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:20260522T150125Z
LOCATION:D220
DTSTART;TZID=America/Chicago:20181111T144500
DTEND;TZID=America/Chicago:20181111T144800
UID:submissions.supercomputing.org_SC18_sess160_ws_whpc127@linklings.com
SUMMARY:Characterization of the Impact of Soft Errors on Iterative Methods
DESCRIPTION:Burcu Mutlu (Pacific Northwest National Laboratory, Polytechni
 c University of Catalonia); Gokcen Kestor (Oak Ridge National Laboratory);
  Osman Unsal (Barcelona Supercomputing Center); and Sriram Krishnamoorthy 
 (Pacific Northwest National Laboratory)\n\nSoft errors caused by transient
  bit flips can lead to silent data corruption which can significantly impa
 ct an application’s behavior. When a transient bit flip affects a hardware
  component, the application is said to be impacted by a soft error. When a
  soft error escapes hardware detection and impacts the application state, 
 it can impact execution by leading to incorrect results or significantly i
 mpacting application execution times. Recent architectural trends, such as
  near-threshold voltage operation and constrained power budgets, exacerbat
 e the frequency and impact of soft errors. This has motivated design of nu
 merous soft error detection and correction techniques, which focuses local
 izing the application vulnerabilities, and then correcting these errors us
 ing micro-architectural, architectural, compilation-based, or application-
 level techniques. \n\nA broad array of techniques has been designed to und
 erstand application behavior under soft errors and to detect, isolate, and
  correct soft-error-impacted application state. The first step toward tole
 rating soft errors involves understanding an application’s behavior under 
 soft errors. This can help elucidate the need for error detection/correcti
 on techniques.\n\nIn this study, we designed a deterministic application l
 evel error injection method that allows us explore the error injection spa
 ce over several iterative solvers. As iterative methods are a crucial comp
 onent of scientific applications, and consume a significant fraction of su
 percomputing time, we performed this study on the iterative methods. We us
 ed real life datasets to evaluate and characterize vulnerabilities and str
 engths of each solver under identical settings.\n\nTag: Diversity, Educati
 on, Hot Topics\n\nRegistration Category: Workshop Reg Pass\n\nSession Chai
 rs: Toni Collis (Women in High Performance Computing); Weronika Filinger (
 Edinburgh Parallel Computing Centre (EPCC); University of Edinburgh, Scotl
 and); and Misbah Mubarak (Amazon Web Services)\n\n
END:VEVENT
END:VCALENDAR