SEARCH · Engineering Papers

Results for “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.

SourceResearch connectionFrom yearThrough year

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 1,297 records · Page 72

All sources NASA NTRS DOE OSTI

DOE OSTI2020

An intra-node HPC network architecture with nanosecond-scale photonic switches

We propose a single-stage network architecture for intra-node connectivity that makes use of nanosecond-scale photonic switches. Although buffering at the switch points is of vital importance for complex multi-stage networks, this is not the case for smaller-scale single-stage networks where the end nodes are located only one hop apart. By limiting the buffering to the end points, the proposed architecture manages to minimize the required electro-optic and opto-electronic conversions, leading in this way to both low end-to-end latency and better energy efficiency. Combining these advantages with nanosecond-scale switching times can allow for high-throughput operation even for frequent switch reconfigurations. The performance of the proposed architecture is evaluated via discrete-event simulations for a wide range of synthetic-traffic cases. The simulation results show that high-throughput operation of ≥90% can be achieved even for small message sizes, i.e., 32 KB for all-to-all communication and 2 KB for uniform random traffic, at a data rate of 400 Gb/s and a switch reconfiguration time of ≤72 ns. Moreover, if 100-ns reconfiguration times are achievable as opposed to 150-ns, then for the all-to-all traffic case a 16% and 28% reduction in completion time can be achieved for message sizes of 8 KB and 1 KB, respectively. In a forthcoming era of optically interfaced processors and accelerators, nanosecond-scale photonic switches appear as a highly promising solution for keeping up with the intra-node bandwidth scaling due to their high-bandwidth, low-latency and fast-switching capabilities.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗

DOE OSTI2022

Broadband polarization-entangled source for C+L-band flex-grid quantum networks

The rising demand for transmission capacity in optical networks has motivated steady interest in expansion beyond the standard C-band (1530–1565 nm) into the adjacent L-band (1565–1625 nm) for an approximate doubling of capacity in a single stroke. However, in the context of quantum networking, the L-band has yet to be fully leveraged with the suite of advanced tools for characterization and management available from classical lightwave communications. In this work, we demonstrate an ultrabroadband two-photon source integrating both C- and L-band wavelength-selective switches for complete control of spectral routing and allocation across 7.5 THz in a single setup. Polarization state tomography of all 150 pairs of 25-GHz-wide channels reveals an average fidelity of 0.98 and total distillable entanglement greater than 181 kebits/s. This source is explicitly designed for flex-grid optical networks and can facilitate optimal utilization of entanglement resources across the full C+L-band.

42 ENGINEERING↗

DOE OSTI2021

Adaptive bandwidth management for entanglement distribution in quantum networks

Flexible grid wavelength division multiplexing is a powerful tool in lightwave communications to maximize spectral efficiency. In the emerging field of quantum networking, the need for effective resource provisioning is particularly acute, given the generally lower power levels, higher sensitivity to loss, and inapplicability of optical detection and retransmission. In this letter, we leverage flex grid technology to demonstrate reconfigurable distribution of quantum entanglement in a four-user tabletop network. By adaptively partitioning bandwidth with a single wavelength-selective switch, we successfully equalize two-party coincidence rates that initially differ by over two orders of magnitude. Our scalable approach introduces loss that is fixed with the number of users, offering a practical path for the establishment and management of quality-of-service guarantees in large quantum networks.

42 ENGINEERING↗

DOE OSTI2024

Investigating the relationship between microbial network features of giant kelp “seedbank” cultures and subsequent farm performance

Microbial inoculants can increase the yield of cultivated crops and are successful in independent trials; however, efficacy drops in large-scale applications due to insufficient consideration of microbial community dynamics. The structure of microbiomes, in addition to the impact of individual taxa, is an important factor to consider when designing growth-promoting inoculants. Here, we investigate the microbial network and community assembly patterns of Macrocystis pyrifera gametophyte germplasm cultures (collectively referred to as a “seedbank”) used to cultivate an offshore farm in Santa Barbara, California, and identify network features associated with increased biomass of mature sporophytes. We found that [1] several network features, such as clustering coefficient and edge ratios, significantly vary with biomass outcomes; [2] gametophytes that become low- or high-biomass sporophytes have different hub taxa; and [3] microbial community assembly of gametophyte germplasm cultures is niche-driven. Overall, this study describes microbial community dynamics in M . pyrifera germplasm cultures and ultimately supports the development of early life stage inoculants that can be used on seaweed cultivars to increase biomass yield.

