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:708
DTSTART;TZID=America/Denver:20231113T153000
DTEND;TZID=America/Denver:20231113T160000
UID:submissions.supercomputing.org_SC23_sess451_misc237@linklings.com
SUMMARY:Who's Winning the Performance Portability Race on GPU Platforms?
DESCRIPTION:Abhinav Bhatele (University of Maryland)\n\nEnsuring high prod
 uctivity in scientific software development necessitates developing and ma
 intaining a single codebase that can run efficiently on a range of acceler
 ator-based supercomputing platforms. This requires the use of performance 
 portability layers such as OpenMP, RAJA, Kokkos and SYCL for developing th
 e compute kernels. In this talk, I will present the results of a comprehen
 sive study of a range of proxy applications implemented in the major progr
 amming models suitable for GPU-based platforms. We collect and analyze per
 formance results across NVIDIA and AMD GPU hardware currently deployed in 
 leadership-class computing facilities using a representative set of scient
 ific codes and several programming models -- CUDA, HIP, Kokkos, RAJA, Open
 MP, OpenACC, and SYCL. Based on the specific characteristics of applicatio
 ns tested, we discuss recommendations to developers on how to choose the r
 ight programming model for their code. These results provide a comprehensi
 ve evaluation of the extent to which each programming model for heterogene
 ous systems provides true performance portability in real-world usage.\n\n
 Tag: Large Scale Systems, Middleware and System Software, Programming Fram
 eworks and System Software\n\nRegistration Category: Workshop Reg Pass\n\n
 Session Chairs: Dhabaleswar K. (DK) Panda (The Ohio State University), Kar
 l Schulz (Advanced Micro Devices (AMD) Inc), Aamir Shafi (The Ohio State U
 niversity), and Hari Subramoni (The Ohio State University)\n\n
END:VEVENT
END:VCALENDAR