Overview of fast ML for detectors and control
Presentation on fast machine learning usage in high energy physics for detectors and control
Rankin, Dylan Sheldon [University of Pennsylvania]↗
SEARCH · Engineering Papers
Results for “High Energy Physics”
Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.
Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.
Improving unfolding and systematic uncertainty estimation using generative diffusion networks (Final Technical Report)
This final technical report summarizes the key accomplishments on the unfolding using diffusion model project, a DOE award received by PI Pierre-Hugues Beauchemin at Tufts University. This project main goal was to investigate the potential of diffusion models for unfolding experimental High Energy Physics data from detector effects while controlling systematics uncertainties. The project accomplished its goals by completing the following objectives: 1) Performing an object-by-object, event-by-event unfolding of various kinematic distributions reconstructed from detector data in HEP in a way that keeps correlations between unfolded observables while demonstrating competitive performance compared to standard algorithms used in the field; 2) Address the generalization problem by developing an unfolding algorithm capable to correctly infer the underlying distributions of observables and processes never seen before, while controlling the dominant theoretical uncertainties affecting the process, therefore increasing the effectiveness, the precision, and the applicability of the developed algorithm; 3) Understand the theoretical foundations between the developed algorithm so to extend it to applications beyond experimental HEP, for broader benefits to the society. This report provides an overview of the accomplishments related to each of these key objectives.
72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
Tachyon: Intelligent Multi-Scale Modeling of Distributed Resilient Infrastructure and Workflows for Data Intensive HEP Analyses
The DOE High Energy Physics (HEP) program in Neutrino and Collider science drives data-intensive science and simulation on extreme-scale platforms. Modeling and optimizing the complex distributed components from experimental to leadership computing facilities are essential for HEP workflows to achieve required response times and resilience under various conditions. Tachyon proposes a framework for scalable modeling, simulation, and validation of key performance characteristics for the distributed infrastructure between FNAL and ALCF, along with associated HEP workflows.
Carothers, Chris [Rensselaer Poly.]↗
Optimizations of a Rectilinear Cooling Channel for a Future Muon Collider
Muon colliders require significant beam cooling to achieve the luminosity needed for high-energy physics experiments. Ionization cooling has emerged as a promising solution. This study optimizes a rectilinear muon cooling channel using a multi-objective optimization framework that integrates beam dynamics simulations. We present novel optimizations of final 6D emittance versus total system length as well as those confirming the theoretical trade-offs between transverse and longitudinal emittance. Our results optimizing all stages of the system simultaneously surpass performance benchmarks reported in the literature, demonstrating possible ways to improve the efficiency of such a cooling system.
Zhang, Aubrey [U. Chicago (main)]↗
Tachyon: Intelligent Multi-Scale Modeling of Distributed Resilient Infrastructure and Workflows for Data Intensive HEP Analyses
The DOE High Energy Physics (HEP) program in Neutrino and Collider science drives data-intensive science and simulation on extreme-scale platforms. Modeling and optimizing the complex distributed components from experimental to leadership computing facilities are essential for HEP workflows to achieve required response times and resilience under various conditions. Tachyon proposes a framework for scalable modeling, simulation, and validation of key performance characteristics for the distributed infrastructure between FNAL and ALCF, along with associated HEP workflows.
Carothers, Chris [Rensselaer Poly.]↗
Optimizations of a Rectilinear Cooling Channel for a Future Muon Collider
Muon colliders require significant beam cooling to achieve the luminosity needed for high-energy physics experiments. Ionization cooling has emerged as a promising solution. This study optimizes a rectilinear muon cooling channel using a multi-objective optimization framework that integrates beam dynamics simulations. We present novel optimizations of final 6D emittance versus total system length as well as those confirming the theoretical trade-offs between transverse and longitudinal emittance. Our results optimizing all stages of the system simultaneously surpass performance benchmarks reported in the literature, demonstrating possible ways to improve the efficiency of such a cooling system.
Zhang, Aubrey [U. Chicago (main)]↗
Developments in Performance and Portability for MadGraph5_aMC@NLO
Event generators simulate particle interactions using Monte Carlo techniques, providing the primary connection between experiment and theory in experimental high energy physics. These software packages, which are the first step in the simulation worflow of collider experiments, represent approximately 5 to 20% of the annual WLCG usage for the ATLAS and CMS experiments. With computing architectures becoming more heterogeneous, it is important to ensure that these key software frameworks can be run on future systems, large and small. In this contribution, recent progress on porting and speeding up the Madgraph5_aMC@NLO event generator on hybrid architectures, i.e. CPU with GPU accelerators, is discussed. The main focus of this work has been in the calculation of scattering amplitudes and "matrix elements", which is the computational bottleneck of an event generation application. For physics processes limited to QCD leading order, the code generation toolkit has been expanded to produce matrix element calculations using C++ vector instructions on CPUs and using CUDA for NVidia GPUs, as well as using Alpaka, Kokkos and SYCL for multiple CPU and GPU architectures. Performance is reported in terms of matrix element calculations per time on NVidia, Intel, and AMD devices. The status and outlook for the integration of this work into a production release usable by the LHC experiments, with the same functionalities and very similar user interfaces as the current Fortran version, is also described.
