Search NASA⌕ Search

SEARCH · Search NASA

Results for “High efficiency acceleration”

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 1,747 records · Page 97

Multi-resolution enhancement for full-spectrum neural representations

Scientific data acquisition continues to outpace storage and analysis capabilities, making voxel-basedrepresentations increasingly intractable. Implicit neural representations (INRs) offer a promising solutionby encoding signals through coordinate-based neural networks, serving as surrogates of data, withcomputational and storage requirements scaling with network complexity rather than data dimensionality.However, smaller INRs struggle to faithfully represent multiscale structures, high-frequency informationand fine textures that constitute a large proportion of scientific measurements. We propose WIEN-INR, atheoretically guided hierarchical INR framework that distributes modelling across resolution scales andenables improved representation capacity through a novel enhancement network to recover subtle details.This multiscale architecture allows smaller networks to retain the full spatial-frequency content of thesignal as well as preserve training efficiency and lower storage cost. Evaluated on distinct raw experimentalmeasurements across scales and complexities, WIEN-INR represents a practical step towards a broaderadoption of neural representations in scientific workflows, delivering compact, robust and high-fidelityrepresentations.

Ni, Yuan [SLAC National Accelerator Laboratory (SL↗

PABO: Pseudo Agent-Based Multi-Objective Bayesian Hyperparameter Optimization for Efficient Neural Accelerator Design

The ever increasing computational cost of Deep Neural Networks (DNN) and the demand for energy efficient hardware for DNN acceleration has made accuracy and hardware cost co-optimization for DNNs tremendously important, especially for edge devices. Owing to the large parameter space and cost of evaluating each parameter in the search space, manually tuning of DNN hyperparameters is impractical. Automatic joint DNN and hardware hyperparameter optimization is indispensable for such problems. Bayesian optimization-based approaches have shown promising results for hyperparameter optimization of DNNs. However, most of these techniques have been developed without considering the underlying hardware, thereby leading to inefficient designs. Further, the few works that perform joint optimization are not generalizable and mainly focus on CMOS-based architectures. In this work, we present a novel pseudo agent-based multiobjective hyperparameter optimization (PABO) for maximizing the DNN performance while obtaining low hardware cost. Compared to the existing methods, our work poses a theoretically different approach for joint optimization of accuracy and hardware cost and focuses on memristive crossbar based accelerators. PABO uses a supervisor agent to establish connections between the posterior Gaussian distribution models of network accuracy and hardware cost requirements. The agent reduces the mathematical complexity of the co-optimization problem by removing unnecessary computations and updates of acquisition functions, thereby achieving significant speed-ups for the optimization procedure. PABO outputs a Pareto frontier that underscores the trade-offs between designing high-accuracy and hardware efficiency. Our results demonstrate a superior performance compared to the state-of-the-art methods both in terms of accuracy and computational speed (~100x speed up).

Parsa, Maryam↗

High Power Electron Gun for Metal Additive Manufacturing Based on Superconducting Technology

E-beam additive manufacturing is a rapidly growingindustry. Electron guns capable of reaching higher energiesand higher average power can enable larger melt pools,higher production speed, and permit broader range ofmaterials for metal additive applications. Such electronguns require technologies beyond those normally achievedwith DC high voltage sources. As an example,superconducting radio frequency (SRF) technology is anoption to achieve higher energy and higher power withexcellent efficiency. However, current SRF technologyrequires complex cryogenic systems to safely handlecryogenics gases or liquids such as Helium. Suchcomplexity severely restricts their usage to R&D facilities.In this talk, we introduce and describe, for the first time, anovel technique based on SRF technology to build a highpower, high energy electron gun for advanced metaladditive manufacturing applications. Our design is based ona technique which eliminates liquid Helium entirely. Ourmethod dramatically reduces the complexity and thusallows building practical, efficient, and modern electronsources for additive manufacturing.

