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:20260422T000605Z
LOCATION:DEF Concourse
DTSTART;TZID=America/Denver:20231116T100000
DTEND;TZID=America/Denver:20231116T170000
UID:submissions.supercomputing.org_SC23_sess300_spostg106@linklings.com
SUMMARY:Case Study for Performance Portability of GPU Programming Framewor
 ks for Hemodynamic Simulations
DESCRIPTION:Aristotle Martin (Duke University)\n\nPreparing for the deploy
 ment of large scientific and engineering codes on GPU-dense exascale syste
 ms is made challenging by the unprecedented diversity of vendor hardware a
 nd programming model alternatives for offload acceleration. To leverage th
 e exaflops of GPUs from Frontier (AMD) and Aurora (Intel), users of high p
 erformance computing (HPC) legacy codes originally written to target NVIDI
 A GPUs will have to make decisions with implications regarding porting eff
 ort, performance, and code maintainability. To facilitate HPC users naviga
 ting this space, we have established a pipeline that combines generalized 
 GPU performance models with proxy applications to evaluate the performance
  portability of a massively parallel computational fluid dynamics (CFD) co
 de in CUDA, SYCL, HIP, and Kokkos with backends on current NVIDIA-based ma
 chines as well as testbeds for Aurora (Intel) and Frontier (AMD). We demon
 strate the utility of predictive models and proxy applications in gauging 
 performance bounds and guiding hand-tuning efforts.\n\nRegistration Catego
 ry: Tech Program Reg Pass, Exhibits Reg Pass\n\n
END:VEVENT
END:VCALENDAR