PRACE Spring School 2014

Europe/Vienna
Castle of Hagenberg, Hagenberg/Linz, Austria

Castle of Hagenberg, Hagenberg/Linz, Austria

Kirchenplatz 5b 4232 Hagenberg im Mühlkreis Austria
Michael Krieger (JKU Linz) , Ondrej Jakl (VSB-TUO) , Wolfgang Schreiner (Johannes Kepler University Linz)
Description
Software Engineering for Supercomputers in Research and Industry

(Poster for the PRACE Spring School 2014 can be found here)

The PRACE Spring School 2014 will take place on 15 – 17 April 2014 at the Castle of Hagenberg in Austria. The PRACE Seasonal School event is being hosted and organised jointly by the Research Institute for Symbolic Computation / Johannes Kepler University Linz (Austria), IT4Innovations / VSB-Technical University of Ostrava (Czech Republic) and PRACE.

The three day program includes:
  • a 1-day HPC usage for Industry track bringing together researchers and attendees from industry and academia to discuss the variety of applications of HPC in Europe.
  • two 2-day tracks on software engineering for parallel & emerging computing architectures and deep insight into solving multiphysical problems with Elmer on large-scale HPC resources with lecturers from industry and PRACE members.
The PRACE 2014 Spring School programme offers a unique opportunity to bring users, developers and industry together to learn more about efficient software development for HPC research infrastructures.

The program is free of charge (not including travel and accommodations). For the hands-on sessions, participants are expected to bring their own laptops.

Applications are open to researchers, academics and industrial researchers residing in PRACE member countries, and European Union Member States and Associated Countries. All lectures and training sessions will be in English.
Castle of Hagenberg
Hotel Sommerhaus
Travel Information
    • 09:15 10:30
      Opening Seminarroom (Castle of Hagenberg)

      Seminarroom

      Castle of Hagenberg

      • 09:15
        Welcome & Opening Speech 15m
      • 09:30
        The XSEDE Global Federated File System (GFFS/GFFS.eu) - Breaking Down Barriers to Secure Resource Sharing 1h
        The GFFS offers scientists a simplified means through which they can interact with and share resources. Currently, many scientists struggle to exploit distributed infrastructures because they are complex, unreliable, and require the use of unfamiliar tools. For many scientists, such obstacles interfere with their research; for others, these obstacles render their research impossible. It is therefore essential to lower the barriers to using distributed infrastructures. The first principle of the GFFS is simplicity. Every researcher is familiar with the directory-based paradigm of interaction; the GFFS exploits this familiarity by providing a global shared namespace. The namespace appears to the user as files and directories so that the scientist can easily organize and interact with a variety of resources types. Resources can include compute clusters, running jobs, directory trees in local file systems, groups, as well as storage resources at geographically dispersed locations. Once mapped into the shared namespace, resources can be accessed by existing applications in a location-transparent fashion, i.e., as if they were local. In this talk I will present the GFFS, its functionality, its motivation, as well as typical use cases. I will demonstrate many of its capabilities, including: 1) how to securely share data, storage, and computational resources with collaborators; 2) how to access data at the centers from campus and vice versa; 3) how to create shared compute queues with collaborators; and 4) how to create jobs and interact with them once started.
        Speaker: Prof. Andrew Grimshaw (University of Virginia)
        Slides
    • 10:30 10:45
      Coffee Break 15m
    • 10:45 12:15
      HPC for industry

      This workshop collects input from different industrial areas which are heavily relying on HPC and their expectations on researchers and future developments in HPC.

      • 10:45
        Simulating Brain Aneurysms with HPC 45m
        Speaker: Dr Wolfgang Fenz (RISC Software GmbH)
        Slides
    • 13:45 15:30
      HPC for industry

      This workshop collects input from different industrial areas which are heavily relying on HPC and their expectations on researchers and future developments in HPC.

      • 13:45
        HPC for Remote Sensing Applications 45m
        Speaker: Mr Wolfgang Planer (Catalysts GmbH)
        Slides
    • 15:30 15:45
      Coffee Break 15m
    • 15:45 16:45
      Panel
    • 09:30 10:30
      Opening
      • 09:30
        OpenSPL - The Open Spatial Programming Language 1h
        Speaker: Dr Oskar Mencer (Maxeler Technologies Inc)
        Slides
    • 10:30 10:45
      Coffee Break 15m
    • 10:45 12:15
      Elmer - Finite Element Software for Multiphysical Problems Seminarroom (Castle of Hagenberg)

      Seminarroom

      Castle of Hagenberg

      Elmer is an open source multiphysical simulation software mainly developed by CSC - IT Center for Science (CSC). Elmer development was started 1995 in collaboration with Finnish Universities, research institutes and industry. After it's open source publication in 2005, the use and development of Elmer has become international. This track provides an introduction to and advanced training for Elmer.

