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:20260422T000712Z
LOCATION:405-406-407
DTSTART;TZID=America/Denver:20231116T153000
DTEND;TZID=America/Denver:20231116T160000
UID:submissions.supercomputing.org_SC23_sess170_pap316@linklings.com
SUMMARY:FuzzyFlow: Leveraging Dataflow to Find and Squash Program Optimiza
 tion Bugs
DESCRIPTION:Philipp Schaad, Timo Schneider, Tal Ben-Nun, Alexandru Calotoi
 u, Alexandros Nikolaos Ziogas, and Torsten Hoefler (ETH Zurich - Swiss Fed
 eral Institute of Technology)\n\nThe current hardware landscape and applic
 ation scale is driving performance engineers toward writing bespoke optimi
 zations. Verifying such optimizations, and generating minimal failing case
 s, is important for robustness in the face of changing program conditions,
  such as inputs and sizes. However, isolation of minimal test-cases from e
 xisting applications and generating new configurations are often difficult
  due to side effects on the system state, mostly related to dataflow. This
  paper introduces FuzzyFlow: a fault localization and test case extraction
  framework designed to test program optimizations. We leverage dataflow pr
 ogram representations to capture a fully reproducible system state and are
 a-of-effect for optimizations to enable fast checking for semantic equival
 ence. To reduce testing time, we design an algorithm for minimizing test i
 nputs, trading off memory for recomputation. We demonstrate FuzzyFlow on e
 xemplary use cases in real-world applications where the approach provides 
 up to 528 times faster optimization testing and debugging compared to trad
 itional approaches.\n\nTag: Compilers, Performance Measurement, Modeling, 
 and Tools, Performance Optimization, Programming Frameworks and System Sof
 tware\n\nRegistration Category: Tech Program Reg Pass\n\nReproducibility B
 adges: Artifact Available, Artifact Functional, Results Reproduced\n\nSess
 ion Chair: Martin Kong (Ohio State University)\n\n
END:VEVENT
END:VCALENDAR