Searcharxiv⌕ Search

arXiv subjects

Nhan Tran

Publications and source records attributed to Nhan Tran.

At least 37 records · Page 2Linked to original sources
arXiv2025

Neural Architecture Codesign for Fast Physics Applications

We develop a pipeline to streamline neural architecture codesign for physics applications to reduce the need for ML expertise when designing models for novel tasks. Our method employs neural architecture search and network compression in a two-stage approach to discover hardware efficient models. This approach consists of a global search stage that explores a wide range of architectures while considering hardware constraints, followed by a local search stage that fine-tunes and compresses the most promising candidates. We exceed performance on various tasks and show further speedup through model compression techniques such as quantization-aware-training and neural network pruning. We synthesize the optimal models to high level synthesis code for FPGA deployment with the hls4ml library. Additionally, our hierarchical search space provides greater flexibility in optimization, which can easily extend to other tasks and domains. We demonstrate this with two case studies: Bragg peak finding in materials science and jet classification in high energy physics, achieving models with improved accuracy, smaller latencies, or reduced resource utilization relative to the baseline models.

cs.LG↗
arXiv2024

Intelligent Pixel Detectors: Towards a Radiation Hard ASIC with On-Chip Machine Learning in 28 nm CMOS

Detectors at future high energy colliders will face enormous technical challenges. Disentangling the unprecedented numbers of particles expected in each event will require highly granular silicon pixel detectors with billions of readout channels. With event rates as high as 40 MHz, these detectors will generate petabytes of data per second. To enable discovery within strict bandwidth and latency constraints, future trackers must be capable of fast, power efficient, and radiation hard data-reduction at the source. We are developing a radiation hard readout integrated circuit (ROIC) in 28nm CMOS with on-chip machine learning (ML) for future intelligent pixel detectors. We will show track parameter predictions using a neural network within a single layer of silicon and hardware tests on the first tape-outs produced with TSMC. Preliminary results indicate that reading out featurized clusters from particles above a modest momentum threshold could enable using pixel information at 40 MHz.

physics.ins-det↗
arXiv2024

Low latency optical-based mode tracking with machine learning deployed on FPGAs on a tokamak

Active feedback control in magnetic confinement fusion devices is desirable to mitigate plasma instabilities and enable robust operation. Optical high-speed cameras provide a powerful, non-invasive diagnostic and can be suitable for these applications. In this study, we process fast camera data, at rates exceeding 100kfps, on $\textit{in situ}$ Field Programmable Gate Array (FPGA) hardware to track magnetohydrodynamic (MHD) mode evolution and generate control signals in real-time. Our system utilizes a convolutional neural network (CNN) model which predicts the $n$=1 MHD mode amplitude and phase using camera images with better accuracy than other tested non-deep-learning-based methods. By implementing this model directly within the standard FPGA readout hardware of the high-speed camera diagnostic, our mode tracking system achieves a total trigger-to-output latency of 17.6$μ$s and a throughput of up to 120kfps. This study at the High Beta Tokamak-Extended Pulse (HBT-EP) experiment demonstrates an FPGA-based high-speed camera data acquisition and processing system, enabling application in real-time machine-learning-based tokamak diagnostic and control as well as potential applications in other scientific domains.

physics.plasm-ph↗
arXiv2024

Reliable edge machine learning hardware for scientific applications

Extreme data rate scientific experiments create massive amounts of data that require efficient ML edge processing. This leads to unique validation challenges for VLSI implementations of ML algorithms: enabling bit-accurate functional simulations for performance validation in experimental software frameworks, verifying those ML models are robust under extreme quantization and pruning, and enabling ultra-fine-grained model inspection for efficient fault tolerance. We discuss approaches to developing and validating reliable algorithms at the scientific edge under such strict latency, resource, power, and area requirements in extreme experimental environments. We study metrics for developing robust algorithms, present preliminary results and mitigation strategies, and conclude with an outlook of these and future directions of research towards the longer-term goal of developing autonomous scientific experimentation methods for accelerated scientific discovery.

