BEGIN:VCALENDAR VERSION:2.0 PRODID:icalendar-ruby CALSCALE:GREGORIAN BEGIN:VEVENT DTSTAMP:20260706T052027Z UID:c1ae8426-da7a-4a74-b737-5fdea7a5c028 DTSTART:20220627T080000Z DTEND:20220629T140000Z DESCRIPTION:Registration\n\nPlease register with your official e-mail addre ss to prove your affiliation.\n\nContents\n\nIn the ever-growing complexit y of computer architectures\, code optimisation has become the main route to keep pace with hardware advancements and effectively make use of curren t and upcoming High Performance Computing systems.\n\nHave you ever asked yourself:\n\n\n Where are the performance bottlenecks of my application?\n What is the maximum speed-up achievable on the architecture I am using?\n Does my code scale well across multiple machines?\n Does my implementatio n match my HPC objectives?\n\n\nIn this workshop\, we will discuss these q uestions and provide a unique opportunity to learn techniques\, methods an d solutions on how to improve code\, how to enable the new hardware featur es and how to use visualise the potential benefits of an optimisation proc ess.\n\nWe will describe the latest micro-processor architectures and how developers can efficiently use modern HPC hardware\, including SIMD vector units and the memory hierarchy. We will also touch upon exploiting intra- node and inter-node parallelism.\n\nAttendees will be guided along the opt imisation process through the incremental improvement of an example applic ation. Through hands-on exercises they will learn how to enable vectorisat ion using simple pragmas and more effective techniques like changing data layout and alignment.\n\nThe work is guided by hints from compiler reports \, and profiling tools such as Intel® Advisor\, Intel® VTune™ Amplifie r\, Intel® Application Performance Snapshot and LIKWID for investigating and improving the performance of an HPC application.\n\nYou can ask the le cturers in the Q&\;A session about how to optimise your code. Please pr ovide a description of your code in the registration form.\n\nLearning Goa ls\n\nThrough a sequence of simple\, guided examples of code modernisation \, the attendees will develop awareness on features of multi and many-core architecture which are crucial for writing modern\, portable and efficien t applications.\n\nA special focus will be dedicated to scalar and vector optimisations for the Intel® Xeon® Scalable processor\, code-named Skyla ke\, utilised in the SuperMUC-NG machine at LRZ. \n\nThe workshop interlea ves lecture and practical sessions. \n\nPreliminary Agenda\n\n\n\n \n \n  \n \n Session \n \n \n \n \n 1st day morning\n (10:00-12 :00)\n \n \n \n Intro (Volker Weinberg)\n Intro to LRZ HPC Syste ms and Software Stack (Gerald Mathias)\n Principles of optimization (Jon athan Coles)\n \n \n \n \n \n 1st day afternoon\n (13:00-16:00 )\n \n \n HPC Architecture\, Vectorization\n\n Example code\n\n Data structures\n\n (Jonathan Coles)\n \n \n \n \n 2nd day morni ng\n (10:00-12:00)\n \n \n \n Profiling:\n\n Code instrumentat ion\, Roofline Model\, Intel Advisor (Jonathan Coles)\n \n \n \n \n \n 2nd day afternoon\n (13:00-16:00)\n \n \n Scaling Code + MP I Tuning (Jonathan Coles)\n\n Additional tools: valgrind and cache simul ators. (Josef Weidendorfer)\n\n Debuggers (TBD)\n\n I/O Considerations (Patrick Böhl)\n \n \n \n \n 3rd day morning\n (10:00-12:00)\n \n \n \n LikWid (Carla Guillen/Thomas Gruber)\n HPC report (Car la Guillen)\n \n \n \n \n \n 3rd day afternoon\n (13:00-16:00) \n \n \n \n Optimisation highlights by LRZ (CXS Group LRZ)\n \n \n \n \n 3rd day Q&\;A\n All\n \n \n\n\n\n \n\nThe workshop i s a PRACE training event organised by LRZ in cooperation with NHR@FAU .\n\ nAbout the Lecturers\n\nPatrick Böhl works in the HPC group at LRZ. He ob tained his PhD in theoretical physics at LMU Munich where he studied nonli near partial differential equations numerically and analytically. In paral lel he was involved in optimizing an advanced Particle-In-Cell Code which was used for conducting huge three-dimensional plasma simulations on Super MUC(-NG). Patrick joined LRZ in Feb 2020 with his main focus working on I/ O related problems.