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:20190719T085743Z
LOCATION:HG D 1.1
DTSTART;TZID=Europe/Stockholm:20190612T130000
DTEND;TZID=Europe/Stockholm:20190612T133000
UID:submissions.pasc-conference.org_PASC19_sess132_msa174@linklings.com
SUMMARY:PSyclone and its Use in LFRic
DESCRIPTION:Minisymposium\nComputer Science and Applied Mathematics, Clima
 te and Weather\n\nPSyclone and its Use in LFRic\n\nKavcic\n\nPSyclone is a
  domain-specific compiler and source-to-source translator developed for us
 e in finite element, finite volume and finite difference codes. Using the 
 information from a supported API, PSyclone can generate code exploiting di
 fferent parallel programming models, enabling deployment of a single sourc
 e science code onto different machine architectures. We will present overv
 iew of PSyclone functionality with its main concepts and support for speci
 fic models through its APIs. We will then focus on PSyclone API for LFRic,
  the new weather and climate modelling system being developed by the UK Me
 t Office to replace the existing Unified Model in preparation for exascale
  computing in the 2020s. LFRic infrastructure provides APIs to support the
  distributed memory parallelism via libraries, for example YAXT, whereas P
 Syclone adds support for the shared memory parallelism. In addition, PSycl
 one can optimize the use of distributed memory parallelism in LFRic via op
 timizations such as asynchronous-halo-exchange transformation and redundan
 t computation into annexed dofs. Preliminary LFRic performance results sho
 w strong scaling and an indication that hybrid MPI/OpenMP performs better 
 than pure MPI. More optimizations will be explored.
END:VEVENT
END:VCALENDAR