hpc-ch forum on User Workflows and Support

Thursday 29 Oct 2026, 09:30 → 15:35 Europe/Zurich

CB building (Aula)

Julia Gustavsen (Agroscope), Maria Grazia Giuffreda (CSCS)

Topic Description

As scientific computing services evolve, HPC centers are increasingly asked to support not only computational infrastructure, but also the diverse workflows, skills, and expectations of their users. Researchers may arrive with very different levels of experience, from first-time HPC users to advanced users developing complex, automated, and reproducible workflows.

Participants will discuss how HPC centers can better understand what users need, what they are actually doing, and how services can support them in ways that are useful for researchers and sustainable for institutions.

Key Questions

Reproducible Research

• How can HPC centers encourage users to adopt practices and tools that support reproducible research? Examples may include version control, containers, workflow managers, environment management, etc.
  
  

User Documentation and Collaborative Documentation

• How can user documentation be provided in a way that is sustainable for service providers and easy for users to find, understand, and apply?
• How can documentation help set appropriate expectations about the competencies required from users and the support provided?
   

User Support

• What support models work well for HPC users beyond traditional ticketing systems?
• Are centers experimenting with office hours, co-working sessions, workshops, user communities, or other formats?
• How can support models balance responsiveness, scalability, and user empowerment?
   

End-User Service Fulfilment

• How can centers simplify and improve processes such as resource allocation, account creation, access to software environments, and service requests?
• What role can automation, container images, predefined environments, accounting systems, and service catalogue items play?
• How can service fulfilment be made efficient for both users and providers?
   

Portfolio Management for End Users in Hybrid or Multi-Service Environments

• How can providers present a coherent service portfolio when users may need access to clusters, cloud resources, storage, applications, portals, or specialized platforms?
• How can complexity be abstracted away while still allowing services to be customized or adapted to specific scientific needs?
   

Training Course Design and Evolution

• How can training courses be designed to meet users at different levels of experience? Or should they? How do we design the appropriate portfolio of courses to offer?
• How can training evolve alongside changing services, tools, and scientific workflows?

Who Should Attend?

The forum is intended for professionals involved in HPC and scientific computing services, including HPC service providers, support teams, research IT groups, system administrators, research software engineers, trainers, service managers, and others interested in improving the user experience of scientific computing.
 
It is especially relevant for those looking to better understand their users, improve support and documentation, design effective training, and develop sustainable services that help researchers work efficiently and reproducibly.

09:30 → 10:00 Welcome Coffee & Registration

10:00 → 10:15 Welcome and Introduction

Conveners: Julia Gustavsen (Agroscope), Maria Grazia Giuffreda (CSCS)

10:15 → 10:30 Welcome to Agroscope

Convener: Baptiste Berger (Agroscope)

10:30 → 11:00 Understanding and Supporting Emerging HPC Users at Agroscope

Agroscope has only recently entered an era in which it provides dedicated computational resources for research through a small HPC cluster. Beyond the technical and infrastructural challenges, this has required us to better understand users’ needs, scientific workflows, skill levels, and expectations.

This presentation will describe our approach to developing user support for scientific computing at Agroscope. We will present several complementary activities, including training courses, co-working sessions, user documentation, and the development of user personas based on ticketing data. The talk will discuss what data we collect, how we use it to understand our user community, and how these insights inform the evolution of our support and services.

Convener: Julia Gustavsen (Agroscope)

11:00 → 11:30 From Confluence to Code: Rethinking Documentation at CSCS

CSCS is introducing a range of changes aimed at simplifying how users interact with its HPC services.

This talk will present a few key developments: our transition to GitHub-hosted documentation that encourages direct contributions from users; the introduction of Status Pages and Slack as more transparent and effective alternatives to traditional email communication; and the HPC Console, a web-based interface that enables users to submit and monitor jobs without requiring SSH access.

We will discuss how these developments aim to improve the user experience, encourage greater user engagement, and make HPC services more accessible and easier to support.

