Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Programmable devices”

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 613 records · Page 34
All sourcesNASA NTRSDOE OSTI
NASA NTRS2005

Reconfiguration of Analog Electronics for Extreme Environments: Problem or Solution?

This paper argues in favor of adaptive reconfiguration as a technique to expand the operational envelope of analog electronics for extreme environments (EE). In addition to hardening-by-process and hardening-by-design, "hardening-by-reconfiguration", when applicable, could be used to mitigate drifts, degradation, or damage on electronic devices (chips) in EE, by using re-configurable devices and an adaptive self-reconfiguration of their circuit topology. Conventional circuit design exploits device characteristics within a certain temperature/radiation range; when that is exceeded, the circuit function degrades. On a reconfigurable device, although component parameters change in EE, as long as devices still operate, albeit degraded, a new circuit design, suitable for new parameter values, may be mapped into the reconfigurable structure to recover the initial circuit function. Partly degraded resources are still used, while completely damaged resources are bypassed. Designs suitable for various environmental conditions can be determined prior to operation or can be determined in-situ, by adaptive reconfiguration algorithms running on built-in digital controllers. Laboratory demonstrations of this technique were performed by JPL in several independent experiments in which bulk CMOS reconfigurable devices were exposed to, and degraded by, low temperatures (approx. 196 C), high temperatures (approx.300 C) or radiation (300kRad TID), and then recovered by adaptive reconfiguration using evolutionary search algorithms. Taking this technology from Technology Readiness Level (TRL) 3 to TRL 5 is the target of a current NASA project.

Field Programmable Transistor Array (FPTA)↗
DOE OSTI2020

Quantitative and Predictive Genetic Parts for Plant Synthetic Biology

Plant synthetic biology aims to harness the natural abilities of plants and to turn them to new purposes. A primary goal of plant synthetic biology is to produce predictable and programmable genetic circuits from simple regulatory elements and well-characterized genetic components. The number of available DNA parts for plants is increasing, and the methods for rapid quantitative characterization are being developed, but the field of plant synthetic biology is still in its early stages. We here describe methods used to describe the quantitative properties of genetic components needed for plant synthetic biology. Once the quantitative properties and transfer function of a variety of genetic parts are known, computers can select the optimal components to assemble into functional devices, such as toggle switches and positive feedback circuits. However, while the variety of circuits and traits that can be put into plants are limitless, doing synthetic biology in plants poses unique challenges. Plants are composed of differentiated cells and tissues, each representing potentially unique regulatory or developmental contexts to introduced synthetic genetic circuits. Further, plants have evolved to be highly sensitive to environmental influences, such as light or temperature, any of which can affect the quantitative function of individual parts or whole circuits. Measuring the function of plant components within the context of a plant cell and, ideally, in a living plant, will be essential to using these components in gene circuits with predictable function. Mathematical modeling will be needed to account for the variety of contexts a genetic part will experience in different plant tissues or environments. With such understanding in hand, it may be possible to redesign plant traits to serve human and environmental needs.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2015

Development of intelligent instruments with embedded HTTP servers for control and data acquisition in a cryogenic setup—The hardware, firmware, and software implementation

