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:20231112T163500
DTEND;TZID=America/Denver:20231112T170000
UID:submissions.supercomputing.org_SC23_sess433_ws_ships104@linklings.com
SUMMARY:Augmenting ML-Based Predictive Modelling with NLP to Forecast a Jo
 b's Power Consumption
DESCRIPTION:Francesco Antici (University of Bologna), Keiji Yamamoto and J
 ens Domke (RIKEN Center for Computational Science (R-CCS)), and Zeynep Kiz
 iltan (University of Bologna)\n\nAs modern High-Performance Computing (HPC
 ) reach exascale performance, their power consumption becomes a serious th
 reat to environmental and energy sustainability. Efficient power managemen
 t in HPC systems is crucial for optimizing workload management, reducing o
 perational costs, and promoting environmental sustainability. Accurate pre
 diction of job power consumption plays an important role in achieving such
  goals.  We apply a technique combining Machine Learning (ML) algorithms w
 ith Natural Language Processing (NLP) tools to predict job power consumpti
 on. The solution is able to predict job maximum and average power consumpt
 ion per node, leveraging only information which is available at the time o
 f job submission. The prediction is performed in an online fashion, and we
  validate the approach using batch system logs extracted from Supercompute
 r Fugaku.  The experimental evaluation shows promising results of outperfo
 rming classical technique while obtaining an R2 score of more than 0.53 fo
 r our two prediction tasks.\n\nTag: Artificial Intelligence/Machine Learni
 ng, Energy Efficiency, Green Computing, Performance Measurement, Modeling,
  and Tools, Sustainability\n\nRegistration Category: Workshop Reg Pass\n\n
 Session Chairs: Andrea Borghesi (University of Bologna; Department of Elec
 trical, Electronic and Information Engineering) and Daniela Loreti (Univer
 sity of Bologna)\n\n
END:VEVENT
END:VCALENDAR