Efficient use of the CPU nodes on Dardel and Arrhenius

This course is a collaboration between NAISS and PDC. The lectures will be provided by Hewlett Packard Enterprise (HPE)

27 - 28 January 2026

KTH logo NAISS logo

This training course will discuss the efficient use of the CPU nodes on NAISS’ Dardel and NAISS’ forthcoming Arrhenius system.

Dardel is NAISS’ current flagship service hosted at PDC in KTH. Arrhenius which is currently being installed will be NAISS’ forthcoming powerful supercomputer that is build together with EuroHPC. Both systems utilise hardware provided by Hewlett Packard Enterprise (HPE). The course will delivered by consultants from HPE.

This course is expected to be of particular interest for research groups developing Scientific software to be deployed on either Dardel or Arrhenius or both.

This course is aimed at users with experience in HPC. It is not a beginners course. Our NAISS introduction week might be more suitable for new users of HPC: https://naiss-training.github.io/NAISS-intro-week/.

This course will not discuss any GPU usage. It is aimed at CPU usage only. We plan to offer a course on GPU usage at a later stage.

Location

Kungliga Tekniska Högskolan (KTH), Stockholm.

The precise location will be announced closer to the event.

Schedule

  • 27th January 2026: 9:00 - 17:30
  • 28th January 2026: 9:00 - 16:00

Note: The time on the second day has been amended from earlier notifications!

Topics

The course covers the following topics (preliminary agenda):

  • CPU part of Dardel and Arrhenius Architecture, programming and execution environment
  • Overview of compilers and parallel programming Models
  • Compilers, libraries, runtime and performance optimization: Improving single-core performance
  • MPI topics on the HPE Cray EX supercomputer
  • Advanced mapping of parallel applications on the CPU partition
  • Python on the HPE Cray EX Supercomputer
  • Introduction to Perftools - Perftools-lite modules
  • Advanced performance analysis
  • Debugging at Scale
  • Optimizing large scale I/O

There will be time in the schedule for hand on exercises on NAISS’ Dardel system.

The performance and debugging tools mentioned above are available on Dardel. Their availability on Arrhenius is still under investigation. The schedule of the course will be adopted to focus on tools that will be available on both Arrhenius and Dardel.

Comments and questions

Comments and questions on these training events should be sent to NAISS using the support form in SUPR. Please select “Question about a training event” as “Problem or Questions Type”.

Registration

Please fill in the registration form (handled by the Lund University Branch of NAISS): https://survey.mailing.lu.se/Survey/62634

Registration is set to close on 15th January 2026. Acceptance on the course will be announced no later than the 19th January 2026.