cs.LG↗
arXiv2024

Smart Pixels: In-pixel AI for on-sensor data filtering

We present a smart pixel prototype readout integrated circuit (ROIC) designed in CMOS 28 nm bulk process, with in-pixel implementation of an artificial intelligence (AI) / machine learning (ML) based data filtering algorithm designed as proof-of-principle for a Phase III upgrade at the Large Hadron Collider (LHC) pixel detector. The first version of the ROIC consists of two matrices of 256 smart pixels, each 25$\times$25 $μ$m$^2$ in size. Each pixel consists of a charge-sensitive preamplifier with leakage current compensation and three auto-zero comparators for a 2-bit flash-type ADC. The frontend is capable of synchronously digitizing the sensor charge within 25 ns. Measurement results show an equivalent noise charge (ENC) of $\sim$30e$^-$ and a total dispersion of $\sim$100e$^-$ The second version of the ROIC uses a fully connected two-layer neural network (NN) to process information from a cluster of 256 pixels to determine if the pattern corresponds to highly desirable high-momentum particle tracks for selection and readout. The digital NN is embedded in-between analog signal processing regions of the 256 pixels without increasing the pixel size and is implemented as fully combinatorial digital logic to minimize power consumption and eliminate clock distribution, and is active only in the presence of an input signal. The total power consumption of the neural network is $\sim$ 300 $μ$W. The NN performs momentum classification based on the generated cluster patterns and even with a modest momentum threshold, it is capable of 54.4\% - 75.4\% total data rejection, opening the possibility of using the pixel information at 40MHz for the trigger. The total power consumption of analog and digital functions per pixel is $\sim$ 6 $μ$W per pixel, which corresponds to $\sim$ 1 W/cm$^2$ staying within the experimental constraints.

physics.ins-det↗
arXiv2024

Architectural Implications of Neural Network Inference for High Data-Rate, Low-Latency Scientific Applications

With more scientific fields relying on neural networks (NNs) to process data incoming at extreme throughputs and latencies, it is crucial to develop NNs with all their parameters stored on-chip. In many of these applications, there is not enough time to go off-chip and retrieve weights. Even more so, off-chip memory such as DRAM does not have the bandwidth required to process these NNs as fast as the data is being produced (e.g., every 25 ns). As such, these extreme latency and bandwidth requirements have architectural implications for the hardware intended to run these NNs: 1) all NN parameters must fit on-chip, and 2) codesigning custom/reconfigurable logic is often required to meet these latency and bandwidth constraints. In our work, we show that many scientific NN applications must run fully on chip, in the extreme case requiring a custom chip to meet such stringent constraints.

cs.LG↗
arXiv2024

ACE Science Workshop Report

We summarize the Fermilab Accelerator Complex Evolution (ACE) Science Workshop, held on June 14-15, 2023. The workshop presented the strategy for the ACE program in two phases: ACE Main Injector Ramp and Target (MIRT) upgrade and ACE Booster Replacement (BR) upgrade. Four plenary sessions covered the primary experimental physics thrusts: Muon Collider, Neutrinos, Charged Lepton Flavor Violation, and Dark Sectors. Additional physics and technology ideas were presented from the community that could expand or augment the ACE science program. Given the physics framing, a parallel session at the workshop was dedicated to discussing priorities for accelerator R\&D. Finally, physics discussion sessions concluded the workshop where experts from the different experimental physics thrusts were brought together to begin understanding the synergies between the different physics drivers and technologies. In December of 2023, the P5 report was released setting the physics priorities for the field in the next decade and beyond, and identified ACE as an important component of the future US accelerator-based program. Given the presentations and discussions at the ACE Science Workshop and the findings of the P5 report, we lay out the topics for study to determine the physics priorities and design goals of the Fermilab ACE project in the near-term.

hep-ex↗
arXiv2024

Robust Anomaly Detection for Particle Physics Using Multi-Background Representation Learning

Anomaly, or out-of-distribution, detection is a promising tool for aiding discoveries of new particles or processes in particle physics. In this work, we identify and address two overlooked opportunities to improve anomaly detection for high-energy physics. First, rather than train a generative model on the single most dominant background process, we build detection algorithms using representation learning from multiple background types, thus taking advantage of more information to improve estimation of what is relevant for detection. Second, we generalize decorrelation to the multi-background setting, thus directly enforcing a more complete definition of robustness for anomaly detection. We demonstrate the benefit of the proposed robust multi-background anomaly detection algorithms on a high-dimensional dataset of particle decays at the Large Hadron Collider.

hep-ex↗
arXiv2023

Differentiable Earth Mover's Distance for Data Compression at the High-Luminosity LHC

