Event Detail

Registration

Sorry, the registration period for this event is over.

Parallel Programming Summer School - 17-19 August 2011

 

CSCS will be hosting a three day summer school on parallel programming aimed at graduate students who are new to the world of high performance computing and who wish to learn the basic skills required to write, develop and maintain parallel applications in scientific computing.

Registration deadline: August 12, 2011. 

Please contact nstring(at)cscs.ch for further technical information and apinna(at)cscs.ch for logistical information.

Instructors

Neil Stringfellow, Matthew Cordery, Andreas Jocksch, Jean-Guillaume Piccinali, Tim Robinson from CSCS

VenueCSCS, Via Cantonale, Galleria 2, 6928 Manno  (Please note that on googlemap CSCS is wrongly posted)
Time09:00 - 17:00 each day
Prerequisites

A good working knowledge of C/C++ or Fortran. Familiarity with a Unix/Linux development environment. 

Participants are expected to bring a laptop for hands-on training.

Maximum number of participants28
Accommodation

Participants are kindly requested to make their own arrangements for accommodation

Workshop Agenda


Topics covered during this summer school include:-

    Introduction to parallel execution concepts
        Distributed and shared memory programming
        Scaling and Efficiency
        Domain decomposition
        Scaling limitations and Amdahl's law   

    Basics of computer architectures for HPC
   
    Distributed memory machine architecture
   
    Fundamentals of distributed memory programming with MPI
        Introduction to MPI
        Point-to-point communications
        Collective communications
        MPI Derived Datatypes
        Communicators and Groups
        Advanced MPI-1 features
   
    Extended MPI features
        MPI one-sided communications
        MPI I/O
        MPI dynamic process management
   
    Performance considerations for MPI
   
    Future MPI extensions
   
    Shared memory machine architecture
   
    Shared memory programming
        Possible approaches to shared memory programming
   
    Introduction to OpenMP for shared memory programming
        OpenMP basics
        Work sharing constructs in OpenMP
        OpenMP tasks
        OpenMP 3.0 features
   
    Performance considerations for shared memory programming and OpenMP
   
    Hybrid programming with MPI and OpenMP.
   
    The use of third-party parallel libraries to improve productivity and performance.
   
    Tools to help with parallel performance improvements.

The summer school will be a mixture of lectures, presentations and hands-on sessions.


Back to listing