17. May 2024

ISC High Performance 2024 in Hamburg

It was a blast! As last year, we were again part of the joint booth of the NHR alliance at ISC High Performance 2024. Our experts were available at the booth for interested visitors and explained everything about our computing time and supercomputer LichtenbergII and CLAIX-2023. They were also part of the conference program and actively shaped the ISC High Performance 2024.

3,227 participants visited this year’s conference in Hamburg. The next ISC High Performance will also take place in Hamburg from June 10 – 13, 2025.

⬅️ Let’s do a quick run through ISC High Performance 2024.

Tutorials on Sunday

Christian Terboven, member of our CSG Parallelism and Performance, organized the half-day tutorial “Advanced OpenMP: Performance and 5.2 Features”. Together with Michael Klemm and Bronis R. de Supinski, he discussed language features in-depth, with emphasis on advanced features like vectorization and compute acceleration. They focused on performance aspects, such as data and thread locality on NUMA architectures, and exploitation of the comparably new language features. They also presented the directives for attached compute accelerators. The tutorial is designed for HPC programmers with some knowledge of OpenMP who want to implement efficient shared-memory code.

More information

Alexander Hück, Joachim Jenke, Adrian Schmitz and Simon Schwitanski organized the second half-day tutorial in the context of NHR4CES: “Continuous Correctness Checking for HPC Applications”. The participants were very engaged and interesting discussions took place. As an additional service, Joachim Jenke and Simon Schwitanski organized a sofa talk on the topic and were available to answer questions. 

More information

Sofa Talks

Together with the other eight NHR centers, we organized the so called sofa talks again. In this format, our experts are talking about topics, which are of interest for our community.

Visualization for and with HPC infrastructure

Together with his guests Jens Henrik Göbbert (JSC), Marcel Krüger (NHR4CES), James Kress (KAUST), Kristi Belcher (Lawrence Livermore National Laboratory), François Mazen (Kitware Europe), Uwe Wössner, Dennis Grieger and Martin Aumüller (High-Performance Computing Center Stuttgart), Tim Gerrits, member of our CSG Visualization, discussed current challenges, failures, and success stories of the community. His Sofa Talk on Visualization for and with HPC infrastructure at the ISC High Performance is quite the tradition by now.

Towards a Certified HPC Curriculum

Julian Kunkel (NHR-Nord@Göttingen), Marc-André Hermanns (NHR4CES), Kevin Lüdemann (NHR-Nord@Göttingen), and Jane Herriman (LLNL) discussed recent changes in the HPC Certification Forum’s skill tree specification and how to improve the evolution both in terms of breadth and depth of the specified competencies. Furthermore, the group talked about paths to increase the overlap between defined skills and available HPC curriculum and how to keep momentum in the path to full convergence in the future.

Fostering Diversity in HPC

As part of NHR’s ongoing commitment to improving diversity within HPC, we discussed with an international group of participants what is still needed to improve the career situation for HPC practitioners from underrepresented groups and how HPC institutions such as NHR can support them on their path. The participants of the Sofa Talk ‘Fostering Diversity in HPC’ discussed and acknowledged that there are still many challenges for a truly diverse workforce in science. Nevertheless, especially the exchange of many personal experiences of the participants made this a fruitful event for brainstorming ideas within the circle of influence of NHR to improve on the status-quo of diversity barriers. Thanks to our participants Monique Drees (NHR-Verein), Christiane Manzano (FZ Jülich), Sarah Neuwirth (NHR@SW), Julian Kunkel (NHR@Göttingen), Silvina Grad-Freilich (Mathworks), Brian Wylie (FZ Jülich), Sarah Beecroft (Pawsey Computing Centre), Estela Suarez (FZ Jülich) and other guests – who joined us spontaneously – for attending!


As part of NHR’s AI strategy Marc-André Hermanns (NHR4CES, HPC.NRW) convened with Robert Schade (NHR@PC2, HPC.NRW), Bert Wesarg (NHR@TUD), and Frank Winkler (NHR@TUD) to discuss how the different contributors of all computing tiers to the weekly AI on Supercomputers consulting hour can improve collaboration to increase user engagement in this joint consulting effort, helping them to find sufficient computing time and support for moving from the Laptop towards supercomputers with their AI workload.

Federal state level HPC competence networks

Thorsten Reimann (NHR4CES, HKHLR) and Marc-André Hermanns (NHR4CES, HPC.NRW) gathered Georg Hager (NHR@FAU, KONWIR), Michaela Bleuel (NHR4CES, HPC.NRW), Christian Köhler (NHR-Nord@Göttingen, HLRN), Robert Schade (NHR@PC2, HPC.NRW) and Robert Barthel (NHR@KIT, bwHPC) to reinstate the group of the initial NHR central project ‘federal state level HPC competence networks’ to engage in discussions and an exchange of ideas among the NHR centers and Germany’s state-level HPC competence centers. The group picked up discussion points that were still open, especially in terms of topics such as training and certification, emphasizing the importance of a regular exchange. After this very fruitful meeting, it is planned to keep momentum with regular meetings at future ISC and NHR conferences.