The power of Ethernet for control and automation technology is being largely understood by the automation industry in recent times. Ethernet with HTTP (Hypertext Transfer Protocol) is one of the most widely accepted communication standards today. Ethernet is best known for being able to control through internet from anywhere in the globe. The Ethernet interface with built-in on-chip embedded servers ensures global connections for crate-less model of control and data acquisition systems which have several advantages over traditional crate-based control architectures for slow applications. This architecture will completely eliminate the use of any extra PLC (Programmable Logic Controller) or similar control hardware in any automation network as the control functions are firmware coded inside intelligent meters itself. Here, we describe the indigenously built project of a cryogenic control system built for linear accelerator at Inter University Accelerator Centre, known as “CADS,” which stands for “Complete Automation of Distribution System.” CADS deals with complete hardware, firmware, and software implementation of the automated linac cryogenic distribution system using many Ethernet based embedded cryogenic instruments developed in-house. Each instrument works as an intelligent meter called device-server which has the control functions and control loops built inside the firmware itself. Dedicated meters with built-in servers were designed out of ARM (Acorn RISC (Reduced Instruction Set Computer) Machine) and ATMEL processors and COTS (Commercially Off-the-Shelf) SMD (Surface Mount Devices) components, with analog sensor front-end and a digital back-end web server implementing remote procedure call over HTTP for digital control and readout functions. At present, 24 instruments which run 58 embedded servers inside, each specific to a particular type of sensor-actuator combination for closed loop operations, are now deployed and distributed across control LAN (Local Area Network). A group of six categories of such instruments have been identified for all cryogenic applications required for linac operation which were designed to build this medium-scale cryogenic automation setup. These devices have special features like remote rebooters, daughter boards for PIDs (Proportional Integral Derivative), etc., to operate them remotely in radiation areas and also have emergency switches by which each device can be taken to emergency mode temporarily. Finally, all the data are monitored, logged, controlled, and analyzed online at a central control room which has a user-friendly control interface developed using LabVIEW{sup ®}. This paper discusses the overall hardware, firmware, software design, and implementation for the cryogenics setup.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗
NASA NTRS1985

Design of high-voltage, high-power, solid state remote power controllers for aerospace applications

Two general types of remote power controllers (RPC's), which combine the functions of a circuit breaker and a switch, were developed for use in dc aerospace systems. Power-switching devices used in the designs are the gate-turnoff thyristor (GTO) and MOSFET. The RPC's can switch dc voltages to 1200 V and currents to 1000 A. Seven different units were constructed and subjected to laboratory and thermal vacuum testing. Two of these were dual units that switch both positive and negative voltages simultaneously. The RPC's using MOSFET's have slow turnon and turnoff times which limit surge currents and voltage spiking from high di/dt. The GTO's have much faster transition times. All RPC's have programmable overload tripout proportional to I sq T and microsecond tripout for large overloads.

Sturman, J. C.↗
DOE OSTI2022

CyOTE ASSET OWNER ENGAGEMENT – SIDE CHANNEL POWER ANALYSIS PROTOTYPE

The U.S. Department of Energy’s (DOE) Office of Cybersecurity, Energy Security, and Emergency Response (CESER), through the Cybersecurity for the Operational Technology Environment (CyOTE) Program, worked with energy sector asset owners and operators (AOOs), partners, and Idaho National Laboratory (INL) to develop capabilities for AOOs to independently identify adversarial tactics, techniques, and procedures (TTPs) within their operational technology (OT) environments. The CyOTE methodology seeks to identify adversarial techniques within an AOO OT environment that could result in physical disruptions to energy flow or damage to equipment. CyOTE provides a general roadmap for AOOs, starting from a triggering event, or the point in time and space they perceive an anomalous event or condition meriting investigation, and culminating when the AOO has sufficient confidence to make a business risk decision on the appropriate resolution. This paper outlines the results of one such engagement with the New York Power Authority (NYPA), where the CyOTE program partnered with an AOO to develop a design specification for a power side channel detector to identify anomalous changes to device load. It describes the goal of developing this capability, the development process, the challenges the technical teams faced and the future steps an AOO will need to take to install and use this detector in its OT environment.

