Category: Interactive HPC

UCloud and Digital Sovereignty in focus during Ministerial Visit

Post author By line
Post date November 12, 2025

Photo: Morten Mossin

On 27 October, Minister for Digital Affairs Caroline Stage Olsen visited the University of Southern Denmark (SDU) to learn more about UCloud and the Interactive HPC Consortium. 

The visit aimed to showcase how Danish research contributes to strengthening Denmark’s digital independence and sovereignty. The Department of Mathematics and Computer Science (IMADA) and the SDU eScience Center at the Faculty of Science were pleased to welcome the Minister to SDU.

During her visit, the Minister was introduced to UCloud, an open-source cloud platform operated by the Interactive HPC Consortium. Originally developed by SDU, UCloud has been available since 2019 via the DeiC Interactive HPC service to all researchers in Denmark. Today, the consortium behind UCloud comprises SDU, AU, and AAU, who jointly develop and operate the platform.

The Minister emphasised that digital sovereignty and the development of cloud solutions under Danish control are key priorities for the government:

”This is something we are increasingly discussing – how we can become more independent and strengthen our control over digital infrastructure. That is part of what I am learning about today,” said Caroline Stage Olsen, Minister for Digital Affairs, during her visit.

Building Bridges Between Research and Society

UCloud serves as Denmark’s national platform for interactive high-performance computing (HPC) and is Europe’s most widely used research supercomputing platform. With more than 18,000 users across universities, public authorities, and private companies, it stands as a clear example of how Danish-developed solutions can promote digital self-reliance.

“True digital sovereignty requires public infrastructure you can inspect, control, and improve. UCloud turns sovereignty from a slogan into a living, open-source public good — Europe’s largest research cloud built in Denmark. Investing in open infrastructure like UCloud is how we can secure our digital future,” said Professor Claudio Pica, Head of the SDU eScience Center.

A Responsibility Towards Society

The visit also prompted a broader dialogue about the responsibility of research institutions in an era where digitalisation permeates every aspect of society – from healthcare and education to the energy sector and public services.

The visit concluded with a tour of SDU’s supercomputing facilities, where the Minister was introduced to the advanced infrastructure that supports Interactive HPC – UCloud.

This article is based on an original story published on SDU’s website.

Interactive HPC Supercomputing

Working together on the future of UCloud

Post author By line
Post date November 4, 2025

Recently, representatives from Center for Humanities Computing at Aarhus University, eScience Center at University of Southern Denmark and CLAADIA at Aalborg University gathered for a two-day workshop filled with collaboration, knowledge sharing, and new ideas for the UCloud platform.

The workshop brought together the consortium’s technical experts, user support teams, and outreach coordinators to discuss the next steps for UCloud. The goal was to align strategies, evaluate recent developments, and ensure that UCloud continues to evolve in line with the needs of the Danish research community.

Collaboration across universities

UCloud is developed and operated in close collaboration between the three universities, and workshops like this play a key role in maintaining strong coordination across the consortium. Since the teams are based at different locations, meeting in person provides an important space to strengthen relations, align efforts, and identify shared priorities for the platform’s future.

Plenary sessions, presentations, and parallel workshops

The two days included a mix of plenary sessions, presentations, and parallel workshops across different working groups. Topics ranged from user support, training, and outreach to technical development and new feature implementation. Updates were also shared on UCloud’s fast-growing user base.

Insights from users

A highlight of the workshop was hearing directly from the users. Researchers Postdoc Kenneth Enevoldsen and Postdoc Simon Aagaard Enni joined us to share their research projects, Lex.llm and Danish Foundation Models. They shared how UCloud supports their research today, while also pointing to areas where the platform can evolve further to better support their work.

Kenneth, who is an experienced UCloud user, emphasized how future inference support on UCloud could help both him and many other researchers in practice. In this context, he specifically referred to the need for running pre-trained large language models (LLMs) directly on the platform.

