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:20260522T150119Z
LOCATION:D175
DTSTART;TZID=America/Chicago:20181111T152400
DTEND;TZID=America/Chicago:20181111T154000
UID:submissions.supercomputing.org_SC18_sess143_ws_drbsd110@linklings.com
SUMMARY:Feature-Relevant Data Reduction for In Situ Workflows
DESCRIPTION:Will Fox (Massachusetts Institute of Technology), Matthew Wolf
  (Oak Ridge National Laboratory), Jeremy Logan (University of Tennessee), 
 Jong Youl Choi and Scott Klasky (Oak Ridge National Laboratory), and Tahsi
 n Kurc (Stony Brook University)\n\nAs the amount of data produced by HPC s
 imulations continues to grow and I/O throughput fails to keep up, in situ 
 data reduction is becoming an increasingly necessary component of HPC work
 flows. Application scientists, however, prefer to avoid reduction in order
  to preserve data fidelity for post-hoc analysis. In an attempt to comprom
 ise between data quality and data quantity, this work introduces the conce
 pt of feature-relevant compression. We explore two scientific datasets in 
 an attempt to quantify the impacts of compression on features of interest 
 by identifying such features and analyzing changes in their properties aft
 er compression. We find that it is indeed possible to compress simulation 
 data in a lossy manner while preserving desired properties within a predet
 ermined error rate. Additionally, we suggest that this error quantificatio
 n could be applied as part of an in situ workflow to dynamically tune comp
 ression parameters during simulation, compressing aggressively when featur
 es are simple but preserving structure where data complexity increases. Fu
 ture work should focus on implementation, extension to additional compress
 ion algorithms, and analysis of these techniques on quantities which are d
 erived from original simulation data.\n\nTag: Data Management, Hot Topics,
  Scientific Computing\n\nRegistration Category: Workshop Reg Pass\n\n
END:VEVENT
END:VCALENDAR