BEGIN:VCALENDAR
VERSION:2.0
PRODID:icalendar-ruby
CALSCALE:GREGORIAN
BEGIN:VEVENT
DTSTAMP:20260705T203034Z
UID:349e0bc3-fdd5-4da8-87fa-3525382520eb
DTSTART:20210301T083000Z
DTEND:20210303T160000Z
DESCRIPTION:The training will be fully remote. In order to participate\, as
  a pre-requisite\, you need a personal computer connected to the internet 
 with a working SSH client.\n\n\n\n\nThe increase in computational power go
 es hand in hand with an increase in the size of the data to be managed\, b
 oth on the input and on the output sides. IO can easily become a bottlenec
 k for large scale architectures. Understanding of parallel file system mec
 hanisms and parallel IO concepts enables users to efficiently use existing
  high level libraries like HDF5\, NetCDF or SIONlib.\n\nWith the increasin
 g performance gap between compute and storage\, even the best use of IO ba
 ndwidth might not be enough. This is especially critical for checkpointing
  in the context of fault tolerance. This course proposes an introduction t
 o FTI\, a library that aims to give computational scientists the means to 
 perform fast and efficient multilevel checkpointing in large scale superco
 mputers.\n\nTo handle the multiplication of libraries\, this course will a
 lso introduce the PDI Data Interface\, a solution to couple simulation cod
 es with libraries for I/O and post-processing (including HDF5\, NetCDF\, S
 IONlib\, FTI\, ...) based on simple annotations. PDI improves code quality
  with a) annotations independent of the library used\, b) a choice of IO s
 trategy at runtime in YAML and c) negligible overhead to access the full p
 ower of the underlying libraries.\n\nAgenda:\n\nMonday:\n\n\n	9h30 - 11h: 
 Storage@TGCC &amp\; Lustre filesystems (Thomas Leibovici - TGCC\, CEA)\n	1
 1h - 12h30: parallel I/O strategies and optimization with a focus on  SIO
 Nlib (Sebastian Lührs - JSC\, FZJ)\n	14h - 14h30: parallel I/O strategies
  and optimization with a focus on SIONlib Contn'd (Sebastian Lührs - JSC\
 , FZJ)\n	14h30 - 17h: Sequential HDF5 (Matthieu Haefele - LMAP\, CNRS)\n\n
 \nTuesday:\n\n\n	9h30 - 12h30: Parallel HDF5 (Matthieu Haefele - LMAP\, CN
 RS)\n	14h - 17h: NetCDF (Olga Abramkina - MdlS/IDRIS\, CNRS)\n\n\nWednesda
 y:\n\n\n	9h30 - 12h30: the FTI fault-tolerance library (Leonardo Bautista 
 Gomez - BSC)\n	14h - 17h: the PDI Data Interface (Julien Bigot - MdlS\, CE
 A)\n\n\nInstructors: Olga Abramkina (MdlS/IDRIS\, CNRS)\, Leonardo Bautist
 a Gomez (BSC)\, Julien Bigot (MdlS\, CEA)\, Matthieu Haefele (LMAP\, CNRS)
 \, Thomas Leibovici (TGCC\, CEA)\, Sebastian Lührs (JSC\, FZJ)\n\nLearnin
 g outcomes: After this course\, participants should understand the trade-o
 ffs implied by using a parallel file-system\, and know how to efficiently 
 use parallel IO libraries. Participants will also have a basic understandi
 ng and practise of FTI and PDI.\n\nPrerequisites: Knowledge of C or Fortra
 n programming languages\, parallel programming with MPI\nhttps://events.pr
 ace-ri.eu/event/1121/
SUMMARY:[ONLINE] High Performance Parallel IO and post-processing @ MdlS
URL;VALUE=URI:https://events.prace-ri.eu/event/1121/
END:VEVENT
END:VCALENDAR