BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:America/Chicago
X-LIC-LOCATION:America/Chicago
BEGIN:DAYLIGHT
TZOFFSETFROM:-0600
TZOFFSETTO:-0500
TZNAME:CDT
DTSTART:19700308T020000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=2SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:-0500
TZOFFSETTO:-0600
TZNAME:CST
DTSTART:19701101T020000
RRULE:FREQ=YEARLY;BYMONTH=11;BYDAY=1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20211207T055412Z
LOCATION:Online
DTSTART;TZID=America/Chicago:20211114T140000
DTEND;TZID=America/Chicago:20211114T143000
UID:submissions.supercomputing.org_SC21_sess429_ws_p3hpc105@linklings.com
SUMMARY:Evaluation of Performance Portability of Applications and Mini-App
 s across AMD, Intel, and NVIDIA GPUs
DESCRIPTION:Workshop\n\nEvaluation of Performance Portability of Applicati
 ons and Mini-Apps across AMD, Intel, and NVIDIA GPUs\n\nKwack, Tramm, Bert
 oni, Ghadar, Homerding...\n\nThis paper evaluates the progress being made 
 on achieving performance portability by ECP applications, or their proxy-a
 pplications, across a diverse spectrum of applications domains and approac
 hes to achieving performance portability. The applications or proxy-apps e
 valuated are AMR-Wind, HACC, SW4, GAMESS RI-MP2, XSBench, and TestSNAP. Th
 ese codes are being redeveloped using the SYCL, OpenMP, RAJA, or Kokkos pr
 ogramming models, or the AMReX frame-work. In this paper we assess their p
 erformance portability across the AMD MI100, Intel Gen9, and NVIDIA A100 G
 PUs. Since each GPU has different performance characteristics we have util
 ized the roofline performance model to compute the performance efficiency 
 and evaluate performance portability across the platforms. The merits of d
 ifferent metrics for quantifying performance portability are considered an
 d a metric based on the standard deviation of roofline efficiencies is pro
 posed as a preferred metric. Finally, observations on developer productivi
 ty are made based on the experience gained working with these applications
 .\n\nTag: Online Only, Heterogeneous Systems, Parallel Programming Languag
 es and Models, Performance, Productivity Tools, Software Engineering\n\nRe
 gistration Category: Workshop Reg Pass
END:VEVENT
END:VCALENDAR