Join Us at ISC High Performance 2026

Event/Agenda:  NVIDIA ISC26 Executive Briefing
Date: Monday, June 22, 2026
Time: 4:30 - 5:10 pm
Location: Venue NORD EVENT Panoramadeck, Emporio (Freiraum I & II), Dammtorwall, Hamburg HH 20355, Germany
Speakers: BSC, CINECA, HLRS, Stockholm Uni. Moderator: Christelle Piechurski

Session: Community Stage Meet Up
Date: Tuesday, June 23, 2026
Time: 5:15 - 6:00 pm
Room: Community Stage, Hall H, ground floor
SwapCard Link: https://app.swapcard.com/event/isc-high-performance-2026/planning/UGxhbm5pbmdfNDQ1ODY0NQ==

Abstract
As high-performance computing (HPC) rapidly evolves from a primarily academic resource into a cornerstone of industrial innovation—powering AI, digital twins, advanced manufacturing, drug discovery, and emerging quantum/HPC workflows—the pressure on HPC centers is fundamentally changing. It is no longer sufficient to provide world-class compute infrastructure alone. Centers are increasingly expected to translate this capability into scalable, repeatable, and economically meaningful impact for industry. Yet, despite significant investments across Europe, North America, and globally, many of these collaborations remain fragmented, ad hoc, and difficult to scale.
 

Session: Research Poster
Date: Tuesday, June 23, 2026
Time: 6:00 - 6:07 pm
Room: Hall Z - 3rd Floor
SwapCard Link: https://isc.app.swapcard.com/widget/event/isc-high-performance-2026/planning/UGxhbm5pbmdfNDQ1NjYxMg==

Abstract
Johannes Gebert and Jonathan Schäfer from HLRS, together with Daniel Renschler and Mark Parsons are among the finalists of the ISC High Performance 2026 Research Poster Awards, for a poster titled "SpMV for the Cerebras Wafer-Scale Engine - exploring how traditional HPC workloads can be adapted to AI-specialized architectures".

The nomination recognizes work in areas spanning HPC architectures, AI-driven applications, and scalable systems, highlighting ongoing efforts to advance high-performance computing and its applications.

Date: Tuesday, June 23, 2026
Time: 6:00 - 6:30 pm
Location: EuroHPC JU booth J30

Abstract
At the EuroHPC JU booth J30, EuroHPC JU will have a presentation corner for EuroHPC projects, AI Factories, quantum and supercomputers' hosting entities to showcase their activities and outcomes. On 23 June at 18:00, Dr. Natalie Lewandowski from the CASTIEL 3 project will present the “HPC in Europe Portal”. The talk will introduce the objectives of this new integrated 3-in-1 platform and highlight how it supports and benefits HPC users across Europe.

Date: Wednesday, June 24, 2026
Time: 3:00 - 6:00 pm
Location: GCS booth K02

Take a break, grab a drink, and connect with the HPC community!

Join the Gauss Centre for Supercomputing (GCS) for a relaxed Happy Hour at our booth.

Session: Community Stage Meet Up
Date: Thursday, June 25, 2026
Time: 1:00 - 1:45 pm
Room: Community Stage, Hall H, ground floor
SwapCard Link: app.swapcard.com/event/isc-high-performance-2026/planning/UGxhbm5pbmdfNDQ1ODY0NA==

Abstract
The EuroHPC Virtual Training Academy (EVITA) project aims to provide a structured online platform and curated material catalogue to showcase high-quality HPC knowledge in Europe. In order for the EVITA project to be able to provide high-standard training material, discussions about structure, style, and applicability to university and industry sectors are essential. Industry, in particular, is of high importance in this discussion in order to broaden perspectives, given that the consortium of the EVITA project is mainly composed of  research institutions and HPC providers.

The EVITA project introduces the Competence and Qualification Framework (CQF) that defines a unified and modular learning structure. In the CQF, modules are basic stand-alone training units, which are peer-reviewed and tagged using the Skill Tree of the HPC Certification Forum (HPC CF). Modules contain knowledge covering one or more skills and can be learned in one to four hours (including independent study time) if the learner meets the prerequisite knowledge mentioned in the module. The CQF defines that one or more modules can be used by instructors or training providers to be turned into courses, which, in addition to the material, also include a curated survey after the course as well as an examination to gain a course certificate. Ultimately, this material is free to be used by any instructor teaching students and learners, be it at universities, at HPC centres, or in the industry.

