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 F 3
DTSTART;TZID=Europe/Stockholm:20190612T153000
DTEND;TZID=Europe/Stockholm:20190612T160000
UID:submissions.pasc-conference.org_PASC19_sess129_msa304@linklings.com
SUMMARY:High-Performance Stencil Codes for Weather Computation in Python u
 sing GT4Py
DESCRIPTION:Minisymposium\nComputer Science and Applied Mathematics, Clima
 te and Weather\n\nHigh-Performance Stencil Codes for Weather Computation i
 n Python using GT4Py\n\nGonzález Paredes\n\nGT4Py is a Python library to w
 rite and run high-performance implementations of stencil kernels using Pyt
 hon. Focusing in weather and climate scientific applications – in a 
 model-agnostic way – it uses the Python language as an embedded
  DSL to define stencils for the regular and block-structured grids typical
 ly found in this domain. GT4Py leverages the GridTools C++ framework to ge
 nerate performance portable implementations of the high-level stencil defi
 nitions written in a subset of Python. This higher level of abstraction in
  the stencil definition, compared to writing pure C++ stencils, enhances s
 cientists' productivity while preserves the excellent performance portabil
 ity of the C++ library on a wide range of computer architectures. The libr
 ary internally works by generating GridTools C++ code which is transparent
 ly compiled and loaded as Python extensions. Additional pure Python implem
 entations can also be generated for debugging or rapid prototyping of new 
 ideas. The library integrates smoothly with key packages of the scientific
  Python ecosystem like NumPy, and also abstracts the differences in data l
 ayout problems by using the Buffer Protocol. In this presentation we will 
 cover the stencil definition interface for GT4Py, its internal architectur
 e allowing the integration of different implementation backends, and its p
 erformance running on different platforms.
END:VEVENT
END:VCALENDAR

