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:20240116T191659Z
LOCATION:605
DTSTART;TZID=America/Denver:20231112T115800
DTEND;TZID=America/Denver:20231112T120600
UID:submissions.supercomputing.org_SC23_sess414_ws_hust104@linklings.com
SUMMARY:PEAK:  A Light-Weight Profiler for HPC Systems
DESCRIPTION:Workshop\n\nYinzhi Wang and Junjie Li (University of Texas)\n\
 nIn the context of the expanding landscape of contemporary High-Performanc
 e Computing (HPC) applications from petascale to exascale, the pursuit of 
 performance optimization emerges as a significant impediment within softwa
 re development endeavors.  In the meantime, the escalating intricacies inh
 erent in parallel architectures and systems serve to compound the challeng
 es associated with performance enhancement.\n\nHere, we introduce PEAK (Pe
 rformance Evaluation and Analysis Kit), a light-weight profiling tool deve
 loped with a specific focus on large-scale HPC applications.  Using Dynami
 c Binary Instrumentation, PEAK is able to profile large-scale multi-thread
 ed, multi-process applications with low overhead and high accuracy.  We an
 alyzed the overhead and accuracy of PEAK using synthetic benchmarks and re
 al applications and compared it against the other widely used HPC profilin
 g tools available.  Our demonstration underscores that PEAK exhibits compa
 rable overhead and accuracy to alternative profiling tools, while preservi
 ng its inherent simplicity.\n\nTag: Programming Frameworks and System Soft
 ware\n\nRegistration Category: Workshop Reg Pass\n\nSession Chairs: Chris 
 Bording (University of Western Australia); Elsa J. Gonsiorowski (Lawrence 
 Livermore National Laboratory (LLNL)); Lev Gorenstein (Globus, University 
 of Chicago); and Karen Tomko (Ohio Supercomputer Center)
END:VEVENT
END:VCALENDAR