99 GENERAL AND MISCELLANEOUS↗
NASA NTRS1992

Electronic Still Camera

A handheld, programmable, digital camera is disclosed that supports a variety of sensors and has program control over the system components to provide versatility. The camera uses a high performance design which produces near film quality images from an electronic system. The optical system of the camera incorporates a conventional camera body that was slightly modified, thus permitting the use of conventional camera accessories, such as telephoto lenses, wide-angle lenses, auto-focusing circuitry, auto-exposure circuitry, flash units, and the like. An image sensor, such as a charge coupled device ('CCD') collects the photons that pass through the camera aperture when the shutter is opened, and produces an analog electrical signal indicative of the image. The analog image signal is read out of the CCD and is processed by preamplifier circuitry, a correlated double sampler, and a sample and hold circuit before it is converted to a digital signal. The analog-to-digital converter has an accuracy of eight bits to insure accuracy during the conversion. Two types of data ports are included for two different data transfer needs. One data port comprises a general purpose industrial standard port and the other a high speed/high performance application specific port. The system uses removable hard disks as its permanent storage media. The hard disk receives the digital image signal from the memory buffer and correlates the image signal with other sensed parameters, such as longitudinal or other information. When the storage capacity of the hard disk has been filled, the disk can be replaced with a new disk.

Holland, S. Douglas↗
NASA NTRS2008

Efficient G(sup 4)FET-Based Logic Circuits

A total of 81 optimal logic circuits based on four-gate field-effect transistors (G(sup 4)4FETs) have been designed to implement all Boolean functions of up to three variables. The purpose of this development was to lend credence to the expectation that logic circuits based on G(sup 4)FETs could be more efficient (in the sense that they could contain fewer transistors), relative to functionally equivalent logic circuits based on conventional transistors. A G(sup 4)FET a combination of a junction field-effect transistor (JFET) and a metal oxide/semiconductor field-effect transistor (MOSFET) superimposed in a single silicon island and can therefore be regarded as two transistors sharing the same body. A G(sup 4)FET can also be regarded as a single device having four gates: two side junction-based gates, a top MOS gate, and a back gate activated by biasing of a silicon-on-insulator substrate. Each of these gates can be used to control the conduction characteristics of the transistor; this possibility creates new options for designing analog, radio-frequency, mixed-signal, and digital circuitry. One such option is to design a G(sup 4)FET to function as a three-input NOT-majority gate, which has been shown to be a universal and programmable logic gate. Optimal NOT-majority-gate, G(sup 4)FET-based logic-circuit designs were obtained in a comparative study that also included formulation of functionally equivalent logic circuits based on NOR and NAND gates implemented by use of conventional transistors. In the study, the problem of finding the optimal design for each logic function and each transistor type was solved as an integer-programming optimization problem. Considering all 81 non-equivalent Boolean functions included in the study, it was found that in 63% of the cases, fewer logic gates (and, hence, fewer transistors) would be needed in the G(sup 4)FET-based implementations.

Vatan, Farrokh↗
DOE OSTI2024

Construction of Reconfigurable and Polymorphic DNA Origami Assemblies with Coiled‐Coil Patches and Patterns

Abstract DNA origami nanodevices achieve programmable structure and tunable mechanical and dynamic properties by leveraging the sequence‐specific interactions of nucleic acids. Previous advances have also established DNA origami as a useful building block to make well‐defined micron‐scale structures through hierarchical self‐assembly, but these efforts have largely leveraged the structural features of DNA origami. The tunable dynamic and mechanical properties also provide an opportunity to make assemblies with adaptive structures and properties. Here the integration of DNA origami hinge nanodevices and coiled‐coil peptides are reported into hybrid reconfigurable assemblies. With the same dynamic device and peptide interaction, it is made multiple higher‐order assemblies (i.e., polymorphic assembly) by organizing clusters of peptides into patches or arranging single peptides into patterns on the surfaces of DNA origami to control the relative orientation of devices. The coiled‐coil interactions are used to construct circular and linear assemblies whose structure and mechanical properties can be modulated with DNA‐based reconfiguration. Reconfiguration of linear assemblies leads to micron scale motions and ≈2.5‐10‐fold increase in bending stiffness. The results provide a foundation for stimulus‐responsive hybrid assemblies that can adapt their structure and properties in response to nucleic acid, peptide, protein, or other triggers.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2015

