BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:America/New_York
X-LIC-LOCATION:America/New_York
BEGIN:DAYLIGHT
TZOFFSETFROM:-0500
TZOFFSETTO:-0400
TZNAME:EDT
DTSTART:19700308T020000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=2SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:-0400
TZOFFSETTO:-0500
TZNAME:EST
DTSTART:19701101T020000
RRULE:FREQ=YEARLY;BYMONTH=11;BYDAY=1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20260422T143140Z
LOCATION:B309
DTSTART;TZID=America/New_York:20241120T163000
DTEND;TZID=America/New_York:20241120T170000
UID:submissions.supercomputing.org_SC24_sess378_pap255@linklings.com
SUMMARY:Learning Generalizable Program and Architecture Representations fo
 r Performance Modeling
DESCRIPTION:Lingda Li, Thomas Flynn, and Adolfy Hoisie (Brookhaven Nationa
 l Laboratory)\n\nPerformance modeling is an essential tool in many areas, 
 including performance characterization/optimization, design space explorat
 ion, and resource allocation problems, to\nname a few. However, existing p
 erformance modeling approaches have limitations, such as high computationa
 l cost for discrete-event simulators, narrow flexibility of hardware emula
 tors, or\nrestricted accuracy/generality of analytical/data-driven models.
 \nTo address these limitations, this paper proposes PerfVec, a novel deep-
 learning-based performance modeling framework that learns high-dimensional
  and independent/orthogonal program and microarchitecture representations.
  Once learned, a program representation can be used to predict its perform
 ance on any microarchitecture, and likewise, a microarchitecture represent
 ation can be applied in the performance prediction of any program. Additio
 nally, PerfVec yields a foundation model that captures the performance ess
 ence of instructions, which can be directly used by developers in numerous
  performance modeling-related tasks without incurring its training cost. T
 he evaluation demonstrates that PerfVec is more general and efficient than
  previous approaches.\n\nTag: Heterogeneous Computing, Linear Algebra, Net
 work, Parallel Programming Methods, Models, Languages and Environments, Pe
 rformance Evaluation and/or Optimization Tools\n\nRegistration Category: T
 ech Program Reg Pass\n\nAward Finalist: Best Paper Finalist\n\nSession Cha
 ir: Sarah Neuwirth (Johannes Gutenberg University Mainz)\n\n
END:VEVENT
END:VCALENDAR