Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “accelerated computing”

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 1,189 records · Page 66
All sourcesNASA NTRSDOE OSTI
DOE OSTI2021

Understanding and Overcoming Water-induced Interfacial Degradation in Si Modules (Final Technical Report)

Moisture ingress is an established issue for photovoltaic module durability. Durability studies probing moisture effects typically evaluate performance losses at the module level, attributing global power losses to the overall humidity condition of the test environment while leaving local module behavior unknown. This project successfully develops an in situ short-wave infrared probe of moisture content in PV modules with demonstrated detection limits better than 100 ppm in EVA-encapsulated Al-BSF and PERC architectures (66 µg/cm3) and applicability across the range of terrestrial conditions. Combining this optical moisture quantification method (water reflectometry detection, WaRD) with peel testing of interfaces, biased photoluminescence imaging of performance, and first principles computation, we correlate module moisture content and cell performance over the course of accelerated damp heat tests. By performing multi-level humidity and thermal accelerated testing over thousands of hours (0, 65 and 85% RH and 25, 65, and 85°C), we are able to attribute effects of individual environmental stresses and the dependence on module architecture. Using first-principles computational chemistry, we find that the decomposition of EVA to acetic acid (Norrish II) is thermodynamically preferred over a pathway to acetaldehyde (Norrish I), elucidating the underlying chemistry of encapsulant degradation. We also find it is likely that water segregates to the stable interfaces of the cell, especially the SiNx/EVA interface. We find that moisture is strongly correlated with reduction in adhesion at the interfaces of the cells, particularly at Ag fingers. We show uniquely that the peel strength required to delaminate the encapsulant from the front of the cell depends not only on the exposure to a high heat and humid environment, but also to the moisture content during the mechanical testing, indicating an important interaction between when mechanical stresses are faced in terms of the moisture content of the module and its durability. We find that the background sheet resistance (that is, the sheet resistance rise not attributable to finger interruptions or cracks) is substantially higher in the backsheet mini-modules vs glass-glass packages when humidity and temperature are faced. This increase in power loss due to the background resistance increase over time in damp heat in glass-backsheet modules resulted in ~3-4% larger decrease in relative PCE compared to glass-glass modules. In glass-backsheet modules, the effect of the moisture dose alone is comparable or greater than that of the combined temperature and water term, suggesting that glass-backsheet modules are overall more susceptible to moisture induced performance loss. Finally, we show that thin film PV modules are amenable to WaRD measurement, based on the ability to detect moisture in POE encapsulants or in the cell stack itself. The ease of WaRD measurement and its few requirements on the bill of materials should enable broad applicability of this technique for studying the effects of moisture in PV.

14 SOLAR ENERGY↗
DOE OSTI2023

Superconducting Hyperdimensional Associative Memory Circuit for Scalable Machine Learning

Here we propose a generalized architecture for the first rapid-single-flux-quantum (RSFQ) associative memory circuit. The circuit employs hyperdimensional computing (HDC), a machine learning (ML) paradigm utilizing vectors with dimensionality in the thousands to represent information. HDC designs have small memory footprints, simple computations, and simple training algorithms compared to superconducting neural network accelerators (SNNAs), making them a better option for scalable SFQ machine learning (ML) solutions. The proposed superconducting HDC (SHDC) circuit uses entirely on-chip RSFQ memory which is tightly integrated with logic, operates at 33.3 GHz, is applicable to general ML tasks, and is manufacturable at practically useful scales given current SFQ fabrication limits. Tailored to a language recognition task, SHDC consists of ~ 2-20 M Josephson junctions (JJs) and consumes up to three times less power than an analogous CMOS HDC circuit while achieving 78-84% higher throughput. SHDC is capable of outperforming the state of the art RSFQ SNNA, SuperNPU, by 48-99% for all benchmark NN architectures tested while occupying up to 90% less area and consuming up to nine times less power. To the best of the authors' knowledge, SHDC is currently the only superconducting ML approach feasible at practically useful scales for real-world ML tasks and capable of online learning.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗
DOE OSTI2021