“UCloud support for inference would be incredibly liberating for research. It would either allow us to avoid spending credits on commercial API providers or significantly reduce the time required to set up LLMs for exploring small research questions. A broad inference API would enable more open experimentation with models, enhancing both the scope and impact of research projects.”

Looking ahead

As the workshop concluded, the participants outlined key priorities and future directions for the platform. The discussions and decisions made will guide upcoming development efforts and ensure that UCloud remains a flexible, secure, and user-friendly environment for researchers across disciplines.

We will also continue to focus on preparing users to work confidently with UCloud through hands-on training sessions held as webinars. During the workshop, we finalised the plans for this semester’s webinar offerings, which will take place on the following dates:

Application Interactive HPC Supercomputing Workshop

Workshop 4/12: Building Intelligent Knowledge Assistants with RAGFlow on UCloud

Post author By line
Post date October 31, 2025

Date: 4 December 2025
Time: 13:00 – 14:00 (CET)
Location: Online, via Zoom

Join us for a hands-on webinar introducing RAGFlow, an advanced Retrieval-Augmented Generation (RAG) platform available on UCloud. RAGFlow enables you to build your own intelligent assistants by combining document retrieval, embeddings, and large language models – all within a user-friendly visual interface.

In this session, you will learn how to:

Launch RAGFlow on UCloud – explore the application options in the job page and understand how to configure your workspace.
Navigate the RAGFlow interface – get familiar with the main dashboard, menus, and workflow structure.
Set up an embedding and chat model – choose from available models and understand their roles in retrieval and dialogue generation.
Create and manage knowledge bases – upload and process your own datasets, configure chunking and embedding strategies, and test retrieval quality.
Build a chat assistant – connect your models and knowledge base to create an interactive, context-aware chatbot.
Design a basic agentic workflow – combine tools and steps in RAGFlow’s visual builder to automate reasoning and responses.
Launch RAGFlow on UCloud – explore the application options in the job page and understand how to configure your workspace.

All workflows will be demonstrated live on UCloud, showing how to go from data ingestion to an operational AI assistant – no coding required.

Target audience: Researchers, students, and AI-curious users interested in document-based question answering, knowledge management, and conversational AI.

Technical Level: Beginner to Intermediate — no prior experience with RAG or model configuration needed. Basic familiarity with UCloud will be helpful. If you’re new to the platform, we recommend signing up for the Getting Started with UCloud workshop on 19 November

Sign up for the RAGFlow on UCloud workshop

About the RAGFlow application:

RAGFlow is a new application being introduced on UCloud this December. It lets you build your own smart assistant using your documents. It combines search and AI to answer questions based on information you provide – no coding required.

Key features:

Create your own knowledge base
Upload PDFs, text files, or notes and let RAGFlow organize them for easy retrieval.
Choose AI models
Pick an embedding model for search and a chat model for conversations, all within the app. You can also add other model types – such as reranking for better answer accuracy, speech-to-text for audio input, image-to-text for scanned documents, or text-to-speech for spoken responses.
Chat with your data
Ask questions and get context-aware answers directly from your uploaded materials.
Simple visual workflow
Set up each step through an easy, guided interface.
Secure on UCloud
All data stays private and runs safely within UCloud’s secure environment.

RAGFlow is perfect for beginners who want to explore how AI can search, summarize, and explain their own research data or learning materials. For advanced users, RAGFlow also offers powerful customization through agentic workflows, where multiple models and tools can work together in sequence.

Keep an eye on updates here or in the UCloud app catalogue to be the first to know when RAGFlow becomes available.

Interactive HPC

Join us on LinkedIn

Post author By line
Post date October 30, 2025

We’re excited to share that you can now follow Interactive HPC on LinkedIn

Our new page will keep you updated with the latest news from our world of high-performance computing – including use cases, upcoming events, webinars, and applications on UCloud.

