Sano Representatives at the High Performance Computing Users
HPC Users’ Conference - KUKDM 2026
Sano Centre for Computational Medicine had a strong presence at this year’s High Performance Computing Users HPC Users’ Conference – KUKDM 2026, where our researchers presented five posters highlighting current work in computational medicine, bioinformatics, and large-scale data analysis.
Sano Poster Presentations
The scientific posters presented by the Sano team reflected the breadth of the centre’s research in high-performance computing applications for medicine and biological sciences.
‘Benchmarking a Heterogeneous Metagenomics Pipeline on Ares and Helios Supercomputers’
The team benchmarked a heterogeneous metagenomics pipeline on Poland’s Ares and Helios supercomputers, demonstrating the capabilities of national HPC infrastructure for genomic data analysis.
HPC in Global Quantification and Personalization of Regulated Cardiovascular Model – a Cohort Study‘
Authors: Karolina Tlałka, Ian Halliday, Andrew Narracott, Maciej Malawski
The presentation demonstrated the application of high-performance computing to personalized cardiovascular modeling across patient cohorts, enabling global quantification and individualization of regulated cardiovascular models.
Automating CFD Workflows on HPC with PyFluent: from Batch Simulations to Large-Scale Study‘
Autors: Karol Zając, Magdalena Otta, Maciej Malawski, Andrew Narracott
This poster presented an approach for automating CFD workflows using PyFluent and HPC systems. The proposed solution enables scalable and reproducible simulation campaigns through Python-based workflow automation. By leveraging HPC resources, the workflow significantly reduces computation time and supports large-scale studies involving thousands of simulations. The work also demonstrates the potential of automated CFD pipelines for scalable and complex UQ and SA studies.
‘PubDB: an Institutional Publication Database for KPI Monitoring and Visibility‘
Authors: Taras Zhyhulin, Adam Nowak, Jan Meizner, Piotr Nowakowski
This work presented an institutional publication database designed to monitor and enhance the visibility of scientific output.
‘Evaluating the Stability of Deep Neural Networks for Brain Microstructural Parameter Estimation under Varying Diffusion Gradient Schemes‘
Authors: Dominika Ciupek, Maciej Malawski, Tomasz Piciąk
The study evaluated the stability of deep neural networks for estimating brain microstructural parameters across different diffusion gradient schemes.
‘High‑Throughput Statistical Shape Modelling of Venous Anatomy Using Containerised Deformetrica on Ares Supercomputer‘
Autors: Magdalena Otta, Piotr Kica, Maciej Malawski
This work presents containerised approaches to high-throughput statistical shape modelling of venous anatomy, leveraging the Ares supercomputer infrastructure for scalable and efficient analysis, particularly in the context of large cohorts of geometries prepared for downstream computational fluid dynamics (CFD) studies.
‘Using High Performance Computing for the Application of PINNs in Cardiovascular Models’
Authors: Elena Gómez del Pozo, Karolina Tlałka, Karol Zając, Maciej Malawski
The presentation explored the application of Physics-Informed Neural Networks (PINNs) in cardiovascular modeling using high-performance computing. Unlike classical neural networks, PINNs incorporate known physics as part of the loss term. In this project, several loss term combinations were evaluated — such as mass conservation and no-negative-volumes — studying which one is most effective for fitting cardiovascular models of different complexity.
Keynote Speakers
This strong poster presence was complemented by the participation of distinguished speakers closely connected to Sano and its partner network. Among the keynote lecturers was Dr Olav Zimmermann from the Jülich Supercomputing Centre in Germany, a representative of one of Sano’s partner institutions, who delivered the talk ‘The stubborn old protein folding problem in the era of ExaFlops, Qbits and AI’. Another keynote address was given by Dr Marian Bubak, Scientific Affairs Director at Sano and affiliated with ACK Cyfronet AGH in Kraków, who presented ‘Towards European Infrastructure for In Silico Medicine’.
About KUKDM Conference
KUKDM itself has been organized by ACK Cyfronet AGH since 2008 as an annual meeting for researchers using high-performance computing systems and software provided by Cyfronet. Its thematic scope includes large-scale computing and simulations, new algorithms in computer science, tools and techniques for HPC systems, artificial intelligence methods, and the analysis of large datasets. The 2026 edition brought together representatives of science, business, and public administration, and featured guest lectures, 31 presentations, and 12 posters by users of the Polish PLGrid infrastructure coordinated by ACK Cyfronet AGH. In addition to its scientific sessions, the conference offers training opportunities and consultations with the PLGrid support team, making it not only a forum for presenting results but also a practical space for improving the use of supercomputing in research.