Upgrades of Diagnostic Techniques and Technologies for JET next D-T Campaigns

JET next D-T campaign is presently scheduled for the year 2017. The main scientific objectives include the assessment of the isotopic effects on various plasma aspects: mainly on confinement, on the threshold to access the H mode and on ELM behaviour. From a technical point of view, the total yield of the entire D-T phase is expected to be 1.7 1021 neutrons, about a factor of six higher than the previous main D-T campaign on JET, DTE1. Therefore the radiation field will be quite relevant for next step devices, since the neutron flux at the first wall (∼10{sup 16} n/cm{sup 2}s), for example, will be comparable to the one in ITER behind the blanket. From a point of view of diagnostics developments, for many years JET diagnostics have been upgraded in order to provide adequate support for the scientific exploitation of a D-T campaign. The main efforts have concentrated on improving three main aspects of JET measuring capability: 1) the quality of the measurements of the electron and ion fluids to support the plasma physics programme 2) the diagnostic for the fusion products 3) diagnostic technologies for ITER. In terms of general diagnostic capability, compared to the previous DTE1, JET diagnostics have a much better spatial and temporal resolution of both the ion and electron fluid (about one order of magnitude improvement for each parameter). The consistency of the various independent measurements of the same parameters has also increased significantly; the three independent measurements of the electron temperature, for example, agree now within 5%. Moreover, solutions are being addressed to operate some cameras, both visible and IR, even during the full D-T phase to provide imaging of the plasma and the first wall. Various upgrades of neutral particle analysis are being considered, mainly to measure the isotopic composition. A new set of reflectometers is expected to provide valuable information about the changes in the turbulence with the different fuel mixtures. With regard to the fusion products, JET now can deploy a consistent set of techniques to measure the neutron yield and neutron spectra and to diagnose the fast particles. A full calibration of the neutron diagnostics with a 14 MeV source is being considered, after the recent very successful calibration for the 2.45 MeV neutrons. Vertical and horizontal lines of sight are foreseen for neutron and gamma spectrometry, in order to better determine the thermal neutron yield and to separate the trapped and passing components of the alphas. Various gamma ray spectrometers are being developed to cover all the various operational scenarios, from trace tritium to 50-50 D-T operation. The redistribution of the alphas will be measured with the gamma ray cameras, recently upgraded with full digital electronics; new detectors are being considered to bring the time resolution of the system in the ten of ms range. The lost alphas will also be diagnosed with improved spatial and temporal resolution, using Faraday cups and a scintillator probe. From a technological perspective, the D-T campaign will provide a unique opportunity to test ITER relevant technologies. From radiation hard detectors, for example Hall probes, to neutron absorbers and to shielding concepts, the potential of various solutions in a realistic 14 MeV radiation field will be assessed. The effects of neutrons and gamma on ancillary technologies and systems, such as fibre optics and electronics circuits, are also expected to be sufficiently high to derive useful information about the competitive advantage of various alternatives.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
DOE OSTI2017

Benchmarking and Evaluating Unified Memory for OpenMP GPU Offloading

Here, the latest OpenMP standard offers automatic device offloading capabilities which facilitate GPU programming. Despite this, there remain many challenges. One of these is the unified memory feature introduced in recent GPUs. GPUs in current and future HPC systems have enhanced support for unified memory space. In such systems, CPU and GPU can access each other's memory transparently, that is, the data movement is managed automatically by the underlying system software and hardware. Memory over subscription is also possible in these systems. However, there is a significant lack of knowledge about how this mechanism will perform, and how programmers should use it. We have modified several benchmarks codes, in the Rodinia benchmark suite, to study the behavior of OpenMP accelerator extensions and have used them to explore the impact of unified memory in an OpenMP context. We moreover modified the open source LLVM compiler to allow OpenMP programs to exploit unified memory. The results of our evaluation reveal that, while the performance of unified memory is comparable with that of normal GPU offloading for benchmarks with little data reuse, it suffers from significant overhead when GPU memory is over subcribed for benchmarks with large amount of data reuse. Based on these results, we provide several guidelines for programmers to achieve better performance with unified memory.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2012

Toward Active-Matrix Lab-On-A-Chip: Programmable Electrofluidic control Enaled by Arrayed Oxide Thin Film Transistors