Automated Generation of Integrated Digital and Spiking Neuromorphic Machine Learning Accelerators

The growing numbers of application areas for artificial intelligence (AI) methods have led to an explosion of domain-specific accelerators that could support every new machine learning (ML) algorithm advancement, clearly highlighting the need for a capability to quickly and automatically transition from algorithm definition to hardware implementation and explore design space along a variety of SWaP (size, weight and Power). The software defined architectures (SODA) synthesizer implements a compiler-based modular infrastructure for the end-to-end generation of machine learning accelerators from high-level frameworks to hardware description language. At the same time, neuromorphic computing, by mimicking how the brain operates, promises to perform artificial intelligence tasks at efficiencies orders of magnitude higher than the current conventional tensor-processing based accelerators, as demonstrated by a variety of specialized designs leveraging Spiking Neural Networks (SNNs). Nevertheless, the mapping of an artificial neural network (ANN) to solutions supporting SNNs is still a non-trivial and very device-specific task, and completely lack the possibility to design hybrid systems that integrate conventional and spiking neural models. In this paper we discuss the support for such an integrated generation leveraging the SODA Synthesizer framework and its modular structure. In particular, we present a new MLIR dialect (part of the SODA frontend) that allows expressing spiking neural network features (e.g., available resources, spiking sequences, analog signal reading, etc.) and illustrate how it enables mapping to Spiking Neurons and deployment to the related specialized hardware (which, in the digital domain, could be generated through the other existing layers of the SODA Synthesizer). We then discuss the opportunities for even deeper integration afforded by the hardware compilation infrastructure, providing a path towards the generation of complex heterogeneous artificial intelligence systems.

Curzel, Serena↗
DOE OSTI2020

Target normal sheath acceleration with a large laser focal diameter

In this work, the dependence of the laser-driven ion acceleration from thin titanium foils in the Target Normal Sheath Acceleration (TNSA) regime on target and laser parameters is explored using two dimensional particle-in-cell simulations. The oblique incidence (θ L = 45°) and large focal spot size (w 0 = 40μm) are chosen to take an advantage of quasi one-dimensional geometry of sheath fields and effective electron heating. This interaction setup also reveals low and achromatic angular divergence of a proton beam. It is shown that the hot electron temperature deviates from the ponderomotive scaling for short laser pulses and small pre-plasmas. This deviation is mainly due to the laser sweeping, as the short duration laser pulse each moment in time effectively heats only a fraction of a focal spot on the foil. This instantaneous partial heating results in an electron temperature deviation from the ponderomotive scaling and, thus, lower maximum proton energies than it could have been expected from the TNSA theory.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
NASA NTRS1995

Modeling the vestibulo-ocular reflex of the squirrel monkey during eccentric rotation and roll tilt

Model simulations of the squirrel monkey vestibulo-ocular reflex (VOR) are presented for two motion paradigms: constant velocity eccentric rotation and roll tilt about a naso-occipital axis. The model represents the implementation of three hypotheses: the "internal model" hypothesis, the "gravito-inertial force (GIF) resolution" hypothesis, and the "compensatory VOR" hypothesis. The internal model hypothesis is based on the idea that the nervous system knows the dynamics of the sensory systems and implements this knowledge as an internal dynamic model. The GIF resolution hypothesis is based on the idea that the nervous system knows that gravity minus linear acceleration equals GIF and implements this knowledge by resolving the otolith measurement of GIF into central estimates of gravity and linear acceleration, such that the central estimate of gravity minus the central estimate of acceleration equals the otolith measurement of GIF. The compensatory VOR hypothesis is based on the idea that the VOR compensates for the central estimates of angular velocity and linear velocity, which sum in a near-linear manner. During constant velocity eccentric rotation, the model correctly predicts that: (1) the peak horizontal response is greater while "facing-motion" than with "back-to-motion"; (2) the axis of eye rotation shifts toward alignment with GIF; and (3) a continuous vertical response, slow phase downward, exists prior to deceleration. The model also correctly predicts that a torsional response during the roll rotation is the only velocity response observed during roll rotations about a naso-occipital axis. The success of this model in predicting the observed experimental responses suggests that the model captures the essence of the complex sensory interactions engendered by eccentric rotation and roll tilt.

