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:20220812T074335Z
LOCATION:Darwin Room
DTSTART;TZID=Europe/Stockholm:20220629T113000
DTEND;TZID=Europe/Stockholm:20220629T120000
UID:submissions.pasc-conference.org_PASC22_sess152_msa185@linklings.com
SUMMARY:New Advances to Prepare GYSELA-X Code for Exascale Global Gyrokine
 tic Plasma Turbulence Simulations: Porting on GPU and ARM Architectures
DESCRIPTION:Minisymposium\n\nNew Advances to Prepare GYSELA-X Code for Exa
 scale Global Gyrokinetic Plasma Turbulence Simulations: Porting on GPU and
  ARM Architectures\n\nGrandgirard, Obrejan, Asahi, Bernard, Bigot...\n\nUn
 derstanding turbulent transport in fusion plasma is a key challenge for fu
 sion devices like ITER. Non-linear flux-driven 5D gyrokinetic codes are th
 e most accurate framework to tackle this problem. They are extremely HPC c
 hallenging. The GYSELA code [V. Grandgirard et al., Comp. Phys. Com. 2017 
 (2016) 35] is one of them based on a Backward Semi-Lagrangian scheme. The 
 code, developed in Fortran and based on a hybrid MPI/OpenMP parallelizatio
 n, runs efficiently on more than 100'000 cores on current standard archite
 ctures. With a consumption of 150 million CPU hours per year, the code alr
 eady makes massive use of petascale computing capacities. It has demonstra
 ted that edge-core turbulent plasma simulations with kinetic electrons wil
 l require exascale HPC capabilities. However, the new architectures planne
 d for exascale, based on heterogeneous accelerated compute nodes, are less
  favourable to applications such as GYSELA which requires a lot of memory 
 and parallel communications. Achieving performance on exascale computers w
 ill be far from trivial. We will present our feedback on the porting of th
 e GYSELA code to ARM-A64FX architectures (the supercomputer Fugaku at RIKE
 N/Japan) and AMD-GPU architectures. We will also discuss the advances made
  to manage the Tbytes of data handled during each simulation.\n\nDomain: C
 omputer Science and Applied Mathematics, Physics
END:VEVENT
END:VCALENDAR