Valassi, Andrea↗
Block encoding bosons by signal processing
Block Encoding (BE) is a crucial subroutine in many modern quantum algorithms, including those with near-optimal scaling for simulating quantum many-body systems, which often rely on Quantum Signal Processing (QSP). Currently, the primary methods for constructing BEs are the Linear Combination of Unitaries (LCU) and the sparse oracle approach. In this work, we demonstrate that QSP-based techniques, such as Quantum Singular Value Transformation (QSVT) and Quantum Eigenvalue Transformation for Unitary Matrices (QETU), can themselves be efficiently utilized for BE implementation. Specifically, we present several examples of using QSVT and QETU algorithms, along with their combinations, to block encode Hamiltonians for lattice bosons, an essential ingredient in simulations of high-energy physics. We also introduce a straightforward approach to BE based on the exact implementation of Linear Operators Via Exponentiation and LCU (LOVE-LCU). We find that, while using QSVT for BE results in the best asymptotic gate count scaling with the number of qubits per site, LOVE-LCU outperforms all other methods for operators acting on up to qubits, highlighting the importance of concrete circuit constructions over mere comparisons of asymptotic scalings. Using LOVE-LCU to implement the BE, we simulate the time evolution of single-site and two-site systems in the lattice theory using the Generalized QSP algorithm and compare the gate counts to those required for Trotter simulation.
Kane, Christopher F↗
Tomographic Muon Imaging of the Great Pyramid of Giza
The pyramids of the Giza plateau have fascinated visitors since ancient times and are the last of the Seven Wonders of the ancient world still standing. It has been half a century since Luiz Alvarez and his team used cosmic-ray muon imaging to look for hidden chambers in Khafre’s Pyramid. Advances in instrumentation for High-Energy Physics (HEP) allowed a new survey, ScanPyramids, to make important new discoveries at the Great Pyramid (Khufu) utilizing the same basic technique that the Alvarez team used, but now with modern instrumentation. Exploring the Great Pyramid Mission plans to field a very large muon telescope system that will be transformational with respect to the field of cosmic-ray muon imaging. We plan to field a telescope system that has upwards of 100 times the sensitivity of the equipment that has recently been used at the Great Pyramid, will image muons from nearly all angles, and will, for the first time, produce a true tomographic image of such a large structure.
47 OTHER INSTRUMENTATION↗
Ps and Qs: Quantization-Aware Pruning for Efficient Low Latency Neural Network Inference
Efficient machine learning implementations optimized for inference in hardware have wide-ranging benefits, depending on the application, from lower inference latency to higher data throughput and reduced energy consumption. Two popular techniques for reducing computation in neural networks are pruning, removing insignificant synapses, and quantization, reducing the precision of the calculations. In this work, we explore the interplay between pruning and quantization during the training of neural networks for ultra low latency applications targeting high energy physics use cases. Techniques developed for this study have potential applications across many other domains. We study various configurations of pruning during quantization-aware training, which we term quantization-aware pruning, and the effect of techniques like regularization, batch normalization, and different pruning schemes on performance, computational complexity, and information content metrics. We find that quantization-aware pruning yields more computationally efficient models than either pruning or quantization alone for our task. Further, quantization-aware pruning typically performs similar to or better in terms of computational efficiency compared to other neural architecture search techniques like Bayesian optimization. Surprisingly, while networks with different training configurations can have similar performance for the benchmark application, the information content in the network can vary significantly, affecting its generalizability.
46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗
A Comparison between Invariant and Equivariant Classical and Quantum Graph Neural Networks
Machine learning algorithms are heavily relied on to understand the vast amounts of data from high-energy particle collisions at the CERN Large Hadron Collider (LHC). The data from such collision events can naturally be represented with graph structures. Therefore, deep geometric methods, such as graph neural networks (GNNs), have been leveraged for various data analysis tasks in high-energy physics. One typical task is jet tagging, where jets are viewed as point clouds with distinct features and edge connections between their constituent particles. The increasing size and complexity of the LHC particle datasets, as well as the computational models used for their analysis, have greatly motivated the development of alternative fast and efficient computational paradigms such as quantum computation. In addition, to enhance the validity and robustness of deep networks, we can leverage the fundamental symmetries present in the data through the use of invariant inputs and equivariant layers. In this paper, we provide a fair and comprehensive comparison of classical graph neural networks (GNNs) and equivariant graph neural networks (EGNNs) and their quantum counterparts: quantum graph neural networks (QGNNs) and equivariant quantum graph neural networks (EQGNN). The four architectures were benchmarked on a binary classification task to classify the parton-level particle initiating the jet. Based on their area under the curve (AUC) scores, the quantum networks were found to outperform the classical networks. However, seeing the computational advantage of quantum networks in practice may have to wait for the further development of quantum technology and its associated application programming interfaces (APIs).
