BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//CERN//INDICO//EN
BEGIN:VEVENT
SUMMARY:GPU Programming Part 2: Special and Advanced Topics
DTSTART:20260629T070000Z
DTEND:20260703T110000Z
DTSTAMP:20260528T162100Z
UID:indico-event-478@events.hpc-portal.eu
CONTACT:j.meinke@fz-juelich.de
DESCRIPTION:This course will take place as an online event. The link to th
 e streaming platform will be provided to the accepted registrants only. Th
 e course will be held in English.\nCourse Content:\nGPU-accelerated comput
 ing drives current scientific research. Writing fast numeric algorithms fo
 r GPUs offers high application performance by offloading compute-intensive
  portions of the code to a GPU.\nThis advanced course will cover special
  and advanced aspects of GPU architectures and programming. Examples of in
 creasing complexity are used to demonstrate the optimisation and tuning of
  scientific applications.\nThe course consists of modules providing more i
 n-depth coverage of multi-GPU programming\, modern CUDA concepts\, CUDA Fo
 rtran\, and portable programming models such as OpenACC and C++ parallel S
 TL algorithms:\nA) Advanced multi-GPU programming with MPIB) Advanced m
 ulti-GPU programming with NVIDIA Collective Communications Library (NCCL) 
 and NVIDIA Shared Memory (NVSHMEM)C) Advanced and modern CUDA concepts
  (Cooperative Groups\, CUB Primitives\, modern C++ programming)D) CUDA F
 ortran and GPU-accelerated standard Fortran or Kokkos for portable GP
 U programmingE) GPU programming with abstractions (OpenACC\, Standard La
 nguage Programming (pSTL))\nAttendees are invited to pick and choose the p
 arts of the advanced course (A - E) they want to attend. The modules are m
 ostly freestanding.\nNote: If you want to learn the basics of GPU programm
 ing\, there is a separate course GPU Programming Part 1: Foundation. It 
 includes an introduction to GPU/parallel computing\, programming with CUDA
 \, GPU libraries\, tools for debugging and profiling\, and performance opt
 imisations. It will take place 24.03.-26.03.2026 on-site at JSC. Please vi
 sit GPU Programming Part 1: Foundations for more information and to regi
 ster.\nPrerequisites:\nParticipants either need to attend the course GPU 
 Programming Part 1: Foundations first or prove equivalent knowledge of GP
 U programming in order to participate in this advanced course.\nProgrammer
 s interested primarily in OpenACC may skip parts D and E of the "Foundatio
 ns" course and still choose part E from the "Special and Advanced Topics" 
 course. Participation in the full "Foundations" course\, however\, is reco
 mmended.\nSome knowledge about Linux\, e.g. make\, command line editor\, L
 inux shell\, experience in C/C++ is also required.\nA personal institution
 al email address (university/research institution\, government agency\, or
 ganisation\, or company) is required to register for JSC training courses.
  If you don't have an institutional email address\, please get in touch wi
 th the contact person for this course.\nTarget Audience:\nScientists who w
 ant to improve their use of GPU systems\nLanguage:\nThis course is given i
 n English.\nDuration:\n5 half days\nDates:\n29.06.-03.07.2026\, 09:00-13:0
 0 each day\nVenue:\nOnline\nNumber of Participants:\nMaximum 30\nInstructo
 rs:\nJan Meinke\, Andreas Herten\, Kaveh Haghighi-Mood\, Jayesh Badwaik (J
 SC)Laura Morgenstern\, Markus Hrywniak (NVIDIA)\n\nhttps://events.hpc-port
 al.eu/event/478/
LOCATION:Online
URL:https://events.hpc-portal.eu/event/478/
END:VEVENT
END:VCALENDAR