Node-Level Performance Engineering @ LRZ
Topic outline
-
This course covers performance engineering approaches on the compute node level. Even application developers who are fluent in OpenMP and MPI often lack a good grasp of how much performance could at best be achieved by their code.
This is because parallelism takes us only half the way to good performance.
Even worse, slow serial code tends to scale very well, hiding the fact that resources are wasted. This course conveys the required knowledge to develop a thorough understanding of the interactions between software and hardware. This process must start at the core, socket, and node level, where the code gets executed that does the actual computational work. We introduce the basic architectural features and bottlenecks of modern processors and compute nodes.
Pipelining, SIMD, superscalarity, caches, memory interfaces, ccNUMA, etc., are covered. A cornerstone of node-level performance analysis is the Roofline model, which is introduced in due detail and applied to various examples from computational science. We also show how simple software tools can be used to acquire knowledge about the system, run code in a reproducible way, and validate hypotheses about resource consumption. Finally, once the architectural requirements of a code are understood and correlated with performance measurements, the potential benefit of code changes can often be predicted, replacing hope-for-the-best optimizations by a scientific process. Hands-on exercises provide an opportunity for participants to apply the concepts right away.
Lecturers: Dr. Georg Hager and Prof. Dr. Gerhard Wellein
Course date: December 4-6, 2023 (9:00 am - 4:30 pm)
This course will be conducted online as a Zoom event. Details have been sent vie e-mail to registered participants.
Course program:- Introduction
- Computer architecture for software developers
- Tools topology & affinity in multicore environments
- Roofline model: basics
- Tools: hardware performance counters
- Roofline case studies
- Optimal use of parallel resources: SIMD and ccNUMA
- Performance Engineering basic skills
- Extending Roofline: The ECM performance model
-
-
Excel 2007 spreadsheet
-
Our HPC book: Hager & Wellein: Introduction to High Performance Computing for Scientists and Engineers
Important links:
IntelĀ® 64 and IA-32 Architectures Optimization Reference Manual
LIKWID tool suite: https://github.com/RRZE-HPC/likwid
LIKWID documentation Wiki: http://tiny.cc/LIKWID
Kerncraft automatic Roofline/ECM modeling tool: https://github.com/RRZE-HPC/kerncraft
Open-Source Architecture Code Analyzer: https://github.com/RRZE-HPC/OSACA
Online layer condition calculator: https://rrze-hpc.github.io/layer-condition/#calculator
Intel page on "Gather Data Sampling" (Downfall): https://www.intel.com/content/www/us/en/developer/articles/technical/software-security-guidance/technical-documentation/gather-data-sampling.html