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:20260522T162631Z
LOCATION:Bldg. 6 - Room 104
DTSTART;TZID=Europe/Stockholm:20260629T133000
DTEND;TZID=Europe/Stockholm:20260629T153000
UID:submissions.pasc-conference.org_PASC26_sess109@linklings.com
SUMMARY:MS1F - What the FORTRAN? – Drowning in Waves Edition
DESCRIPTION:Organizer(s): Ahmad Nawab, Michael Lange, and Balthasar Reuter
  (ECMWF)\n\nFortran, the primary programming language underpinning many op
 erational weather and climate codes, was built around the fundamental prin
 ciple that performance optimisation is left to the compiler. However, to f
 ully utilise modern HPC architectures and GPUs, additional programming par
 adigms as well as invasive code changes are often needed. While high-level
  DSLs or C-based abstraction layers like Kokkos provide a disruptive path 
 to performance portability, for Fortran such abstractions are lacking. Thi
 s leads to the fundamental question “What the actual FORTRAN can we do to 
 achieve performance portability for complex physical models?” In this mini
 symposium we continue this discussion by looking at cross-architecture por
 tability of ecWAM, ECMWF’s operational wave model. ecWAM is a key componen
 t of ECMWF’s Integrated Forecasting System (IFS), and is also used operati
 onally in all the IFS-derived Destination Earth digital twins. We highligh
 t the complexities and challenges that its diverse range of computing patt
 erns pose, and together with industrial and academic partners, discuss dif
 ferent strategies towards a maintainable cross-platform adaptation strateg
 y that runs efficiently across different HPC architectures. As ecWAM share
 s many similarities with other scientifically state-of-the-art third gener
 ation wave models, this minisymposium potentially offers a wide-ranging im
 pact across the weather and climate communities.\n\nYou Never Really Cared
  About Double Precision, Did You?\n\nAs scientific applications, especiall
 y large Fortran simulations, are increasingly expected to run efficiently 
 on hardware originally designed for machine learning workloads, performanc
 e portability has become essential. At the same time, the reduced precisio
 n supported by such hardware poses a majo...\n\n\nAlexandru Calotoiu, Prat
 yai Mazumder, and Torsten Hoefler (ETH Zurich)\n---------------------\nFro
 m Parallelism to Bytes in Flight: Modernising ecWAM for Memory-Bandwidth-D
 ominated AI-Era GPUs\n\nWeather and climate models are complex scientific 
 applications, often written in Fortran requiring performance portability a
 cross diverse high-performance computing architectures. As these codes are
  mainly maintained by domain scientists, extensive hardware-specific rewri
 tes are often impractical. D...\n\n\nAnastasia Stulova and Lukas Mosimann 
 (NVIDIA Inc.)\n---------------------\nAre CPUs Still Relevant for Weather 
 Codes?\n\nWe discuss the establishment of well-defined and relevant perfor
 mance challenges for the ecwam dwarf. We investigate code refactoring oppo
 rtunities, and we will try to draw parallels to the opportunities found in
  other dwarfs from ECMWF. Finally, we present performance results obtained
  on specific I...\n\n\nJacob Poulsen and Igor Vorobtsov (Intel Corporation
 )\n---------------------\nMaking Waves on AMD GPUs with the AMD Next-Gener
 ation Fortran Compiler\n\nFortran remains a key language in scientific com
 puting, especially in domains such as weather and climate modelling. In th
 at context, AMD has created a next-generation Fortran compiler based on th
 e open-source LLVM Flang project, enabling HPC developers to exploit AMD G
 PUs through OpenMP target synt...\n\n\nPaul Mullowney (AMD, Inc)\n\nDomain
 : Climate, Weather, and Earth Sciences\n\nSession Chair: Ahmad Nawab (ECMW
 F)
END:VEVENT
END:VCALENDAR
