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:20260422T000713Z
LOCATION:603
DTSTART;TZID=America/Denver:20231112T152000
DTEND;TZID=America/Denver:20231112T154500
UID:submissions.supercomputing.org_SC23_sess433_ws_ships102@linklings.com
SUMMARY:An End-to-End HPC Framework for Dynamic Power Objectives
DESCRIPTION:Daniel C. Wilson (Boston University, Intel Corporation); Fatih
  Acun (Boston University); Siddhartha Jana, Federico Ardanaz, and Jonathan
  M. Eastep (Intel Corporation); and Ioannis Ch. Paschalidis and Ayse K. Co
 skun (Boston University)\n\nHigh-Performance Computing (HPC) centers deman
 d a lot of power, and continue to grow through the exascale era. This work
  establishes the need for a multi-tiered, feedback-driven power management
  framework to follow dynamic power objectives while maximizing job perform
 ance, highlighting the need to respond to external factors (e.g., power co
 nstraints), and internal factors (e.g., performance variation). We present
  a practical implementation of this framework on a real-world cluster in a
 ddition to conducting simulations for larger data centers. We accurately t
 rack a moving power target for demand response while reacting to incomplet
 e or inaccurate prior knowledge about job power and performance properties
 . We demonstrate that online performance feedback from a job runtime enabl
 es a cluster power management policy to recover most of the performance de
 gradation introduced by job-type misclassification.\n\nTag: Artificial Int
 elligence/Machine Learning, Energy Efficiency, Green Computing, Performanc
 e Measurement, Modeling, and Tools, Sustainability\n\nRegistration Categor
 y: Workshop Reg Pass\n\nSession Chairs: Andrea Borghesi (University of Bol
 ogna; Department of Electrical, Electronic and Information Engineering) an
 d Daniela Loreti (University of Bologna)\n\n
END:VEVENT
END:VCALENDAR
