Topics

Research (Meta)data Management at PNM. A Story of Vacuum cars, QR codes and Four-dimensional Images

With rapid advances in detector technology and computing resources, modern electron microscopes produce massive amounts of data. Moreover, advanced sample preparation techniques often involve various steps utilizing multiple devices each generating data as well. Therefore, data management should be considered a crucial part of today’s research infrastructure. This led to the publication of guiding principles for scientific data management describing four fundamental cornerstones: Findabilty, Accessibility, Interoperability and Re-usability (FAIR).

The University of Vienna is running one of the world's top electron microscopes embedded in a worldwide unique ultra-high vacuum system built and operated by the workgroup Physics of Nanostructured Materials (PNM). The high degree of integration of various advanced devices in the system made it necessary to deal with (meta)data management from the beginning. Therefore, PNM created a system capable of handling most of the usual data management tasks in an automatic fashion without getting in the way. This is implemented by a file storage infrastructure automatically collecting data from various experimental devices including electron microscopes, optical microscopes, sample preparation devices and mechanical testing systems. Raw research data is considered immutable and is therefore stored read-only on a scalable distributed storage system. The data can be accessed via a front-end server by a secure connection using asymmetric cryptography.

Metadata is handled via an electronic lab-book system. This system is accessible via standard web browsers and is used for documenting experimental steps and persons, devices and samples involved in carrying them out. The lab-book system is also used for keeping a database of samples allowing an automatic sample-tracking workflow based on QR codes and a distributed database populated from the lab-book entries.

To setup a future-proof, independent system and to avoid possible vendor lock-in scenarios, care was taken to only use free and open-source software.

Speaker: Clemens Mangler

Implementing Electronic Lab Notebooks in the Life Sciences

Electronic lab notebooks (ELNs) are becoming increasingly important in digitized research environments, and their adoption is essential for improving data integrity, collaboration, and reproducibility. In my talk, I will present the selection process undertaken at the Centre for Microbiology and Environmental Systems Science (CeMESS) to choose an ELN that meets the diverse needs of our research community. Additionally, I will outline the upcoming implementation plan for the chosen ELN, RSpace, at CeMESS and the Faculty of Life Sciences.

Speaker: Michael Feichtinger

EuroCC / EuroHPC-JU

EuroCC Austria plays a key role in the European landscape of supercomputing, connecting Austria to the broader ambitions of the EuroHPC Joint Undertaking (EuroHPC-JU). The EuroHPC-JU is a collaborative initiative aimed at positioning Europe as a leader in high-performance computing (HPC), with a network of world-class supercomputers that drive scientific and industrial advancements.

Through this, Austria participates in utilizing these advanced systems, contributing to the shared European goal of cutting-edge developments in HPC.

At EuroCC Austria, we provide comprehensive services designed to meet the needs of both academia and non-academic sectors. Our consulting services cover every stage of a project, offering tailored advice to industries, research institutions, and SMEs. Whether it’s optimizing an HPC related business process, conducting advanced research, or implementing large-scale simulations, we provide the expertise to make it happen. Our services are open to a wide range of sectors, from technical disciplines to industries like healthcare, manufacturing, and beyond, encouraging also non-academic bodies to harness the power of HPC.

An essential part of our offering is facilitating access to EuroHPC systems. These supercomputers represent some of the most powerful computational resources in the world, capable of tackling enormous data sets, training complex AI models, and accelerating scientific simulations.

Artificial Intelligence (AI) is another area where we see tremendous development and need for computational resources. AI is not just for tech fields like physics or informatics; it is being used in a broad range of disciplines, creating transformative effects across various sectors. We showcase AI use-cases from Austrian universities, to demonstrate its far-reaching potential. These examples highlight interdisciplinary collaboration where AI is being applied in fields such as biology, social sciences, economics, and even the humanities. This breadth of application demonstrates how AI is evolving into a vital tool for all researchers, not just those in traditionally technology-focused areas.

To encourage more widespread use of AI, we invite all researchers, regardless of technical expertise, to explore the developments happening right now . AI is rapidly becoming a powerful tool in research across the board, offering new insights and capabilities that can revolutionize fields as diverse as medicine, sociology, and the arts. Our mid-term goal is to make AI accessible to all academic users, creating services and tools that simplify its usage. By offering training programs and user-friendly platforms, we aim to make AI as easy to use as possible for researchers from any discipline, enabling them to fully harness its potential without needing deep technical knowledge.

Contact us at info@eurocc-austria.at and join us in shaping the future of high-performance computing and AI.

Speaker: Markus Stöhr

Vienna Scientific Cluster and MUSICA

The Vienna Scientific Cluster (VSC) and the MUlti Site Computer Austria (MUSICA) are Austria's leading high-performance computing (HPC) infrastructures. These clusters are designed to meet the diverse computational needs of scientific research across multiple disciplines, including physics, chemistry, biology, engineering, and artificial intelligence (AI). Together, VSC and MUSICA provide state-of-the-art hardware, a well-developed software ecosystem, a fair access policy, and comprehensive training programs to enable Austrian researchers to address complex computational problems.

VSC, with its currently operated systems VSC-4 and VSC-5, which offer over 100,000 AMD EPYC cores, more than 38,000 Intel Cores, high-speed Infiniband interconnects, and GPU nodes featuring 210 NVIDIA A40 and A100 accelerators, supports large-scale simulations and data-intensive projects. The upcoming system MUSICA, designed for data-driven research and artificial intelligence (AI) applications, complements VSC with specialized hardware optimized for machine learning and AI workloads.

Access to both clusters is managed through a fair-share model, with applications evaluated based on peer reviewed project with high computational demands. Researchers from Austrian universities and research institutions which are part of the VSC or MUSICA consortium can apply for resources via a web portal.

Both clusters provide a robust software infrastructure, with Linux-based environments, optimized libraries, and support for parallel computing frameworks such as MPI, OpenMP, and CUDA. Domain-specific software for disciplines like computational chemistry, physics, and AI is also available.

To ensure efficient use of the resources, VSC and MUSICA offer comprehensive training programs, including introductory courses for new users, advanced workshops on parallel programming, GPU computing, and performance optimization, along with domain-specific training tailored to the needs of various scientific disciplines. These programs ensure that users can fully leverage the capabilities of both HPC systems for their research.

More information about accessing the systems can be found on the VSC homepage https://vsc.ac.at//access/ or you can contact us directly via support@vsc.ac.at.

Speaker: Jan Zabloudil

Funding for Research with Azure Services

In order to support research activities in Azure, the ZID offers financial support. After successfully funding numerous projects in 2024, the funding programme will be continued in 2025. Applications are possible between 28.10. and 31.12.2024.

Speaker: Fabian Jusufi

Data Center Strategy of the ZID. New Challenges to Infrastructure when using AI

In many fields machine learning and AI has become a vital part of research. The hardware that is needed for this type of computing resources poses a lot of new challenges to IT infrastructure. This presentation gives a brief overview of those challenges and the strategies of the ZID to cope with them.

Speaker: Ulrich Kiermayr

Page tree

Opening Remarks