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:20260202T201338Z
LOCATION:242
DTSTART;TZID=America/Chicago:20251116T090000
DTEND;TZID=America/Chicago:20251116T123000
UID:submissions.supercomputing.org_SC25_sess216@linklings.com
SUMMARY:ISAV25: In Situ AI, Analysis, and Visualization
DESCRIPTION:As HPC platforms and applications increase significantly in si
 ze, complexity, and heterogeneity, one major challenge is the widening gap
  between computation and our ability to gain insight from extreme-scale da
 ta and make timely, data-driven decisions. A well-known, yet challenging, 
 approach to visual analysis, learning and processing is performing the tas
 ks in situ, i.e., while the computed data is still resident in memory. Sin
 ce 2015, ISAV has fostered a community of in situ developers, practitioner
 s, researchers, and users of in situ methods and infrastructure, connectin
 g industry, government laboratories, and academia across all career levels
 . Through presentations, discussions, and peer-reviewed proceedings of res
 earch findings, lessons learned, and early ideas, ISAV illuminates new req
 uirements and gaps driven by science and engineering applications and fost
 ers the application of in situ methods.\n\n
In-Situ Visualization and Stee
 ring of Digital Twins for Particle Accelerators\n\nDigital twins are emerg
 ing as a transformative concept for the design, operation, and optimisatio
 n of complex scientific instruments. In this plenary, I will present recen
 t progress on integrating in-situ visualization and interactive steering i
 nto the open-source accelerator simulation framework OP...\n\n\nAndreas Ad
 elmann (Paul Scherrer Institute PSI)\n---------------------\nIntroduction 
 by ISAV OC\n\nAxel Huebl (Lawrence Berkeley National Laboratory)\n--------
 -------------\nLightning Talk: Adaptive In-Situ Sampling Framework for Acc
 elerated Online Change Point Detection in Scientific Workflows\n\nHigh-per
 formance computing (HPC) systems at the\nExascale generate monitoring and 
 simulation data at rates that\nexceed the capacity of traditional offline 
 analysis, especially for\nchange point detection (CPD) in large-scale scie
 ntific workflows.\nWe present an adaptive in-situ sampling framework that 
 ...\n\n\nVijayalakshmi Saravanan (University of Texas at Tyler, UTD); Sai 
 Karthik Navuluru (UTD); Seetharam Vanam Rao (UTT); Tamil Lakshman (UTD); a
 nd Khaled Z Ibrahim (Lawrence Berkeley National Laboratory (LBNL))\n------
 ---------------\nLightning Talk: Predicting Optimal Rendering Frequency fo
 r In Situ Visualization in GPU–CPU Workflows\n\nHigh Performance Computing
  (HPC) simulations often run on GPUs while in situ rendering tasks are off
 loaded to CPUs. Deciding how frequently to perform these in situ rendering
 s is a challenge: render too frequently, and the simulation, or producer, 
 oversaturates the rendering pipeline, or consumer; r...\n\n\nColleen Heine
 mann (University of Illinois at Urbana-Champaign); Jens Henrik Göbbert and
  Mathis Bode (Forschungzentrum Juelich, Jülich Supercomputing Centre (JSC)
 ); Mengjiao Han, Joseph A. Insley, Silvio Rizzi, and Janet Knowles (Argonn
 e National Laboratory (ANL)); Michael E. Papka (Argonne National Laborator
 y (ANL), University of Illinois Chicago); and Victor A. Mateevitsi (Argonn
 e National Laboratory (ANL))\n---------------------\nLightning Talk: Diffe
 rentiable In-Situ Diagnostics for Particle Accelerator Digital Twins\n\nSi
 mulation codes are extensively used to design and operate particle acceler
 ators. \nSignificant effort is invested towards digital twins that closely
  mirror a physical system and inform tuning of particle accelerators in re
 al time, during operation. Some aspects of these digital twins benefit fro
 m u...\n\n\nAxel Huebl, Samuel Barber, and Christopher Doss (Lawrence Berk
 eley National Laboratory (LBNL)); Auralee Edelen (SLAC National Laboratory
 ); Arianna Formenti and Chad E Mitchell (Lawrence Berkeley National Labora
 tory (LBNL)); Ryan Roussel (SLAC National Laboratory); and Jean-Luc Vay, D
 an Wang, and Remi Lehe (Lawrence Berkeley National Laboratory (LBNL))\n---
 ------------------\nBest Paper Award and Closing Thoughts\n\nAxel Huebl (L
 awrence Berkeley National Laboratory)\n---------------------\nFrom Explora
 tion to Explanation: ML-Driven Causal Discovery for Datacenter Reliability
  at Scale\n\nModern datacenters operate at unprecedented scale, supporting
  HPC and AI workloads while consuming hundreds of megawatts of power. Thei
 r reliability is challenged by complex interdependencies across cooling, p
 ower, and network subsystems, where failures can cascade into downtime and
  degraded perform...\n\n\nPavana Prakash, Rolando P. Hong Enriquez, and Se
 rgey Serebryakov (Hewlett Packard Labs); David Grant and Wesley Brewer (Oa
 k Ridge National Laboratory (ORNL)); and Dejan Milojicic (Hewlett Packard 
 Labs)\n---------------------\nIn-Transit Data Transport Strategies for Cou
 pled AI-Simulation Workflow Patterns\n\nCoupled AI-Simulation workflows ar
 e becoming the major workloads for HPC facilities, and their increasing co
 mplexity necessitates new tools for performance analysis and prototyping o
 f new in-situ workflows. We present SimAI-Bench, a tool designed to both p
 rototype and evaluate these coupled workflow...\n\n\nHarikrishna Tummalapa
 lli, Riccardo Balin, and Christine M. Simpson (Argonne National Laboratory
  (ANL)); Andrew Park and Aymen Alsaadi (Rutgers University); Andrew E. Sha
 o (Hewlett Packard Enterprise (HPE)); Wesley Brewer (Oak Ridge National La
 boratory (ORNL)); and Shantenu Jha (Rutgers University)\n-----------------
 ----\nMorning Break - ISAV25\n---------------------\nA Data-Size Adaptive 
 Approach to I/O of Poorly Load Balanced In Situ Data Extracts\n\nEfficient
  parallel I/O is essential for large-scale scientific workflows, particula
 rly in in situ visualization pipelines where output sizes and data distrib
 ution can vary significantly over time.\nWe present a new data-size based 
 aggregation strategy for the ADIOS2 BP5 engine, designed to improve pa...\
 n\n\nCaitlin Ross and Scott Wittenburg (Kitware Inc.), Greg Eisenhauer (Ge
 orgia Institute of Technology), and Corey Wetterer-Nelson (Kitware Inc.)\n
 \nRecording: Partially Livestreamed, Partially Recorded\n\nRegistration Ca
 tegory: Technical Program Reg Pass, Workshop Reg Pass\n\nSession Chairs: A
 xel Huebl (Lawrence Berkeley National Laboratory (LBNL)); Will Usher (Lumi
 nary Cloud Inc); Nicola Ferrier (Argonne National Laboratory (ANL), Univer
 sity of Chicago); Estelle Dirand (TotalEnergies, France); Victor Mateevits
 i (Argonne National Laboratory (ANL), University of Illinois Chicago); and
  Jens Henrik Göbbert (Forschungszentrum Jülich, Jülich Supercomputing Cent
 re (JSC))
END:VEVENT
END:VCALENDAR