BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:America/Chicago
X-LIC-LOCATION:America/Chicago
BEGIN:DAYLIGHT
TZOFFSETFROM:-0600
TZOFFSETTO:-0500
TZNAME:CDT
DTSTART:19700308T020000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=2SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:-0500
TZOFFSETTO:-0600
TZNAME:CST
DTSTART:19701101T020000
RRULE:FREQ=YEARLY;BYMONTH=11;BYDAY=1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20211207T055405Z
LOCATION:Online
DTSTART;TZID=America/Chicago:20211119T083000
DTEND;TZID=America/Chicago:20211119T120000
UID:submissions.supercomputing.org_SC21_sess334@linklings.com
SUMMARY:12th Workshop on Latest Advances in Scalable Algorithms for Large
Scale Systems
DESCRIPTION:Workshop\n\nUsability of Markov Chain Monte Carlo Precondition
ers in Practical Problems\n\nLebedev, Alexandrov, Sahin\n\nIn this paper w
e present the results of our exploration of applicability of preconditione
rs computed using the Markov Chain Monte Carlo Matrix Inversion ((MC) 2 MI
) method to a variety of linear systems from the domain of quantum chromod
ynamics, plasma physics and engineering. The latter two are rep...\n\n----
-----------------\nOptimized Cascadic Multigrid Parareal Method for Explic
it Time-Marching Scheme\n\nChen, Nakajima\n\nHigh-performance computing re
search is entering the exascale computing era. Large-scale simulations hav
e more than enough parallel resources but reach a saturation point in spat
ial parallelization due to the communication cost and synchronization over
head. Parallel-in-time (PinT) methods are a solut...\n\n------------------
---\nInvited Talk: Intelligent Simulations – How Combining AI and HPC Can
Enable New Discoveries\n\nFoster\n\nThe search for ever-more accurate and
detailed simulations of physical phenomenon has driven decades of improve
ments in both supercomputer architecture and computational methods. It see
ms increasingly likely that the next several orders of magnitude improveme
nts are likely to come, at least in part,...\n\n---------------------\nMor
ning Break\n\n\n\n---------------------\nWorkshop Opening\n\nAlexandrov, E
ngelmann\n\n---------------------\nWorkshop Closing\n\nAlexandrov, Engelma
nn\n\n---------------------\n12th Workshop on Latest Advances in Scalable
Algorithms for Large-Scale Systems\n\nAlexandrov, Dongarra, Geist, Engelma
nn\n\nNovel scalable scientific algorithms are needed to enable key scienc
e applications to exploit the computational power of large scale systems.
These extreme scale algorithms need to hide network and memory latency, ha
ve very high computation/communication overlap and minimal communication a
nd have no...\n\n---------------------\nUnleashing the Performance of bmSp
arse for the Sparse Matrix Multiplication in GPUs\n\nBerger, Freire, Marin
i, Dufrechou, Ezzatti\n\nThe evolution of data science and machine learnin
g has increased the applicability of the sparse matrix multiplication (SPG
EMM) kernel. Unlike more well-known operations such as the SPMV, in the SP
GEMM the nonzero pattern of the result is determined by the interaction be
tween the nonzero patterns of...\n\n---------------------\nBatched Sparse
Iterative Solvers for Computational Chemistry Simulations on GPUs\n\nAggar
wal, Kashi, Nayak, Balos, Woodward...\n\nThis paper presents batched itera
tive solvers for GPU architectures. We elaborate on the design of the batc
hed functionality aiming for optimal performance while still giving the us
er some flexibility in terms of choosing a sparse matrix format, a precond
itioner optimized for the distinct items of t...\n\n---------------------\
nPassel: Improved Scalability and Efficiency of Distributed SVM Using a Ca
cheless PGAS Migrating Thread Architecture\n\nPage, Kogge\n\nStochastic Gr
adient Descent (SGD) is a valuable algorithm for large-scale machine learn
ing, but has proven difficult to parallelize on conventional architectures
because of communication and memory access issues. The HogWild series of
mixed logically distributed and physically multi-threaded algorit...\n\n--
-------------------\nIterative Methods with Mixed-Precision Preconditionin
g for Ill-Conditioned Linear Systems in Multiphase CFD Simulations\n\nIna,
Idomura, Imamura, Yamashita, Onodera\n\nA new mixed-precision preconditio
ner based on the iterative refinement (IR) method is developed for the pre
conditioned conjugate gradient (P-CG) solver and the multigrid preconditio
ned conjugate gradient (MGCG) solver in the multi-phase thermal-hydraulic
CFD code JUPITER. In the IR preconditioner, m...\n\n\nTag: Online Only, Al
gorithms, Extreme Scale Computing\n\nRegistration Category: Workshop Reg P
ass
END:VEVENT
END:VCALENDAR