Agile micro- and nano-fluidic control is critical to numerous life science and chemical science synthesis as well as kinetic and thermodynamic studies. To this end, we have demonstrated the use of thin film transistor arrays as an active matrix addressing method to control an electrofluidic array. Because the active matrix method minimizes the number of control lines necessary (m + n lines for the m x n element array), the active matrix addressing method integrated with an electrofluidic platform can be a significant breakthrough for complex electrofluidic arrays (increased size or resolution) with enhanced function, agility and programmability. An amorphous indium gallium zinc oxide (a-IGZO) semiconductor active layer is used because of its high mobility of 1-15 cm{sup 2} V{sup -1} s{sup -1}, low-temperature processing and transparency for potential spectroscopy and imaging. Several electrofluidic functionalities are demonstrated using a simple 2 x 5 electrode array connected to a 2 x 5 IGZO thin film transistor array with the semiconductor channel width of 50 {mu}m and mobility of 6.3 cm{sup 2} V{sup -1} s{sup -1}. Additionally, using the TFT device characteristics, active matrix addressing schemes are discussed as the geometry of the electrode array can be tailored to act as a storage capacitor element. Finally, requisite material and device parameters are discussed in context with a VGA scale active matrix addressed electrofluidic platform.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2021

Toward Evaluating High-Level Synthesis Portability and Performance between Intel and Xilinx FPGAs

Offloading computation from a CPU to a hardware accelerator is becoming a more common solution for improving performance because traditional gains enabled by Moore’s law and Dennard scaling have slowed. GPUs are often used as hardware accelerators, but field-programmable gate arrays (FPGAs) are gaining traction. FPGAs are beneficial because they allow hardware specific to a particular application to be created. However, they are notoriously difficult to program. To this end, two of the main FPGA manufacturers, Intel and Xilinx, have created tools and frameworks that enable the use of higher level languages to design FPGA hardware. Although Xilinx kernels can be designed by using C/C++, both Intel and Xilinx support the use of OpenCL C to architect FPGA hardware. However, not much is known about the portability and performance between these two device families other than the fact that it is theoretically possible to synthesize a kernel meant for Intel to Xilinx and vice versa.In this work, we evaluate the portability and performance of Intel and Xilinx kernels. We use OpenCL C implementations of a subset of the Rodinia benchmarking suite that were designed for an Intel FPGA and make the necessary modifications to create synthesizable OpenCL C kernels for a Xilinx FPGA. We find that the difficulty of porting certain kernel optimizations varies, depending on the construct. Once the minimum amount of modifications is made to create synthesizable hardware for the Xilinx platform, more nontrivial work is needed to improve performance. However, we find that constructs that are known to be performant for an FPGA should improve performance regardless of the platform; the difficulty comes in deciding how to invoke certain kernel optimizations while also abiding by the constraints enforced by a given platform’s hardware compiler.

Cabrera, Anthony↗
DOE OSTI2014

Diagnostics for machine protection of DEMO

DEMO aims to (i) integrate, demonstrate and validate all relevant technology necessary to convert fusion energy to electrical energy and (ii) that the machine and its operations are economically and environmentally acceptable. To maintain the efficiency and availability of the machine, there are several physics and combined physics/technology issues as well as the engineering issues. Machine Protection (also known as Protection of Investment) addresses both the risks to plant (to avoid costly repair or replacement) and the risks to normal operating time (to avoid loss of productivity and the return on investment). The plasma-related Machine Protection issues involve measurement and control of plasma stability, plasma purity, and plasma-wall interactions. Machine Protection aims to avoid hitting catastrophic limits by using early warning alarm systems, and controlled termination or avoidance, involving coordinated actions of the magnets, gas and auxiliary heating or current-drive systems. This article outlines the key processes, some of which are used in present-day tokamaks and some of which are new specifically for DEMO (e.g. First wall and divertor power handling) and reveals the need to research and develop new science and technology for Machine Protections in DEMO's high radiation and thermal fields. This work was funded by the RCUK Energy Programme under grant EP/I501045 and the European Communities under the contract of Association between EURATOM and CCFE and conducted partly under EFDA PPPT (WP13-DAS04). The views and opinions expressed herein do not necessarily reflect those of the European Commission.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
NASA NTRS1999