The Earth mover's distance (EMD) is a useful metric for image recognition and classification, but its usual implementations are not differentiable or too slow to be used as a loss function for training other algorithms via gradient descent. In this paper, we train a convolutional neural network (CNN) to learn a differentiable, fast approximation of the EMD and demonstrate that it can be used as a substitute for computing-intensive EMD implementations. We apply this differentiable approximation in the training of an autoencoder-inspired neural network (encoder NN) for data compression at the high-luminosity LHC at CERN. The goal of this encoder NN is to compress the data while preserving the information related to the distribution of energy deposits in particle detectors. We demonstrate that the performance of our encoder NN trained using the differentiable EMD CNN surpasses that of training with loss functions based on mean squared error.

hep-ex↗
arXiv2023

Beyond PID Controllers: PPO with Neuralized PID Policy for Proton Beam Intensity Control in Mu2e

We introduce a novel Proximal Policy Optimization (PPO) algorithm aimed at addressing the challenge of maintaining a uniform proton beam intensity delivery in the Muon to Electron Conversion Experiment (Mu2e) at Fermi National Accelerator Laboratory (Fermilab). Our primary objective is to regulate the spill process to ensure a consistent intensity profile, with the ultimate goal of creating an automated controller capable of providing real-time feedback and calibration of the Spill Regulation System (SRS) parameters on a millisecond timescale. We treat the Mu2e accelerator system as a Markov Decision Process suitable for Reinforcement Learning (RL), utilizing PPO to reduce bias and enhance training stability. A key innovation in our approach is the integration of a neuralized Proportional-Integral-Derivative (PID) controller into the policy function, resulting in a significant improvement in the Spill Duty Factor (SDF) by 13.6%, surpassing the performance of the current PID controller baseline by an additional 1.6%. This paper presents the preliminary offline results based on a differentiable simulator of the Mu2e accelerator. It paves the groundwork for real-time implementations and applications, representing a crucial step towards automated proton beam intensity control for the Mu2e experiment.

cs.LG↗
arXiv2023

Smartpixels: Towards on-sensor inference of charged particle track parameters and uncertainties

The combinatorics of track seeding has long been a computational bottleneck for triggering and offline computing in High Energy Physics (HEP), and remains so for the HL-LHC. Next-generation pixel sensors will be sufficiently fine-grained to determine angular information of the charged particle passing through from pixel-cluster properties. This detector technology immediately improves the situation for offline tracking, but any major improvements in physics reach are unrealized since they are dominated by lowest-level hardware trigger acceptance. We will demonstrate track angle and hit position prediction, including errors, using a mixture density network within a single layer of silicon as well as the progress towards and status of implementing the neural network in hardware on both FPGAs and ASICs.

hep-ex↗
arXiv2023

Neural Architecture Codesign for Fast Bragg Peak Analysis

We develop an automated pipeline to streamline neural architecture codesign for fast, real-time Bragg peak analysis in high-energy diffraction microscopy. Traditional approaches, notably pseudo-Voigt fitting, demand significant computational resources, prompting interest in deep learning models for more efficient solutions. Our method employs neural architecture search and AutoML to enhance these models, including hardware costs, leading to the discovery of more hardware-efficient neural architectures. Our results match the performance, while achieving a 13$\times$ reduction in bit operations compared to the previous state-of-the-art. We show further speedup through model compression techniques such as quantization-aware-training and neural network pruning. Additionally, our hierarchical search space provides greater flexibility in optimization, which can easily extend to other tasks and domains.

cs.LG↗
arXiv2023

Accelerating Machine Learning Inference with GPUs in ProtoDUNE Data Processing

We study the performance of a cloud-based GPU-accelerated inference server to speed up event reconstruction in neutrino data batch jobs. Using detector data from the ProtoDUNE experiment and employing the standard DUNE grid job submission tools, we attempt to reprocess the data by running several thousand concurrent grid jobs, a rate we expect to be typical of current and future neutrino physics experiments. We process most of the dataset with the GPU version of our processing algorithm and the remainder with the CPU version for timing comparisons. We find that a 100-GPU cloud-based server is able to easily meet the processing demand, and that using the GPU version of the event processing algorithm is two times faster than processing these data with the CPU version when comparing to the newest CPUs in our sample. The amount of data transferred to the inference server during the GPU runs can overwhelm even the highest-bandwidth network switches, however, unless care is taken to observe network facility limits or otherwise distribute the jobs to multiple sites. We discuss the lessons learned from this processing campaign and several avenues for future improvements.

hep-ex↗
arXiv2023

Smart pixel sensors: towards on-sensor filtering of pixel clusters with deep learning

