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:20260202T201804Z
LOCATION:242
DTSTART;TZID=America/Chicago:20251116T112000
DTEND;TZID=America/Chicago:20251116T114500
UID:submissions.supercomputing.org_SC25_sess216_ws_isav105@linklings.com
SUMMARY:A Data-Size Adaptive Approach to I/O of Poorly Load Balanced In Si
 tu Data Extracts
DESCRIPTION:Caitlin Ross and Scott Wittenburg (Kitware Inc.), Greg Eisenha
 uer (Georgia Institute of Technology), and Corey Wetterer-Nelson (Kitware 
 Inc.)\n\nEfficient parallel I/O is essential for large-scale scientific wo
 rkflows, particularly in in situ visualization pipelines where output size
 s and data distribution can vary significantly over time.\nWe present a ne
 w data-size based aggregation strategy for the ADIOS2 BP5 engine, designed
  to improve parallel I/O performance under load imbalance. Whereas existin
 g aggregation strategies for ADIOS group writers based on compute node ass
 ignment,  our approach dynamically balances subfile sizes according to the
  amount of data each process will write. We evaluate this strategy using a
  synthetic workload under low and severe load imbalance. Results show that
  the data-size based aggregation matches or outperforms existing strategie
 s. These findings highlight the potential of adaptive aggregation strategi
 es to improve I/O performance for imbalanced scientific workloads.\n\nTag:
  AI, Machine Learning, & Deep Learning, Clouds & Distributed Computing, Pe
 rformance Evaluation, Scalability, & Portability, Scientific & Information
  Visualization\n\nRecording: Partially Livestreamed, Partially Recorded\n\
 nRegistration Category: Technical Program Reg Pass, Workshop Reg Pass\n\nS
 ession Chairs: Axel Huebl (Lawrence Berkeley National Laboratory (LBNL)); 
 Will Usher (Luminary Cloud Inc); Nicola Ferrier (Argonne National Laborato
 ry (ANL), University of Chicago); Estelle Dirand (TotalEnergies, France); 
 Victor Mateevitsi (Argonne National Laboratory (ANL), University of Illino
 is Chicago); and Jens Henrik Göbbert (Forschungszentrum Jülich, Jülich Sup
 ercomputing Centre (JSC))\n\n
END:VEVENT
END:VCALENDAR