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:401-402
DTSTART;TZID=America/Denver:20231117T110000
DTEND;TZID=America/Denver:20231117T113000
UID:submissions.supercomputing.org_SC23_sess460_misc343@linklings.com
SUMMARY:The Future of Machine Learning is Sparse
DESCRIPTION:Tal Ben-Nun (Lawrence Livermore National Laboratory (LLNL))\n\
 nAs machine learning models become prohibitively expensive to train and us
 e, sparsity is increasingly essential for neural networks such as large la
 nguage models. If performed correctly, sparsity does not come at the cost 
 of performance, and pruned deep neural networks have shown benefits in bot
 h throughput and generalization. In this talk, we will review the many fac
 es of sparsity in deep learning and how it can be leveraged, from input re
 presentation to specialized hardware. We will discuss the differences in s
 parsity structure from other HPC applications, learning schedules that ena
 ble near-lossless sparse training, and libraries that aid in introducing h
 igh-performance sparsity to existing machine learning models. Lastly, we w
 ill outline the current limitations of sparsity in deep learning and discu
 ss future opportunities in pushing those boundaries.\n\nTag: Graph Algorit
 hms and Frameworks, Linear Algebra, Programming Frameworks and System Soft
 ware, State of the Practice\n\nRegistration Category: Workshop Reg Pass\n\
 nSession Chairs: Xing Cai (University of Oslo, Norway; Simula Research Lab
 oratory/University of Oslo); Aleksandar Ilic (INESC-ID, IST, University of
  Lisbon; University of Lisbon, Portugal); Kamer Kaya (Sabancı University, 
 Turkey); and Didem Unat (Koç University, Turkey)\n\n
END:VEVENT
END:VCALENDAR