Search NASA⌕ Search

SEARCH · Search NASA

Results for “High Performance Networking”

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
All sourcesNASA NTRSDOE OSTI
DOE OSTI2025

Complementing Dynamical Downscaling With Super‐Resolution Convolutional Neural Networks

Despite advancements in Artificial Intelligence (AI) methods for climate downscaling, significant challenges remain for their practicality in climate research. Current AI-methods exhibit notable limitations, such as limited application in downscaling Global Climate Models (GCMs), and accurately representing extremes. To address these challenges, we implement an AI-based methodology using super-resolution convolutional neural networks (SRCNN), trained and evaluated on 40 years of daily precipitation data from a reanalysis and a high-resolution dynamically downscaled counterpart. The dynamical downscaled simulations, constrained using spectral nudging, enable the replication of historical events at a higher resolution. This allows the SRCNN to emulate dynamical downscaling effectively. Modifications, such as incorporating elevation data and data pre-processing enhances overall model performance, while using exponential and quantile loss functions improve the simulation of extremes. Our findings show SRCNN models efficiently and skillfully downscale precipitation from GCMs. Future work will expand this methodology to downscale additional variables for future climate projections.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2020

Prediction of organic homolytic bond dissociation enthalpies at near chemical accuracy with sub-second computational cost

Bond dissociation enthalpies (BDEs) of organic molecules play a fundamental role in determining chemical reactivity and selectivity. However, BDE computations at sufficiently high levels of quantum mechanical theory require substantial computing resources. In this paper, we develop a machine learning model capable of accurately predicting BDEs for organic molecules in a fraction of a second. We perform automated density functional theory (DFT) calculations at the M06-2X/def2-TZVP level of theory for 42,577 small organic molecules, resulting in 290,664 BDEs. A graph neural network trained on a subset of these results achieves a mean absolute error of 0.58 kcal mol-1 (vs DFT) for BDEs of unseen molecules. We further demonstrate the model on two applications: first, we rapidly and accurately predict major sites of hydrogen abstraction in the metabolism of drug-like molecules, and second, we determine the dominant molecular fragmentation pathways during soot formation.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2024

Hybrid polymer network cathode-enabled soluble-polysulfide-free lithium–sulfur batteries

Among the emerging 'beyond lithium-ion' technologies for maximized sustainability, lithium-sulfur (Li-S) is a favoured chemistry because of its exceptional energy density from the conversion of sulfur, an element in abundant supply. However, the dissolution of several intermediate polysulfides formed during conversion leads to rapid performance degradation over cycling. Here, in this study, we address this issue by sulfurizing a hybrid polymer network with polyphosphazene and carbon as a cathode for Li-S batteries. With rich sites to re-bond and adsorb dissociative sulfur species, this hybrid polymer network circumvents the formation of soluble polysulfides and enables a unique, reversible inserting conversion reaction. Thus, our cathode delivers both high capacity (~900 mAh g -1 cathode ) and excellent cycling stability in Li-S coin cells, with a pouch cell demonstration of projected energy density of ~300 Wh kg -1 and 84.9% capacity retention after 150 cycles. The strategy can be extended to other cost-effective, recyclable polymers, advancing sulfur-based batteries towards practical energy storage application.

25 ENERGY STORAGE↗
DOE OSTI2015

Binder-free highly conductive graphene laminate for low cost printed radio frequency applications

In this paper, we demonstrate realization of printable radio frequency identification (RFID) antenna by low temperature processing of graphene ink. The required ultra-low resistance is achieved by rolling compression of binder-free graphene laminate. With compression, the conductivity of graphene laminate is increased by more than 50 times compared to that of as-deposited one. Graphene laminate with conductivity of 4.3 × 10{sup 4 }S/m and sheet resistance of 3.8 Ω/sq (with thickness of 6 μm) is presented. Moreover, the formation of graphene laminate from graphene ink reported here is simple and can be carried out in low temperature (100 °C), significantly reducing the fabrication costs. A dipole antenna based on the highly conductive graphene laminate is further patterned and printed on a normal paper to investigate its RF properties. The performance of the graphene laminate antenna is experimentally measured. The measurement results reveal that graphene laminate antenna can provide practically acceptable return loss, gain, bandwidth, and radiation patterns, making it ideal for low cost printed RF applications, such as RFID tags and wearable wireless sensor networks.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
DOE OSTI2024

Assessment of Cloud-based Applications for Enabling a Scalable Riskinformed Predictive Maintenance Strategy

