BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:Europe/Stockholm
X-LIC-LOCATION:Europe/Stockholm
BEGIN:DAYLIGHT
TZOFFSETFROM:+0100
TZOFFSETTO:+0200
TZNAME:CEST
DTSTART:19700308T020000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=-1SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:+0200
TZOFFSETTO:+0100
TZNAME:CET
DTSTART:19701101T020000
RRULE:FREQ=YEARLY;BYMONTH=10;BYDAY=-1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20210916T132448Z
LOCATION:Jean Calvin
DTSTART;TZID=Europe/Stockholm:20210705T160000
DTEND;TZID=Europe/Stockholm:20210705T163000
UID:submissions.pasc-conference.org_PASC21_sess132_msa194@linklings.com
SUMMARY:Lessons Learned in Optimizing Weather Models for GPUs with Directi
 ves
DESCRIPTION:Minisymposium\n\nLessons Learned in Optimizing Weather Models 
 for GPUs with Directives\n\nGovett, Flynt, Middlecoff, Stringer, Yu\n\nDir
 ective-based parallelization is a widely-used approach used to run weather
  prediction models on GPUs.  Directives are inserted into the code an
 d used by the compiler to guide parallelization of the application.  
 Two types of industry standard directives commonly used are OpenACC and Op
 enMP. NOAA’s Global Systems Laboratory has primarily used OpenACC di
 rectives for parallelization but recently began exploring OpenMP as its ca
 pabilities have grown. In this presentation, I will describe our expe
 riences using OpenACC to parallelize portions of several weather predictio
 n models.  Select routines will be shown to highlight the strengths a
 nd weaknesses of these parallelization efforts. I will also report on prog
 ress using openMP to run select kernels on the GPU.\n\nDomain: CS and Math
 , Climate and Weather, Physics
END:VEVENT
END:VCALENDAR