Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Field-Programmable Gate Arrays”

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.

At least 73 records · Page 4

Rapid Inference of Logic Gate Neural Networks for Anomaly Detection in High Energy Physics

The increasing data rates and complexity of detectors at the Large Hadron Collider (LHC) necessitate fast and efficient machine learning models, particularly for rapid selection of what data to store, known as triggering. Building on recent work in differentiable logic gates, we present a public implementation of a Convolutional Differentiable Logic Gate Neural Network (CLGN). We apply this to detecting anomalies at the Level-1 Trigger at CMS using public data from the CICADA project. We demonstrate that the CLGN achieves physics performance on par with or superior to conventional quantized neural networks. We also synthesize an LGN for a Field-Programmable Gate Array (FPGA) and show highly promising FPGA characteristics, notably zero Digital Signal Processor (DSP) resource usage. This work highlights the potential of logic gate networks for high-speed, on-detector inference in High Energy Physics and beyond.

FOS: Physical sciences↗

hls4ml: A Flexible, Open-Source Platform for Deep Learning Acceleration on Reconfigurable Hardware

We present hls4ml, a free and open-source platform that translates machine learning (ML) models from modern deep learning frameworks into high-level synthesis (HLS) code that can be integrated into full designs for field-programmable gate arrays (FPGAs) or application-specific integrated circuits (ASICs). With its flexible and modular design, hls4ml supports a large number of deep learning frameworks and can target HLS compilers from several vendors, including Vitis HLS, Intel oneAPI and Catapult HLS. Together with a wider eco-system for software-hardware co-design, hls4ml has enabled the acceleration of ML inference in a wide range of commercial and scientific applications where low latency, resource usage, and power consumption are critical. In this paper, we describe the structure and functionality of the hls4ml platform. The overarching design considerations for the generated HLS code are discussed, together with selected performance results.

FOS: Computer and information sciences↗

First application of a digital mirror Langmuir probe for real-time plasma diagnosis

For the first time, a digital Mirror Langmuir probe (MLP) has successfully sampled plasma temperature, ion saturation current, and floating potential together on a single probe tip in real time in a radio-frequency driven helicon linear plasma device. This is accomplished by feedback control of the bias sweep to ensure a good fit to I-V characteristics with a high frequency, high power digital amplifier and field-programmable gate array (FPGA) controller. Measurements taken by the MLP were validated by a low speed I-V characteristic manually collected during static plasma conditions. Plasma fluctuations, induced by varying the axial magnetic field (f̃ = 10 Hz), were also successfully monitored with the MLP. Further refinement of the digital MLP pushes it towards a turn-key system that minimizes the time to deployment and lessens the learning curve, positioning the digital MLP as a capable diagnostic for the study of low radio-frequency plasma physics. These demonstrations bolster confidence in fielding such digital MLP diagnostics in magnetic confinement experiments with high spatial and adequate temporal resolution such as edge plasma, scrape-off layer, and divertor probes.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Searching for Clues for a Matter Dominated Universe in Liquid Argon Time Projection Chambers

Liquid Argon Time Projection Chambers (LArTPCs) represent one of the most widely utilized neutrino detection techniques in neutrino experiments, for instance, in the Short Baseline Neutrino (SBN) program and the future large-scale LArTPC: Deep Underground Neutrino Experiment (DUNE). The high-end technique, facilitating excellent spatial and calorimetric reconstruction resolution, also enables testing exotic Beyond Standard Model (BSM) theories, such as baryon number violation (BNV) processes (e.g., proton-decay, neutron-antineutron oscillation). At the same time, Machine Learning (ML) techniques have demonstrated their ubiquitous use in recent decades; ML techniques have also become some of the most powerful tools in high-energy physics (HEP) analyses. Furthermore, the development of algorithms to cater to the needs of problems in HEP (i.e., triggering, reconstruction, improving sensitivity, etc.) has also become an active area of research. By developing a combined approach using Convolutional Neural Network (CNN) and Boosted Decision Tree (BDT) techniques, the sensitivity of neutron-antineutron oscillation in DUNE is evaluated for a projected exposure of 400kton·years. Additionally, to meet the triggering requirement to select such rare events in DUNE, such a search is only supported with highly efficient self-triggering algorithms. An ML-based self-triggering scheme for large-scale LArTPCs, such as DUNE, is also developed with the intention of implementation on field-programmable gate arrays (FPGAs). The ML-based approach for searching for neutron-antineutron oscillation can be demonstrated and validated on the current LArTPC MicroBooNE. The analysis in MicroBooNE represents the first-ever search for neutron-antineutron oscillation in a LArTPC. DUNE's projected 90% C.L. sensitivity to the neutron antineutron oscillation lifetime is 6.45×10³² years, assuming 1.327×10³⁵ neutron·years, equivalent to 10 years of DUNE far detector exposure (400kton·years). For MicroBooNE, assuming 372 seconds of exposure (equivalent to 3.13×10³⁶ neutron·years), the 90% C.L. lifetime sensitivity is found at 3.07×10²⁵ yrs, after accounting for Monte-Carlo statistical uncertainty and systematic uncertainty from detector effects.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