The current light-water reactor fleet uses time-based maintenance strategies to achieve high-capacity factors. But to make nuclear more competitive in the energy market, these reactors could utilize emerging artificial intelligence (AI) and cloud computing technologies to achieve a cost-effective, predictive-maintenance strategy. This paper presents discussion and results on the application of cloud computing in the nuclear industry. The technical viability of cloud computing was analyzed using data from a boiling-water reactor’s safety relief valve. The models were hosted on three different systems: a local personal computer, Idaho National Laboratory’s high-performance computer system, and Microsoft Azure. The data were loaded and processed, and two types of models were trained in an A/B fashion. Based on the speed at which these actions were completed, it was determined that cloud computing affords adequate computing resources. Additionally, the computing power can scale with the demanded load. To enable cloud computing in the existing fleet, additional sensors, networks, and other requirements must be implemented to ensure a smooth transition from current maintenance strategies. However, the benefit is that the plants no longer need to manage their own servers, software, cybersecurity, and information technology support staff for in-house data analytics purpose. Many of these features can be offloaded to the cloud provider for a potential cost savings. Demonstrating how AI can improve the maintenance and operation of non-safety-related systems seems the likely path forward for implementing AI and cloud computing resources inside nuclear power plants.

azure↗
DOE OSTI2024

Super Resolving Unrolled Neural Networks for Remote Sensing

In remote sensing systems, the capabilities of the system are constrained by the complex interactions between size, weight, and power (SWAP) of potential designs. In electro-optical (EO) systems, examples of these critical parameters include the system’s sensitivity and resolution. Those parameters can be increased by ever larger optical apertures and focal planes but at the cost of more SWAP. Multi-image super resolution (MISR) techniques allow resolution to be enhanced via computation rather than more sophisticated optical hardware. These algorithms combine multiple images together into a single, higher resolution image, trading temporal resolution and computation for spatial resolution. Fielded MISR techniques, such as Drizzle, can require several hundred images to create a single super resolved image, implying reduced temporal resolution, increased data acquisition load, and limiting mission applications. Iterative techniques, such as model-based image reconstruction and compressive sensing, have been shown to create super resolved images using fewer images than Drizzle. They do this by posing an optimization problem that balances accuracy between a highly accurate physical model and an image model. In the case of super resolution, the physical model is defined by the relation between low resolution input images and the desired high resolution output image. The image model encodes some assumptions about the super resolved image. These assumptions are meant to suppress reconstruction artifacts that arise due to deterministic physical model error, stochastic measurement noise, and potential undersampling. In practice, the performance of iterative methods are limited by imaging models compatible with optimization. Deep learning-based methods can effectively learn image models of arbitrary complexity, but lack the theoretical explainability and robustness of iterative techniques. Consensus equilibrium (CE) generalizes the iterative techniques beyond optimization, enabling blackbox algorithms such as traditional and neural image denoisers to be used as the image model. CE-based approaches retain much of the explainability and robustness of iterative techniques while allowing the expressiveness of machine learning image models to be used. Additionally, by unrolling iterations of CE with an embedded image denoiser, the image denoiser can be further trained and specialized to the specific application with potentially higher quality reconstructions. Under this project, we demonstrated the feasibility of training an unrolled neural network based upon CE. While we didn’t train one, we showed that the CE process is differentiable and its gradient can be tractably computed. We also explored the usage of a variants of CE akin to generative neural works. Most importantly, we applied the CE framework to a number of problems including non-blind deconvolution, upsampling, single-image super resolution, MISR, event-based sensing, and saturated deconvolution. Our MISR prototype creates high quality reconstructions with an order of magnitude fewer images than previous approaches and, critically, produces these reconstructions fast enough for practical usage.

47 OTHER INSTRUMENTATION↗
DOE OSTI2023

Abisko: Deep codesign of an architecture for spiking neural networks using novel neuromorphic materials

The Abisko project aims to develop an energy-efficient spiking neural network (SNN) computing architecture and software system capable of autonomous learning and operation. The SNN architecture explores novel neuromorphic devices that are based on resistive-switching materials, such as memristors and electrochemical RAM. Equally important, Abisko uses a deep codesign approach to pursue this goal by engaging experts from across the entire range of disciplines: materials, devices and circuits, architectures and integration, software, and algorithms. Here, the key objectives of our Abisko project are threefold. First, we are designing an energy-optimized high-performance neuromorphic accelerator based on SNNs. This architecture is being designed as a chiplet that can be deployed in contemporary computer architectures and we are investigating novel neuromorphic materials to improve its design. Second, we are concurrently developing a productive software stack for the neuromorphic accelerator that will also be portable to other architectures, such as field-programmable gate arrays and GPUs. Third, we are creating a new deep codesign methodology and framework for developing clear interfaces, requirements, and metrics between each level of abstraction to enable the system design to be explored and implemented interchangeably with execution, measurement, a model, or simulation. As a motivating application for this codesign effort, we target the use of SNNs for an analog event detector for a high-energy physics sensor.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2015

