Goethe-HLR Cluster and Batch Usage
- Anja Gerbes
The Goethe-HLR is a general purpose HPC cluster based on Intel Xeon Skylake CPU architecture. The system was installed in “Industriepark Höchst” in 2019. At the 484 compute server are more than 19 000 cores provided. The nodes are connected via EDR-Infiniband. At each compute node are at least 192 GB RAM available. The measured HPL performance is 958.67 TFlop/s. Slurm is the chosen Batch System (Workload Manager) that is used on Goethe-HLR. The talk will give you an introduction how to write batch script to run them later on a cluster.
Managing HPC Software Complexity with Spack
Spack is an open-source tool for HPC package management, used on some of the fastest supercomputers in the world. It allows developers to write simple recipes in Python to automate builds with arbitrary combinations of compilers, MPI versions, build options, and dependency libraries. With Spack, users can install over 3000 community-maintained packages without knowing how to build them, developers can automate builds of tens or hundreds of dependency libraries, and staff at HPC centers can deploy many versions of software for thousands of users. We provide a thorough introduction to Spack’s capabilities: basic software installation, creating new packages, and development workflows using Spack. Attendees are encouraged to bring a laptop with Docker installed if they wish to follow along with the live demonstrations.
Gregory Becker is a computer scientist at Lawrence Livermore National Laboratory. His focus is on bridging the gap between research code and production use. He is one of the lead developers of the Spack project.
OpenACC Support for Clang and LLVM
- Joel E. Denny
This talk will present the Clacc project, which is developing OpenACC support for Clang and LLVM. It will address the following questions. What is OpenACC? Why is it important for HPC? How does it compare to OpenMP? What is the current state of OpenACC support? What is Clacc? How is it designed? What is the current state of Clacc development? How do I use it?
Joel E. Denny is a Computer Scientist in the Future Technologies Group at Oak Ridge National Laboratory. His research currently focuses on LLVM-based compiler projects. He received his Ph.D. in Computer Science from Clemson University in 2010.
Performance Evaluation using the TAU Performance System
- Sameer Shende
The TAU Performance System provides HPC and AI researchers a powerful profiling and tracing toolkit. This tutorial will focus on performance data collection, analysis, and performance optimization of parallel applications. The tutorial will introduce profiling and debugging support in TAU, cover performance evaluation of parallel programs written in Python, Fortran, C++, and/or C. The tutorial will also cover parallel performance analysis of applications using MPI, OpenMP, and other parallel runtime environments. We will demonstrate different techniques for program instrumentation including compiler-based instrumentation for LLVM and highlight TAU's support for memory debugging and I/O evaluation. The tutorial will guide the developers through the instrumentation, measurement, and analysis process steps in TAU. Performance data will include MPI timings, I/O and memory, and hardware performance counters from PAPI.
Sameer Shende serves as the Director of the Performance Research Laboratory at the University of Oregon. His research interests include performance evaluation tools, runtime systems, instrumentation, measurement, and analysis tools, and optimizing compilers.
|Bachelor / Master Students from Germany||40,00|
|PhD students from Germany||40,00|
|Members of German universities and public research institutes||40,00|
|Others (non academic)||40,00|
The registration fee includes:
- snacks and beverages during the coffee breaks (without lunch)
Lunch is available in various locations in the vicinity of Westend campus (self-paid).