Lab Home | Phone | Search | ||||||||
|
||||||||
LIKWID (Like I Knew What I'm Doing) is a toolsuite for performance-oriented users and developers maintained since 2009. It provides tools and a library for system topology lookup, affinity control, micro-benchmarking, system feature manipulation and profiling through hardware performance counters. It supports x86, ARM and POWER CPUs as well as accelerators from Nvidia and AMD. In this talk, the main developer of LIKWID will give an overview over some of the tools and how they can be used to run and analyze own codes on CPUs and GPUs. The talk showcases how you can derive the empirical Roofline model for your system and application. Speaker Bio: During his apprenticeship at the Erlangen Regional Computing Center (RRZE), the IT service provider for the Friedrich-Alexander-Universitat Erlangen-Nurnberg (FAU), Thomas Gruber collected experience with all kinds of clustering approaches. Afterwards, he studied Computer Science at RWTH Aachen University with emphasis on parallel programming and operating system kernel development. At the same time, he worked as a research assistant for the HPC group of the RWTH IT center. After receiving his M. Sc. degree, he went back to RRZE to work for the HPC group. Since 2021, he works in the Software & Tools division at Erlangen National High Performance Computing Center (NHR@FAU). Thomas Gruber leads the development of the performance tool suite LIKWID, which comprises easy-to-use tools for hardware performance monitoring, affinity control and micro-benchmarking. He also works on projects involving monitoring and analysis of hardware performance data. He is actively involved in tutorials and lectures about performance analysis and optimization, where he presents the LIKWID tools and leads hands-on exercises. Point of contact: Anna Matsekh (matsekh@lanl.gov). Seminar details: The Institutional Computing Science Seminar Series features scientific research that employs Institutional Computing HPC systems at LANL Host: Jointly hosted by CNLS and the Institutional Computing Applications Team |