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:20241120T082409Z
LOCATION:HG F 26.3
DTSTART;TZID=Europe/Stockholm:20240604T123000
DTEND;TZID=Europe/Stockholm:20240604T130000
UID:submissions.pasc-conference.org_PASC24_sess116_msa191@linklings.com
SUMMARY:Scientific Machine Learning to Optimize Plasma Turbulence Simulati
 ons
DESCRIPTION:Minisymposium\n\nVirginie Grandgirard (CEA); David Zarzoso (CN
 RS); Robin Varennes (National University of Singapore); and Feda Almuhisen
 , Kevin Obrejan, and Julien Bigot (CEA)\n\nControlled fusion offers the pr
 omise of sustainable and safe energy production on Earth. In magnetic fusi
 on devices, the power gain increases nonlinearly with the energy confineme
 nt time. The quality of the plasma energy confinement thus largely determi
 nes the size and therefore the cost of a fusion reactor. Unfortunately, sm
 all-scale turbulence limits the quality of confinement in most fusion devi
 ces. Hence modelling of turbulent transport is mandatory to find routes to
 wards improved confinement regimes. \nNumerical simulations are based on a
  kinetic description of the plasma that can only be performed on most powe
 rful supercomputers. The gyrokinetic GYSELA code runs efficiently on sever
 al hundred thousand CPU cores. With a consumption of 150 million CPU hours
  per year, the code makes massive use of petascale computing capacities an
 d manipulates Petabytes of data. However, there are still many challenges 
 to overcome in order to achieve a full description of electron dynamics th
 at will require exascale simulations. \nIn this context, scientific machin
 e learning could play an important role to optimize the storage and the nu
 mber of CPU hours consumed. We will present here our on-going work on inte
 grating in-situ diagnostics using Artificial Intelligence techniques for d
 ata compression and automatic detection of anomalies or rare events.\n\nDo
 main: Physics, Computational Methods and Applied Mathematics\n\nSession Ch
 airs: Stephan Brunner (EPFL); Eric Sonnendrücker (Max Planck Institute for
  Plasma Physics, Technical University of Munich); and Laurent Villard (EPF
 L)
END:VEVENT
END:VCALENDAR