\n\nJonathan Coles works in the AstroLab of the Computa tional X Support (CXS) group at LRZ. He holds a B.Sc. and M.Sc. in Compute r Science from RIT in the United States and a PhD in Computational Astroph ysics from the University of Zurich. After completing a Postdoc position i n Zurich\, Jonathan moved to Paris to work at the University of Versailles in close collaboration with the CEA and Intel Corp. There he developed a highly parallel and distributed implementation of the Fast Multipole Metho d using OpenMP and MPI for molecular dynamics applications. In 2016\, he m oved to Munich to work at the TUM in the Biophysics department where he co ntinued to develop the FMM code on SuperMUC. Since September 2021\, Jonath an is employed at LRZ.\n\nThomas Gruber (né Röhl) works in the HPC grou p of NHR@FAU. He leads the development of the performance tool suite LIKWI D\, which comprises easy-to-use tools for hardware performance monitoring\ , affinity control and micro-benchmarking. He also works on projects invol ving monitoring and analysis of hardware performance data.\n\nCarla Guille n works as a researcher in the application support group at the LRZ. She o btained her PhD in computer science at the Technische Universitaet Muenche n in 2015. She joined the LRZ in 2009\, and has been working in the fields of system-wide performance monitoring and energy optimisation of large sc ale clusters.\n\nGerald Mathias leads the Computational X Support (CXS) gr oup at LRZ. After his PhD in Theoretical Biopyhsics he joined the chair of Theoretical Chemistry at the RUB in Bochum as a postdoc\, followed by a h abilition at LMU Munich. He is experienced in the development and optimisa tion of highly parallel ab initio and force field based molecular dynamics codes\, both in Fortran and C.\n\nJosef Weidendorfer leads the Future Com puting Group at LRZ\, which is developing smooth migration strategies for future HPC systems and evaluating novel technologies. This includes improv ement of system level and workload analysis tools as well as novel paralle l programming models. Previous research involved best use of accelerators\ , heterogeneous computing\, and tuning strategies for parallel code includ ing dynamic code generation techniques. Josef completed his habilitation a t TUM in 2016 on simulation-driven performance analysis for parallel code\ , especially looking at capturing bottlenecks in the memory hierarchy of m odern architectures and presenting them in a way to hint at adequate perfo rmance optimizations. He received his Ph.D. from TUM in 2003 for studying load balancing issues in car crash simulation on industrial code at BMW AG .\n\nPrerequisites\n\nAttendees should be comfortable with either C/C++ or Fortran programming language and basic Linux command\, like make and ssh. No previous experience in vectorisation\, parallelisation and profiling t ools is required.\n\nHands-On\n\n\n Exercises will be done on the CooLMUC2 Cluster @ LRZ with 28-way Haswell-based nodes and FDR14 Infiniband interc onnect\n Please use your own laptop or PC with X11 support and an ssh clie nt installed for the hands-on sessions.\n \n \n Under Windows\n\n \n We recommend to install the comfortable tool MobaXterm (https://mobaxterm. mobatek.net/download-home-edition.html) which also includes an X11 client. \n Alternatively install and run the Xming X11 Server for Windows: https ://sourceforge.net/projects/xming/ and then install and run the terminal s oftware putty: https://www.chiark.greenend.org.uk/~sgtatham/putty/latest.h tml \n \n \n Under macOS\n \n Install X11 support for macOS XQuartz :  https://www.xquartz.org/\n \n \n Under Linux\n \n ssh and X11 su pport comes with all distributions\n \n \n \n \n\n\nLanguage\n\nEnglish\ n\nLecturers\n\nDr. Patrick Böhl\, Dr. Jonathan Coles\, Dr. Gerald Mathia s \, Dr. Carla Guillen\, Dr. Josef Weidendorfer (LRZ)\n\nThomas Gruber (NH R@FAU)\n\nPrices and Eligibility\n\nThe course is open and free of charge for people from academia and industry from the Member States (MS) of the E uropean Union (EU) and Associated Countries to the Horizon 2020 programme. \n\nWithdrawal Policy\n\nSee Withdrawal\nhttps://events.prace-ri.eu/event/ 1394/ SUMMARY:[ONLINE] HPC Code Optimisation Workshop URL;VALUE=URI:https://events.prace-ri.eu/event/1394/ END:VEVENT END:VCALENDAR