BEGIN:VCALENDAR
VERSION:2.0
PRODID:icalendar-ruby
CALSCALE:GREGORIAN
BEGIN:VEVENT
DTSTAMP:20260717T180250Z
UID:8627d089-9f7e-4495-b9c7-2b437b7cf60f
DTSTART:20170112T090000Z
DTEND:20170112T160000Z
DESCRIPTION:Most HPC systems are clusters of shared memory nodes. Such SMP 
 nodes can be small multi-core CPUs up to large many-core CPUs. Parallel pr
 ogramming may combine the distributed memory parallelization on the node i
 nterconnect (e.g.\, with MPI) with the shared memory parallelization insid
 e of each node (e.g.\, with OpenMP or MPI-3.0 shared memory). This course 
 analyzes the strengths and weaknesses of several parallel programming mode
 ls on clusters of SMP nodes. Multi-socket-multi-core systems in highly par
 allel environments are given special consideration. MPI-3.0 has introduced
  a new shared memory programming interface\, which can be combined with in
 ter-node MPI communication. It can be used for direct neighbor accesses si
 milar to OpenMP or for direct halo copies\, and enables new hybrid program
 ming models. These models are compared with various hybrid MPI+OpenMP appr
 oaches and pure MPI.  Numerous case studies and micro-benchmarks demonstr
 ate the performance-related aspects of hybrid programming.\n\nTools for hy
 brid programming such as thread/process placement support and performance 
 analysis are presented in a "how-to" section. This course provides scienti
 fic training in Computational Science\, and in addition\, the scientific e
 xchange of the participants among themselves.\n\nThe course is a PRACE Adv
 anced Training Center event.\n\nhttps://events.prace-ri.eu/event/533/
SUMMARY:Introduction to hybrid programming in HPC @ LRZ
URL;VALUE=URI:https://events.prace-ri.eu/event/533/
END:VEVENT
END:VCALENDAR