Weekly outline

  • General

    The seminar covers optimization and parallelization techniques for modern multi- and manycore systems. The topics are chosen from interesting contemporary problems in High Performance Computing on modern hardware like multicore processors, accelerators (e.g., GPGPUs or Xeon Phi), and clusters. 

    Lecturer: Prof. G. Wellein (gerhard.wellein@fau.de), Martensstr. 1, Room 01.131. Phone -28136
    Location: 2.037 (e-Studio), Martensstr. 1 (RRZE), 2nd floor
    Time: Monday 16:00-17:30

    First seminar: Monday, October 23

    Either 2.5 or 5 ECTS credits will be granted, depending on whether the student gives one or two talks. In either case, a written seminar report is mandatory.

    Possible topics can be found in the intro talk (see below).


  • 16 October - 22 October

    No seminar this week

    • 23 October - 29 October

      Monday seminar: Christopher Bross: Relaxed Thread Synchronization (part 1)

      Please note: Next week's seminar was preponed to this Friday, October 27, 10 am in Room 2.049 (big seminar room at RRZE).

      Friday seminar: Kushal Srivastava and Nivesh Dommaraju: Performance Analysis of the Himeno Benchmark (part 2)

    • 30 October - 5 November

      The seminar is preponed to Friday, October 27, 10 am.

      • 6 November - 12 November

        Thomas Röhl (HPC group, RRZE): The LIKWID multicore tool suite

      • 13 November - 19 November

        Jan Eitzinger: Crash Course Performance Engineering

      • 20 November - 26 November

        G. Hager: The ECM performance model

      • 27 November - 3 December

        Julian Hammer: Recap of the ECM model and introduction to the kerncraft Roofline/ECM modeling tool

      • 4 December - 10 December

        This week's seminar was shifted to Thursday, December 7, 14:15-15:45 in Room 2.049 (big seminar room at RRZE).

        Jan Laukemann: Bachelor thesis final talk: Design and Implementation of a Framework for the Prediction of Instruction Throughput

        • 11 December - 17 December

          Q&A session! You advisor(s) will be present to discuss the current state of your seminar project.

          • 18 December - 24 December

            Today's seminar is canceled.

            Instead, we suggest you visit the public part of the PhD defense of our student Moritz Kreutzer, titled Performance Engineering for Exascale-Enabled Sparse Linear Algebra Building Blocks.

            When: Monday, December 18, 2017, 14:00 s.t.

            Where: RRZE Room 2.049 (large seminar room)

            If you want to attend, please don't be late.

            • 8 January - 14 January

              Talks by RRZE HPC members about recently submitted papers: G. Wellein, G. Hager, M. Witmann

              • 15 January - 21 January

                Q&A session

                • 22 January - 28 January

                  Markus Wittmann: Report on fw/bw substitution in the PARDISO solver

                  • 29 January - 4 February

                    January 29: Thomas Köster, initial master thesis talk: Porting Physical Parameterizations from a Climate Model to Accelerators. Slides 

                    February 1, 4:15 pm, Room 2.009: First seminar talk by Lukas Liebischer and Chaitanya Dev about ECM model on Skylake

                    • 5 February - 11 February

                      Invited talk by Prof. Martin Schulz, TU München: A Case for More Adaptivity in HPC.

                      Abstract: Current HPC environments and applications are rather rigid and inflexible: applications run only with fixed numbers of processors, scheduling is done without considering the context of other jobs and their impact, network and I/O congestion are often only accepted as a matter of fact, and power is in most cases still fully provisioned to avoid the need for any adaptivity. This leads to inefficient usage of HPC systems with resources fragmentation, high variability and suboptimal performance. In this talk, I will discuss several examples of such inefficient usage of resources and will show how adaptivity has to be an important part to help solve this challenge. In particular, I will focus on efficient power usage both on the application and system level, as well as approaches to support malleable applications using new programming abstractions. These efforts provide first steps towards more adaptive systems, which will enable us to exhaust HPC systems to their full capabilities.