Security in NHR

Gerrit Toehgiono, member of our CSG Parallelism and Performance, talked about Security in NHR. He contributed his perspective as a system administrator on IT security at HPC and discussed the current state of security with his guests: For example, what measures have been introduced and what risks exist for users.

Continuous Correctness Checking for HPC Applications

After their tutorial ‘Continuous Correctness Checking for HPC Applications’. Simon Schwitanski and Joachim Jenke, member of our CSG Parallelism and Performance, offered participants the opportunity to clarify further questions at the NHR booth during a sofa talk as an additional offer. We are very happy that the Sofa Talks – a format we introduced in 2022 – are going so well and look forward to continuing them at ISC High Performance 2025.

Workshop Presentations

Marcel Krüger (see below), member of our CSG Visualization, presented their recent work at the 8th International Workshop on In Situ Visualization.In this talk the CSG Visualization presented InsitUE – a Catalyst2-compatible library that enables hybrid in-situ/in-transit workflows with Unreal Engine focusing on efficiency and portability.With the introduction of InsitUE, they address the lack of in-situ workflows for modern game engines. An issue raised in one of their recent journal submissions where they discuss and share insights on how game engines such as Unreal Engine can transform the development process for scientific visualizations and report the experiences gathered over several years in the immersive visualization lab at RWTH Aachen University.InsitUE enables users to combine the Unreal Engine with Catalyst2 instrumented applications which are essential for handling large datasets or obtaining early simulation insights. The paper on InsitUE will be published in the Springer Lecture Notes of Computer Science series soon – stay tuned.

Simon Schwitanski presented at the Workshop for Compiler-based Correctness Analysis and Performance Optimization (C3PO’24) how compiler analysis can accelerate the search for data races in programs with remote memory accesses. The corresponding papers will be published in Lecture Notes in Computer Science (Springer LNCS) in the coming weeks.

BoF Talks

Felix Wolf, director of NHR4CES, was part of the BoF session ‘Go with the (Energy) Flow: Adaptive Capacity Computing’. This BoF continued the discussion started at SC23 on approaches for the HPC community at large to provide adaptive capacity computing and how to minimize the challenges created for HPC application developers and users. ➡️ More information

On Tuesday, Matthias Müller, director of NHR4CES, spoke at the BoF ‘Synergistically Integrating HPC and Cloud’ This BoF focused on the convergence of HPC and Cloud from the perspective of a cloud provider, a public HPC center, an AI research center, a large scientific project, and industry. ➡️ More information

Christian Terboven, member of our CSG Parallelism and Performance, was an invited speaker at two BoF talks. As an expert on OpenMP, he shared his thoughts in the session ‘What to Expect from OpenMP API Version 6.0 ‘. ➡️ More information

He also spoke at the BoF-talk ‘AI Service Centers: Pioneering AI Research and Infrastructure in Germany’. This session focused on the four BMBF-funded national AI service centers and their goals. Christian talked about the AI service center WestAI.  ➡️ More information

Poster & Presentations

We awere represented with several posters at the poster session of ISC High Performance 2024 – have a look:

Title: ENSIMA: Energy-Efficient Simulation Methods for Application-Oriented Computational Problems – Authors: Seyed Ali Mohammadi, Semih Burak, Jan André Reuter, Lukas Moj, Kai Diethelm, Koutaiba Kassem-Manthey, Bernd Mohr, Christian Terboven, Felix Wolf, Christopher Woll  ➡️ More information

Title: targetDart – Authors: Adrian Schmitz, Christian Terboven, Christian Siebert, Josè Gracia, David Schneller, Mario Wille, Michael Bader, Tobias Weinzierl ➡️ More information

Title: EE-HPC – Authors: Radita Liem, Christian Terboven, Jan Eitzinger, Thomas Gruber, Rafael Ravedutti, Jose Gracia, Kingshuk Haldar, Pay Giesselmann, Jan Frederik Engels, Christian Simmendinger, Torsten Wilde, David Brayford, Marcel Marquardt, Jan Maeder  ➡️ More information

Title: Revisiting IO500 Benchmark for the System’s Changes Visibility – Authors: Subeen Kim, Radita Liem, Julian Kunkel, Jay Lofstead. ➡️ More information

Title: Compiler-Assisted Correctness Checking and Performance Optimization for HPC – Authors: Simon Schwitanski, Yussur Mustafa Oral, Cornelius Pätzold, Joachim Henke, Matthias S. Müller ➡️ More information