43 PARTICLE ACCELERATORS↗

Ion Acceleration by Laser Plasma Interaction from Cryogenic Micro Jets - Oral Presentation [Slides]

Processes that occur in extreme conditions, such as in the center of stars and large planets, can be simulated in the laboratory using facilities such as SLAC National Accelerator Laboratory and the Jupiter Laser Facility (JLF) at Lawrence Livermore National Laboratory (LLNL). These facilities allow scientists to investigate the properties of matter by observing their interactions with high power lasers. Ion acceleration from laser plasma interaction is gaining greater attention today due to its widespread potential applications, including proton beam cancer therapy and fast ignition for energy production. Typically, ion acceleration is achieved by focusing a high power laser on thin foil targets through a mechanism called Target Normal Sheath Acceleration. Based on research and recent experiments, we hypothesized that a pure liquid cryogenic jet would be an ideal target for this type of interaction, capable of producing the highest proton energies possible with today’s laser technologies. Furthermore, it would provide a continuous, pure target, unlike metal foils which are consumed in the interaction and easily contaminated. In an effort to test this hypothesis and investigate new, potentially more efficient mechanisms of ion acceleration, we used the 527 nm split beam, frequency-doubled TITAN laser at JLF. Data from the cryogenic jets was limited due to the flow of current up the jet into the nozzle during the interaction, heating the jet and damaging the orifice. However, we achieved a pure proton beam with an indication of a monoenergetic feature. Furthermore, data from gold and carbon wires showed surprising and interesting results. Preliminary analysis of data from two ion emission diagnostics, Thomson parabola spectrometers (TPs) and radio chromic films (RCFs), suggests that shockwave acceleration occurred rather than target normal sheath acceleration, the standard mechanism of ion acceleration. Upon completion of the experiment at TITAN, I researched the possibility of transforming our liquid cryogenic jets into droplet streams. This type of target should solve our problems with the jet as it will prevent the flow of exocurrent into the nozzle. It is also highly effective as it is even more mass-limited than standard cryogenic jets. Furthermore, jets break up spontaneously anyway. If we can control the breakup, we can synchronize the droplet emission with the laser pulses. In order to assist the team prepare for an experiment later this year, I familiarized myself with the physics and theory of droplet formation, calculated values for the required parameters, and ordered the required materials for modification of the jet. Future experiments will test these droplet streams and continue towards the goal of ion acceleration using cryogenic targets.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

A Spatiotemporal Indexing Approach for Efficient Processing of Big Array-Based Climate Data with MapReduce

Climate observations and model simulations are producing vast amounts of array-based spatiotemporal data. Efficient processing of these data is essential for assessing global challenges such as climate change, natural disasters, and diseases. This is challenging not only because of the large data volume, but also because of the intrinsic high-dimensional nature of geoscience data. To tackle this challenge, we propose a spatiotemporal indexing approach to efficiently manage and process big climate data with MapReduce in a highly scalable environment. Using this approach, big climate data are directly stored in a Hadoop Distributed File System in its original, native file format. A spatiotemporal index is built to bridge the logical array-based data model and the physical data layout, which enables fast data retrieval when performing spatiotemporal queries. Based on the index, a data-partitioning algorithm is applied to enable MapReduce to achieve high data locality, as well as balancing the workload. The proposed indexing approach is evaluated using the National Aeronautics and Space Administration (NASA) Modern-Era Retrospective Analysis for Research and Applications (MERRA) climate reanalysis dataset. The experimental results show that the index can significantly accelerate querying and processing (10 speedup compared to the baseline test using the same computing cluster), while keeping the index-to-data ratio small (0.0328). The applicability of the indexing approach is demonstrated by a climate anomaly detection deployed on a NASA Hadoop cluster. This approach is also able to support efficient processing of general array-based spatiotemporal data in various geoscience domains without special configuration on a Hadoop cluster.

big data↗

Similarity Metric for Data Optimization and Efficient Training of Reactive Machine Learning Force Fields for Hydrocarbon Radiolysis