Model-Based Integrated High Penetration Renewables Planning and Control Analysis

Increasing adoption of Solar Photovoltaic (PV) generation at the distribution level poses several changes for the reliable operation of electrical power distribution systems. The addition of a significant amount of PV to a distribution network can introduce a variety of operational problems, including steady-state overvoltages, reverse flows, voltage flicker and excessive controller movement among others. These adverse impacts can be mitigated through a variety of equipment upgrades which represent a cost to either the electric utility or the owner of the PV site. The study performed here aims to quantify the levels of PV generation which present operational problems on a distribution circuit and how those problems might be alleviated. The study performed here included 20 distribution feeders selected from Pepco Holdings, Inc. (PHI) service territory. These feeders are located in the states of Delaware, Maryland and New Jersey. A hosting capacity study was performed on each feeder to determine how much additional PV it could support in its current configuration. Several improvements were then performed on these circuits including phase balancing, capacitor redesign, reducing the voltage regulator set points, fixed power factor operation on the PV inverters and the installation of battery storage. After each of these improvements the hosting capacity of the circuit was reevaluated in order to determine how that particular improvement impacted the amount of PV that could be hosted by the circuit. Each of these improvements represents a real cost in terms of labor and equipment in order to be implemented. They are expected to provide a benefit in terms of the amount of additional PV generation which can be safely interconnected to the distribution feeder. A cost benefit analysis was performed in order to evaluate the expected costs of each feeder improvement and how each one was able to increase the PV hosting capacity of each feeder. It is hoped that these results can be utilized by other distribution utilities in order to understand how they can improve the hosting capacity of their feeders and facilitate the deployment of more PV generation at the distribution level. Admittedly other utility companies will most likely have different feeder architectures and differing labor and equipment costs so the realized cost benefit numbers may be significantly different.

14 SOLAR ENERGY↗
DOE OSTI2025

Multi-frequency progressive refinement for learned inverse scattering

Interpreting scattered acoustic and electromagnetic wave patterns is a computational task that enables remote imaging in a number of important applications, including medical imaging, geophysical exploration, sonar and radar detection, and nondestructive testing of materials. However, accurately and stably recovering an inhomogeneous medium from far-field scattered wave measurements is a computationally difficult problem, due to the nonlinear and non-local nature of the forward scattering process. We design a neural network, called Multi-Frequency Inverse Scattering Network (MFISNet), and a training method to approximate the inverse map from far-field scattered wave measurements at multiple frequencies. We consider three variants of MFISNet, with the strongest performing variant inspired by the recursive linearization method — a commonly used technique for stably inverting scattered wavefield data — that progressively refines the estimate with higher frequency content. MFISNet outperforms past methods in regimes with high-contrast, heterogeneous large objects, and inhomogeneous unknown backgrounds.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2024

Anomaly detection with flow-based fast calorimeter simulators

Recently, several normalizing flow-based deep generative models have been proposed to accelerate the simulation of calorimeter showers. Using caloflow as an example, we show that these models can simultaneously perform unsupervised anomaly detection with no additional training cost. As a demonstration, we consider electromagnetic showers initiated by one (background) or multiple (signal) photons. The caloflow model is designed to generate single-photon showers, but it also provides access to the shower likelihood. We use this likelihood as an anomaly score and study the showers tagged as being unlikely. As expected, the tagger struggles when the signal photons are nearly collinear but is otherwise effective. This approach is complementary to a supervised classifier trained on only specific signal models using the same low-level calorimeter inputs. While the supervised classifier is also highly effective at unseen signal models, the unsupervised method is more sensitive in certain regions, and, thus, we expect that the ultimate performance will require a combination of these approaches.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
DOE OSTI2020

Foliar functional traits from imaging spectroscopy across biomes in eastern North America