NASA Discipline Neuroscience↗
DOE OSTI2025

OLCF’s Advanced Computing Ecosystem (ACE): FY25 Update for Ongoing Efforts

The advent of widespread use of artificial intelligence (AI) and machine learning (ML) models in science, coupled with fast data production rates of scientific instruments strain the traditional batch-oriented high-performance computing (HPC) environment. As scientific exploration continues to require more data and faster processing and analysis, new emerging technologies and capabilities to enable cross-facility and time-sensitive workflows are required for seamless integration of HPC and experimental facilities. The Advanced Computing Ecosystem (ACE) is a strategic initiative within the Oak Ridge Leadership Computing Facility (OLCF) established in 2024 to support the development of cutting-edge technologies to advance computational research and infrastructure at OLCF and across the Department of Energy (DOE). Several DOE initiatives are spearheading the evolution of the scientific landscape by blurring facility boundaries and connecting the user facilities to advance scientific capabilities and ensure energy dominance. The DOE Integrated Research Infrastructure (IRI) program is one example that is laying a foundation to support complex cross-facility workflows. The IRI program aims to integrate diverse computational resources, data infrastructures, and scientific instruments to facilitate collaboration and accelerate scientific discovery. The Interconnected Science Ecosystem (INTERSECT) initiative at Oak Ridge National Laboratory (ORNL) is another example that aims to revolutionize scientific research through AI-driven, interconnected autonomous laboratories and research facilities. Finally, the American Science Cloud (AmSC), recently announced in the “One Big Beautiful Bill”, aims to leverage prior infrastructure efforts of the IRI and automation and AI efforts of INTERSECT (and others) to build a federated, AI-augmented AmSC platform to unify the DOE’s computing, experimental, and data resources to catalyze scientific innovation.

97 MATHEMATICS AND COMPUTING↗
NASA NTRS2013

Orion MPCV Touchdown Detection Threshold Development and Testing

A robust method of detecting Orion Multi-Purpose Crew Vehicle (MPCV) splashdown is necessary to ensure crew and hardware safety during descent and after touchdown. The proposed method uses a triple redundant system to inhibit Reaction Control System (RCS) thruster firings, detach parachute risers from the vehicle, and transition to the post-landing segment of the Flight Software (FSW). An in-depth trade study was completed to determine optimal characteristics of the touchdown detection method resulting in an algorithm monitoring filtered, lever-arm corrected, 200 Hz Inertial Measurement Unit (IMU) vehicle acceleration magnitude data against a tunable threshold using persistence counter logic. Following the design of the algorithm, high fidelity environment and vehicle simulations, coupled with the actual vehicle FSW, were used to tune the acceleration threshold and persistence counter value to result in adequate performance in detecting touchdown and sufficient safety margin against early detection while descending under parachutes. An analytical approach including Kriging and adaptive sampling allowed for a sufficient number of finite element analysis (FEA) impact simulations to be completed using minimal computation time. The combination of a persistence counter of 10 and an acceleration threshold of approximately 57.3 ft/s2 resulted in an impact performance factor of safety (FOS) of 1.0 and a safety FOS of approximately 2.6 for touchdown declaration. An RCS termination acceleration threshold of approximately 53.1 ft/s(exp)2 with a persistence counter of 10 resulted in an increased impact performance FOS of 1.2 at the expense of a lowered under-parachutes safety factor of 2.2. The resulting tuned algorithm was then tested on data from eight Capsule Parachute Assembly System (CPAS) flight tests, showing an experimental minimum safety FOS of 6.1. The formulated touchdown detection algorithm will be flown on the Orion MPCV FSW during the Exploration Flight Test 1 (EFT-1) mission in the second half of 2014.

