Performance monitoring with TAU and E4S on LUMI and Cloud HPC platforms

Oct 29, 2025, 8:45 AM → Oct 30, 2025, 5:00 PM Europe/Stockholm

RISE office at KTH Campus (Hybrid)

Overview

KTH, ParaTools, the University of Oregon, and ENCCS (EuroCC National Competence Centre Sweden) come together for the organization of this hybrid workshop. The event brings together code developers, researchers, and research software engineers working on high-performance computing (HPC). It creates an amazing opportunity for sharing innovative ideas and best practices for performance monitoring and software management on supercomputers and cloud platforms. This workshop includes:

• Performance monitoring and analysis using the TAU Performance System.
• Deploying scientific software stacks on supercomputing environments like LUMI and cloud HPC.
• Hands-on instrumentation of MPI and GPU (ROCm and CUDA) applications.
• Analyzing performance data with ParaProf and PerfExplorer.
• Using containers and package managers in HPC with E4S.

The theoretical optimization of complex code is somewhat limited, and great progress in scientific computing has been achieved through performance analysis tools. Indeed, advances in massively parallel systems like the LUMI supercomputer and cloud HPC platforms have further boosted our understanding of complex phenomena at unprecedented resolutions. To leverage these resources effectively, robust tools for performance monitoring, debugging, and software management are required. These tools enable the development of key applications for science, industry, and society. knowledge and enabling key applications for science, industry, and society.

Performance Monitoring with TAU

TAU (Tuning and Analysis Utilities) is the state-of-the-art performance evaluation tool covered in this workshop. It is a portable and versatile profiling and tracing toolkit for analyzing the performance of parallel programs written in languages like Fortran, C++, C, UPC, Java, and Python. In this workshop, you will learn how TAU can be used to instrument, measure, and analyze applications on both the LUMI supercomputer and cloud HPC platforms. It enables a deep understanding of application behavior by gathering detailed performance data from various runtime systems, including MPI and GPUs (ROCm and CUDA). The novelty is that by providing detailed, multi-level performance information, developers can pinpoint bottlenecks with high precision. Due to its comprehensive approach, TAU’s utility in optimizing complex codes is immense, and this workshop will feature hands-on sessions using its powerful analysis tools like ParaProf, TAUdb, and PerfExplorer.

Prerequisites

• PhD students, postdocs, industry engineers
• Basic familiarity with Unix shell
• Basic familiarity with HPC or cloud environments (including MPI and GPUs)

More events & contact

Check out more upcoming events from ENCCS and our European network at https://enccs.se/events.

For questions regarding this workshop or general questions about ENNCS training events, please contact training@enccs.se

Schedules can change!

To ensure that everyone has the opportunity to participate, we kindly request that you let us know as soon as possible if you are unable to attend an event after registering.

Please send us an email at training@enccs.se to cancel your attendance.

We understand things can change, but repeated cancellations without notice may unfortunately result in your name being removed from future event registration lists.

---

Regulations

Due to EuroCC2 regulations, we CAN NOT ACCEPT generic or private email addresses. Please use your official university or company email address for registration.

This training is for users who live and work in the European Union or a country associated with Horizon 2020. You can read more about the countries associated with Horizon2020 HERE.