:https://indico.cern.ch/event/625043/contributions/2614518/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Quark-antiquark excited flux tube
DTSTART;VALUE=DATE-TIME:20170627T143000Z
DTEND;VALUE=DATE-TIME:20170627T150000Z
DTSTAMP;VALUE=DATE-TIME:20210121T153628Z
UID:indico-contribution-625043-2615893@indico.cern.ch
DESCRIPTION:Speakers: Nuno Cardoso (IST)\nWe present color field profiles
for some of the first SU(3) gluonic excitations of the flux tube in the pr
esence of a static quark-antiquark pair.\n\nhttps://indico.cern.ch/event/6
25043/contributions/2615893/
LOCATION:
URL:https://indico.cern.ch/event/625043/contributions/2615893/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Numerical Stochastic Perturbation Theory in $\\varphi^4$ Theory
DTSTART;VALUE=DATE-TIME:20170628T091500Z
DTEND;VALUE=DATE-TIME:20170628T094500Z
DTSTAMP;VALUE=DATE-TIME:20210121T153628Z
UID:indico-contribution-625043-2628611@indico.cern.ch
DESCRIPTION:Speakers: Marco Garofalo (Higgs Centre for Theoretical Physics
\, The University of Edinburgh)\nNumerical stochastic perturbation theory
(NSPT) is a powerful tool for estimating high-order perturbative expansion
s in lattice field theory. The standard NSPT is based on the Langevin equa
tion. In this contribution\, we investigate in $\\varphi^4$ theory some al
ternative methods. In particular\, we present a study of the recently prop
osed Instantaneous Stochastic Perturbation Theory\, as well as a formulati
on of numerical stochastic perturbation theory based on Generalised Hybrid
Molecular Dynamics algorithms.\n\nhttps://indico.cern.ch/event/625043/con
tributions/2628611/
LOCATION:
URL:https://indico.cern.ch/event/625043/contributions/2628611/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Lattice QCD with mixed action - Borici-Creutz valence quarks on st
aggered sea
DTSTART;VALUE=DATE-TIME:20170627T140000Z
DTEND;VALUE=DATE-TIME:20170627T143000Z
DTSTAMP;VALUE=DATE-TIME:20210121T153628Z
UID:indico-contribution-625043-2631514@indico.cern.ch
DESCRIPTION:Speakers: Subhasish Basak (NISER)\nWe used Borici-Creutz fermi
on to study discrete chiral symmetry breaking\nat strong coupling in 2-dim
Gross-Neveu model and mass spectra in 2-dim\nfield theories. Mixed actio
n lattice QCD study with Borici-Creutz valence\nquarks on staggered sea qu
arks is carried out. The counter terms are fixed\nnonperturbatively to res
tore the broken symmetries. The effect of partial\nquenching and unitarity
violation is investigated.\n\nhttps://indico.cern.ch/event/625043/contrib
utions/2631514/
LOCATION:
URL:https://indico.cern.ch/event/625043/contributions/2631514/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Lattice Quantum Field Theory: Numerical Integration in an Infinite
Number of Dimensions
DTSTART;VALUE=DATE-TIME:20170627T130000Z
DTEND;VALUE=DATE-TIME:20170627T140000Z
DTSTAMP;VALUE=DATE-TIME:20210121T153628Z
UID:indico-contribution-625043-2631634@indico.cern.ch
DESCRIPTION:Speakers: Anthony Kennedy (Higgs Centre for Theoretical Physic
s\, The University of Edinburgh)\nRelativistic Quantum Field Theory (QFT)
is the formalism upon which the Standard Model of particle physics is buil
t\, and it is remarkably successful and accurate. When the interactions ar
e weak then the techniques of renormalized perturbation theory using Feynm
an diagrams works beautifully\, but when the interactions are strong we ha
ve to turn to numerical evaluation of the functional integrals that define
the quantum theory. The only way we know of to evaluate such infinite-dim
ensional integrals is to use Markov Chain Monte Carlo (MCMC) techniques. I
shall attempt to give an overview of the functional integral formalism of
QFT\, and how MCMC integration works. If time permits I will discuss how
fermions (half-integer spin) are dealt with\, and maybe introduce the Hybr
id (or Hamiltonian) Monte Carlo algorithm.\n\nhttps://indico.cern.ch/event
/625043/contributions/2631634/
LOCATION:
URL:https://indico.cern.ch/event/625043/contributions/2631634/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Nonperturbative anisotropy calibration in lattice QCD at strong co
upling
DTSTART;VALUE=DATE-TIME:20170628T101500Z
DTEND;VALUE=DATE-TIME:20170628T104500Z
DTSTAMP;VALUE=DATE-TIME:20210121T153628Z
UID:indico-contribution-625043-2631943@indico.cern.ch
DESCRIPTION:Speakers: Helvio Jose Caleiro Vairinhos (Fisica Teorica e Met.
