Annotation
Applications need to be effectively parallelised, using techniques such as MPI message-passing and OpenMP multithreading, to fully exploit the capabilities of modern systems for high-performance computing. The course introduces the open-source community-developed Score-P instrumentation and measurement infrastructure, and the Scalasca and Vampir tools, to provide a practical basis for portable performance analysis of parallel applications.
The course will be delivered as a series of presentations with associated hands-on practical exercises using the Salomon computer cluster. It starts with basic application instrumentation and measurement to generate execution profiles, then improves measurement quality via customization capabilities, and progresses to interactive and automated analyses of execution traces.
While analysis of provided example codes will be used to guide the class through the relevant steps and familiarise with usage of the tools, coaching will also be available to assist participants to analyse their own parallel application codes and may suggest opportunities for improving their execution performance and scalability.
This is especially true for users interested in a deeper analysis and optimization of their codes, which will be offered on 14-15 December 2016 during a Performance Optimization and Productivity Tutorial organized at IT4Innovations jointly with POP (Performance Optimisation and Productivity), a Centre of Exellence in Computing Applications. Programmers, developers, consult the POP web pages for details and do not miss this unique opportunity.
Purpose of the course (benefits for the attendees)
On completion participants should be familiar with common performance analysis techniques and how they can be employed in practice (on a range of HPC systems). Those who prepared their own application test cases will have been coached in the tuning of their measurement and analysis, and provided optimization suggestions.
Level
Language
English
About the tutor(s)
Brian Wylie is a research scientist at Jülich Supercomputing Centre of Forschungszentrum Jülich since 2004, where he is a developer of the Scalasca toolset for scalable performance analysis of large-scale parallel applications. Additional responsibilities for application engagement and training include the series of VI-HPS Tuning Workshops and hands-on tutorials. He has a further 13 years of experience researching and developing parallel performance tools, and as applications scientist working with a spectrum of academic and commercial applications. His 1990 PhD in computational physics was awarded by the University of Edinburgh.
Ronny Brendel received his Diploma in Computer Science at TU Dresden where he now works as a researcher. His work revolves around Software Performance Analysis. For over eight years Ronny has been researching and developing tools for software performance analysis, e.g., Vampir, Score-P, OTF and VampirTrace. Over the years he contributed to many levels of the performance analysis tool stack jointly developed by TU Dresden, Jülich Supercomputing Centre, RWTH Aachen, TU Munich, and TU Darmstadt.