Osborne, Melisa G. (ORCID:0000000154980057)↗

DOE OSTI2023

Experimental Observations of the Topology of Convolutional Neural Network Activations

Topological data analysis (TDA) is a branch of computational mathematics, bridging algebraic topology and data science, that provides compact, noise-robust representations of complex structures. Deep neural networks (DNNs) learn millions of parameters associated with a series of transformations defined by the model architecture resulting in high-dimensional, difficult to interpret internal representations of input data. As DNNs become more ubiquitous across multiple sectors of our society, there is increasing recognition that mathematical methods are needed to aid analysts, researchers, and practitioners in understanding and interpreting how these models' internal representations relate to the final classification. In this paper we apply cutting edge techniques from TDA with the goal of gaining insight towards interpretability of convolutional neural networks used for image classification. We use two common TDA approaches to explore several methods for modeling hidden layer activations as high-dimensional point clouds, and provide experimental evidence that these point clouds capture valuable structural information about the model's process. First, we demonstrate that a distance metric based on persistent homology can be used to quantify meaningful differences between layers and discuss these distances in the broader context of existing representational similarity metrics for neural network interpretability. Second, we show that a mapper graph can provide semantic insight as to how these models organize hierarchical class knowledge at each layer. These observations demonstrate that TDA is a useful tool to help deep learning practitioners unlock the hidden structures of their models.

topological data analysis, deep learning↗

DOE OSTI2020

Recovery and Calibration of Legacy Underground Nuclear Test Seismic Data from the Leo Brady Seismic Network

The Leo Brady Seismic Network (LBSN, originally the Sandia Seismic Network) was established in 1960 by Sandia National Laboratories to monitor underground nuclear tests (UGTs) at the Nevada National Security Site (NNSS, formerly named the Nevada Test Site). The LBSN has been in various configurations throughout its existence, but it has generally been comprised of four to six stations at regional distances (~150–400 km) from the NNSS with approximately evenly spaced azimuthal coverage. Between 1962 and the end of nuclear testing in 1992, the LBSN—and a sister network operated by Lawrence Livermore National Laboratories—was the most comprehensive United States source of regional seismic data of UGTs. Approximately 75% of all UGTs performed by the United States occurred in the predigital era. At that time, LBSN data were transmitted as frequency-modulated (FM) audio over telephone lines to a central location and recorded as analog waveforms on high-fidelity magnetic audio tapes. These tapes have been in dry temperature-stable storage for decades and contain the sole record of this irreplaceable data; full waveforms of LBSN-recorded UGTs from this era were not routinely digitized or otherwise published. We have developed a process to recover and calibrate data from these tapes. First, we play back and digitize the tapes as audio. Next, we demodulate the FM “audio” into individual waveforms. We then estimate the various instrument constants through careful measurement of “weight-lift” tests performed prior to each UGT on each instrument. Finally, these coefficients allow us to scale and shape the derived instrument response of the seismographs and compute poles and zeros. Finally, the result of this process is a digital record of the recorded seismic ground motion in a modern data format, stored in a searchable database. To date, we have digitized tapes from 592 UGTs.

58 GEOSCIENCES↗

DOE OSTI2025

Reducing the Parameter Dependency of Phase-Picking Neural Networks with Dice Loss

Training a neural network for picking seismic phase arrivals has been commonly posed as a segmentation problem. It is a highly imbalanced segmentation problem in the sense that the background vastly dominates the foreground because we are trying to pick the optimal single sample point that represents the arrival of a seismic phase in a many seconds long time window. Here, we test the Dice loss, which is a preferred loss function for highly imbalanced image segmentation problems. We show that phase-picking neural networks trained on the Dice loss behave in a binary fashion for which the prediction output is almost always either nearly 1 or nearly 0. This feature removes the strong dependence of data processing workflows on the prediction score threshold, which is an otherwise critical parameter to determine when using neural networks trained on the cross-entropy loss. When strategically used, models trained on the Dice loss can reduce the parameter dependency of machine learning-based seismic monitoring.

58 GEOSCIENCES↗

DOE OSTI2022

Fluid – Fluid Interfacial Area and Its Impact on Relative Permeability - A Pore Network Modeling Study

