Node-Level Performance Engineering


This online course organised in cooperation with NHR@FAU 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.

  • Introduction
    • Basic architecture of multicore systems: pipelines, SIMD, caches, sockets, memory
    • The important role of system topology
  • Tools: topology & affinity in multicore environments
    • likwid-topology and likwid-pin, alternatives
  • Roofline model basics         
    • Model assumptions and construction
    • Simple examples
    • Limitations of the Roofline model
  • Tools: hardware performance counters
    • Why hardware performance counters?
    • likwid-perfctr
    • Validating performance models
  • Optimal use of parallel resources
    • Single Instruction Multiple Data (SIMD)
    • Cache-coherent Non-Uniform Memory Architecture (ccNUMA)
  • Roofline case studies
    • Tall & skinny dense matrix-matrix multiplication
    • Sparse matrix-vector multiplication
    • Jacobi (stencil) smoother
  • Basics of performance engineering
    • PE process
    • Code profiling
    • Proper benchmarking
    • Reproducibility and documentation
  • Beyond Roofline: The ECM performance model (optional)



You have to be able to handle a Linux command line and file editing remotely. Basic knowledge of C, C++, or Fortran programming and of OpenMP is required.


Exercises will be done on a cluster at NHR@FAU. For the exercises you need an SSH client on your local computer.




Prof. Gerhard Wellein, Dr. habil. Georg Hager (NHR@FAU, Uni. Erlangen)

Prices and Eligibility

The course is open for people from academia and industry.

The following categories can be selected during registration:

  • [Students without Diploma/Master] Students without Diploma/Master: 30 EUR
  • [Academia Germany] Students with Diploma/Master (PhD students) and members of German universities and German public research institutes: 60 EUR
  • [Academia EU/Associated Countries] Students with Diploma/Master (PhD students) and members of universities and public research institutes within EU or Associated/Other Countries to the Horizon 2020 programme: 120 EUR
  • [Academia non-EU/Associated Countries] Students with Diploma/Master (PhD students) and members of other universities and public research institutes: 240 EUR
  • [Others] Non-academic participants: 600 EUR


Please register with your official e-mail address to prove your affiliation. Following your successful registration, you will receive an invoice approx. 1-2 weeks before the course. After paying the invoice, you will not receive a receipt. If you require proof of payment (e.g., for reimbursement) please use a copy of the invoice together with your bank statement indicating the payment.

Withdrawal Policy

See Withdrawal

Legal Notices

For registration for LRZ courses and workshops we use the service edoobox from Etzensperger Informatik AG ( Etzensperger Informatik AG acts as processor and we have concluded a Data Processing Agreement with them.

See Legal Notices

Online Course Node-Level Performance Engineering
Number hnpf1w23
Available places 64
Date 04.12.2023 – 06.12.2023
Price € 30.00 – 600.00
Location ONLINE

Registration deadline 27.11.2023 23:55
No. Date Time Leader Location Room Description
1 04.12.2023 09:00 – 16:00 Georg Hager
Gerhard Wellein
ONLINE Lecture
2 05.12.2023 09:00 – 16:00 Georg Hager
Gerhard Wellein
ONLINE Lecture
3 06.12.2023 09:00 – 16:00 Georg Hager
Gerhard Wellein
ONLINE Lecture