For years, ACC Cyfronet AGH has been one of the computing centers within the WLCG (Worldwide LHC Computing Grid), the global computing network of the Large Hadron Collider (LHC). The network utilizes computing resources from many different units, enabling centralized commissioning and distribution to dedicated processors of tasks that would otherwise be beyond the capabilities of a single center. In the CMS experiment, these are computational campaigns involving data reconstruction and Monte Carlo simulations that model particle interactions in the detector and the resulting responses of the measurement systems.
Since 2025, Cyfronet's scope of operations has further expanded. Our infrastructure has been used to store data generated between 2022 and 2025 by one of the CMS subsystems, namely the PPS, which is the primary data source for researchers at AGH University of Krakow.
Due to the storage of hundreds of petabytes of data per year (after filtering), data from previous periods is quickly transferred to magnetic tapes at CERN. Therefore, thanks to Cyfronet's support, researchers have easier, immediate access to data, which is a significant advantage over having to extract it from CERN's archive media. Data stored in Cyfronet allows scientists to calculate detector calibrations and ensure adequate data quality.
Furthermore, software packages used by Polish researchers in the CMS experiment are installed on our clusters, and the PLGrid Helpdesk team ensures their updates and full availability.
Employees of ACC Cyfronet AGH co-authored an article published in Nature
Cyfronet's contribution extends beyond providing and operating the infrastructure needed for data processing at CERN. Thanks to their specializations, Cyfronet's R&D staff actively support experiments conducted at the LHC.
At the end of 2025, the results of the CMS experiment were published in the prestigious journal Nature. A condensed description of the research can be found on the AGH Faculty of Computer Science website.
The list of co-authors includes Prof. AGH Maciej Malawski, and dr Leszek Grzanka, both employees of the AGH Faculty of Computer Science in Kraków and specialists employed at the ACC Cyfronet AGH.
Our AGH team's contribution began with the TOTEM experiment, and in the following years we also joined the CMS experiment. At that time, we were working on distributed (parallel) data analysis, expanding the ROOT platform widely used at CERN.
- says dr Leszek GrzankaFor testing, we also used the PLGrid infrastructure, in particular the Prometheus and Ares supercomputers operating at Cyfronet. The ROOT platform was later used to conduct all the data analysis needed to obtain the CMS experiment results published in Nature.
For those interested in the ROOT platform, we recommend the following publications: [1], [2].
It is worth remembering that Cyfronet's collaboration with CERN has been ongoing for many years. A prime example was the European CrossGrid project (Prof. Michał Turała, Dr. Marian Bubak), coordinated by Cyfronet from 2002 to 2005. As part of the project, in collaboration with, among others, physicists from the Institute of Nuclear Physics of the Polish Academy of Sciences in Krakow, systems for distributed data processing were developed. The direct result of this work was the creation of the PLGrid infrastructure and modern computing clusters in Cyfronet.
- adds Prof. Maciej Malawski
Currently, data analysis using machine learning is becoming increasingly important for CERN, which fits very well into the current topic of research on artificial intelligence methods.
The yaptide Platform
An additional result of the collaboration with CERN is the yaptide platform for commissioning calculations of matter-interaction simulations. It is integrated with PLGrid services, allowing Polish scientists to access it conveniently through the PLGrid portal. Currently, computational tasks are being performed on the Ares supercomputer installed at Cyfronet.
Yaptide users have access to radiation-matter interaction simulators developed at CERN, including the Geant4 package, which was used in the research described in Nature, and the Fluka package.