Radiolysis is a common approach to sterilize polymers, chemically modify them for upcycling, and accelerate their decomposition for recycling purposes. Reactive molecular dynamics (MD) simulations provide a powerful tool to generate atomic-level trajectories of the reactive processes and quantify radiolytic chemical degradation pathways. For this, machine learning (ML) surrogate models for reactive force fields with quantum mechanical accuracy are now widely used, which require ML training data sets that can provide information on atomic environments for target chemical systems. However, radiolysis chemistry can be highly complex and diverse, which poses significant challenges for generating training data to parametrize ML models. In this regard, we developed a method for optimizing the training data set using a cosine similarity metric to help guide training set selection for radiolysis of polyethylene, a model hydrocarbon polymer, as well as to enhance the transferability of our reactive ML force field (MLFF) to a variety of molecular and polymeric systems. Our approach performs atom-by-atom comparisons between local atomic environments to pinpoint important data points associated with rare and localized events, such as radiolysis damage within structures. We apply this approach to train the Chebyshev Interaction Model for Efficient Simulation (ChIMES) MLFF model, which expresses the atomic interaction potentials in terms of linear combinations of many-body Chebyshev polynomials. We first show that our method can reduce our training set size by ∼70% while improving overall accuracy compared to more standard MD model fitting approaches. We then validate our optimum model against diverse hydrocarbon simulation data, including simple alkanes and systems with unsaturated carbon bonds, over a wide range of thermodynamic conditions. Finally, we use our ChIMES model to perform MD simulations of radiolytic damage with large-scale systems that help avoid system size effects. Overall, our approach yields an MD force field that retains most of the accuracy of the underlying quantum method while yielding many orders of improvement in computational efficiency. In conclusion, our efforts will have impact on future hydrocarbon polymer radiolysis studies, where the chemical details of the polymer–radiation interactions can have a strong effect on the resulting products observed in experiments.

Hydrocarbons↗

SU-F-T-266: Dynalogs Based Evaluation of Different Dose Rate IMRT Using DVH and Gamma Index

Purpose: This work investigates the impact of low and high dose rate on IMRT through Dynalogs by evaluating Gamma Index and Dose Volume Histogram. Methods: The Eclipse™ treatment planning software was used to generate plans on prostate and head and neck sites. A range of dose rates 300 MU/min and 600 MU/min were applied to each plan in order to investigate their effect on the beam ON time, efficiency and accuracy. Each plan had distinct monitor units per fraction, delivery time, mean dose rate and leaf speed. The DVH data was used in the assessment of the conformity and plan quality.The treatments were delivered on Varian™ Clinac 2100C accelerator equipped with 120 leaf millennium MLC. Dynalogs of each plan were analyzed by MATLAB™ program. Fluence measurements were performed using the Sun Nuclear™ 2D diode array and results were assessed, based on Gamma analysis of dose fluence maps, beam delivery statistics and Dynalogs data. Results: Minor differences found by adjusted R-squared analysis of DVH’s for all the plans with different dose rates. It has been also found that more and larger fields have greater time reduction at high dose rate and there was a sharp decrease in number of control points observed in dynalog files by switching dose rate from 300 MU/min to 600 MU/min. Gamma Analysis of all plans passes the confidence limit of ≥95% with greater number of passing points in 300 MU/min dose rate plans. Conclusion: The dynalog files are compatible tool for software based IMRT QA. It can work perfectly parallel to measurement based QA setup and stand-by procedure for pre and post delivery of treatment plan.

60 APPLIED LIFE SCIENCES↗

Accelerating high-order mesh optimization using finite element partial assembly on GPUs

In this paper we present a new GPU-oriented mesh optimization method based on high order finite elements. Our approach relies on node movement with fixed topology, through the Target-Matrix Optimization Paradigm (TMOP) and uses a global nonlinear solve over the whole computational mesh, i.e., all mesh nodes are moved together. A key property of the method is that the mesh optimization process is recast in terms of finite element operations, which allows us to utilize recent advances in the field of GPU-accelerated high order finite element algorithms. For example, we reduce data motion by using tensor factorization and matrix-free methods, which have superior performance characteristics compared to traditional full finite element matrix assembly and offer advantages for GPU based HPC hardware. Furthermore, we describe the major mathematical components of the method along with their efficient GPU-oriented implementation. In addition, we propose an easily reproducible mesh optimization test that can serve as a performance benchmark for the mesh optimization community.