High performance context switching for virtualized FPGA accelerators

A hardware context manager in a field-programmable gate array (FPGA) device includes configuration logic configured to program one or more programming regions in the FPGA device based on configuration data for implementing a target configuration of the one or more programming regions. Context management logic in the hardware context manager is coupled with the configuration logic and saves a first context corresponding to the target configuration by retrieving first state information from the set of one or more programming regions, where the first state information is generated based on the target configuration, and storing the retrieved first state information in a context memory. The context management logic restores the first context by transferring the first state information from the context memory to the one or more programming regions, and causing the configuration logic to program the one or more programming regions based on the configuration data.

42 ENGINEERING↗

Accelerating Transformer-based Deep Learning Models on FPGAs using Column Balanced Block Pruning

Although Transformer-based language representations achieve state-of-the-art accuracy on various natural language processing (NLP) tasks, the large model size has been challenging the resource constrained computing platforms. Weight pruning, as a popular and effective technique in reducing the number of weight parameters and accelerating the Transformer, has been investigated on GPUs. However, the Transformer acceleration using weight pruning on field-programmable gate array (FPGAs) remains unexplored. This paper investigates the column balanced block-wise pruning on Transformer and designs an FPGA acceleration engine to customize the balanced blockwise matrix multiplication. We implement the Transformer model with proper hardware scheduling, and the experiments show that the Transformer inference on FPGA achieves 10.35 ms latency with the batch size of 32, which is 10.96 × speed up comparing to CPU platform and 2.08 × speed up comparing to GPU platform.

Peng, Hongwu↗

Open-source FPGA-ML codesign for the MLPerf Tiny Benchmark

We present our development experience and recent results for the MLPerf Tiny Inference Benchmark on field-programmable gate array (FPGA) platforms. We use the open-source hls4ml and FINN workflows, which aim to democratize AI-hardware codesign of optimized neural networks on FPGAs. We present the design and implementation process for the keyword spotting, anomaly detection, and image classification benchmark tasks. The resulting hardware implementations are quantized, configurable, spatial dataflow architectures tailored for speed and efficiency and introduce new generic optimizations and common workflows developed as a part of this work. The full workflow is presented from quantization-aware training to FPGA implementation. The solutions are deployed on system-on-chip (Pynq-Z2) and pure FPGA (Arty A7-100T) platforms. The resulting submissions achieve latencies as low as 20 $\mu$s and energy consumption as low as 30 $\mu$J per inference. We demonstrate how emerging ML benchmarks on heterogeneous hardware platforms can catalyze collaboration and the development of new techniques and more accessible tools.

Borras, Hendrik↗

Modbus RTU for Embedded Cyber Secure Inverter Controller

The Modbus communication protocol is a widely adopted communication standard in industrial control systems. This communication protocol is known for being reliable and straightforward to implement while being versatile in terms of its operating parameters while supporting multiple formats over various hardware infrastructures and architectures. Many intelligent devices such as Programmable Logic Controllers (PLCs), Human-Machine Interfaces (HMIs), Internet-of-Things (IoT), and various Operational Technologies (OT) utilize Modbus for their communication systems. These types of systems must communicate with each other through a standardized and central communication process. To support the integration of these modular systems, a Field-Programmable Gate Array (FPGA) can act as an embedded central routing fabric for this communication to take place. Embedded systems are versatile enough to interface with various devices and systems to accomplish various goals. Additionally, embedded systems require relatively small physical designs to minimize the required resources to facilitate the intended application by providing low-level system access. This minimization of system resources goes hand in hand with reducing the financial cost of a proposed solution or system. As remotely collaborating researchers often use FPGAs to prototype designs that are required to have a method for data transmission among systems, it is imperative to provide a baseline standard for communications among devices and systems. A typical method of implementing the Modbus RTU communication protocol in an embedded environment is using integrated logic architectures within the FPGA called “Intellectual Property (IP) cores.” IP cores can be designed using integrated logic or circuit designs to function as an embedded processor. These IP cores can then perform the required computational actions to support the Modbus RTU communication protocol by utilizing high-level programming languages such as the C programming language. The hardware description language of Very High-Speed Integrated Circuit Hardware Description Language (VHDL) allows for the control of real hardware at the logic gate and signal level. These logic gates and signals can be designed and controlled to perform desired actions based on the system design. Programming an FPGA using VHDL allows an individual to access the lowest abstraction level of the system during FPGA development. This level of abstraction is referred to as the register-transfer level (RTL), which gives access to manipulating values and variables at the register level. This register-level manipulation provides precision over creating the logical circuit within the FPGA, thus minimizing the required code to perform desired operations. The Modbus RTU communication protocol can be implemented within an FPGA using VHDL programming to establish a standardized and embedded serial communication pathway. This implementation provides a standardized communication protocol to streamline research efforts among researchers, thus increasing the efficiency of research efforts. Additionally, this Modbus RTU implementation requires fewer resources when compared to typical communication protocol implementations that utilize an IP core, reducing the hardware requirement for effective research efforts.