Summary Foliar functional traits are widely used to characterize leaf and canopy properties that drive ecosystem processes and to infer physiological processes in Earth system models. Imaging spectroscopy provides great potential to map foliar traits to characterize continuous functional variation and diversity, but few studies have demonstrated consistent methods for mapping multiple traits across biomes. With airborne imaging spectroscopy data and field data from 19 sites, we developed trait models using partial least squares regression, and mapped 26 foliar traits in seven NEON (National Ecological Observatory Network) ecoregions (domains) including temperate and subtropical forests and grasslands of eastern North America. Model validation accuracy varied among traits (normalized root mean squared error, 9.1–19.4%; coefficient of determination, 0.28–0.82), with phenolic concentration, leaf mass per area and equivalent water thickness performing best across domains. Across all trait maps, 90% of vegetated pixels had reasonable values for one trait, and 28–81% provided high confidence for multiple traits concurrently. Maps of 26 traits and their uncertainties for eastern US NEON sites are available for download, and are being expanded to the western United States and tundra/boreal zone. These data enable better understanding of trait variations and relationships over large areas, calibration of ecosystem models, and assessment of continental‐scale functional diversity.

Wang, Zhihui↗
DOE OSTI2018

Multi-Task Convolutional Neural Networks for Natural Text Classification

Information extraction and coding of free-text pathology reports is an important activity for cancer registries to support national cancer surveillance. Cancer registrars must process high volumes of pathology reports on an annual basis. We investigated an automated approach using a convolutional neural networks (CNNs) for extracting the primary site, laterality, histology, behavior, and histological grade from unstructured cancer pathology text reports. Our multi-task learning (MTL) with hard parameter sharing approach is used to train a multi-task MT-CNN model for all the tasks. The performance of our proposed approach was compared against a state of-the-art CNN and the commonly used SVM classifier. We observed that the proposed model consistently outperformed the baseline models, especially for the less prevalent classes. Also, the results demonstrate the potential of the proposed method for handling class imbalance within each task.

Yoon, HongJun↗
DOE OSTI2019

Scripts for Running Multiple Lines of Evidence Analysis

The software code titled “Scripts for Running Multiple Lunes of Evidence Analysis” enables high throughput network analysis; efficient analysis of large multi-omic networks in order to rank genes by connectivity. This code creates a set of scripts that take omic networks as input as well as list of known genes and calculates scores for all other genes in genome based on connectivity to those genes and does so accurately and efficiently. In addition, it aids the user in performing a validation check of observed data to random. The code outputs a table of all genes in the genome with LOE scores and scores observed data versus random.

Furches, AnnaK.↗
DOE OSTI2018

Lg-Wave Cross Correlation and Epicentral Double-Difference Location in and near China

In this paper, we perform epicentral relocations for a broad area using cross-correlation measurements made on Lg waves recorded at regional distances on a sparse station network. Using a two-step procedure (pairwise locations and cluster locations), we obtain final locations for 5623 events—3689 for all of China from 1985 to 2005 and 1934 for the Wenchuan area from May to August 2008. These high-quality locations comprise 20% of a starting catalog for all of China and 25% of a catalog for Wenchuan. Of the 1934 events located for Wenchuan, 1662 (86%) were newly detected. The final locations explain the residuals 89 times better than the catalog locations for all of China (3.7302–0.0417 s) and 32 times better than the catalog locations for Wenchuan (0.8413–0.0267 s). The average semimajor axes of the 95% confidence ellipses are 420 m for all of China and 370 m for Wenchuan. The average azimuthal gaps are 205° for all of China and 266° for Wenchuan. 98% of the station distances for all of China are over 200 km. The mean and maximum station distances are 898 and 2174 km. The robustness of our location estimates and various trade-offs and sensitivities is explored with different inversion parameters for the location, such as starting locations for iterative solutions and which singular values to include. Finally, our results provide order-of-magnitude improvements in locations for event clusters, using waveforms from a very sparse far-regional network for which data are openly available.

58 GEOSCIENCES↗
DOE OSTI2021

Machine Learning Training for HOM reduction and Emittance Preservation in a TESLA-type Cryomodule at FAST

Low emittance electron beams are of high importance at facilities like the LCLS-II at SLAC. Emittance dilution effects due to off-axis beam transport for a TESLA-type cryomodule (CM) have been shown at the Fermilab Accelerator Science and Technology facility. The results showed the correlation between the electron beam-induced cavity high-order modes (HOMs) and submacropulse centroid slewing and oscillation downstream of the CM. Mitigation of emittance dilution can be achieved by reducing the HOM signals and the variances in the submacropulse beam positions downstream of the CM. Here we present a Machine Learning based optimization and model construction for HOM signal level reduction using Neural Networks and Gaussian Processes. To gather training data we performed experiments using single bunch and 50 bunch electron beams with charges up to 125 pC/b. We measured HOM signals of all cavities and beam position with a set of BPMs downstream of the CM. The beam trajectory was changed using V/H125 corrector set located upstream of the CM. The results presented here will inform the LCLS-II injector commissioning and will serve as a prototype for HOM reduction and emittance preservation.

