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:20200129T163558Z
LOCATION:704-706
DTSTART;TZID=America/Denver:20191117T161500
DTEND;TZID=America/Denver:20191117T163000
UID:submissions.supercomputing.org_SC19_sess114_ws_prot109@linklings.com
SUMMARY:CHAMPVis: Comparative Hierarchical Analysis of Microarchitectural 
 Performance
DESCRIPTION:Workshop\n\nCHAMPVis: Comparative Hierarchical Analysis of Mic
 roarchitectural Performance\n\nPentecost, Gupta, Ngan, Beyer, Wei...\n\nPe
 rformance analysis and optimization are essential tasks for hardware and s
 oftware engineers.  In the age of datacenter-scale computing, it is partic
 ularly important to conduct comparative performance analysis to understand
  discrepancies and limitations among different hardware systems and applic
 ations.\nHowever, there is a distinct lack of productive visualization too
 ls for these comparisons.\n\nWe present CHAMPVis, a web-based, interactive
  visualization tool that leverages the hierarchical organization of hardwa
 re systems to enable productive performance analysis.  With CHAMPVis, user
 s can make definitive performance comparisons across applications or hardw
 are platforms.  In addition, CHAMPVis provides methods to rank and cluster
  based on performance metrics to identify common optimization opportunitie
 s.  Our thorough task analysis reveals three types of datacenter-scale per
 formance analysis tasks: summarization, detailed comparative analysis, and
  interactive performance bottleneck identification.  We propose techniques
  for each class of tasks including (1) 1-D feature space projection for si
 milarity analysis; (2) Hierarchical parallel co-ordinates for comparative 
 analysis; and (3) User interactions for rapid diagnostic queries to identi
 fy optimization targets.  We evaluate CHAMPVis by analyzing standard datac
 enter applications and machine learning benchmarks in two different case s
 tudies.\n\nTag: Workshop Reg Pass, Performance, Programming Systems, Visua
 lization\n\nRegistration Category: Workshop Reg Pass, Performance, Program
 ming Systems, Visualization
URL:https://sc19.supercomputing.org/presentation/?id=ws_prot109&sess=sess1
 14
END:VEVENT
END:VCALENDAR