97 MATHEMATICS AND COMPUTING↗

Towards performance portability in the Spark astrophysical magnetohydrodynamics solver in the Flash-X simulation framework

Simulations of core-collapse supernovae, and other astrophysical phenomena, are quintessential extreme-scale computing challenges. For core-collapse supernova simulations to be carried out by the ExaStar project under the Exascale Computing Project umbrella, a robust, efficient, and state-of-the-art magnetohydrodynamics solver is a critical requirement. In Flash-X, the primary software instrument for ExaStar, a new magnetohydrodynamics solver has been designed and implemented from the ground up to achieve accuracy and efficiency for simulations of complex astrophysical flows. This new solver, dubbed Spark, uses high-order spatial reconstruction, Runge-Kutta time integration, and an efficient cell-centered approach to satisfying the divergence-free condition for the magnetic fields. Spark was written to be optimized for data locality in cache hierarchy of CPUs. Since data locality optimizations for cache hierarchy are not directly compatible with those of accelerators, we have taken the approach of using program synthesis to avoid massive amounts of code replication that would be necessary if we were to maintain two different versions of the solver. Our program synthesis relies on a simple key-dictionary approach, implemented in python, that enables us to assemble the version of the solver suitable for the target hardware from code fragments identified by specific keys. In this work, we describe the data locality optimizations of the solver for CPUs and accelerators and the program synthesis tools that enable this portability. We also detail the parallel performance of Spark for both CPUs and accelerators.

97 MATHEMATICS AND COMPUTING↗

Bunch characteristics of an electron beam generated by a diamond secondary emitter amplifier

Electron bunches for high performance free electron lasers are subject to constraints on charge per bunch and pulse shape. A Diamond secondary emitter used in conjunction with a photocathode and drive laser has potential to enable a high brightness, high peak current photoinjector by increasing the effective quantum efficiency of the photocathode. A theoretical characterization of the bunches so produced has been heretofore absent. Using a combination of Monte Carlo and analytical models, the shape of the bunches, their transit time, and emission time constants are determined and shown to be sensitive to the accelerating field in the diamond flake, incident beam profile, doping, and surface conditions. Methods to allow for extension to regimes of technological interest in terms of diamond thickness, external field, and primary pulse shape are given.

07 ISOTOPE AND RADIATION SOURCES↗

Calibration of a microchannel plate based extreme ultraviolet grazing incident spectrometer at the Advanced Light Source

We present the design and calibration of a microchannel plate based extreme ultraviolet spectrometer. Calibration was performed at the Advance Light Source (ALS) at the Lawrence Berkeley National Laboratory (LBNL). This spectrometer will be used to record the single shot spectrum of radiation emitted by the tapered hybrid undulator (THUNDER) undulator installed at the LOASIS GeV-class laser-plasma-accelerator. The spectrometer uses an aberration-corrected concave grating with 1200 lines/mm covering 11-62 nm and a microchannel plate detector with a CsI coated photocathode for increased quantum efficiency in the extreme ultraviolet. A touch screen interface controls the grating angle, aperture size, and placement of the detector in vacuum, allowing for high-resolution measurements over the entire spectral range.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

Comparison of bulk and pitcher-catcher targets for laser-driven neutron production

Laser-driven d(d, n)-{sup 3}He beam-target fusion neutron production from bulk deuterated plastic (CD) targets is compared with a pitcher-catcher target scheme using an identical laser and detector arrangement. For laser intensities in the range of (1-3) x 10{sup 19} W cm{sup -2}, it was found that the bulk targets produced a high yield (5 x 10{sup 4} neutrons per steradian) beamed preferentially in the laser propagation direction. Numerical modeling shows the importance of considering the temperature adjusted stopping powers to correctly model the neutron production. The bulk CD targets have a high background target temperature leading to a reduced stopping power for the deuterons, which increases the probability of generating neutrons by fusion. Neutron production from the pitcher-catcher targets was not as efficient since it does not benefit from the reduced stopping power in the cold catcher target. Also, the inhibition of the deuteron acceleration by a proton rich contamination layer significantly reduces the pitcher-catcher neutron production.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Design and Modeling of Vacuum Pumping for Steady-State Pellet Fueling Systems

