Nhan Tran

Bio: Nhan Tran is an academic researcher from Fermilab. The author has contributed to research in topics: Artificial neural network & CMOS. The author has an hindex of 14, co-authored 54 publications receiving 938 citations. Previous affiliations of Nhan Tran include University of Melbourne & Kyung Hee University.

Fast inference of deep neural networks in FPGAs for particle physics

Abstract: Recent results at the Large Hadron Collider (LHC) have pointed to enhanced physics capabilities through the improvement of the real-time event processing techniques. Machine learning methods are ubiquitous and have proven to be very powerful in LHC physics, and particle physics as a whole. However, exploration of the use of such techniques in low-latency, low-power FPGA (Field Programmable Gate Array) hardware has only just begun. FPGA-based trigger and data acquisition systems have extremely low, sub-microsecond latency requirements that are unique to particle physics. We present a case study for neural network inference in FPGAs focusing on a classifier for jet substructure which would enable, among many other physics scenarios, searches for new dark sector particles and novel measurements of the Higgs boson. While we focus on a specific example, the lessons are far-reaching. A companion compiler package for this work is developed based on High-Level Synthesis (HLS) called hls4ml to build machine learning models in FPGAs. The use of HLS increases accessibility across a broad user community and allows for a drastic decrease in firmware development time. We map out FPGA resource usage and latency versus neural network hyperparameters to identify the problems in particle physics that would benefit from performing neural network inference with FPGAs. For our example jet substructure model, we fit well within the available resources of modern FPGAs with a latency on the scale of 100 ns.

Fast inference of deep neural networks in FPGAs for particle physics

Abstract: Recent results at the Large Hadron Collider (LHC) have pointed to enhanced physics capabilities through the improvement of the real-time event processing techniques. Machine learning methods are ubiquitous and have proven to be very powerful in LHC physics, and particle physics as a whole. However, exploration of the use of such techniques in low-latency, low-power FPGA hardware has only just begun. FPGA-based trigger and data acquisition (DAQ) systems have extremely low, sub-microsecond latency requirements that are unique to particle physics. We present a case study for neural network inference in FPGAs focusing on a classifier for jet substructure which would enable, among many other physics scenarios, searches for new dark sector particles and novel measurements of the Higgs boson. While we focus on a specific example, the lessons are far-reaching. We develop a package based on High-Level Synthesis (HLS) called hls4ml to build machine learning models in FPGAs. The use of HLS increases accessibility across a broad user community and allows for a drastic decrease in firmware development time. We map out FPGA resource usage and latency versus neural network hyperparameters to identify the problems in particle physics that would benefit from performing neural network inference with FPGAs. For our example jet substructure model, we fit well within the available resources of modern FPGAs with a latency on the scale of 100 ns.

Ultrananocrystalline diamond-CMOS device integration route for high acuity retinal prostheses

Abstract: High density electrodes are a new frontier for biomedical implants. Increasing the density and the number of electrodes used for the stimulation of retinal ganglion cells is one possible strategy for enhancing the quality of vision experienced by patients using retinal prostheses. The present work presents an integration strategy for a diamond based, high density, stimulating electrode array with a purpose built application specific integrated circuit (ASIC). The strategy is centered on flip-chip bonding of indium bumps to create high count and density vertical interconnects between the stimulator ASIC and an array of diamond neural stimulating electrodes. The use of polydimethylsiloxane (PDMS) housing prevents cross-contamination of the biocompatible diamond electrode with non-biocompatible materials, such as indium, used in the microfabrication process. Micro-imprint lithography allowed edge-to-edge micro-scale pattering of the indium bumps on non-coplanar substrates that have a form factor that can conform to body organs and thus are ideally suited for biomedical applications. Furthermore, micro-imprint lithography ensures the compatibility of lithography with the silicon ASIC and aluminum contact pads. Although this work focuses on 256 stimulating diamond electrode arrays with a pitch of 150 μm, the use of indium bump bonding technology and vertical interconnects facilitates implants with tens of thousands electrodes with a pitch as low as 10 μm, thus ensuring validity of the strategy for future high acuity retinal prostheses, and bionic implants in general.