      • 10:45
        Introduction to Elmer finite element software 1h 30m
        An overview of the structure and capabilities of Elmer software is given. Also the graphical user interface ElmerGUI is demonstrated setting up a simple case.
        Speaker: Dr Peter Råback (CSC - IT Center for Science)
        Slides
    • 10:45 12:15
      Software Engineering for Parallel & Emerging Computing Architectures

      This track focuses on software engineering techniques for HPC environments. Apart from parallel programming techniques also accelerator cards and profiling techniques and tools are in the focus of this track. Attendees will acquire deep insight into efficiently programming and developing optimized software for x86-CPUs, Intel Xeon Phi and GPGPUs with OpenMP, OpenCL and CUDA.

      • 10:45
        Modern CPU architectures 1h 30m
        This session gives at first the motivation for parallel computing. After an introduction to modern CPU architecture deeper insight into memory hierarchies is provided. Based on the roofline model participants will learn the principles for optimizing parallel applications (compute bound vs. memory bound).
        Speaker: Mr Alexander Leutgeb (RISC Software GmbH)
        Slides
    • 13:30 15:00
      Elmer - Finite Element Software for Multiphysical Problems

      Elmer is an open source multiphysical simulation software mainly developed by CSC - IT Center for Science (CSC). Elmer development was started 1995 in collaboration with Finnish Universities, research institutes and industry. After it's open source publication in 2005, the use and development of Elmer has become international. This track provides an introduction to and advanced training for Elmer.

      • 13:30
        Hands-on session using ElmerGUI 1h 30m
        The participants will go through simple exercices using ElmerGUI. Some of the cases can be chosen individually to best suite the profile of the participants.
    • 13:30 15:00
      Software Engineering for Parallel & Emerging Computing Architectures

      This track focuses on software engineering techniques for HPC environments. Apart from parallel programming techniques also accelerator cards and profiling techniques and tools are in the focus of this track. Attendees will acquire deep insight into efficiently programming and developing optimized software for x86-CPUs, Intel Xeon Phi and GPGPUs with OpenMP, OpenCL and CUDA.

      • 13:30
        Introduction to Vectorization 1h 30m
        This session gives an introduction into vetorization on Intel x86 CPUs. The participants will learn how to make use of the vector units via the auto vectorization support of the C/C++ compiler and about the requirements for a successful auto vectorization. Portability is also an issue if we want to use the latest vector unit capabilites of Intel x86 CPUs. Finally other ways of using the vector units complete the session. Using practical examples, the acquired knowledge is deepened.
        Speaker: Mr Alexander Leutgeb (RISC Software GmbH)
        Slides
    • 15:00 15:15
      Coffee Break 15m
    • 15:15 16:45
      OpenLab
    • 15:15 16:45
      Software Engineering for Parallel & Emerging Computing Architectures Seminarroom (Castle of Hagenberg)

      Seminarroom

      Castle of Hagenberg

      This track focuses on software engineering techniques for HPC environments. Apart from parallel programming techniques also accelerator cards and profiling techniques and tools are in the focus of this track. Attendees will acquire deep insight into efficiently programming and developing optimized software for x86-CPUs, Intel Xeon Phi and GPGPUs with OpenMP, OpenCL and CUDA.

      • 15:15
        Programming fo the Intel Xeon Phi 1h 30m
        This session starts with an introduction to the Intel Xeon Phi architecture. The different memory models and execution models will be presented. The attendees will apply their acquired knowledge on efficiently using this architecture to practical examples and gain first experiences with Xeon Phi.
        Speaker: Mr Michael Hava (RISC Software GmbH)
        Slides
    • 17:30 22:30
      Social Event Ars Electroncia Center Linz

      Ars Electroncia Center Linz

      • 17:30
        Reception 30m Foyer (Ars Electronica Center Linz)

        Foyer

        Ars Electronica Center Linz

      • 18:00
        Guided Tour throught the AEC 1h Ars Electronica Center Linz

        Ars Electronica Center Linz

      • 19:00
        Dinner 3h 30m Skyloft (Ars Electronica Center Linz)