Daum, Jared↗
DOE OSTI2023

Using Numba for GPU acceleration of Neutron Beamline Digital Twins

Digital twins of neutron instruments using Monte Carlo ray tracing have proven to be useful in neutron data analysis and verifying instrument and sample designs. However, these simulations can become quite complex and computationally demanding with tens of billions of neutrons. In this paper, we present a GPU accelerated version of MCViNE using Python and Numba to balance user extensibility with performance. Numba is an open-source just-in-time (JIT) compiler for Python using LLVM to generate efficient machine code for CPUs and GPUs with NVIDIA CUDA. The JIT nature of Numba allowed complex instrument kernels to be generated easily. Initial simulations have shown a speedup between 200-1000x over the original CPU implementation. The performance gain with Numba enables more sophisticated data analysis and impacts neutron scattering science and instrument design.

Kendrick, Coleman↗
DOE OSTI2022

Multinode Multi-GPU Two-Electron Integrals: Code Generation Using the Regent Language

The computation of two-electron repulsion integrals (ERIs) is often the most expensive step of integral-direct self-consistent field methods. Formally it scales as O(N 4 ), where N is the number of Gaussian basis functions used to represent the molecular wave function. In practice, this scaling can be reduced to O(N 2 ) or less by neglecting small integrals with screening methods. The contributions of the ERIs to the Fock matrix are of Coulomb (J) and exchange (K) type and require separate algorithms to compute matrix elements efficiently. We previously implemented highly efficient GPU-accelerated J-matrix and K-matrix algorithms in the electronic structure code TeraChem. Although these implementations supported the use of multiple GPUs on a node, they did not support the use of multiple nodes. This presents a key bottleneck to cutting-edge ab initio simulations of large systems, e.g., excited state dynamics of photoactive proteins. We present our implementation of multinode multi-GPU J- and K-matrix algorithms in TeraChem using the Regent programming language. Regent directly supports distributed computation in a task-based model and can generate code for a variety of architectures, including NVIDIA GPUs. We demonstrate multinode scaling up to 45 GPUs (3 nodes) and benchmark against hand-coded TeraChem integral code. Finally, we also outline our metaprogrammed Regent implementation, which enables flexible code generation for integrals of different angular momenta.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2022

AI-enhanced Codesign for Next-Generation Neuromorphic Circuits and Systems

This report details work that was completed to address the Fiscal Year 2022 Advanced Science and Technology (AS&T) Laboratory Directed Research and Development (LDRD) call for “AI-enhanced Co-Design of Next Generation Microelectronics.” This project required concurrent contributions from the fields of 1) materials science, 2) devices and circuits, 3) physics of computing, and 4) algorithms and system architectures. During this project, we developed AI-enhanced circuit design methods that relied on reinforcement learning and evolutionary algorithms. The AI-enhanced design methods were tested on neuromorphic circuit design problems that have real-world applications related to Sandia’s mission needs. The developed methods enable the design of circuits, including circuits that are built from emerging devices, and they were also extended to enable novel device discovery. We expect that these AI-enhanced design methods will accelerate progress towards developing next-generation, high-performance neuromorphic computing systems.

