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:20260422T000711Z
LOCATION:601
DTSTART;TZID=America/Denver:20231113T091000
DTEND;TZID=America/Denver:20231113T100000
UID:submissions.supercomputing.org_SC23_sess443_misc151@linklings.com
SUMMARY:AI-Driven Performance Metaprogramming
DESCRIPTION:Torsten Hoefler (ETHZ)\n\nRecent advances in artificial intell
 igence methods show the enormous potential of AI methods. The underlying c
 oncepts are embedding spaces to represent real-world information. These em
 bedding spaces have been used to represent, transform, and work with compl
 ex information in large-language models but also many other domains such a
 s climate sciences or automated driving systems. In this talk, we focus on
  embedding spaces for programs and use those primarily to assess, analyze,
  and improve program performance. We start by deriving a first embedding f
 rom textual LLWM internal representation (IR) and show that it successfull
 y predicts GPU execution times of programs. We then show that textual repr
 esentations bear the danger is missing context and being overly sensitive 
 to specific strings. Using a graph-based representation, we improve the em
 bedding to capture relationships such as data dependencies and flows in LL
 VM IR. Finally, we discuss DaCe's performance metaprogramming capabilities
  and it's programmable graph-based IR. We then demonstrate how a graph-neu
 ral network (GNN)-based embedding can capture general performance properti
 es. Those properties form the concept of Performance Embeddings for Transf
 er Tuning and can be used to select optimization metaprograms to apply to 
 transform the IR graph.\n\nTag: Artificial Intelligence/Machine Learning, 
 Software Engineering\n\nRegistration Category: Workshop Reg Pass\n\nSessio
 n Chairs: Giorgis Georgakoudis (Lawrence Livermore National Laboratory (LL
 NL)), Ignacio Laguna (Lawrence Livermore National Laboratory (LLNL)), and 
 Konstantinos Parasyris (Lawrence Livermore National Laboratory (LLNL))\n\n
END:VEVENT
END:VCALENDAR