To meet the demands of the Beam Energy Scan phase-II (BES-II) program, the STAR experiment at the Relativistic Heavy Ion Collider (RHIC) developed a dual real-time framework consisting of a High Level Trigger (HLT) and an Express Data Production system (xProduction). The HLT operates online within the Data Acquisition (DAQ) chain on a dedicated multi-core CPU cluster with the option to offload compute-intensive kernels to Xeon Phi coprocessors. It uses parallelized algorithms, such as the Cellular Automaton (CA) Track Finder, to perform rapid tracking, vertexing, and event filtering. This allows it to select events of interest in real time and provide immediate feedback on detector and beam conditions. In contrast, the xProduction workflow runs concurrently and independently of the DAQ loop. It applies near offline-quality calibration and reconstruction within hours of data collection. The xProduction input is the express data stream, whose content can be enriched by HLT trigger/priority selections under DAQ/HLT resource constraints, and it uses the STAR calibration/conditions framework, incorporating online calibration/QA information when available. This enables early preliminary physics analysis, including the reconstruction of rare signals, such as hyperons and hypernuclei. It also provides collaboration-wide access to analysis-ready datasets. Together, the HLT and xProduction systems form a complementary architecture: the HLT performs online event selection while the xProduction chain delivers high-quality results within a short amount of time. This integrated framework has enabled the prompt reconstruction of the 5He hypernucleus with high statistical significance and the efficient processing of hundreds of millions of heavy-ion collision events. Its demonstrated scalability and robustness establish a model for future high-luminosity experiments requiring both online event filtering and rapid access to analysis-quality data.

Modern high-rate experiments require rare physics signatures to be identified in real time from continuous streams of reconstructed events under stringent data-throughput and storage constraints. We present a convolutional-neural-network-based trigger concept for selecting events associated with quark-gluon plasma (QGP) formation. Events are encoded as compact multidimensional histograms of reconstructed particle content, including particle species, momentum magnitude, and angular information. The method is first evaluated within the Parton-Hadron-String Dynamics (PHSD) framework, where microscopic QGP-related labels are available. As an independent validation, the same event representation and network architecture are applied to Ultra-relativistic Quantum Molecular Dynamics (UrQMD) simulations, providing a distinct description of the collision dynamics. Cross-checks between PHSD and UrQMD are used to assess the stability of the learned response against generator-dependent effects and to quantify model-transfer robustness. For realistic deployment, a lightweight C++ inference package, ANN4FLES, is employed at the physics-analysis stage after tracking and topology reconstruction. For Au+Au collisions at 30 AGeV, the classification accuracy decreases from 95.1

Machine learning algorithms have become essential tools in modern physics experiments, enabling the precise and efficient analysis of large-scale experimental data. The Compressed Baryonic Matter (CBM) experiment at the Facility for Antiproton and Ion Research (FAIR) demands innovative methods for processing the vast data volumes generated at high collision rates of up to 10 MHz. This study presents a deep-learning-based approach to enhance the signal/background (S/B) ratio for Λ particles within the Kalman Filter (KF) Particle Finder framework. Using the Artificial Neural Networks for First Level Event Selection (ANN4FLES) package of CBM, a multi-layer perceptron model was designed and trained on simulated data to classify Λ particle candidates as signal or background. The model achieved over 98% classification accuracy, enabling significant reductions in background—in particular, a strong suppression of the combinatorial background that lacks physical meaning—while preserving almost the whole Λ particle signal. This approach improved the S/B ratio by a factor of 10.97, demonstrating the potential of deep learning to complement existing particle reconstruction techniques and contribute to the advancement of data analysis methods in heavy-ion physics.

The CBM experiment at FAIR (GSI, Germany) is among the most significant upcoming projects in heavy-ion physics. It is designed to investigate the properties of dense baryonic matter under extreme conditions. A key feature of the experiment is the high interaction rate, reaching up to 1 0 7 collisions per second, resulting in the production of substantial volumes of experimental data that must be processed and analyzed in real time. To meet this computational challenge, the CBM experiment employs a high-performance tracking algorithm based on the Cellular Automaton for track finding and the Kalman Filter for track fitting. The algorithm is designed for efficient parallel execution on modern multicore and GPU architectures. We evaluate the performance and energy efficiency of the Kalman Filter-based fitting algorithm on both CPUs and GPUs. The GPU implementation demonstrates up to a threefold improvement in energy efficiency, resulting in a proportional reduction in power consumption and associated CO2 emissions during data processing. These results highlight the significance of energy-efficient computing in high-rate heavy-ion experiments. The analysis provides a quantitative estimate of the carbon footprint associated with track reconstruction and demonstrates how hardware choices influence overall emissions in large-scale data processing workflows.

The CBM experiment at FAIR is being designed to study heavy-ion collisions at extremely high interaction rates (up to 10 MHz) and high track multiplicities (up to 1000). Since the beam in the CBM experiment will have no bunch structure, but continuous, the groups of events may be close or overlapped in time. Measurements in this case will be 4D (x, y, z, t). Thus, the reconstruction of time slices rather than events will be needed. In addition to such high input rate and complicated event topology, the full event reconstruction and selection will be done at the First Level Even Selection (FLES) stage. In this respect, both the speed of the reconstruction algorithms and their efficiency are crucial. The Cellular Automaton (CA) [1] track finder is fast and robust and thereby is used both for the online and offline track reconstruction in the CBM experiment. The algorithm creates short track segments in each three neighbouring stations, then combines them into track-candidates and selects the best tracks according to the maximum length and minimum χ criteria. The algorithm was further optimized for the case of high track multiplicity with respect to time: additional sorting of found hits according to 2dimensional (y, z) grid was introduced in order to speed up the search for the next hit. The standalone FLES pack-