Advanced Design and Implementation of a Control Architecture for Long Range Autonomous Planetary Rovers

An advanced design and implementation of a Control Architecture for Long Range Autonomous Planetary Rovers is presented using a hierarchical top-down task decomposition, and the common structure of each design is presented based on feedback control theory. Graphical programming is presented as a common intuitive language for the design when a large design team is composed of managers, architecture designers, engineers, programmers, and maintenance personnel. The whole design of the control architecture consists in the classic control concepts of cyclic data processing and event-driven reaction to achieve all the reasoning and behaviors needed. For this purpose, a commercial graphical tool is presented that includes the mentioned control capabilities. Messages queues are used for inter-communication among control functions, allowing Artificial Intelligence (AI) reasoning techniques based on queue manipulation. Experimental results show a highly autonomous control system running in real time on top the JPL micro-rover Rocky 7 controlling simultaneously several robotic devices. This paper validates the sinergy between Artificial Intelligence and classic control concepts in having in advanced Control Architecture for Long Range Autonomous Planetary Rovers.

Martin-Alvarez, A.↗
NASA NTRS2003

ASRC Aerospace Corporation Selects Dynamically Reconfigurable Anadigm(Registered Trademark) FPAA For Advanced Data Acquisition System

Anadigm(registered trademark) today announced that ASRC Aerospace Corporation has designed Anadigm's dynamically reconfigurable Field Programmable Analog Array (FPAA) technology into an advanced data acquisition system developed under contract for NASA. ASRC Aerospace designed in the Anadigm(registered trademark) FPAA to provide complex analog signal conditioning in its intelligent, self-calibrating, and self-healing advanced data acquisition system (ADAS). The ADAS has potential applications in industrial, manufacturing, and aerospace markets. This system offers highly reliable operation while reducing the need for user interaction. Anadigm(registered trademark)'s dynamically reconfigurable FPAAs can be reconfigured in-system by the designer or on the fly by a microprocessor. A single device can thus be programmed to implement multiple analog functions and/or to adapt on-the-fly to maintain precision operation despite system degradation and aging. In the case of the ASRC advanced data acquisition system, the FPAA helps ensure that the system will continue to operating at 100% functionality despite changes in the environment, component degradation, and/or component failures.

Mata, Carlos T.↗
NASA NTRS2015

Single Event Effects in FPGA Devices 2014-2015

This presentation provides an overview of single event effects in FPGA devices 2014-2015 including commercial Xilinx V5 heavy ion accelerated testing, Xilinx Kintex-7 heavy ion accelerated testing. Mitigation study, and investigation of various types of triple modular redundancy (TMR) for commercial SRAM based FPGAs.

Single event upset (SEU)↗
DOE OSTI2012

The Windscale Advanced Gas Cooled Reactor (WAGR) Decommissioning Project A Close Out Report for WAGR Decommissioning Campaigns 1 to 10 - 12474