New applications and tools

Use cases showcasing how researchers from all research areas apply HPC in their work

Events and webinars, where we share knowledge and best practices with UCloud users

Join us on LinkedIn

Application Interactive HPC Supercomputing UCloud

OMERO Now Available on UCloud

Post author By line
Post date September 4, 2025

A Powerful Tool for Biological and Health Science Image Data Management

The DeiC Interactive HPC – UCloud Consortium introduces OMERO on UCloud. OMERO is a robust open-source image data management system tailored for biological research and health science. Designed to handle vast volumes of microscopy images and their associated metadata, OMERO empowers researchers to store, organize, visualize, and analyze imaging data with precision and ease.

What is OMERO?

OMERO (Open Microscopy Environment Remote Objects) is a comprehensive solution for managing biological and health science imaging data. Whether you’re working in digital pathology, cellular biology, or any other imaging-intensive field, OMERO provides a secure repository to support your research workflows.

“The microscopic images that biologist and health scientists typically work with have a lot of pixels and can be extremely heavy. With Omero on UCloud, you not only have the possibility of storing your images on UCloud, but you can also create storage groups and give different permissions. This is extremely helpful for large projects where the PI can divide images into smaller groups or projects and see statistics of how each project is progressing,” says Dr. Federica Lo Verso, Computational Scientist at the SDU eScience Center’s research support team. Federica has been the developer in charge of implementing the OMERO app on UCloud.

Key Features of OMERO

Image Management
OMERO serves as a secure repository for diverse imaging datasets, ensuring that your data is organised and accessible from anywhere.

Multi-Format Support
Powered by Bio-Formats, OMERO supports over 150 proprietary and open-source image formats, making it a versatile choice for labs with varied imaging technologies.

Secure Collaboration
OMERO includes robust permission controls, enabling safe sharing and collaboration across teams and institutions.

Data Visualisation and Analysis
Built-in tools allow users to display, annotate, and explore image data interactively, enhancing interpretation and insight.

Publication Tools
OMERO.figure simplifies the creation of high-quality figures for presentations and publications, directly from your image datasets.

Image source

Why Use OMERO on UCloud?

Deploying OMERO on UCloud offers significant advantages:

Scalable Storage Resources
UCloud provides access to vast and flexible storage, ideal for handling the large datasets typical in biological imaging.

Suitable for Sensitive Data
UCloud’s secure architecture makes it well-suited for handling sensitive or confidential research data, ensuring compliance with data protection regulations.

Integrated Research Environment
OMERO on UCloud integrates seamlessly with other tools and services, supporting collaborative and reproducible research.

Accessibility and Performance
With UCloud’s high-performance infrastructure, researchers can access and process their data efficiently from anywhere.

Whether you’re managing terabytes of microscopy images or preparing figures for your next publication, OMERO on UCloud is a game-changer for biological research. Start exploring its capabilities today and elevate your data management workflow.

Tags Interactive HPC, Supercomputing

UCloud 2025.4.0 Release

Post author By line
Post date August 15, 2025

Command palette, integrated editor & terminal, faster file search and a simpler Syncthing UI. Today we’re releasing UCloud 2025.4.0! This update introduces a keyboard-driven command palette, a built-in text editor, an integrated terminal, a fully revamped file search, and a simplified Syncthing experience.

While these new tools will be the most visible changes for users, they are also part of a broader, long-term effort to refine UCloud’s foundations.

“Software development on UCloud started back in 2017. Since then, we have make many architectural and design changes. This progress inevitably caused a build-up of inconsistensies and inefficiencies. With our understanding of the goal now being a lot clearer than in 2017, the aim is to simplify the internal code structure and get rid of old legacy code to ensure stability, performance and faster development of the service.”

Dan Sebastian Thrane, Senior Software Architect for Research Infrastructure

With that context in mind, here’s a closer look at what’s new in UCloud 2025.4.0.

Navigation and shortcuts

We have added a fast way to move around UCloud without leaving the keyboard.

Keyboard-first workflow: Use it to access shortcuts to common actions, jump to your favorite applications, or quickly switch projects.
Command palette: Open the palette with Ctrl+P on Windows/Linux or Cmd+P on macOS.