42 ENGINEERING↗
DOE OSTI2026

ORCHA: A performance portability system for extreme heterogeneity

Heterogeneity is the prevalent trend in the rapidly evolving high-performance computing (HPC) landscape in both hardware and application software. The diversity in hardware platforms, currently comprising various accelerators and a future possibility of specializable chiplets, poses a significant challenge for scientific software developers aiming to harness optimal performance across different computing platforms while maintaining the quality of solutions when their applications are simultaneously growing more complex. Code synthesis and code generation can provide mechanisms to mitigate this challenge. We have developed a divide and conquer approach where different aspects of performance are handled by different stand-alone tools that are interfaced with the application through generated code. This portability system, ORCHA, enables users to configure and orchestrate their computations among available resources on a platform by specifying a high-level recipe, thereby permitting a many-to-many paradigm where each recipe results in a different variant of the application. The core design goal is to let users decide the application’s hardware mapping and orchestration by editing only the high-level recipe—without modifying the maintained source code or binding the application to a particular runtime system. Tools in ORCHA distribution are: CG-Kit for translating the recipe into an execution graph; Milhoja to execute the graph by orchestrating data and task movement among hardware resources; and Macroprocessor that enables users to define their own code-shorthand for higher composability and easier management of code variants. Additionally, the design of ORCHA permits tools to work in a plug-and-play mode where the application can build and run without CG-Kit and Milhoja, and either tool can be swapped out for other tools with similar capabilities by modifying the code generation portion of ORCHA. In this paper, we describe the design of ORCHA and the role that code-generation plays in isolating applications from tools. We demonstrate the breadth of configurations ORCHA enables with a case study in which an application configuration is realized on three distinct hardware mappings—a GPU-centric, a CPU/GPU balanced, and a CPU/GPU concurrent layouts by using different recipes.

Lee, Youngjun↗
NASA NTRS1978

Lunar tidal acceleration obtained from satellite-derived ocean tide parameters

Observation equations for the M2 ocean tide are computed from Geos 3 data for the long periodic variations of the inclination and node of the orbit. M2 ocean tide parameter values C22+ = 3.23 + or - 0.25 cm, epsilon 22+ = 331 + or - 6 deg, and epsilon 42+ = 113 + or - 6 deg are determined. With the assumption of zero solid tide phase lag, the lunar tidal acceleration is mostly (85%) due to the C22+ term in the expansion of the M2 tide with additional small contributions from the O1 and N2 tides. The calculated value for the tidal acceleration in lunar longitude is -27.4 + or - 3 arc sec/sq (100 yr) which is similar to values determined from astronomical data. The mean elements of Geos 3 are presented in tabular form.

Goad, C. C.↗
DOE OSTI2022

Accelerating $\mathrm{LHC}$ event generation with simplified pilot runs and fast $\mathrm{PDF}$s

Poor computing efficiency of precision event generators for LHC physics has become a bottleneck for Monte-Carlo event simulation campaigns. We provide solutions to this problem by focusing on two major components of general-purpose event generators: The PDF evaluator and the matrix-element generator. For a typical production setup in the ATLAS experiment, we show that the two can consume about 80% of the total runtime. Using NLO simulations of pp→ℓ + ℓ - +jets and pp→$t\bar{t}$+jets as an example, we demonstrate that the computing footprint of LHAPDF and SHERPA can be reduced by factors of order 10, while maintaining the formal accuracy of the event sample. The improved codes are made publicly available.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
NASA NTRS2024

Improved Modelling of SEP Event Onset Within the WSA–Enlil–SEPMOD Framework

Multi-spacecraft observations of solar energetic particle (SEP) events not only enable a deeper understanding and development of particle acceleration and transport theories but also provide important constraints for model validation efforts. However, because of computational limitations, a given physics-based SEP model is usually best suited to capture a particular phase of an SEP event, rather than its whole development from onset through decay. For example, magnetohydrodynamic (MHD) models of the heliosphere often incorporate solar transients only at the outer boundary of their so-called coronal domain – usually set at a heliocentric distance of 20–30 R⊙. This means that particle acceleration at coronal mass ejection (CME)-driven shocks is also computed from this boundary onwards, leading to simulated SEP event onsets that can be many hours later than observed, since shock waves can form much lower in the solar corona. In this work, we aim to improve the modelled onset of SEP events by inserting a “fixed source” of particle injection at the outer boundary of the coronal domain of the coupled WSA–Enlil 3D MHD model of the heliosphere. The SEP model that we employ for this effort is Solar Energetic Particle MODel (SEPMOD), a physics-based test-particle code based on a field line tracer and adiabatic invariant conservation. We apply our initial tests and results of SEPMOD’s fixed-source option to the 2021 October 9 SEP event, which was detected at five well-separated locations in the inner heliosphere – Parker Solar Probe, STEREO-A, Solar Orbiter, BepiColombo, and near-Earth spacecraft.