communication↗

A Graph Neural Network Surrogate Model for hls4ml

Recent advancements in use of machine learning (ML) techniques on field-programmable gate arrays (FPGAs) have allowed for the implementation of embedded neural networks with extremely low latency. This is invaluable for particle detectors at the Large Hadron Collider, where latency and used area are strictly bounded. The hls4ml framework is a procedure that converts trained ML model software to a synthesis result to can be used on an FPGA. However, running the pipeline is a time-consuming procedure, and there is a strong risk of failure. In particular, it may not be possible to successfully convert a model into a synthesis result, or the resource consumption of the model may exceed the resources of the target FPGA. To aid with this development, we introduce wa-hls4ml, a surrogate model using a graph neural network to emulate the structure of the source models. The goal is to estimate the chance of success and resource consumption of a given model when passed through the hls4ml pipeline, without needing to run the pipeline.

Plotnikov, Dennis↗

hls4ml: A Flexible, Open-Source Platform for Deep Learning Acceleration on Reconfigurable Hardware

We present hls4ml, a free and open-source platform that translates machine learning (ML) models from modern deep learning frameworks into high-level synthesis (HLS) code that can be integrated into full designs for field-programmable gate arrays (FPGAs) or application-specific integrated circuits (ASICs). With its flexible and modular design, hls4ml supports a large number of deep learning frameworks and can target HLS compilers from several vendors, including Vitis HLS, Intel oneAPI and Catapult HLS. Together with a wider eco-system for software-hardware co-design, hls4ml has enabled the acceleration of ML inference in a wide range of commercial and scientific applications where low latency, resource usage, and power consumption are critical. In this paper, we describe the structure and functionality of the hls4ml platform. The overarching design considerations for the generated HLS code are discussed, together with selected performance results.

Schulte, Jan-Frederik [Purdue U.] (ORCID:000000034↗

Advancing \textit{otsdaq}: Enhancements for Usability, Accuracy, and Robustness

High-energy physics (HEP) experiments require data acquisition (DAQ) systems that can orchestrate complex detector operations, high data throughput, and responsive, real-time feedback to operators. Traditional DAQ stacks, which are often bespoke, command-line driven and highly specific, impose large learning curves on users. The Off-The-Shelf Data Acquisition (\textit{otsdaq}) framework was created to address these issues by offering a highly customizable and scalable browser-based ’desktop’ environment, in which experiment-specific control and monitoring applications can be easily deployed and integrated. Although the initial development of the \textit{otsdaq} software was aimed at the Fermilab Test Beam Facility, \textit{otsdaq} is now being leveraged for broader deployment, including the upcoming Mu2e experiment, where real-time monitoring of field-programmable gate array (FPGA)-based Data Transfer Controllers (DTCs), Clock and Fanout (CFO) boards, and several other subsystems are critical. We contribute a set of targeted improvements to \textit{otsdaq}: bitmap visualization functionality for configured data, improved and corrected delta-based DTC throughput metrics, version control (VC)-backed source navigation for console messages, custom navigation hooks to eliminate disruptive user interface glitches, and copy-to-clipboard support for macro execution history. These changes improve usability, reduce debugging time, and increase accuracy in performance data as Mu2e moves toward commissioning.

Mohammed, A. [Unlisted, US]↗

Graphical User Interface for a Biasing Board for the PSEC6

The PSEC6 is an application-specific integrated circuit (ASIC) designed for a readout system for a large area picosecond photodetector (LAPPD). The PSEC6 is currently in fabrication and pending testing. The testing system for the PSEC5, the previous iteration of the ASIC, required expensive and non-portable equipment, because the ASIC needs twelve adjustable reference voltages. The new testing system consists of an low-cost, open-source, cross-platform graphical user interface (GUI), a digital system, and a biasing board. The digital system is the interface between the GUI and biasing board, and can be implemented on a microcontroller or field-programmable gate array (FPGA). The biasing board contains twelve digital-to-analog converters (DACs) that are configurable via the GUI, which gives users the ability to write voltage values to all or specific DACs. The GUI was developed in C on Linux using the widget library GTK4 and cross-compiled for Windows compatibility. I2C and SPI protocols were implemented on an Adafruit Feather ESP32-S3 microcontroller to write commands to the DACs and PSEC6. A hardware implementation of the I2C protocol is in development on an FPGA. Since LAPPDs will be used by the Accelerator Neutrino Neutron Interaction Experiment (ANNIE) at Fermilab, the PSEC6 testing system in this internship project can potentially benefit future neutrino research. The project is relevant to the Department of Energy’s microelectronics mission, because the PSEC6 is an ASIC that will handle fast time signals arriving from the detector for readout. It also provided experience with building a cross-platform user interface, practicing digital design and implementation in hardware description language (HDL), and using simulations to inform new design iterations.

