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:20260522T150111Z
LOCATION:C2/3/4 Ballroom
DTSTART;TZID=America/Chicago:20181115T083000
DTEND;TZID=America/Chicago:20181115T170000
UID:submissions.supercomputing.org_SC18_sess324_post144@linklings.com
SUMMARY:Detection of Silent Data Corruptions in Smooth Particle Hydrodynam
 ics Simulations
DESCRIPTION:Aurélien Cavelan, Florina M. Ciorba, and Ruben M. Cabezón (Uni
 versity of Basel)\n\nSoft errors, such as silent data corruptions (SDCs) h
 inder the correctness of large-scale scientific applications. Ghost replic
 ation (GR) is proposed herein as the first SDCs detector relying on the fa
 st error propagation inherent to applications that employ the smooth parti
 cle hydrodynamics (SPH) method. GR follows a two-steps selective replicati
 on scheme. First, an algorithm selects which particles to replicate on a d
 ifferent process. Then, a different algorithm detects SDCs by comparing th
 e data of the selected particles with the data of their ghost. The overhea
 d and scalability of the proposed approach are assessed through a set of s
 trong-scaling experiments conducted on a large HPC system under error-free
  conditions, using upwards of 3, 000 cores. The results show that GR achie
 ves a recall and precision similar to that of full replication methods, at
  only a fraction of the cost, with detection rates of 91−99.9%, no false-p
 ositives, and an overhead of 1−10%.\n\nRegistration Category: Tech Program
  Reg Pass, Exhibits Reg Pass\n\n
END:VEVENT
END:VCALENDAR