Solar eruptions↗
DOE OSTI2022

Strategies for Computer-Aided Discovery of Novel Open-Shell Polymers

Organic π-conjugated polymers with a triplet ground state have been the focus of recent research for their interesting and unique electronic properties, arising from the presence of the two unpaired electrons. These compounds are usually built from alternating electron-donating and electron-accepting monomer pairs which lower the HOMO–LUMO gap and yield a triplet state instead of the typical singlet ground state. In this paper, we use density functional theory calculations to explore the design rules that govern the creation of a ground-state triplet conjugated polymer and find that a small HOMO–LUMO gap in the singlet state is the best predictor for the existence of a triplet ground state, compared to previous use of a pro-quinoidal bonding character. Furthermore, this work can accelerate the discovery of new stable triplet materials by reducing the computational resources needed for electronic-state calculations and the number of potential candidates for synthesis.

36 MATERIALS SCIENCE↗
DOE OSTI2024

Computational simulations and beamline optimizations for an electron beam degrader at CEBAF

An electron beam degrader is under development with the objective of measuring the transverse and longitudinal acceptance of the Continuous Electron Beam Accelerator Facility (CEBAF) at Jefferson Lab. This project is in support of the CE+BAF positron capability. Computational simulations of beam-target interactions and particle tracking were performed integrating the GEANT4 and Elegant toolkits. A solenoid was added to the setup to control the beam's divergence. Parameter optimization of the solenoid field and magnetic quadrupoles gradient was also performed to further reduce particle loss through the rest of the injector beamline.

Lizárraga-Rubio, V.↗
DOE OSTI2019

Convergence in nonlinear laser wakefield accelerators modeling in a Lorentz-boosted frame

Laser wakefield acceleration modeling using the Lorentz-boosted frame technique in the particle-in-cell code has demonstrated orders of magnitude speedups. A convergence study was previously conducted in cases with external injection in the linear regime and without injection in the nonlinear regime, and the obtained results have shown a convergence within the percentage level. Here, a convergence study is carried out to model electron self-injection in a 2D–3V configuration. It is observed that the Lorentz-boosted frame technique is capable of modeling complex particle dynamics with a significant speedup. This result is crucial to curtail the computational time of the modeling of future chains of laser wakefield accelerator stages with high accuracy.

43 PARTICLE ACCELERATORS↗
NASA NTRS1994

Peculiar Behavior of He II Disturbances due to Sloshing Dynamics Driven by Jitter Accelerations Associated with Slew Motion in Microgravity

The mathematical formulation of Dewar container sloshing dynamics when partially filled with superfluid liquid He 11 driven by the gravity jitter acceleration associated with slew motion in a microgravity environment is studied. Numerical computation of the sloshing dynamics is based on the non-inertia container bound co-ordinate, and solving time-dependent, three-dimensional formulations of partial differential equations subject to initial and boundary conditions. This study discloses the capillary effect of the sloshing dynamics governing the liquid-vapour interface fluctuations driven by the gravity jitter acceleration associated with slew motion, which affects the stability of the fluid system in a microgravity environment. The peculiar behaviour of superfluid helium when reacting to the sloshing dynamics driven by the gravity jitter acceleration associated with slew motion in a microgravity environment is also investigated.

Hung, R. J.↗