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:20231112T090100
DTEND;TZID=America/Denver:20231112T093000
UID:submissions.supercomputing.org_SC23_sess426_ws_llvmf101@linklings.com
SUMMARY:OpenMP Kernel Language Extensions for Performance Portable GPU Cod
 es
DESCRIPTION:Shilei Tian (Stony Brook University); Tom Scogland (Lawrence L
 ivermore National Laboratory (LLNL)); Barbara Chapman (Stony Brook Univers
 ity, Hewlett Packard Enterprise (HPE)); and Johannes Doerfert (Lawrence Li
 vermore National Laboratory (LLNL))\n\nIn this work, we introduce extensio
 ns to LLVM OpenMP, transforming it into a versatile and performance portab
 le kernel language for GPU programming.  These extensions allow for the se
 amless porting of programs written in kernel languages to high-performance
  OpenMP GPU programs with minimal modifications.  To evaluate our extensio
 n, we implemented a proof-of-concept prototype that contains a subset of e
 xtensions we proposed.  We ported six established CUDA proxy and benchmark
  applications and evaluated their performance on both AMD and NVIDIA platf
 orms.  By comparing with native versions (HIP and CUDA), our results demon
 strate that OpenMP, augmented with our extensions, can not only match but 
 also in some cases exceed the performance of kernel languages, thereby off
 ering performance portability with minimal effort from application develop
 ers.\n\nTag: Compilers, Heterogeneous Computing, Performance Optimization\
 n\nRegistration Category: Workshop Reg Pass\n\nSession Chairs: James Brodm
 an (AMD); Ryan Kabrick (Tactical Computing Laboratories LLC, University of
  Delaware); Patrick S. McCormick (Los Alamos National Laboratory (LANL)); 
 and Alexis Perry-Holby (Los Alamos National Laboratory (LANL))\n\n
END:VEVENT
END:VCALENDAR
