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:20250822T115804Z
LOCATION:Campussaal - Plenary Room
DTSTART;TZID=Europe/Stockholm:20250616T102000
DTEND;TZID=Europe/Stockholm:20250616T105000
UID:submissions.pasc-conference.org_PASC25_sess149_pos148@linklings.com
SUMMARY:P22 - GT4Py: A Python Framework for the Development of High-Perfor
 mance Weather and Climate Applications
DESCRIPTION:Mauro Bianco (ETH Zurich / CSCS); Yilu Chen (ETH Zurich); Till
  Ehrengruber (ETH Zurich / CSCS); Sara Faghih-Naini (ECMWF); Nicoletta Far
 abullini (ETH Zurich); Abishek Gopal (NCAR, ETH Zurich); Rico Häuselmann (
 ETH Zurich / CSCS); Samuel Kellerhals (ETH Zurich); Christos Kotsalos and 
 Ioannis Magkanaris (ETH Zurich / CSCS); Magdalena Luz (ETH Zurich); Christ
 oph Müller (MeteoSwiss); Philip Müller, Edoardo Paone, and Enrique Gonzále
 z Paredes (ETH Zurich / CSCS); David Strassmann (ETH Zurich); and Felix Th
 aler, Hannes Vogt, and Thomas Schulthess (ETH Zurich / CSCS)\n\nGT4Py is a
  Python framework for weather and climate applications simplifying the dev
 elopment and maintenance of high-performance codes in prototyping and prod
 uction environments. \nGT4Py separates model development from hardware-dep
 endent optimizations, instead of intermixing them in source code, as regul
 arly done in lower-level languages like Fortran or C. Domain scientists fo
 cus solely on numerical modeling using a declarative embedded domain speci
 fic language supporting common computational patterns of dynamical cores a
 nd physical parametrizations. An optimizing toolchain then transforms this
  high-level representation into a finely tuned implementation for the targ
 et hardware architecture. This separation of concerns allows performance e
 ngineers to add optimizations or support new hardware architectures withou
 t modifying the application itself, increasing productivity for both domai
 n scientists and performance engineers. \nWe present recent developments f
 rom the past year, including a new performance backend based on DaCe, a fr
 amework for high-performance parallel programming that uses a data-centric
  intermediate representation (IR) based on Stateful DataFlow multiGraphs (
 SDFG). Performance results on Nvidia GH200 and AMD MI300 are presented for
  two weather models: an ICON-based model ported to GT4Py as part of the EX
 CLAIM project at ETH, and PMAP, a portable atmospheric model developed at 
 ECMWF, covering applications from large-eddy simulations to global numeric
 al weather prediction.\n\nSession Chair: Chris Cantwell (Imperial College 
 London)\n\n
END:VEVENT
END:VCALENDAR