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:20181221T160727Z
LOCATION:D165
DTSTART;TZID=America/Chicago:20181112T093000
DTEND;TZID=America/Chicago:20181112T100000
UID:submissions.supercomputing.org_SC18_sess161_ws_pmbsf104@linklings.com
SUMMARY:Exploring and Quantifying How Communication Behaviors in Proxies R
 elate to Real Applications
DESCRIPTION:Workshop\nBenchmarks, Parallel Programming Languages, Librarie
 s, and Models, Performance, Simulation, Workshop Reg Pass\n\nExploring and
  Quantifying How Communication Behaviors in Proxies Relate to Real Applica
 tions\n\nAaziz, Cook, Cook, Vaughan\n\nProxy applications, or proxies, are
  simple applications meant to exercise systems in a way that mimics real a
 pplications (their parents). However, characterizing the relationship betw
 een the behavior of parent and proxy applications is not an easy task. In 
 prior work, we presented a data-driven methodology to characterize the rel
 ationship between parent and proxy applications based on collecting runtim
 e data from both and then using data analytics to find their correspondenc
 e or divergence. We showed that it worked well for hardware counter data, 
 but our initial attempt using MPI function data was less satisfactory. In 
 this paper, we present an exploratory effort at making an improved quantif
 ication of the correspondence of communication behavior for proxies and th
 eir respective parent applications. We present experimental evidence of po
 sitive results using four proxy applications from the current ECP Proxy Ap
 plication Suite and their corresponding parent applications (in the ECP ap
 plication portfolio). Results show that each proxy analyzed is representat
 ive of its parent with respect to communication data. In conjunction with 
 our method presented earlier (correspondence between computation and memor
 y behavior), we get a strong understanding of how well a proxy predicts th
 e comprehensive performance of its parent.
URL:https://sc18.supercomputing.org/presentation/?id=ws_pmbsf104&sess=sess
 161
END:VEVENT
END:VCALENDAR