The HiDALGO2 Centre of Excellence is showcasing its latest research at the 16th International Conference on Parallel Processing and Applied Mathematics (PPAM 2026) in Poznań, Poland!
A central highlight of our participation is the Advancements of Global Challenges Applications (AGCA) workshop, which brings together cutting-edge work across wildfire dynamics, urban microclimates, data analytics placement, and hybrid HPC execution environments.
AGCA Workshop Sessions Schedule:
Session 1: Monday, August 31 · 11:00–12:40 · Track C
Session 2: Tuesday, September 1 · 15:50–18:20 · Track E
Our researchers will deliver eight technical presentations spanning plenary and workshop tracks – scroll down for hte abstracts:
Plenary Presentation
– Data-driven prediction of optimal MPI rank placement and node count for memory-bound CFD on AMD EPYC clusters, Marcin Lawenda, Łukasz Szustak, Aleksandra Krasicka
An explainable machine-learning surrogate to predict runtime and energy consumption for OpenFOAM CFD workloads across Zen 2 to Zen 5 AMD EPYC architectures.
AGCA Workshop Highlights
– WUI Fire Simulation with FDS/WFDS on EuroHPC Supercomputers, David Caballero, Luis Torres, Ángela Rivera
An end-to-end pipeline connecting procedural Unreal Engine 5 vegetation generation to FDS/WFDS fire simulations on EuroHPC systems (LUMI, MeluXina) and back to real-time volumetric rendering.
– Benchmarking Data-Local Analytics for Hybrid HPC/HPDA Scientific Workflows, by László Környei, Attila Ádám Balics, Nikolaos Chalvantzis, Vasiliki Kostoula, Dimitrios Tsoumakos
Evaluation of simulation-adjacent data placement and conversion for the Urban Air Pollution workflow on Komondor.
– Adaptive Scheduling of Hybrid HPC–FaaS Workflows with Multicriteria Decision Support, by Áron Ákos Németh, László Környei
A two-layer decision engine integrating Slurm, K3s/Nuclio, and MinIO to minimise latency for short workflow tasks using an on-premises FaaS layer.
– Ktirio Urban Building: Reproducible Scenario-Based City-Scale Energy Simulation on EuroHPC Systems, by Juliette Antonczak, Vincent Chabannes, Javier Cladellas, Gwennolé Chappron, Philippe Pincon, Christophe Prud’Homme
Scalable, reproducible building energy framework using Feel++, ReFrame-HPC, Apptainer, and Slurm for urban microclimate studies.
– Interpretable Patch‑Based Deep Learning for Wildfire Spread Prediction from Ensemble Simulations, by Marcin Lawenda, Aleksandra Krasicka, David Caballero, Luis Torres, Łukasz Szustak
Coupling geospatial analytics with deep learning (U-Net) and attribution methods to predict 30-minute wildfire evolution.
– High-Resolution Urban Climate Modelling via HPC: Application to Poznań, by Zoltán Horváth, Mátyás Constans
Demonstrating the RedSim framework running street-level scale simulations faster than real-time over a 6 km × 6 km domain in Poznań.
Scaling the EULAG solver for Grand Challenges problems on petascale machines, by Michal Kulczewski
Performance optimisations for the Fortran 77-based EULAG solver in the AQURES toolkit for micro-scale wind modelling.
🔗 Abstracts of the HiDALGO2 team: https://www.hidalgo2.eu/wp-content/uploads/2026/08/List-of-presentations-at-PPAM-–-AGCA-2026-HiDALGO2.docx
🔗 AGCA Workshop Details: https://users.man.poznan.pl/lawenda/ppam-agca/2026/
🔗 Full PPAM Programme: https://ppam.edu.pl/program#program