43 PARTICLE ACCELERATORS↗
DOE OSTI2025

Track Reconstruction using Graph Neural Networks in the EMPHATIC Experiment

Track reconstruction is essential for extracting physics observables from detector data in high-energy and nuclear physics experiments. In this work, we investigate the use of graph neural networks (GNNs) to reconstruct particle momentum in the EMPHATIC experiment using simulated data. The model takes raw hit information from the silicon strip detectors (SSDs) as input and is trained to predict momentum components and the scattering angle of the particle. We describe the GNN architecture, training procedure, and performance metrics, and present results showing improved resolution in momentum reconstruction. These results demonstrate the potential of GNN-based approaches in track reconstruction tasks within complex detector environments like EMPHATIC.

Bhattarai, Aayush [Notre Dame U.]↗
DOE OSTI2017

Scaling Deep Learning on GPU and Knights Landing clusters

The speed of deep neural networks training has become a big bottleneck of deep learning research and development. For example, training GoogleNet by ImageNet dataset on one Nvidia K20 GPU needs 21 days. To speed up the training process, the current deep learning systems heavily rely on the hardware accelerators. However, these accelerators have limited on-chip memory compared with CPUs. To handle large datasets, they need to fetch data from either CPU memory or remote processors. We use both self-hosted Intel Knights Landing (KNL) clusters and multi-GPU clusters as our target platforms. From an algorithm aspect, current distributed machine learning systems are mainly designed for cloud systems. These methods are asynchronous because of the slow network and high fault-tolerance requirement on cloud systems. We focus on Elastic Averaging SGD (EASGD) to design algorithms for HPC clusters. Original EASGD used round-robin method for communication and updating. The communication is ordered by the machine rank ID, which is inefficient on HPC clusters. First, we redesign four efficient algorithms for HPC systems to improve EASGD's poor scaling on clusters. Async EASGD, Async MEASGD, and Hogwild EASGD are faster \textcolor{black}{than} their existing counterparts (Async SGD, Async MSGD, and Hogwild SGD, resp.) in all the comparisons. Finally, we design Sync EASGD, which ties for the best performance among all the methods while being deterministic. In addition to the algorithmic improvements, we use some system-algorithm codesign techniques to scale up the algorithms. By reducing the percentage of communication from 87% to 14%, our Sync EASGD achieves 5.3x speedup over original EASGD on the same platform. We get 91.5% weak scaling efficiency on 4253 KNL cores, which is higher than the state-of-the-art implementation.

60 APPLIED LIFE SCIENCES↗
DOE OSTI2017

Comoving Stars in Gaia DR1: An Abundance of Very Wide Separation Comoving Pairs

The primary sample of the Gaia Data Release 1 is the Tycho-Gaia Astrometric Solution (TGAS): ≈2 million Tycho-2 sources with improved parallaxes and proper motions relative to the initial catalog. This increased astrometric precision presents an opportunity to find new binary stars and moving groups. We search for high-confidence comoving pairs of stars in TGAS by identifying pairs of stars consistent with having the same 3D velocity using a marginalized likelihood ratio test to discriminate candidate comoving pairs from the field population. Although we perform some visualizations using (bias-corrected) inverse parallax as a point estimate of distance, the likelihood ratio is computed with a probabilistic model that includes the covariances of parallax and proper motions and marginalizes the (unknown) true distances and 3D velocities of the stars. We find 13,085 comoving star pairs among 10,606 unique stars with separations as large as 10 pc (our search limit). Some of these pairs form larger groups through mutual comoving neighbors: many of these pair networks correspond to known open clusters and OB associations, but we also report the discovery of several new comoving groups. Most surprisingly, we find a large number of very wide (>1 pc) separation comoving star pairs, the number of which increases with increasing separation and cannot be explained purely by false-positive contamination. Our key result is a catalog of high-confidence comoving pairs of stars in TGAS. We discuss the utility of this catalog for making dynamical inferences about the Galaxy, testing stellar atmosphere models, and validating chemical abundance measurements.

79 ASTRONOMY AND ASTROPHYSICS↗