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:20230831T095745Z
LOCATION:Davos
DTSTART;TZID=Europe/Stockholm:20230626T112000
DTEND;TZID=Europe/Stockholm:20230626T115000
UID:submissions.pasc-conference.org_PASC23_sess104_pos146@linklings.com
SUMMARY:P18 - Directive-Based, Fortran/C++ Interoperable Approach to GPU O
 ffloading of the High Performance Gyrokinetic Turbulence Code GENE-X
DESCRIPTION:Poster\n\nJordy Trilaksono, Philipp Ulbl, and Andreas Stegmeir
  (Max Planck Institute for Plasma Physics) and Frank Jenko (Max Planck Ins
 titute for Plasma Physics, University of Texas at Austin)\n\nThe achieveme
 nt of high plasma confinement is the key to realize commercially attractiv
 e energy production by magnetic confinement fusion (MCF) devices. Turbulen
 ce plays a significant role in maintaining the plasma confinement within M
 CF devices. The GENE-X code is based on an Eulerian (continuum) approach t
 o the discretization of the five dimensional gyrokinetic equation that des
 cribes plasma turbulence. Our discretization is specialized to simulate pl
 asma turbulence anywhere within MCF devices, from the hot plasma core to t
 he cold wall. GENE-X is written in object-oriented modern Fortran 2008 lev
 eraging MPI+OpenMP parallelization to facilitate large scale turbulence si
 mulations. Here, we present our development efforts to further extend the 
 parallelization scheme to GPUs, which is essential for scalability support
  towards simulations of larger, reactor-relevant fusion devices. The curre
 nt implementation in GENE-X provides a proof of concept of our native Fort
 ran/C++ interoperability approach by successfully supporting several GPU b
 ackends such as OpenACC, OpenMP offload and CUDA. We present first benchma
 rks of our directive-based OpenACC implementation of the most computationa
 lly expensive part of GENE-X. A significant performance increase was achie
 ved on the GPU, compared to equivalent CPU benchmarks.\n\nSession Chair: E
 laine M. Raybourn (Sandia National Laboratories)
END:VEVENT
END:VCALENDAR