Virtual Institute – High Productivity Supercomputing

8th VI-HPS Tuning Workshop (GRS, Germany)

Date

September 5-9, 2011

Location

German Research School for Simulation Sciences, Aachen, Germany

Goals

This workshop will:

  • give an overview of the VI-HPS programming tools suite
  • explain the functionality of individual tools, and how to use them effectively
  • offer hands-on experience and expert assistance using the tools

The Paraver toolset from Barcelona Supercomputing Center is also being featured.

 

The workshop will be held in English and run from 09:00 to not later than 18:00 each day, with breaks for lunch and refreshments. There is no fee for participation, however, participants are responsible for their own travel and accommodation.

Schedule

 

Day 1 Monday 5 Sep
09:00 (registration & set-up of course accounts on workshop computers)
[Optional] Individual preparation of participants' own codes.
12:00 (lunch)
13:00 (registration)
13:30 Welcome & Introduction to VI-HPS
  • Introduction to parallel performance analysis & engineering [Moore, UTK-ICL]
  • 15:00 (break)
    15:30 Overview of VI-HPS tools [Wylie, JSC]
    16:30 Lab setup
  • Computer systems and software environment
  • Building and running NPB-MZ-MPI/BT-MZ
  • 17:30 (adjourn)
    19:00 Social dinner sponsored by Bull, Im Alten Zollhaus

    Day 2 Tuesday 6 Sep
    09:00 Scalasca performance analysis toolset [Wylie, JSC]
  • Scalasca hands-on tutorial exercises
  • Scalasca performance properties
  • Scalasca case studies
  • 10:30 (break)
    11:00 Periscope automatic performance analysis tool [Oleynik, TUM]
  • Periscope hands-on tutorial exercises (instructions)
  • 12:30 (lunch)
    13:30 Hands-on coaching to apply tools to analyze participants' own code(s).
    17:00 Review of day and schedule for remainder of workshop
    17:30 (adjourn)

    Day 3 Wednesday 7 Sep
    09:00 TAU performance system [Shende, UOregon]
  • TAU hands-on tutorial exercises
  • 10:30 (break)
    11:00 KCachegrind toolset [Weidendorfer, TUM]
  • KCachegrind hands-on tutorial exercises
  • 12:30 (lunch)
    13:30 Hands-on coaching to apply tools to analyze participants' own code(s).
    17:00 Review of day and schedule for remainder of workshop
    17:30 (adjourn)

    Day 4 Thursday 8 Sep
    09:00 Vampir trace analysis toolset [Hilbrich, TUD-ZIH]
  • Vampir hands-on tutorial exercises
  • 10:30 (break)
    11:00 Paraver trace analysis toolset [Labarta/Gimenez, BSC]
  • Paraver hands-on tutorial exercises
  • 12:30 (lunch)
    13:30 Hands-on coaching to apply tools to analyze participants' own code(s).
    17:00 Review of day and schedule for remainder of workshop
    17:30 (adjourn)

    Day 5 Friday 9 Sep
    09:00 MUST/Marmot correctness checking tools [Hilbrich/Protze, TUD-ZIH]
  • MUST/Marmot hands-on tutorial exercises
  • 10:30 (break)
    11:00 VI-HPS libraries
  • Hardware counter analysis with PAPI [Moore, UTK-ICL]
  • SIONlib scalable native parallel file I/O [Frings, JSC]
  • 12:30 (lunch)
    13:30 Hands-on coaching to apply tools to analyze participants' own code(s).
    15:00 (break)
    15:30 (adjourn or continue with work to 16:30)

     

    Classroom capacity is limited, therefore priority will be given to applicants with parallel codes already running on the workshop computer systems, and those bringing codes from similar systems to work on. Participants are therefore encouraged to prepare their own MPI, OpenMP and hybrid OpenMP/MPI parallel application codes for analysis.

    VI-HPS Tools

     

    • KCachegrind is a free cache-utilization visualization tool developed by TUM.
    • MUST & Marmot are free correctness checking tools for MPI programs developed by TUD-ZIH and partners.
    • PAPI is a free library interfacing to hardware performance counters developed by UTK-ICL, used by many tools.
    • Periscope is an automatic performance analysis tool using a distributed online search for performance bottlenecks being developed by TUM.
    • Scalasca is an open-source toolset developed by JSC & GRS that can be used to analyze the performance behaviour of MPI & OpenMP parallel applications and automatically identify inefficiencies.
    • Vampir is a commercial framework and graphical analysis tool developed by TUD-ZIH to display and analyze trace files, such as those produced by the open-source VampirTrace library.
    • TAU is a performance system for measurement and analysis of parallel programs written in Fortran, C, C++, Java & Python developed by the University of Oregon.

     

    Hardware and Software Platforms

    The local systems are expected to be the primary platforms for the workshop, with priority for improved job turnaround and local system support. Course accounts will be provided for those who need them.

    • Sun/Bull Nehalem cluster (JSC Juropa / HPC-FF): Intel Xeon X5570 quad-core processors, SLES Linux, ParaStation MPI, Intel compilers
    • Sun Nehalem cluster (RWTH): Nehalem 8-core & Westmere 6-core processors, Scientific Linux, Intel MPI, Intel compilers
    • IBM BlueGene/P (JSC Jugene): PowerPC 450 quad-core processors, BG-Linux compute kernel, IBM BG-MPI library, IBM BG-XL compilers

     

    The VI-HPS tools support and are also installed on a range of HPC platforms, including:

     

    Other systems where up-to-date versions of the tools are installed can also be used when preferred, though support may be limited. Participants are expected to already possess user accounts on non-local systems they intend to use, and should be familiar with the procedures for compiling and running parallel applications on the systems.

    Contact

    Brian Wylie (JSC), phone +49 2461 61-6589
    Marc-Andre Hermanns (GRS), phone +49 241 80-99753