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:20260422T000712Z
LOCATION:603
DTSTART;TZID=America/Denver:20231112T093100
DTEND;TZID=America/Denver:20231112T094600
UID:submissions.supercomputing.org_SC23_sess420_ws_ss109@linklings.com
SUMMARY:Comparing Power Signatures of HPC Workloads: Machine Learning vs S
 imulation
DESCRIPTION:Anish Govind (University of California, San Diego (UCSD)) and 
 Sridutt Bhalachandra, Zhengji Zhao, Ermal Rrapaj, Brian Austin, and Hai Ah
  Nam (Lawrence Berkeley National Laboratory (LBNL))\n\nPower is a limiting
  factor for supercomputers limiting their scale and operation. Characteriz
 ing the power signatures of different application types can enable data ce
 nters to operate efficiently, even when power constrained. This paper inve
 stigates power profiles of diverse scientific applications, spanning both 
 traditional simulations and modern machine learning (ML) running on the Pe
 rlmutter supercomputer at the National Energy Research Scientific Computin
 g Center (NERSC). Our findings indicate that traditional simulations typic
 ally consume more power on average than ML workloads. Furthermore, ML appl
 ications exhibit periodic power fluctuations attributed to epoch transitio
 ns during training. Finally, we discuss the potential implications of the 
 research insights toward automatic demand response (ADR) and consideration
 s for designing future systems.\n\nTag: Energy Efficiency, Green Computing
 , Sustainability\n\nRegistration Category: Workshop Reg Pass\n\nSession Ch
 airs: Kimmo Koski (CSC – IT Center for Science Ltd, Finland); James H. Rog
 ers (Oak Ridge National Laboratory (ORNL)); Fumiyoshi Shoji (RIKEN Center 
 for Computational Science (R-CCS), Center for Computational Science); Will
 iam W. Thigpen (NASA); Michèle Weiland (EPCC, The University of Edinburgh;
  The University of Edinburgh); and Mike Woodacre (Hewlett Packard Enterpri
 se (HPE))\n\n
END:VEVENT
END:VCALENDAR