We report that future long-pulse magnetic confinement fusion reactors will require density and isotopic mixture control using steady-state repeating pellet injectors. For high-energy density burning plasmas, pellet velocities of 1 km/s and above will be required for sufficient plasma penetration to achieve high fueling efficiency. Currently, steady-state repeating injection systems utilize cryogenic extruder systems to produce an extrusion of solid deuterium or deuterium-tritium. In repeating light gas gun injectors, the solid extrusion is cut and simultaneously loaded into a barrel. Once loaded, a fast operating gas valve delivers a high pressure burst of gas to accelerate the pellet down the barrel and into the machine. This process takes ~10 ms to achieve. Adequate gas pumping of the extruder exhaust and injection line propellant gas collection chambers is necessary for optimal operation of the pellet fueling system. Excess solid from the extruder sublimates in an exhaust chamber. The gas pressure in the extruder exhaust chamber must remain low to maintain low heating loads on the cooling mechanism (cryorefrigerators or liquid helium flow) and to reduce thermal conduction to the extrusion. Pumping the injection line chambers is necessary to limit propellant gas flow into the machine. A numerical simulation code was created to predict temporal pumping performance for these repeating pellet injection systems. Lastly, this paper outlines the methods and assumptions used to create this model and compares results to the pellet injection system currently employed on DIII-D, the steady-state pellet injection system planned for the Wendelstein 7-X, and a brief analysis of the ITER conceptual pellet fueling system.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Low-Temperature Ozone Exposure Technique to Modulate the Stoichiometry of WO(x) Nanorods and Optimize the Electrochromic Performance

A low-temperature ozone exposure technique was employed for the post-treatment of WO{sub x} nanorod thin films fabricated from hot-wire chemical vapor deposition (HWCVD) and ultrasonic spray deposition (USD) techniques. The resulting films were characterized with x-ray diffraction (XRD), transmission electron microscopy (TEM), Raman spectroscopy, UV-vis-NIR spectroscopy and x-ray photoelectron spectroscopy (XPS). The stoichiometry and surface crystallinity of the WO{sub x} thin films were subsequently modulated upon ozone exposure and thermal annealing without particle growth. The electrochromic performance was studied in a LiClO{sub 4}-propylene carbonate electrolyte, and the results suggest that the low-temperature ozone exposure technique is superior to the traditional high-temperature thermal annealing (employed to more fully oxidize the WO{sub x}). The optical modulation at 670 nm was improved from 35% for the as-deposited film to 57% for the film after ozone exposure at 150 C. The coloration efficiency was improved and the switching speed to the darkened state was significantly accelerated from 18.0 s for the as-deposited film to 11.8 s for the film after the ozone exposure. The process opens an avenue for low-temperature and cost-effective manufacturing of electrochromic films, especially on flexible polymer substrates.

36 MATERIALS SCIENCE↗

IceCube constraints on fast-spinning pulsars as high-energy neutrino sources

Relativistic winds of fast-spinning pulsars have been proposed as a potential site for cosmic-ray acceleration from very high energies (VHE) to ultrahigh energies (UHE). We re-examine conditions for high-energy neutrino production, considering the interaction of accelerated particles with baryons of the expanding supernova ejecta and the radiation fields in the wind nebula. We make use of the current IceCube sensitivity in diffusive high-energy neutrino background, in order to constrain the parameter space of the most extreme neutron stars as sources of VHE and UHE cosmic rays. We demonstrate that the current non-observation of 10{sup 18} eV neutrinos put stringent constraints on the pulsar scenario. For a given model, birthrates, ejecta mass and acceleration efficiency of the magnetar sources can be constrained. When we assume a proton cosmic ray composition and spherical supernovae ejecta, we find that the IceCube limits almost exclude their significant contribution to the observed UHE cosmic-ray flux. Furthermore, we consider scenarios where a fraction of cosmic rays can escape from jet-like structures piercing the ejecta, without significant interactions. Such scenarios would enable the production of UHE cosmic rays and help remove the tension between their EeV neutrino production and the observational data.