Forestano, Roy T. (ORCID:0000000203552076)↗
Overview of CMOS Sensors for Future Tracking Detectors
Depleted Complementary Metal-Oxide-Semiconductor (CMOS) sensors are emerging as one of the main candidate technologies for future tracking detectors in high luminosity colliders. Their capability of integrating the sensing diode into the CMOS wafer hosting the front-end electronics allows for reduced noise and higher signal sensitivity, due to the direct collection of the sensor signal by the readout electronics. They are suitable for high radiation environments due to the possibility of applying high depletion voltage and the availability of relatively high resistivity substrates. The use of a CMOS commercial fabrication process leads to their cost reduction and allows faster construction of large area detectors. In this contribution, a general perspective of the state of the art of CMOS detectors for High Energy Physics experiments is given. The main developments carried out with regard to these devices in the framework of the CERN RD50 collaboration are summarized.
46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗
A Comprehensive Survey of Readout Strategies for SiPMs Used in Nuclear Imaging Systems
Silicon photomultipliers (SiPMs) offer advantages such as lower relative cost, smaller size, and lower operating voltages compared to photomultiplier tubes. A SiPM’s readout circuit topology can significantly affect the characteristics of an imaging array. In nuclear imaging and detection, energy, timing, and position are the primary characteristics of interest. Nuclear imaging has applications in the medical, astronomy, and high energy physics fields, making SiPMs an active research area. This work is focused on the circuit topologies required for nuclear imaging. We surveyed the readout strategies including the front end preamplification topology choices of transimpedance amplifier, charge amplifier, and voltage amplifier. In addition, a review of circuit topologies suitable for energy, timing, and position information extraction was performed along with a summary of performance limitations and current challenges.
42 ENGINEERING↗
LGAD-Based Silicon Sensors for 4D Detectors
Low-Gain Avalanche Diodes (LGAD) are a class of silicon sensors developed for the fast detection of Minimum Ionizing Particles (MIPs). The development was motivated by the need of resolving piled-up tracks of charged particles emerging from several vertexes originating from the same bunch-crossing in High-Energy Physics (HEP) collider experiments, which, however, are separated not only in space but also in time by a few tens of picoseconds. Built on thin silicon substrates and featuring an internal moderate gain, they provide fast signals for excellent timing performance, which are therefore useful to distinguish the different tracks. Unfortunately, this comes at the price of poor spatial resolution. To overcome this limitation, other families of LGAD-based silicon sensors which can deliver in the same substrate both excellent timing and spatial information are under development. Such devices are, to name a few, capacitively coupled LGADs (AC-LGAD), deep-junction LGADs (DJ-LGAD) and trench-isolated LGADs (TI-LGADs). These devices can be fabricated by even small-scale research-focused clean rooms for faster development within the scientific community. However, to scale up production, efforts towards integrating these sensor concepts in CMOS substrates, with the obvious advantage of the possibility of integrating part of the read-out electronics in the same substrate, have begun.
spatial resolution↗
Novel Sensors for Particle Tracking: a Contribution to the Snowmass Community Planning Exercise of 2021
Five contemporary technologies are discussed in the context of their potential roles in particle tracking for future high energy physics applications. These include sensors of the 3D configuration, in both diamond and silicon, submicron-dimension pixels, thin film detectors, and scintillating quantum dots in gallium arsenide. Drivers of the technologies include radiation hardness, excellent position, vertex, and timing resolution, simplified integration, and optimized power, cost, and material.
Detectors↗
Programs Enabling Deep Technology Transfer from National Labs
To maximize the technology transfer potential, it is important to create an ecosystem for the inventors to adapt the technologies developed for basic science to successful commercial ventures. In this white paper we present a brief overview of technology transfer programs at high energy physics (HEP) laboratories with a focus on the programs at Fermilab and KEK, and identify opportunities and recommendations for increasing partnerships and commercialization at HEP-centric laboratories.
72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
TF07 Snowmass Report: Theory of Collider Phenomena
Theoretical research has long played an essential role in interpreting data from high-energy particle colliders and motivating new accelerators to advance the energy and precision frontiers. Collider phenomenology is an essential interface between theoretical models and experimental observations, since theoretical studies inspire experimental analyses while experimental results sharpen theoretical ideas. Here, this report -- from the Snowmass 2021 Theory Frontier topical group for Collider Phenomenology (TF07) -- showcases the dynamism, engagement, and motivations of collider phenomenologists by exposing selected exciting new directions and establishing key connections between cutting-edge theoretical advances and current and future experimental opportunities. By investing in collider phenomenology, the high-energy physics community can help ensure that theoretical advances are translated into concrete tools that enable and enhance current and future experiments, and in turn, experimental results feed into a more complete theoretical understanding and motivate new questions and explorations.
Maltoni, Fabio↗
Forward dijet and vector meson production at the LHC
We first describe the predictions concerning jet production in the very forward region, especially in the FOCAL kinematic region of the ALICE collaboration at CERN as a test of saturation. We also compare vector meson production cross sections at the LHC in Pb Pb and p Pb collisions with the Balitsky Fadin Kuraev Lipatov and the Balitsky Kovchegov evolution equations and data favor the observation of saturation at the LHC.
73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