The proposal of this Community Stage Meetup targets training providers as the main audience but also reaches out to companies seeking to deliver training material or receive training. The EVITA project addresses the need for modules and appeals to potential module providers to contribute to the funded call for modules. Moreover, in this meetup, the EVITA project also aims to have open floor discussions to gather feedback on the proposed CQF and optimise the available list of current and future modules, as well as, the current module templates. A collaboration tool such as Google Docs or Hedgedoc will be used to collect feedback, and GitLab issues on code.europa.eu can be used to address requests for the available modules and templates. The key insight will be the feedback from the community for the project and the publication of the call for modules to the community.

The current catalogue of modules, available on a public GitLab (https://code.europa.eu/eurohpc-ju/evita), contains a few prototype modules that were created by members of the project, which act as guidance for future module providers. The EVITA project provides funding for any training provider in Europe that wants to either turn already existing material into modules or submit entirely new material covering new technologies or as-yet unexplored areas of the HPC CF Skill Tree. A presentation and open floor discussion about the running call for modules will be included in the
session. The first call opened in spring 2026, and there will be two additional calls: one in 2027 and one in 2028. Moreover, the session will also highlight that the EVITA project is open to collaboration and contributions from other EuroHPC JU projects, including the AI Factories.

Attendees will not only to learn about the EVITA project, but also recognize their capacity to contribute meaningfully to its development through their insights and engagement. The meetup will begin with a short presentation about the CQF and pilot modules, followed by an exploration of the existing modules with the audience. Next, an open discussion on the existing modules and which modules and skills should be included in the follow-up calls will take place. Active discussion is the most important part since the EVITA project needs to get in touch with the community to gain insight into what is actually needed. This will allow the project to steer the follow-up proposals into relevant domains since the initial call is for general and core HPC technology.

Session: Invited Talk
Date: Thursday, June 25, 2026
Time: 1:00 - 2:00 pm
Room: Hall Z - 3rd Floor
SwapCard Link: https://isc.app.swapcard.com/widget/event/isc-high-performance-2026/planning/UGxhbm5pbmdfNDQ1NjUyNQ==

Abstract
Physical AI is connecting physical systems like cameras, sensors, digital twins and even robots with AI models and systems to perceive, reason and act in real environments.
Physics-based simulations are essential for the training of autonomous systems that shall interact safely between humans and machines.
The first talk in this session explains the important role of simulation in connecting the physical and digital world.
The second talk will focus on the AI requirements from industrial robotics use cases and how AI factories are supporting such projects.

Session: Workshop
Date: Friday, June 26, 2026
Time: 9:00 am - 6:00 pm
Room: Hall X6 - 1st Floor
SwapCard Link: https://isc.app.swapcard.com/widget/event/isc-high-performance-2026/planning/UGxhbm5pbmdfNDM5NTcxMw==

Abstract
How can AI workloads be engineered for optimal performance in modern HPC environments?

The rapid advancement of Artificial Intelligence (AI) and Machine Learning (ML) has positioned High-Performance Computing (HPC) systems as indispensable platforms for developing, training, and executing these workloads. However, the architectural complexity and batch-oriented design of traditional HPC systems pose unique challenges distinct from those encountered in resource-elastic environments such as clouds.
The parallelization characteristics, input/output requirements, and dynamic workflows of AI workloads demand innovative techniques for efficient utilization of HPC resources. Moreover, the performance engineering of such workloads is crucial to achieve scalability, portability, and reproducibility across diverse system architectures.
This workshop aims to bring together researchers, practitioners, and system developers to discuss engineering challenges, performance optimization, and emerging opportunities at the intersection of AI and HPC. It invites among others, papers that present experimental results, architectural insights, performance studies, and best practices advancing the convergence of these domains.

Further information on the workshop can be found here: https://ai-on-hpc.github.io/