Files and search

Finding things is now much quicker and more reliable and edit text files directly in UCloud without needing to start a job.

Revamped file search: File search has been completely rebuilt and is now significantly faster. This requires your files to be indexed. Indexing of files may take up to 24 hours after the file has been created.
Edit-in-place: Double-click any text file in the file browser to open it in the new integrated editor.
Syntax highlighting: The editor supports basic syntax highlighting.
Tabs & file tree: Work across multiple tabs and use the sidebar to navigate your project’s file tree quickly.

Integrated terminal

A lightweight terminal is now available across Kubernetes-based service providers.

Auto-shutdown: Sessions shut down automatically when inactive. (Do not use it for long-running or background tasks.)
Availability: The integrated terminal is now available on the SDU/K8s provider. It works similar to the terminal released last year for the SDU/Hippo provider. You can open a terminal in any folder via “Open terminal” button.
For small tasks: Ideal for quick edits and running common utilities.
Resource limits: The terminal has low CPU and memory and (is not intended for running jobs).

Jobs and Syncthing

New real-time metrics have been added to jobs and the Syncthing has been made easier to understand and operate.

Simplified interface: A cleaner, more approachable Syncthing UI for managing synchronization.
Real-time job metrics: You can now monitor real time metrics about CPU, memory and GPU utilization from the job view.

Improved data transfer

The integration module has received a large upgrade to the upload protocol.

This protocol is used both by end-users and between service providers.
Data transfer between the SDU and AAU provider is now much easier and faster.
The upload protocol is smart and attempts to avoid sending files which are already present. It also utilizes multiple TCP streams and threads to maximize performance.

As always, you can visit UCloud’s documentation at https://docs.cloud.sdu.dk for more information.

Interactive HPC Teaching Tutorial Webinars & Tutorials - video

Webinar Recording: Transcribing and editing audio transcriptions with Transcriber and Speech Analyzer apps

Post author By line
Post date July 11, 2025

In this video we will guide you through the complete pipeline of transcribing audio files from speech to text and editing and classifying transcription segments.

In this session, you’ll learn how to:

Use Transcriber for transcribing audio/video files. Transcriber is based on Open AI’s Whisper language model. The app can transcribe speech audio to text in various formats and uses the WhisperX package to perform speaker recognition.
Navigate the new, simple, drag and drop Transcriber user interface to make it easier for you to use AI to transcribe audio files.
Edit and classify the transcriptions with Speech Analyzer. Speech Analyzer is an application built on top of Label Studio, specifically optimized for dialogue analysis. It enables you to label, edit, and annotate transcriptions generated using Transcriber.
Perform a comprehensive dialogue analysis on UCloud involving transcribing audio files using Transcriber, followed by transcription analysis with Speech Analyzer.

All workflows will be executed inside a UCloud project environment with access to GPU resources.

Target audience: Researchers across all Departments, particularly Digital Humanities and Social Science, Students, AI interested.

Interactive HPC Supercomputing Tutorial UCloud Webinars & Tutorials - video Workshop

Webinar recording: Fine-Tuning and Deploying Large Language Models

Post author By line
Post date July 7, 2025

In this video we will guide you through the complete pipeline of fine-tuning large language models (LLMs) for specialised tasks such as medical question-answering using NeMo Framework and Triton Inference Server.

Prepare and preprocess open-source datasets for fine-tuning.
Apply Parameter-Efficient Fine-Tuning (PEFT) using LoRA with NVIDIA NeMo Framework.
Deploy optimised LLMs using NVIDIA Triton Inference Server and TensorRT-LLM.
Generate a synthetic Q&A dataset using Label Studio connected to a live inference backend.
Fine-tune and evaluate your customised LLM for domain-specific applications.

All workflows will be executed inside a UCloud project environment with access to GPU resources.

