LUMI Profiling and Optimization Workshop

This comprehensive 3-day on-site course in Stockholm, Sweden provides an in-depth exploration of the LUMI supercomputer architecture and programming environment. Designed for experienced HPC users, the course combines theoretical knowledge with practical hands-on exercises to maximize your efficiency on both CPU and GPU partitions.

What You’ll Learn

LUMI Architecture: Deep dive into system design and component
Performance Tools: Compiler optimization, performance analysis, and profiling
Environment variables: Understanding Cray MPI on Slingshot, GPU support and rank reordering
I/O Optimization: Efficient data handling and storage strategies
AMD GPU Tools: Specialized profiling and optimization for GPU workloads

Event Details

Location: Fantum room, KTH, Lindstedtsvägen 24, Stockholm, Sweden
Format: On-Site training
Cost: Free

Registration Information

Deadline:  May 3, 2026 at 23:59 CEST 

Prerequisites

Solid knowledge of HPC cluster computing
Familiarity with Linux command line
Programming experience in C, Fortran, or Python
Good understanding of LUMI supercomputer
Active LUMI project or application in progress

Post-Registration Requirements

Join the course-specific training project
Set up a LUMI account (instructions provided after registration)

Instructors

LUMI User Support Team (LUST)
HPE experts
AMD specialists

Resources

Training Materials: https://lumi-supercomputer.github.io/LUMI-training-materials/
Help: https://lumi-supercomputer.eu/user-support/need-help/

Important Notes

Participants arrange their own travel and accommodation.
Course details will be emailed shortly after the registration deadline. 
Please cancel promptly if plans change to free seats for the waiting list