Abstract Relative permeability (kr) is commonly modeled as an empirical function of phase saturation. Although current empirical models can provide a good match of one or two measured relative permeabilities using saturation alone, they are unable to predict relative permeabilities well when there is hysteresis or when physical properties such as wettability change. Further, current models often result in relative permeability discontinuities that can cause convergence and accuracy problems in simulation. To overcome these problems, recent research has modeled relative permeability as a state function of both saturation (S) and phase connectivity (X). Pore network modeling (PNM) data, however, shows small differences in relative permeability for the same S-X value when approached from a different flow direction. This paper examines the impact of one additional Minkowski parameter (Mecke and Arns, 2005), the fluid-fluid interfacial area, on relative permeability to identify if that satisfactorily explains this discrepancy. We calculate the total fluid-fluid interfacial areas (IA) during two-phase (oil/water) flow in porous media using pore network modeling. The area is calculated from PNM simulations using the areas associated with corners and throats in pore elements of different shapes. The pore network is modeled after a Bentheimer sandstone, using square, triangular, and circular pore shapes. Simulations were conducted for numerous primary drainage and imbibition cycles at a constant contact angle of 0° for the wetting phase. Simultaneous measurements of capillary pressure, relative permeability, saturation, and phase connectivity are made for each displacement. Fluid-fluid interfacial area is calculated from the PNM capillary pressure, the fluid location in the pore elements, and the pore element dimensional data. The results show that differences in the relative permeability at the same (S,X) point is explained well by differences in the fluid-fluid interfacial area (IA). That is, for a larger change in IA at these intersection points, the permeability difference is greater. That difference in relative permeability approaches zero as the difference in IA approaches zero. This confirms that relative permeability can be modeled better as a unique function of S, X, and IA. The results also show that an increase in IA restricts flow decreasing the nonwetting (oil) phase permeability. This decrease is caused by an increase in the throat area fraction compared to the corner area as the total area IA increases. The wetting phase relative permeability, however, shows the inverse trend, in that its relative permeability is greater when IA becomes larger owing to a greater fraction of the total area associated with the corners. The area IA, however, impacts the nonwetting phase relative permeability more than the wetting phase relative permeability. Corner flow improves the wetting phase relative permeability because the wetting phase is continuous there. Finally, a sensitivity analysis shows that relative permeability a is more sensitive to change in S than they are for IA for the case studied implying that if only two parameters are used to model relative permeability it is better to choose S and X.

Engineering↗

DOE OSTI2024

How to understand limitations of generative networks

Well-trained classifiers and their complete weight distributions provide us with a well-motivated and practicable method to test generative networks in particle physics. We illustrate their benefits for distribution-shifted jets, calorimeter showers, and reconstruction-level events. In all cases, the classifier weights make for a powerful test of the generative network, identify potential problems in the density estimation, relate them to the underlying physics, and tie in with a comprehensive precision and uncertainty treatment for generative networks.

Physics↗

DOE OSTI2024

Multipartite edge modes and tensor networks

Holographic tensor networks model AdS/CFT, but so far they have been limited by involving only systems that are very different from gravity. Unfortunately, we cannot straightforwardly discretize gravity to incorporate it, because that would break diffeomorphism invariance. In this note, we explore a resolution. In low dimensions gravity can be written as a topological gauge theory, which can be discretized without breaking gauge-invariance. However, new problems arise. Foremost, we now need a qualitatively new kind of “area operator,” which has no relation to the number of links along the cut and is instead topological. Secondly, the inclusion of matter becomes trickier. We successfully construct a tensor network both including matter and with this new type of area. Notably, while this area is still related to the entanglement in “edge mode” degrees of freedom, the edge modes are no longer bipartite entangled pairs. Instead they are highly multipartite. Along the way, we calculate the entropy of novel subalgebras in a particular topological gauge theory. We also show that the multipartite nature of the edge modes gives rise to non-commuting area operators, a property that other tensor networks do not exhibit.

Akers, Chris (ORCID:0000000227929827)↗

DOE OSTI2020

TRANSENSOR: Transformer Real-time Assessment INtelligent System with Embedded Network of Sensors and Optical Readout. Final Report