79 ASTRONOMY AND ASTROPHYSICS↗

GPU-Accelerated Asynchronous Error Correction for Mixed Precision Iterative Refinement

In hardware-aware high performance computing, block- asynchronous iteration and mixed precision iterative refinement are two techniques that are applied to leverage the computing power of SIMD accelerators like GPUs. Although they use a very different approach for this purpose, they share the basic idea of compensating the convergence behaviour of an inferior numerical algorithm by a more efficient usage of the provided computing power. In this paper, we want to analyze the potential of combining both techniques. Therefore, we implement a mixed precision iterative refinement algorithm using a block-asynchronous iteration as an error correction solver, and compare its performance with a pure implementation of a block-asynchronous iteration and an iterative refinement method using double precision for the error correction solver. For matrices from theUniversity of FloridaMatrix collection,we report the convergence behaviour and provide the total solver runtime using different GPU architectures.

97 MATHEMATICS AND COMPUTING↗

High Energy Observations of X-Ray Binaries and Gamma-Ray Blazars

The report discusses the CGRO observation of X-ray binary systems and studies of gamma-ray blasars. Numerous authors have suggested mechanisms for particle acceleration within X-Ray Binary (XRB) systems. Among the acceleration mechanisms that have been applied are pulsar acceleration, shock acceleration at an accretion shock front, shock acceleration at a pulsar wind termination shock, plasma turbulence excited by the accretion flow, and a number of electrodynamic mechanisms. There are therefore many mechanisms which are capable of generating very energetic particles in the XRB environment. If the reports of TeV/PeV gamma-ray generation in XRBs are correct, then one can show that the accelerated particles must be hadrons and that the most likely gamma-ray production mechanism is the decay of collisionally-produced (or photoproduced) neutral pions. At these ultra-high energies, the emission is so strongly beamed that the target conditions are constrained by the requirement that the column depth be large enough to efficiently generate gamma-rays, but not so large that the gamma-rays are absorbed. These constraints naturally lead to models that explain the periodic, narrow duty-cycle pulses observed at TeV/PeV energies as arising from interactions with, either, the atmosphere of the binary companion, an accretion column, or an accretion disk. The production of these TeV/PeV gamma-rays by the decay of pions from "leading isobars" must also be accompanied by a more isotropic emission component in the EGRET energy band from the decay of slower pions (i.e. the "pionization" component). Since the attenuation of 35 MeV-1 GeV photons by photon-photon pair production is not likely to be significant in most XRBs, the TeV/PeV reports therefore strongly suggest sporadic emission in the EGRET energy band. One of the key unresolved issues for understanding AGN is the relationship between XBLs and RBLs.To test the "reunification" hypothesis, authors conducted a multiwavelength campaign during Cycle 7 of the CGRO observing program. The analysis of the cycle 4 through cycle 6 CGRO measurements was completed. Two previously unknown EGRET sources was discovered.

Vestrand, W. Thomas↗

Muon Collider Progress: Accelerators

A muon collider would be a powerful tool for exploring the energy-frontier with leptons, and would complement the studies now under way at the LHC. Such a device would offer several important benefits. Muons, like electrons, are point particles so the full center-of-mass energy is available for particle production. Moreover, on account of their higher mass, muons give rise to very little synchrotron radiation and produce very little beamstrahlung. The first feature permits the use of a circular collider that can make efficient use of the expensive rf system and whose footprint is compatible with an existing laboratory site. The second feature leads to a relatively narrow energy spread at the collision point. Designing an accelerator complex for a muon collider is a challenging task. Firstly, the muons are produced as a tertiary beam, so a high-power proton beam and a target that can withstand it are needed to provide the required luminosity of ~1 × 10{sup 34} cm{sup –2}s{sup –1}. Secondly, the beam is initially produced with a large 6D phase space, which necessitates a scheme for reducing the muon beam emittance (“cooling”). Finally, the muon has a short lifetime so all beam manipulations must be done very rapidly. The Muon Accelerator Program, led by Fermilab and including a number of U.S. national laboratories and universities, has undertaken design and R&D activities aimed toward the eventual construction of a muon collider. Design features of such a facility and the supporting R&D program are described.

43 PARTICLE ACCELERATORS↗