BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:Europe/Stockholm
X-LIC-LOCATION:Europe/Stockholm
BEGIN:DAYLIGHT
TZOFFSETFROM:+0100
TZOFFSETTO:+0200
TZNAME:CEST
DTSTART:19700308T020000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=-1SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:+0200
TZOFFSETTO:+0100
TZNAME:CET
DTSTART:19701101T020000
RRULE:FREQ=YEARLY;BYMONTH=10;BYDAY=-1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20210916T132447Z
LOCATION:Henry Dunant
DTSTART;TZID=Europe/Stockholm:20210709T123000
DTEND;TZID=Europe/Stockholm:20210709T130000
UID:submissions.pasc-conference.org_PASC21_sess188_pap_jun101@linklings.co
 m
SUMMARY:Extreme-Scale Task-Based Cholesky Factorization Toward Climate and
  Weather Prediction Applications
DESCRIPTION:Paper\n\nExtreme-Scale Task-Based Cholesky Factorization Towar
 d Climate and Weather Prediction Applications\n\nCao, Pei, Akbudak, Mikhal
 ev, Bosilca...\n\nClimate and weather can be predicted statistically via g
 eospatial Maximum Likelihood Estimates (MLE), as an alternative to running
  large ensembles of forward models. The MLE-based iterative optimization p
 rocedure requires factorization of large-scale linear systems that consist
  of a symmetric positive-definite covariance matrix-a demanding dense fact
 orization in terms of memory footprint and computation. We propose a novel
  solution to this problem that results in fast and accurate predictions: a
 t the mathematical level, we reduce the computational requirement by explo
 iting data sparsity structure of the matrix off-diagonal tiles by means of
  low-rank approximations; and, at the programming-paradigm level, we integ
 rate PaRSEC, a dynamic, task-based runtime to reach unparalleled levels of
  efficiency for solving extreme-scale problems for environmental applicati
 ons. The resulting solution leverages fine-grained computations to facilit
 ate asynchronous execution while providing flexible data distribution to m
 itigate load imbalance. Performance results are reported using 3D datasets
  up to 42M geospatial locations on 130,000 cores.\n\nDomain: CS and Math, 
 Climate and Weather
END:VEVENT
END:VCALENDAR