        Skyloft

        Ars Electronica Center Linz

    • 09:30 10:30
      Opening
      • 09:30
        About the importance of HPC for Life Sciences 1h
        Speaker: Prof. Oswaldo Trelles (University of Malaga)
        Slides
    • 10:30 10:45
      Coffee Break 15m
    • 10:45 12:15
      Elmer - Finite Element Software for Multiphysical Problems Seminarroom (Castle of Hagenberg)

      Seminarroom

      Castle of Hagenberg

      Elmer is an open source multiphysical simulation software mainly developed by CSC - IT Center for Science (CSC). Elmer development was started 1995 in collaboration with Finnish Universities, research institutes and industry. After it's open source publication in 2005, the use and development of Elmer has become international. This track provides an introduction to and advanced training for Elmer.

      • 10:45
        Advanced use of Elmer 1h 30m
        In this session the more advanced operation of Elmer is presented. The structure of the Elmer simulation, MATC language, and UDFs is discussed and the command line operation of the software is presented. Also parallel operation is demonstrated. Operation with other pre- and postprocessing tools is shortly discussed.
    • 10:45 12:15
      Software Engineering for Parallel & Emerging Computing Architectures

      This track focuses on software engineering techniques for HPC environments. Apart from parallel programming techniques also accelerator cards and profiling techniques and tools are in the focus of this track. Attendees will acquire deep insight into efficiently programming and developing optimized software for x86-CPUs, Intel Xeon Phi and GPGPUs with OpenMP, OpenCL and CUDA.

      • 10:45
        Profiling techniques for parallel applications 1h 30m
        This session gives an introduction to effective strategies for analyzing program performance. The focus will lie on profiling tools utilizing event-based sampling and hardware counters (such as HPCToolkit, Amplifier, etc.). After a short overview over useful metrics for different problems, practical examples will provide attendants with knowlegde for improving their applications.
        Speaker: Mr Thomas Ponweiser (RISC / Johannes Kepler University Linz)
        Slides
    • 13:30 14:30
      Elmer - Finite Element Software for Multiphysical Problems Seminarroom (Castle of Hagenberg)

      Seminarroom

      Castle of Hagenberg

      Elmer is an open source multiphysical simulation software mainly developed by CSC - IT Center for Science (CSC). Elmer development was started 1995 in collaboration with Finnish Universities, research institutes and industry. After it's open source publication in 2005, the use and development of Elmer has become international. This track provides an introduction to and advanced training for Elmer.

      • 13:30
        Hands-on session with advanced features and parallel computation 1h
        The participants can continue to study the cases using the command line interface. Simple modifications and parallel operation can be tested. Finally, users can get some guidance how their own problems could be solved using Elmer finite element software.
    • 13:30 14:30
      Software Engineering for Parallel & Emerging Computing Architectures

      This track focuses on software engineering techniques for HPC environments. Apart from parallel programming techniques also accelerator cards and profiling techniques and tools are in the focus of this track. Attendees will acquire deep insight into efficiently programming and developing optimized software for x86-CPUs, Intel Xeon Phi and GPGPUs with OpenMP, OpenCL and CUDA.

      • 13:30
        Introduction to OpenCL Programming in Heterogeneous Architectures 1h
        Speaker: Mr Bernhard Gruber (RISC Software GmbH)
        Slides
    • 14:30 14:45
      Coffee Break 15m
    • 14:45 15:45
      Elmer - Finite Element Software for Multiphysical Problems

      Elmer is an open source multiphysical simulation software mainly developed by CSC - IT Center for Science (CSC). Elmer development was started 1995 in collaboration with Finnish Universities, research institutes and industry. After it's open source publication in 2005, the use and development of Elmer has become international. This track provides an introduction to and advanced training for Elmer.

    • 14:45 15:45
      Software Engineering for Parallel & Emerging Computing Architectures

      This track focuses on software engineering techniques for HPC environments. Apart from parallel programming techniques also accelerator cards and profiling techniques and tools are in the focus of this track. Attendees will acquire deep insight into efficiently programming and developing optimized software for x86-CPUs, Intel Xeon Phi and GPGPUs with OpenMP, OpenCL and CUDA.

      • 14:45
        Introduction to GPGPU Programming with CUDA 1h
        Speaker: Mr Torsten Welsch (RISC Software GmbH)
    • 15:45 16:15
      Closing