Mat.-Inst. Nacional de Invest. Cientif.-Un)\nWe propose a simple criterio
n for the nonperturbative renormalization of the anisotropy coupling in la
ttice QCD with massless staggered fermions\, in the strong coupling limit.
We compute numerically and to high precision the renormalised anisotropy\
, and the analogue of Karsch’s coefficients\, using diagrammatic Monte C
arlo algorithms and multi-histogram reweighting. We observe a large nonper
turbative correction to the mean field anisotropy\, and we analyse the imp
lications of such a correction on the continuous time limits of the phase
diagram of lattice QCD at strong coupling\, and of the baryon mass.\n\nhtt
ps://indico.cern.ch/event/625043/contributions/2631943/
LOCATION:
URL:https://indico.cern.ch/event/625043/contributions/2631943/
END:VEVENT
BEGIN:VEVENT
SUMMARY:GPU Computing to the Exascale and Beyond
DTSTART;VALUE=DATE-TIME:20170628T130000Z
DTEND;VALUE=DATE-TIME:20170628T140000Z
DTSTAMP;VALUE=DATE-TIME:20210121T153628Z
UID:indico-contribution-625043-2633073@indico.cern.ch
DESCRIPTION:Speakers: Kate Clark (NVIDIA)\nWith the demise of Denard scali
ng\, it is well known that we cannot go faster for greater throughput\, ra
ther we have to go wider. However\, with the imminent demise of Moore's L
aw\, there lies continued challenges in reaching and exceeding the Exascal
e. We discuss how and why GPU computing provides a solution to take compu
tational science to this next level. We consider some of the software and
algorithm challenges that the Exascale will bring. Lastly\, we take a di
ve into the newly-launched Volta GPU architecture\, the architecture that
will drive the next-generation CORAL Supercomputers.\n\nhttps://indico.cer
n.ch/event/625043/contributions/2633073/
LOCATION:
URL:https://indico.cern.ch/event/625043/contributions/2633073/
END:VEVENT
BEGIN:VEVENT
SUMMARY:DD-αAMG on QPACE 3
DTSTART;VALUE=DATE-TIME:20170627T153000Z
DTEND;VALUE=DATE-TIME:20170627T163000Z
DTSTAMP;VALUE=DATE-TIME:20210121T153628Z
UID:indico-contribution-625043-2633093@indico.cern.ch
DESCRIPTION:Speakers: Daniel Ritchmann (University of Regensburg )\, Peter
Georg (University of Regensburg)\nWe describe our experiences porting the
Regensburg implementation of\nthe DD-αAMG solver from QPACE 2 to QPACE 3
. We first review porting from\nthe first generation Intel Xeon Phi proces
sor (Knights Corner) to its\nsuccessor (Knights Landing). Secondly\, we pr
esent an overview of\nOmni-Path\, comparing it to the well-known competito
r InfiniBand.\nFinally\, we present the performance of the code on a singl
e processor as\nwell as the scaling on many nodes.\n\nhttps://indico.cern.
ch/event/625043/contributions/2633093/
LOCATION:
URL:https://indico.cern.ch/event/625043/contributions/2633093/
END:VEVENT
BEGIN:VEVENT
SUMMARY:HMC\, high level structures and architecture support in Grid
DTSTART;VALUE=DATE-TIME:20170628T140000Z
DTEND;VALUE=DATE-TIME:20170628T144500Z
DTSTAMP;VALUE=DATE-TIME:20210121T153628Z
UID:indico-contribution-625043-2641604@indico.cern.ch
DESCRIPTION:Speakers: Guido Cossu (The University of Edinburgh)\nI will pr
esent the strategies we developed in Grid for the HMC sector\, in order to
support a variety of behaviours without code replication. I will also dis
cuss the current status of the architecture support in view of the upcomin
g machines.\n\nhttps://indico.cern.ch/event/625043/contributions/2641604/
LOCATION:
URL:https://indico.cern.ch/event/625043/contributions/2641604/
END:VEVENT
BEGIN:VEVENT
SUMMARY:Analyzing AMA data on 48^3 × 96 lattices
DTSTART;VALUE=DATE-TIME:20170628T144500Z
DTEND;VALUE=DATE-TIME:20170628T151000Z
DTSTAMP;VALUE=DATE-TIME:20210121T153628Z
UID:indico-contribution-625043-2641597@indico.cern.ch
DESCRIPTION:Speakers: Oliver Witsel ()\nAs an example for analyzing data o
n RBC-UKQCD's 48^3 x 96 ensemble with physical pions\, I investigate corre
lations between propagator sources placed on the same configuration and di
scuss options for analyzing this data set to extract the Bs-meson decay co
nstant. The data are generated taking advantage of all-mode averaging (AMA
) to reduce the numerical costs.\n\nhttps://indico.cern.ch/event/625043/co
ntributions/2641597/
LOCATION:
URL:https://indico.cern.ch/event/625043/contributions/2641597/
END:VEVENT
END:VCALENDAR