Guerrero, Sasha Camila [North Central Coll.]↗

Cooling and Timing Tests of the ATLAS Fast TracKer VME Boards

The Fast TracKer (FTK) is an ATLAS trigger upgrade built for full-event, low-latency, high-rate tracking. The FTK core, made of 9U VME boards, performs the most demanding computational task. The associative memory board (AMB) serial link processor and the auxiliary card (AUX), plugged on the front and back sides of the same VME slot, constitute the processing unit (PU), which finds tracks using hits from eight layers of the inner detector. The PU works in pipeline with the second stage board (SSB), which finds 12-layer tracks by adding extra hits to the identified tracks. In the designed configuration, 16 PUs and four SSBs are installed in a VME crate. The high power consumption of the AMB, AUX, and SSB (respectively, of about 250, 70, and 160 W per board) required the development of a custom cooling system. Even though the expected power consumption for each VME crate of the FTK system is high compared with a common VME setup, the 8 FTK core crates will use ≈60 kW, which is just a fraction of the power and the space needed for a CPU farm performing the same task. We report on the integration of 32 PUs and eight SSBs inside the FTK system, on the infrastructures needed to run and cool them, and on the tests performed to verify the system processing rate and the temperature stability at a safe value.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

Overview and first measurements of the MAST Upgrade bolometer diagnostic

Here, a suite of multi-channel resistive bolometers has been implemented to measure the total radiation from Mega Amp Spherical Tokamak Upgrade plasmas, with cameras covering the core plasma and lower divertor chamber. Data are digitized and processed using novel field-programmable gate array-based electronics, offering improved compactness and new operational capabilities. A synthetic diagnostic has been developed to explore the quality of 2D reconstructions available from the system and to quantify the uncertainty on quantities such as the total radiated power. Measurements in the first campaign have demonstrated correct functioning of the diagnostic while also highlighting issues with electrical noise and some failure mechanisms of the detectors, as well as significant neutral beam fast-particle losses.

47 OTHER INSTRUMENTATION↗

Design and Testing of the Address in Real-Time Data Driver Card for the Micromegas Detector of the ATLAS New Small Wheel Upgrade

The Address in Real Time Data Driver Card (ADDC) is designed to transmit the trigger data in the Micro-Mesh Gaseous Structure (Micromegas) detector of the ATLAS New Small Wheel (NSW) upgrade. The Address in Real Time (ART) signals are generated by the front-end ASIC, named VMM chip, to indicate the address of the first above-threshold event. A custom ASIC (ART ASIC) is designed to receive the ART signals from the VMM chip and implement the hit-selection. The processed data from the ART ASIC will be transmitted out of the NSW through the GBTx serializer, the unidirectional Versatile Twin Transmitter (VTTx) and fiber optical links. The ART signal is critical for the ATLAS experiment trigger selection thus the performance and stability of the ADDC is very important. To ensure extensive testing of the ADDC, an FPGA Mezzanine Card (FMC) based testing platform and a special designed firmware/software are developed. We find that this test platform works with the commercial Xilinx VC707 FPGA development kit, independent of the other electronics of the NSW it can test all the functions of the ADDC and its long-term stability. This paper will introduce the design, testing procedure and results of the ADDC and the FMC testing platform.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗

Distance-Weighted Graph Neural Networks on FPGAs for Real-Time Particle Reconstruction in High Energy Physics

Graph neural networks have been shown to achieve excellent performance for several crucial tasks in particle physics, such as charged particle tracking, jet tagging, and clustering. An important domain for the application of these networks is the FGPA-based first layer of real-time data filtering at the CERN Large Hadron Collider, which has strict latency and resource constraints. We discuss how to design distance-weighted graph networks that can be executed with a latency of less than one μs on an FPGA. To do so, we consider a representative task associated to particle reconstruction and identification in a next-generation calorimeter operating at a particle collider. We use a graph network architecture developed for such purposes, and apply additional simplifications to match the computing constraints of Level-1 trigger systems, including weight quantization. Using the hls4ml library, we convert the compressed models into firmware to be implemented on an FPGA. Performance of the synthesized models is presented both in terms of inference accuracy and resource usage.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