Graph Neural Networks for Particle Reconstruction in High Energy Physics detectors

Abstract: Author(s): Ju, Xiangyang; Farrell, Steven; Calafiura, Paolo; Murnane, Daniel; Prabhat; Gray, Lindsey; Klijnsma, Thomas; Pedro, Kevin; Cerati, Giuseppe; Kowalkowski, Jim; Perdue, Gabriel; Spentzouris, Panagiotis; Tran, Nhan; Vlimant, Jean-Roch; Zlokapa, Alexander; Pata, Joosep; Spiropulu, Maria; An, Sitong; Aurisano, Adam; Hewes, Jeremy; Tsaris, Aristeidis; Terao, Kazuhiro; Usher, Tracy | Abstract: Pattern recognition problems in high energy physics are notably different from traditional machine learning applications in computer vision. Reconstruction algorithms identify and measure the kinematic properties of particles produced in high energy collisions and recorded with complex detector systems. Two critical applications are the reconstruction of charged particle trajectories in tracking detectors and the reconstruction of particle showers in calorimeters. These two problems have unique challenges and characteristics, but both have high dimensionality, high degree of sparsity, and complex geometric layouts. Graph Neural Networks (GNNs) are a relatively new class of deep learning architectures which can deal with such data effectively, allowing scientists to incorporate domain knowledge in a graph structure and learn powerful representations leveraging that structure to identify patterns of interest. In this work we demonstrate the applicability of GNNs to these two diverse particle reconstruction problems.

Development of Long-Range UHF-band RFID Tag chip Using Schottky Diodes in Standard CMOS Technology

Abstract: We present the design of three key building blocks for UHF-band passive RFID tag chip, i.e., voltage multiplier, ASK demodulator, and internal clock generator. An analysis on a simple equivalent circuit of RFID tag chip for long reading range is presented taking into account the finite turn-on voltage of tag chip. The Schottky diodes used in the passive RFID tag chip were fabricated using titanium (Ti/Al/Ta/Al)-silicon (n-type) junction in 0.35 mum CMOS process, and the effect of size of Schottky diode on the turn-on voltage and the input impedance of the voltage multiplier was investigated. For 300 mV RF input voltage, the fabricated voltage multiplier using Schottky diodes generated output voltages of 1.5 V and corresponding voltage conversion efficiency of 45%. In addition, we propose an example circuit for internal oscillator of tag chip with digital calibration, which can generate precise copy of RFID reader timing signals.

The ATLAS Experiment at the CERN Large Hadron Collider

Abstract: This paper describes the ATLAS experiment as installed in i ts experimental cavern at point 1 at CERN. It also presents a brief overview of the expec ted performance of the detector.

Image Processing

Abstract: MUCKE aims to mine a large volume of images, to structure them conceptually and to use this conceptual structuring in order to improve large-scale image retrieval. The last decade witnessed important progress concerning low-level image representations. However, there are a number problems which need to be solved in order to unleash the full potential of image mining in applications. The central problem with low-level representations is the mismatch between them and the human interpretation of image content. This problem can be instantiated, for instance, by the incapability of existing descriptors to capture spatial relationships between the concepts represented or by their incapability to convey an explanation of why two images are similar in a content-based image retrieval framework. We start by assessing existing local descriptors for image classification and by proposing to use co-occurrence matrices to better capture spatial relationships in images. The main focus in MUCKE is on cleaning large scale Web image corpora and on proposing image representations which are closer to the human interpretation of images. Consequently, we introduce methods which tackle these two problems and compare results to state of the art methods. Note: some aspects of this deliverable are withheld at this time as they are pending review. Please contact the authors for a preview.

Machine learning and the physical sciences

Abstract: Machine learning (ML) encompasses a broad range of algorithms and modeling tools used for a vast array of data processing tasks, which has entered most scientific disciplines in recent years. This article reviews in a selective way the recent research on the interface between machine learning and the physical sciences. This includes conceptual developments in ML motivated by physical insights, applications of machine learning techniques to several domains in physics, and cross fertilization between the two fields. After giving a basic notion of machine learning methods and principles, examples are described of how statistical physics is used to understand methods in ML. This review then describes applications of ML methods in particle physics and cosmology, quantum many-body physics, quantum computing, and chemical and material physics. Research and development into novel computing architectures aimed at accelerating ML are also highlighted. Each of the sections describe recent successes as well as domain-specific methodology and challenges.

Parton distributions for the LHC

Abstract: We present a preliminary set of updated NLO parton distributions. For the first time we have a quantitative extraction of the strange quark and antiquark distributions and their uncertainties determined from CCFR and NuTeV dimuon cross sections. Additional jet data from HERA and the Tevatron improve our gluon extraction. Lepton asymmetry data and neutrino structure functions improve the flavour separation, particularly constraining the down quark valence distribution.