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:20181221T160731Z
LOCATION:C141/143/149
DTSTART;TZID=America/Chicago:20181115T143000
DTEND;TZID=America/Chicago:20181115T150000
UID:submissions.supercomputing.org_SC18_sess192_pap430@linklings.com
SUMMARY:PRISM: Predicting Resilience of GPU Applications Using Statistical
  Methods
DESCRIPTION:Paper\nAlgorithms, Architectures, GPUs, Linear Algebra, Networ
 ks, Resiliency, Tech Program Reg Pass\n\nPRISM: Predicting Resilience of G
 PU Applications Using Statistical Methods\n\nKalra, Previlon, Li, Rubin, K
 aeli\n\nAs Graphics Processing Units (GPUs) become more pervasive in HPC a
 nd safety-critical domains, ensuring that GPU applications can be protecte
 d from data corruption grows in importance. Despite prior efforts to mitig
 ate errors, we still lack a clear understanding of how resilient these app
 lications are in the presence of transient faults.  Due to the random natu
 re of these faults, predicting whether they will alter the program output 
 is a challenging problem. In this paper, we build a framework named PRISM,
  which uses a systematic approach to predict failures in GPU programs. PRI
 SM extracts micro-architecture agnostic features to characterize program r
 esiliency, which serve as predictors in our statistical model. PRISM enabl
 es us to predict failures in applications without running exhaustive fault
 -injection campaigns on a GPU, thereby reducing the error estimation effor
 t. PRISM can also be used to gain insight into potential architectural sup
 port required to improve the reliability of GPU applications.
URL:https://sc18.supercomputing.org/presentation/?id=pap430&sess=sess192
END:VEVENT
END:VCALENDAR

