BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:America/Denver
X-LIC-LOCATION:America/Denver
BEGIN:DAYLIGHT
TZOFFSETFROM:-0700
TZOFFSETTO:-0600
TZNAME:MDT
DTSTART:19700308T020000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=2SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:-0600
TZOFFSETTO:-0700
TZNAME:MST
DTSTART:19701101T020000
RRULE:FREQ=YEARLY;BYMONTH=11;BYDAY=1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20260422T000713Z
LOCATION:405-406-407
DTSTART;TZID=America/Denver:20231116T140000
DTEND;TZID=America/Denver:20231116T143000
UID:submissions.supercomputing.org_SC23_sess175_pap585@linklings.com
SUMMARY:Breaking Boundaries: Distributed Domain Decomposition with Scalabl
 e Physics-Informed Neural PDE Solvers
DESCRIPTION:Arthur Feeney, Zitong Li, Ramin Bostanabad, and Aparna Chandra
 mowlishwaran (University of California, Irvine)\n\nMosaic Flow is a novel 
 domain decomposition method designed to scale physics-informed neural PDE 
 solvers to large domains. Its unique approach leverages pre-trained networ
 ks on small domains to solve partial differential equations on large domai
 ns purely through inference, resulting in high reusability. This paper pre
 sents an end-to-end parallelization of Mosaic Flow, combining data paralle
 l training and domain parallelism for inference on large-scale problems. B
 y optimizing the network architecture and data parallel training, we signi
 ficantly reduce the training time for learning the Laplacian operator to m
 inutes on 32 GPUs. Moreover, our distributed domain decomposition algorith
 m enables scalable inferences for solving the Laplace equation on domains 
 4096x larger than the training domain, demonstrating strong scaling while 
 maintaining accuracy on 32 GPUs. The reusability of Mosaic Flow, combined 
 with the improved performance achieved through the distributed-memory algo
 rithms, makes it a promising tool for modeling complex physical phenomena 
 and accelerating scientific discovery.\n\nTag: Artificial Intelligence/Mac
 hine Learning, Applications, Modeling and Simulation, State of the Practic
 e\n\nRegistration Category: Tech Program Reg Pass\n\nSession Chair: Wei Xu
  (Brookhaven National Laboratory)\n\n
END:VEVENT
END:VCALENDAR