Target audience: Machine learning practitioners, researchers, and engineers interested in LLM customisation, domain adaptation, or scalable model deployment.

Technical Level: Intermediate to Advanced.

Notebooks: https://github.com/emolinaro/ucloud-workshop-28-05-2025

Application Interactive HPC Supercomputing Teaching Tutorial Webinars & Tutorials - video

Webinar recording: UCloud courses hands-on

Post author By line
Post date July 7, 2025

In this video we will go through the process of developing a course on UCloud, using UCloud Courses – a tool for hosting and managing university courses on UCloud. Want to know more about this new feature, check out our webinar recording introducing UCloud Courses.

Introduction

00:00 – Outline of the workshop agenda
03:10 – Introduction to the UCloud Courses concept
04:23 – Advantages of using UCloud Courses
05:56 – Outline of the steps involved in developing a UCloud Course
09:54 – Requesting and planning a UCloud Course
10:20 – Example of the planning and structuring of an existing UCloud Course

Developing a UCloud Course step by step

Setting up the environment
15:53 – Preliminary remarks
17:15 – Showing the existing UCloud Course that will be re-developed in the workshop
18:02 – Showing the UCloud-Courses GitHub repository
18:40 – Software requirements and -recommendations
20:22 – Short introduction to Git and GitHub
24:27 – Cloning the repository
25:15 – Creating a working branch
26:12 – What should go in the UCloud-Courses repository and what shouldn’t
27:24 – Installing the required dependencies

Creating a new UCloud Course and modifying the templates
28:00 – Creating a UCloud Course using a prepared script
30:31 – Walk-through of the different auto-generated files/folders in the course folder
33:35 – Short introduction to Docker
38:15 – Modifying the Dockerfile
42:32 – What should go in the UCloud-Courses repository and what shouldn’t
45:28 – Building the Docker image locally using a prepared script
47:54 – Modifying the starting script
01:00:50 – What can, shouldn’t, and mustn’t be changed while the course is running

Building and testing the course locally
01:05:00 – Re-building the Docker image locally using a prepared script
01:05:25 – Running a Docker container locally using a prepared script
01:10:58 – Opening the JupyterLab interface on localhost

Finalising the course app
01:12:48 – Opening a pull request and requesting code review
01:19:44 – Testing the course app on UCloud before it’s deployed

Closing remarks

01:25:44 – Reusing/updating an existing UCloud course
01:28:19 – The financial model in brief
01:29:38 – Useful links and resources

UCloud-Courses GitHub repository:
https://github.com/SDU-eScience/UCloud-Courses

README in the UCloud-Courses GitHub repository:
https://github.com/SDU-eScience/UCloud-Courses/blob/main/README.md

Wiki page in the UCloud-Courses GitHub repository:
https://github.com/SDU-eScience/UCloud-Courses/wiki

The UCloud course app redeveloped in the workshop (UCloud login required):
https://cloud.sdu.dk/app/jobs/create?app=nlp-demo-course_147222U005

eScience Servicedesk (point of first contact):
https://support.escience.sdu.dk/

Application Interactive HPC

New App: AlphaFold 3 – a game-changing tool for predicting protein structure

Post author By line
Post date July 4, 2025

AlphaFold 3, the new AI model developed by Google DeepMind and Isomorphic Labs, can predict the structures of complexes formed by different biomolecules, including protein-DNA, protein-RNA, and protein-ligand interactions.

AlphaFold 3 deployment on Interactive HPC – UCloud is optimized for use with GPUs, and includes the necessary multiple genetic (sequence) protein and RNA databases to run.

Beside the Batch Mode, additional app flavors (Lab Mode and Visualization Mode) allow respectively to use the software in a JupyterLab environment and to visualize protein structures obtained by running AlphaFold 3.

AlphaFold 3 has been implemented on Interactive HPC – UCloud to support specific type of work, e.g., accurate structure prediction and modelling, and is ready for use by all Interactive HPC – UCloud users.

For more details, visit the UCloud documentation for the app.