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:605
DTSTART;TZID=America/Denver:20231112T155500
DTEND;TZID=America/Denver:20231112T162000
UID:submissions.supercomputing.org_SC23_sess434_ws_ftxs108@linklings.com
SUMMARY:Recovery from Silent Data Corruption via Spatial Data Prediction
DESCRIPTION:Kristen Guernsey, Sarah Placke, Alexandra Poulos, and Jon Calh
 oun (Clemson University)\n\nHigh-performance computing applications are ce
 ntral to advancement in many fields of science and engineering. Central to
  this advancement is the supposed reliability of the HPC system. However, 
 as system size grows and hardware components run with near-threshold volta
 ges, transient upset events become more likely. Many works have explored t
 he problem of detecting silent data corruption; however, recovery is often
  left to checkpoint-restart or application-specific techniques. Recovering
  from a checkpoint incurs overhead due to reading a checkpoint and recompu
 ting lost work. Allowing the application to recover just the corrupted dat
 a enables faster and more efficient recovery. This paper explores using sp
 atial similarities to recover from silent data corruption. We explore seve
 ral reconstruction methods and evaluate their effectiveness at recovering 
 corrupted entries in data arrays. Results show that the Lorenzo 1-Layer pr
 ediction method yields the best results, with over half of its reconstruct
 ions having less than 1% relative error across all applications.\n\nTag: F
 ault Handling and Tolerance, Large Scale Systems\n\nRegistration Category:
  Workshop Reg Pass\n\nSession Chairs: John Daly (US Department of Defense)
 , Scott Levy (Sandia National Laboratories), and Keita Teranishi (Oak Ridg
 e National Laboratory (ORNL))\n\n
END:VEVENT
END:VCALENDAR