The reactor core of the Windscale Advanced Gas-Cooled Reactor (WAGR) has been dismantled as part of an ongoing decommissioning project. The WAGR operated until 1981 as a development reactor for the British Commercial Advanced Gas cooled Reactor (CAGR) power programme. Decommissioning began in 1982 with the removal of fuel from the reactor core which was completed in 1983. Subsequently, a significant amount of engineering work was carried out, including removal of equipment external to the reactor and initial manual dismantling operations at the top of the reactor, in preparation for the removal of the reactor core itself. Modification of the facility structure and construction of the waste packaging plant served to provide a waste route for the reactor components. The reactor core was dismantled on a 'top-down' basis in a series of 'campaigns' related to discrete reactor components. This report describes the facility, the modifications undertaken to facilitate its decommissioning and the strategies employed to recognise the successful decommissioning of the reactor. Early decommissioning tasks at the top of the reactor were undertaken manually but the main of the decommissioning tasks were carried remotely, with deployment systems comprising of little more than crane like devices, intelligently interfaced into the existing structure. The tooling deployed from the 3 tonne capacity (3te) hoist consisted either purely mechanical devices or those being electrically controlled from a 'push-button' panel positioned at the operator control stations, there was no degree of autonomy in the 3te hoist or any of the tools deployed from it. Whilst the ATC was able to provide some tele-robotic capabilities these were very limited and required a good degree of driver input which due to the operating philosophy at WAGR was not utilised. The WAGR box proved a successful waste package, adaptable through the use of waste box furniture specific to the waste-forms generated throughout the various decommissioning campaigns. The use of low force compaction for insulation and soft wastes provided a simple, robust and cost effective solution as did the direct encapsulation of LLW steel components in the later stages of reactor decommissioning. Progress through early campaigns was good, often bettering the baseline schedule, especially when undertaking the repetitive tasks seen during Neutron Shield and Graphite Core decommissioning, once the operators had become experienced with the equipment, though delays became more pronounced, mainly as a result of increased failures due to the age and maintainability of the RDM and associated equipment. Extensive delays came about as a result of the unsupported insulation falling away from the pressure vessel during removal and the inability of the ventilation system to manage the sub micron particulate generated during IPOPI cutting operations, though the in house development of revised and new methodologies ultimately led to the successful completion of PV and I removal. In a programme spanning over 12 years, the decommissioning of the reactor pressure vessel and core led to the production 110 ILW and 75 LLW WAGR boxes, with 20 LLW ISO freight containers of primary reactor wastes, resulting in an overall packaged volume of approximately 2500 cubic metres containing the estimated 460 cubic metres of the reactor structure. (authors)

12 MANAGEMENT OF RADIOACTIVE AND NON-RADIOACTIVE W↗
DOE OSTI2024

HDBind: encoding of molecular structure with hyperdimensional binary representations

Traditional methods for identifying “hit” molecules from a large collection of potential drug-like candidates rely on biophysical theory to compute approximations to the Gibbs free energy of the binding interaction between the drug and its protein target. These approaches have a significant limitation in that they require exceptional computing capabilities for even relatively small collections of molecules. Increasingly large and complex state-of-the-art deep learning approaches have gained popularity with the promise to improve the productivity of drug design, notorious for its numerous failures. However, as deep learning models increase in their size and complexity, their acceleration at the hardware level becomes more challenging. Hyperdimensional Computing (HDC) has recently gained attention in the computer hardware community due to its algorithmic simplicity relative to deep learning approaches. The HDC learning paradigm, which represents data with high-dimension binary vectors, allows the use of low-precision binary vector arithmetic to create models of the data that can be learned without the need for the gradient-based optimization required in many conventional machine learning and deep learning methods. This algorithmic simplicity allows for acceleration in hardware that has been previously demonstrated in a range of application areas (computer vision, bioinformatics, mass spectrometery, remote sensing, edge devices, etc.). To the best of our knowledge, our work is the first to consider HDC for the task of fast and efficient screening of modern drug-like compound libraries. We also propose the first HDC graph-based encoding methods for molecular data, demonstrating consistent and substantial improvement over previous work. We compare our approaches to alternative approaches on the well-studied MoleculeNet dataset and the recently proposed LIT-PCBA dataset derived from high quality PubChem assays. We demonstrate our methods on multiple target hardware platforms, including Graphics Processing Units (GPUs) and Field Programmable Gate Arrays (FPGAs), showing at least an order of magnitude improvement in energy efficiency versus even our smallest neural network baseline model with a single hidden layer. Our work thus motivates further investigation into molecular representation learning to develop ultra-efficient pre-screening tools. We make our code publicly available at https://github.com/LLNL/hdbind.

59 BASIC BIOLOGICAL SCIENCES↗