BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:America/New_York
X-LIC-LOCATION:America/New_York
BEGIN:DAYLIGHT
TZOFFSETFROM:-0500
TZOFFSETTO:-0400
TZNAME:EDT
DTSTART:19700308T020000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=2SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:-0400
TZOFFSETTO:-0500
TZNAME:EST
DTSTART:19701101T020000
RRULE:FREQ=YEARLY;BYMONTH=11;BYDAY=1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20250626T234542Z
LOCATION:B301
DTSTART;TZID=America/New_York:20241118T092000
DTEND;TZID=America/New_York:20241118T094000
UID:submissions.supercomputing.org_SC24_sess746_ws_esp102@linklings.com
SUMMARY:A Scalable Training-Free Diffusion Model for Uncertainty Quantific
 ation
DESCRIPTION:Ali Haisam Muhammad Rafid (Virginia Tech), Junqi Yin (Oak Ridg
 e National Laboratory (ORNL)), Yuwei Geng (University of South Carolina), 
 Siming Liang and Feng Bao (Florida State University), Lili Ju (University 
 of South Carolina), and Guannan Zhang (Oak Ridge National Laboratory (ORNL
 ))\n\nGenerative artificial intelligence extends beyond its success in ima
 ge/text synthesis, proving itself a powerful uncertainty quantification (U
 Q) technique through its capability to sample from complex high-dimensiona
 l probability distributions. However, existing methods often require a com
 plicated training process, which greatly hinders their applications to rea
 l-world UQ problems. To alleviate this challenge, we developed a scalable,
  training-free score-based diffusion model for high-dimensional sampling. 
 We incorporate a parallel-in-time method into our diffusion model to use a
  large number of GPUs to solve the backward stochastic differential equati
 on and generate new samples of the target distribution. Moreover, we also 
 distribute the computation of the large matrix subtraction used by the tra
 ining-free score estimator onto multiple GPUs available across all nodes. 
 We showcase the remarkable strong and weak scaling capabilities of the pro
 posed method on the Frontier supercomputer, as well as its uncertainty red
 uction capability in hurricane predictions when coupled with AI-based foun
 dation models.\n\nTag: Applications and Application Frameworks, Algorithms
 , Performance Evaluation and/or Optimization Tools\n\nRegistration Categor
 y: Workshop Reg Pass\n\nSession Chairs: Tiernan Casey (Sandia National Lab
 oratories) and Antigoni Georgiadou (Oak Ridge National Laboratory (ORNL))\
 n\n
END:VEVENT
END:VCALENDAR