PRACE Winter School 2012, Bologna, Italy

CET
Description

The PRACE Winter School is an intense, 5 day, graduate level course in high performance computing, organized in the framework of the European project PRACE (http://www.prace-ri.eu).

The school will be focused on hybrid programming  for the best exploitation of massively parallel architectures. The facility available at CINECA for exercises is called PLX and is the largest public GPU cluster in Europe. It is made of 274 compute nodes, each containing 2 NVIDIA® Tesla® M2070 and 2 Intel(R) Xeon(R) Westmere six-core E5645 processors.
The school will be held on 6-10 February and the official language will be English.


The school is aimed at PRACE users, final year master students, PhD students, and young researchers in computational sciences and engineering, with different backgrounds, interested in applying the emerging technologies on high performance computing to their research. The school will cover the following topics:

  • Massively parallel architectures: BlueGene, hybrid clusters
  • Accelerators (NVIDIA GPU, Intel MIC)
  • Advanced MPI
  • Hybrid programming with OpenMP
  • Hybrid programming with CUDA
  • Towards new hybrid solutions with MPI: TBB.

Some use cases will also be presented.

Slides
    • 09:10 10:50
      Introduction to massively parallel architectures and hybrid programming
    • 10:50 11:10
      Coffee Break 20m
    • 11:10 13:00
      BlueGene/Q
    • 13:00 14:30
      Lunch 1h 30m
    • 14:30 16:10
      Accelerators: MIC, GPU
    • 16:10 16:30
      Coffee Break 20m
    • 16:30 18:20
      Hybrid cluster: PLX
    • 09:10 10:50
      OpenMP
    • 10:50 11:10
      Coffee Break 20m
    • 11:10 13:00
      OpenMP
    • 11:10 13:00
      Towards new solutions: OpenACC
    • 13:00 14:30
      Lunch 1h 30m
    • 14:30 16:10
      Introduction to MPI+OpenMP
    • 16:10 16:30
      Coffee Break 20m
    • 16:30 18:20
      Introduction to MPI+OpenMP
    • 16:30 18:20
      Use Case: OpenFOAM
    • 09:10 10:50
      MPI3
    • 09:10 10:50
      Overview on debuggers
    • 10:50 11:10
      Coffee Break 20m
    • 11:10 13:00
      Overview on profilers Scalasca
    • 13:00 14:30
      Lunch 1h 30m
    • 14:30 16:10
      MPI+OpenMP Exercises
    • 16:10 16:30
      Coffee Break 20m
    • 16:30 18:20
      MPI+OpenMP Exercises
    • 09:10 10:50
      CUDA & Exercises
    • 10:50 11:10
      Coffee Break 20m
    • 11:10 13:00
      CUDA & Exercises
    • 13:00 14:30
      Lunch 1h 30m
    • 14:30 16:10
      Introduction to MPI+CUDA
    • 16:10 16:30
      Coffee Break 20m
    • 16:30 18:20
      MPI+CUDA Exercises
    • 09:10 10:50
      MPI+OpenMP+CUDA Use Case: QE
    • 10:50 11:10
      Coffee Break 20m
    • 11:10 13:00
      Towards new solutions: TBB
    • 13:00 14:30
      Lunch 1h 30m
    • 14:30 16:10
      TBB Exercises