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:20250822T115805Z
LOCATION:Room 5.0B15 & 16
DTSTART;TZID=Europe/Stockholm:20250618T113000
DTEND;TZID=Europe/Stockholm:20250618T120000
UID:submissions.pasc-conference.org_PASC25_sess171_pap129@linklings.com
SUMMARY:OpenACC and OpenMP-Accelerated Fortran/C++ Gyrokinetic Fusion Code
  GENE-X for Heterogeneous Architectures
DESCRIPTION:Jordy Trilaksono and Philipp Ulbl (Max Planck Institute for Pl
 asma Physics), Jeremy Williams (KTH Royal Institute of Technology), Carl-M
 artin Pfeiler and Marion Finkbeiner (Max Planck Institute for Plasma Physi
 cs), Tilman Dannert and Erwin Laure (Max Planck Computing and Data Facilit
 y), Stefano Markidis (KTH Royal Institute of Technology), and Frank Jenko 
 (Max Planck Institute for Plasma Physics)\n\nAchieving net-positive fusion
  energy and its commercialization requires not only engineering marvels bu
 t also state-of-the-art, massively parallel codes that can handle reactor-
 scale simulations. The GENE-X code is a global continuum gyrokinetic turbu
 lence code designed to predict energy confinement and heat exhaust for fut
 ure fusion reactors. GENE-X is capable of simulating plasma turbulence fro
 m the core region to the wall of a magnetic confinement fusion (MCF) devic
 e. Originally written in Fortran 2008, GENE-X leverages MPI+OpenMP for par
 allel computing. In this paper, we augment the Fortran-based compute opera
 tors in GENE-X to a C++-17 layer exposing them to a wide array of C++-comp
 atible tools. Here we focus on offloading the augmented operators to GPUs 
 via directive-based programming models such as OpenACC and OpenMP offload.
  The performance of GENE-X is comprehensively characterized, e.g., by roof
 line analysis on a single GPU and scaling analysis on multi-GPUs. The majo
 r compute operators achieve significant performance improvements, shifting
  the bottleneck to inter-GPU communications. We discuss additional opportu
 nities to enhance further the performance, such as by reducing memory traf
 fic and improving memory utilization efficiency.\n\nDomain: Engineering, P
 hysics, Computational Methods and Applied Mathematics\n\nSession Chair: An
 dreas Lintermann (Forschungszentrum Jülich GmbH)\n\n
END:VEVENT
END:VCALENDAR