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:20211207T055347Z
LOCATION:Online
DTSTART;TZID=America/Chicago:20211115T120000
DTEND;TZID=America/Chicago:20211115T123000
UID:submissions.supercomputing.org_SC21_sess422_ws_pmbsf116@linklings.com
SUMMARY:Understanding Power Variation and its Implications on Performance 
 Optimization on the Cori Supercomputer
DESCRIPTION:Workshop\n\nUnderstanding Power Variation and its Implications
  on Performance Optimization on the Cori Supercomputer\n\nBhalachandra, Au
 stin, Wright\n\nPower is increasingly becoming a limiting factor in superc
 omputing.  The performance and scale of future high-performance computing 
 systems will be determined by how efficiently they manage their power budg
 ets. Therefore, any amount of unused power is forsaken performance. Regard
 less of the processors chosen for a future system,  it will be necessary t
 o understand power variation and its implications on performance optimizat
 ion. In this paper, we identify and quantify the factors that affect the p
 ower consumption of the NERSC Cori supercomputer at different levels of th
 e system hierarchy. Our study presents node-level power-performance trade-
 offs for fundamental computational patterns. We show that I/O activity and
  load imbalance are common causes of job-level power variation among jobs 
 in Cori’s production workload. We quantitatively attribute system-level po
 wer variation to three sources and find that 86% of the variation is due t
 o temporal variation in application behavior over the duration of a job. F
 urthermore, our analysis reveals that under typical workload conditions, t
 he Cori system’s power budget could accommodate up to 60% more nodes.\n\nT
 ag: Online Only, Accelerator-based Architectures, Applications, Computatio
 nal Science, Emerging Technologies, Extreme Scale Computing, File Systems 
 and I/O, Heterogeneous Systems, Parallel Programming Languages and Models,
  Performance, Scientific Computing, Software Engineering\n\nRegistration C
 ategory: Workshop Reg Pass
END:VEVENT
END:VCALENDAR