Speaker: Tim Robinson (CSCS)

11:30 → 12:00 Beyond Hardware: Creating a Nice User Experience

This talk describes how we design the user experience for the Euler cluster. It covers different aspects of the user journey, including communication with users, an almost fully automated onboarding process, and making the system as easy to use as possible.

We will also present our documentation based on MkDocs, discuss how we organize and handle user support, and give an overview of the training courses we provide. The talk will share some of our approaches and experiences in providing a user-friendly HPC environment beyond the hardware itself.

Speaker: Samuel Fux (ETH Zurich)

12:00 → 12:30 Galaxy, an Open Source Platform for Accessible, Reproducible, and Transparent Computational Research

The Galaxy project (https://galaxyproject.org/) offers an open-source platform that allows researchers to analyze and share scientific data
using interoperable APIs and various user-friendly web-based interfaces.

One of the key features of the Galaxy platform is its emphasis on transparency, reproducibility, and reusability. Galaxy is a multi-user environment which facilitates sharing of e.g. tools, workflows, notebooks, visualizations, and data with others.

The Galaxy project was launched in 2005 and has since become a powerful framework for researchers across a wide range of research fields, including *omics, biodiversity, machine learning, cheminformatics, NLP, material science, climate research, and humanities.

The flexible nature of the platform allows its installation on any kind of hpc hardware from a simple, single box to big compute centres using SLURM, HtCondor, etc (e.g.: usegalaxy.eu, where several million of jobs are scheduled every month).

We have been using Galaxy at the Friedrich Miescher Institute (www.fmi.ch) since 2008 providing easy access to the world of reproducibility, workflows, and high performance computing.

Speakers: Björn Grüning (Albert-Ludwigs-Universität Freiburg), Hans-Rudolf Hotz (Friedrich Miescher Institute for Biomedical Research, Basel)

12:30 → 13:30 Lunch and Networking

13:30 → 14:00 Science IT at UZH: Supporting Research in the AI Era

Science IT (S3IT) provides computational infrastructure, support, and consulting to researchers at the University of Zurich.

Across our three computing services, we offer various types of CPU and GPU resources with tools, interfaces, and platforms tailored for research.

This talk introduces our services and how we support researchers. We will discuss how our documentation, training, and support enable independent, reproducible work across a diverse user-base.

Finally, we will explore adapting to growing AI demands, from scientific discovery to coding assistants, while keeping our services scalable, sustainable, and researcher-focused.

Speaker: Philip Shemella (University of Zurich)

14:00 → 14:30 Rethinking User Onboarding at the University of Bern

At the University of Bern, we recently redesigned the user onboarding workflow, moving from a ticket-driven, manual process towards a more automated approach.

In this lightning talk, we will present the motivation behind the change, give a brief overview of the new onboarding workflow, and share our initial experiences.

Speaker: Christian Lehmann (University of Bern)

14:30 → 15:30 Guided Tour

• Tour I (in French and German only, NO English)
  Title: Tour of the Posieux Experimental Farm
  Description: Discover a rather unique farm: one where production and research go hand in hand. During the tour, we’ll introduce you to the animals, the crops, the farm’s key statistics, and some of the facilities used for agricultural research.

• Tour II (in English)
  Title: Agroforestry for Forage Production: Providing Supplementary Feed Supply for Livestock during Summer Drought Events
  Description: Climate change is an important factor in yield and forage quality losses, mainly due to the increased periods of drought during the growing season. Agroforestry for forage production could be a promising solution, particularly in summer when the risk of reduced forage production from drought is highest in grasslands. The AgroForageTree project aims to test the potential of fodder tree hedgerows for providing supplementary tree leaf-based forage in summer while also providing other important ecosystem services (carbon storage, biodiversity, microclimate regulation, animal health). During the visit of a fodder tree hedgerow, the results of the project will be presented, including some information on tree leaf-based forage yield and quality, animal consumption and species preference, and impacts on insect biodiversity.

• Tour III (in English)
  Experimental Cheese Factory

15:30 → 15:35 Farewell and End of the Meeting