BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:Europe/Stockholm
X-LIC-LOCATION:Europe/Stockholm
BEGIN:DAYLIGHT
TZOFFSETFROM:+0100
TZOFFSETTO:+0200
TZNAME:CEST
DTSTART:19700308T020000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=-1SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:+0200
TZOFFSETTO:+0100
TZNAME:CET
DTSTART:19701101T020000
RRULE:FREQ=YEARLY;BYMONTH=10;BYDAY=-1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20250822T115806Z
LOCATION:Room 5.2D02
DTSTART;TZID=Europe/Stockholm:20250617T163000
DTEND;TZID=Europe/Stockholm:20250617T170000
UID:submissions.pasc-conference.org_PASC25_sess129_msa116@linklings.com
SUMMARY:Design and Use of Energy-Efficient Systems in the Deep Learning Er
 a
DESCRIPTION:Pamela Delgado (HES-SO)\n\nModern GPUs, together with larger d
 atasets, facilitate the exponential growth and adoption of deep learning m
 odels. The training and deployment of deep neural networks in widely used 
 large-scale data centers, on the other hand, exhibit low GPU hardware util
 ization, barely reaching 50%, as shown by studies done on Microsoft and Al
 ibaba clusters. This is a waste of hardware resources, especially for expe
 nsive GPUs, and contributes to the unsustainable carbon footprint of AI. I
 n fact, less than 15% of large-language model training can lead to a carbo
 n footprint equivalent to the average yearly energy consumption of a US ho
 usehold.\nThis talk will discuss the reasons, challenges and opportunities
  for designing energy-efficient computing infrastructures as well as its p
 ractical applications.\n\nDomain: Computational Methods and Applied Mathem
 atics\n\nSession Chair: Florina Ciorba (University of Basel)\n\n
END:VEVENT
END:VCALENDAR