BEGIN:VCALENDAR
VERSION:2.0
PRODID:icalendar-ruby
CALSCALE:GREGORIAN
BEGIN:VEVENT
DTSTAMP:20260715T133841Z
UID:712919a3-d1c1-4b9d-8f77-ecf66ec473ea
DTSTART:20150520T080000Z
DTEND:20150521T163000Z
DESCRIPTION:ARCHER\, the UK's national supercomputing service\, offers trai
 ning in software development and high-performance computing to scientists 
 and researchers across the UK. As part of our training service we will be 
 running a 2 day ‘Single-sided PGAS Communications Libraries’ training 
 session.\n\nIn some applications\, the overheads associated with the funda
 mentally two-sided (send and receive) nature of MPI message-passing can ad
 versely affect performance. This is of particular concern for scaling up t
 o extremely large systems. There is a renewed interest in simpler single-s
 ided communications models where data can be written/read directly to/from
  remote processes.\nThis two-day course covers two single-sided PGAS libra
 ries: the OpenSHMEM standard http://www.openshmem.org/ on day 1\, and the 
 new open-source GASPI library http://www.gaspi.de/ on day 2. Hands-on prac
 tical sessions will play a central part in the course\, illustrating key i
 ssues such as the need for appropriate synchronisation to ensure program c
 orrectness. All the exercises can be undertaken using C or Fortran. The Op
 enSHMEM material will be delivered by the ARCHER CSE team\; the GASPI mate
 rial will be delivered by Christian Simmendinger (T-Systems Solutions for 
 Research). Further details of the GASPI training are available below.Effic
 ient Parallel Programming with GASPI\nThe HPC programmers of tomorrow will
  have to write codes\, which are able to deal with systems hundreds of tim
 es larger than the top supercomputers of today. In this Tutorial we presen
 t an asynchronous dataflow programming model for Partitioned Global Addres
 s Spaces (PGAS) as an alternative to the programming model of MPI.\nGASPI\
 , which stands for Global Address Space Programming Interface\, is a parti
 tioned global address space (PGAS) API. The GASPI API is designed as a C/C
 ++/Fortran library and focused on three key objectives: scalability\, flex
 ibility and fault tolerance. In order to achieve its significantly improve
 d scaling behaviour GASPI aims at asynchronous dataflow with remote comple
 tion\, rather than bulk-synchronous message exchanges. GASPI follows a sin
 gle/multiple program multiple data (SPMD/MPMD) approach and offers a small
 \, yet powerful API (also see http://www.gaspi.de). GASPI today is used in
  academic and industrial simulation applications.\nThe Tutorial gives an o
 verview over the key concepts of elements of GASPI\, such as synchronizati
 on primitives\, synchronous and asynchronous collectives\, fine grained co
 ntrol over one-sided read and write communication primitives\, global atom
 ics\, passive receives\, communication groups and communication queues. GA
 SPI aims at multi-threaded execution\, offers a thread-safe API and can be
  used in combination with all current threading models (OpenMP\, Pthreads\
 , MCTP\, and others). GASPI provides its partitioned global address space 
 in the form of configurable memory segments and features support for heter
 ogeneous memory architectures. All GASPI segments can directly read and wr
 ite from/to each other. By spawning a GASPI segment across e.g. the main m
 emory of a distributed Xeon Phi system and a segment across the memory of 
 the corresponding x86 host\, the GASPI API hence can provide a consistent 
 and cohesive view of this hybrid distributed memory architecture. The flex
 ibility of the configurable GASPI segments also allows developers to both 
 leverage multiple memory models within a single application and/or to glob
 ally tightly couple different applications (e.g. multi-physics solvers). G
 ASPI is failure tolerant and allows for a dynamic (shrinking or growing) n
 ode set. All non-local procedures feature timeout parameters and provide a
  well defined exit status.\nThe Tutorial will provide a hands-on introduct
 ion (in C and Fortran) which features examples and use-cases for all its k
 ey concepts (segment creation\, synchronization primitives\, read and writ
 e communication primitives\, global atomics\, passive receives\, collectiv
 es\, communication groups and communication queues). Case studies which de
 monstrate the various aspects of the GASPI API and application categories 
 that can take advantage of these aspects are identified. This Tutorial als
 o includes a discussion of the current GPI-2 release\, the first release w
 hich supports the GASPI standard. Benchmark results on different platforms
  are discussed. Tools for programming with GASPI/GPI such as profiling too
 ls are presented in a "howto" section.\n\nhttps://events.prace-ri.eu/event
 /382/
SUMMARY:Single-sided PGAS Communications Libraries @ EPCC
URL;VALUE=URI:https://events.prace-ri.eu/event/382/
END:VEVENT
END:VCALENDAR