Highly granular pixel detectors allow for increasingly precise measurements of charged particle tracks. Next-generation detectors require that pixel sizes will be further reduced, leading to unprecedented data rates exceeding those foreseen at the High Luminosity Large Hadron Collider. Signal processing that handles data incoming at a rate of O(40MHz) and intelligently reduces the data within the pixelated region of the detector at rate will enhance physics performance at high luminosity and enable physics analyses that are not currently possible. Using the shape of charge clusters deposited in an array of small pixels, the physical properties of the traversing particle can be extracted with locally customized neural networks. In this first demonstration, we present a neural network that can be embedded into the on-sensor readout and filter out hits from low momentum tracks, reducing the detector's data volume by 54.4-75.4%. The network is designed and simulated as a custom readout integrated circuit with 28 nm CMOS technology and is expected to operate at less than 300 $μW$ with an area of less than 0.2 mm$^2$. The temporal development of charge clusters is investigated to demonstrate possible future performance gains, and there is also a discussion of future algorithmic and technological improvements that could enhance efficiency, data reduction, and power per area.

physics.ins-det↗
arXiv2023

Photon-rejection Power of the Light Dark Matter eXperiment in an 8 GeV Beam

The Light Dark Matter eXperiment (LDMX) is an electron-beam fixed-target experiment designed to achieve comprehensive model independent sensitivity to dark matter particles in the sub-GeV mass region. An upgrade to the LCLS-II accelerator will increase the beam energy available to LDMX from 4 to 8 GeV. Using detailed GEANT4-based simulations, we investigate the effect of the increased beam energy on the capabilities to separate signal and background, and demonstrate that the veto methodology developed for 4 GeV successfully rejects photon-induced backgrounds for at least $2\times10^{14}$ electrons on target at 8 GeV.

hep-ex↗
arXiv2023

Current Status and Future Prospects for the Light Dark Matter eXperiment

The constituents of dark matter are still unknown, and the viable possibilities span a vast range of masses. The physics community has established searching for sub-GeV dark matter as a high priority and identified accelerator-based experiments as an essential facet of this search strategy. A key goal of the accelerator-based dark matter program is testing the broad idea of thermally produced sub-GeV dark matter through experiments designed to directly produce dark matter particles. The most sensitive way to search for the production of light dark matter is to use a primary electron beam to produce it in fixed-target collisions. The Light Dark Matter eXperiment (LDMX) is an electron-beam fixed-target missing-momentum experiment that realizes this approach and provides unique sensitivity to light dark matter in the sub-GeV range. This contribution provides an overview of the theoretical motivation, the main experimental challenges, how LDMX addresses these challenges, and projected sensitivities. We further describe the capabilities of LDMX to explore other interesting new and standard physics, such as visibly-decaying axion and vector mediators or rare meson decays, and to provide timely electronuclear scattering measurements that will inform the modeling of neutrino-nucleus scattering for DUNE.

hep-ex↗
arXiv2023

On-Sensor Data Filtering using Neuromorphic Computing for High Energy Physics Experiments

This work describes the investigation of neuromorphic computing-based spiking neural network (SNN) models used to filter data from sensor electronics in high energy physics experiments conducted at the High Luminosity Large Hadron Collider. We present our approach for developing a compact neuromorphic model that filters out the sensor data based on the particle's transverse momentum with the goal of reducing the amount of data being sent to the downstream electronics. The incoming charge waveforms are converted to streams of binary-valued events, which are then processed by the SNN. We present our insights on the various system design choices - from data encoding to optimal hyperparameters of the training algorithm - for an accurate and compact SNN optimized for hardware deployment. Our results show that an SNN trained with an evolutionary algorithm and an optimized set of hyperparameters obtains a signal efficiency of about 91% with nearly half as many parameters as a deep neural network.

cs.NE↗
arXiv2023

Structural Re-weighting Improves Graph Domain Adaptation

In many real-world applications, graph-structured data used for training and testing have differences in distribution, such as in high energy physics (HEP) where simulation data used for training may not match real experiments. Graph domain adaptation (GDA) is a method used to address these differences. However, current GDA primarily works by aligning the distributions of node representations output by a single graph neural network encoder shared across the training and testing domains, which may often yield sub-optimal solutions. This work examines different impacts of distribution shifts caused by either graph structure or node attributes and identifies a new type of shift, named conditional structure shift (CSS), which current GDA approaches are provably sub-optimal to deal with. A novel approach, called structural reweighting (StruRW), is proposed to address this issue and is tested on synthetic graphs, four benchmark datasets, and a new application in HEP. StruRW has shown significant performance improvement over the baselines in the settings with large graph structure shifts, and reasonable performance improvement when node attribute shift dominates.

cs.LG↗