Utilities across the world are wrestling with evolving market dynamics, population growth and climate change. Distributed Energy Resources (DERs) such as solar photovoltaics, distributed generators and energy storage systems are becoming important parts of the U.S. energy mix. These factors are driving an increasing need for low-cost grid asset monitoring. Aside from being costly, traditional utility monitoring systems are not sufficiently robust and do not provide real-time visibility into the condition of grid assets such as transformers. Lack of accurate real-time measurements on performance has resulted in the use of lagging indicators, such as oil sample analysis. To address this need, an innovative embedded optical sensing technology, Transformer Real-time Assessment Intelligent System (TRANSENSOR) was developed, validated and demonstrated in this project. To date, TRANSENSOR has focused on transformers but is extendable to other grid assets. In addition to the technology being embedded into new transformers during manufacturing, a retrofit configuration that can be installed on existing transformers in the field was also developed. It is anticipated that the ability to retrofit existing transformers will help accelerate adoption of the underlying TRANSENSOR technology. Phase 1 of the project focused on laboratory development and qualification of the technology. Following iterations and exploration of relevant optical sensing modalities and multiplexed configurations of interest for the transformer environment, an effective candidate configuration was agreed upon, down-selected and custom-designed for embedding into General Electric (GE) network transformers. Two new GE 500 kilovolt ampere (kVA), 27 kilovolt (kV) distribution network transformers were built with embedded fiber-optic (FO) sensors and successfully qualified per industry standards at GE’s Shreveport, Louisiana facility during Phase 1 of the project. Following the successful completion of Phase 1, the team proceeded to Phase 2, which focused on a field demonstration of the technology. Over Phase 2, the first new GE transformer built with embedded fiber-optic sensors was installed in an above-ground cage and connected to the grid at Con Edison’s Astoria facility. A second transformer equipped with TRANSENSOR was installed in an underground vault. Additionally, an older (1982 year model) GE transformer in an above-ground cage was retrofitted with fiber- optic sensors and reconnected to the grid at ConEd’s facility. Analysis of data acquired from the sensors showed interesting correlations with transformer loading. Additionally, key events such as transformer low-voltage network connection, primary-side energizing, and a pressure loss event from an oil sampling were detected by the TRANSENSOR system. Online data processing/feature extraction algorithms for the second transformer in the underground vault detected key features and event alerts that were transmitted through a wireless 4G connection. The remote deployment concept showed promising results with data collection running for a total of 8 months across the three (3) GE transformers instrumented for the Phase 2 field demonstration. This sets the stage well for further development and commercialization.

24 POWER TRANSMISSION AND DISTRIBUTION↗

DOE OSTI2020

Electric Power Grid and Natural Gas Network Operations and Coordination

In this report, we present findings from three studies related to the coordination of natural gas and electricity system operations. We first propose and demonstrate a modeling platform for examining the interdependence of natural gas and electricity networks based on a direct current unit-commitment and economic dispatch model for the power system and a transient hydraulic gas model for the gas system. We use this platform to analyze the value of day-ahead coordination of power and natural gas network operations and to show the importance of considering gas system constraints when analyzing power systems operation with high penetration of gas generators and variable renewable energy sources. In the second study, we utilize our modeling platform to consider the U.S. Federal Energy Regulatory Commission (FERC) Order 809, issued in 2015 to improve day-ahead and intraday coordination of power and gas systems. Finally, in the third study we expand our modeling platform to focus on market-based coordination of electricity and natural gas system operations for a real system, namely a subset of the power and gas networks in the Front Range region of Colorado. We use real system data to evaluate the benefits of coordination operations under different conditions, including different levels of renewable penetration and the use of time-variant, shaped flow nominations. Our results indicate that coordination at various timescales can contribute to a reduction in curtailed gas in high-stress periods (such as those with large ramps in gas offtakes) and a reduction in energy consumption of gas compressor stations. We find that intraday coordination can reduce total power system production costs and natural gas deliverability constraints, yielding cost and reliability benefits. We observe these benefits for the test system as well as in the Colorado case study, where we find that coordination and shaped flows may provide additional value for systems with high penetration of variable renewable energy. Together, these three studies demonstrate a pathway for integrated gas and electricity grid modeling and for studying the benefits of coordinated operations of these increasingly interdependent energy systems.

29 ENERGY PLANNING, POLICY, AND ECONOMY↗

DOE OSTI2020

MPI Session: External Network Transport Implementation (V.1.0)

The MPI Sessions extensions to the MPI standard have been accepted by the MPI Forum and will be included in the upcoming MPI 4 version of the standard. MPI Sessions has the potential to address several limitations of MPI’s current specification: MPI cannot be initialized within an MPI process from different application components without a priori knowledge or coordination; MPI cannot be initialized more than once; and, MPI cannot be reinitialized after MPI finalization. MPI Sessions also offers the possibility for more flexible ways for individual components of an application to express the capabilities they require from MPI at a finer granularity than is presently possible. A prototype of MPI Sessions, based on the Open MPI implementation of the MPI standard, was developed to facilitate acceptance of the Sessions proposal by the Forum. The initial implementation had some limitations, one of the more significant ones being that it was limited in its ability to fully exploit modern network APIs such as OFI libfabric and OpenUCX and underlying network hardware. This report presents enhancements to the prototype implementation of MPI Sessions that removes this restriction for the networks to be used in the next generation of DOE exa–scale systems. Open MPI was used as the implementation vehicle, but results here are also relevant to other middleware stacks.

97 MATHEMATICS AND COMPUTING↗

DOE OSTI2020

Neural Network Interatomic Potentials

In this project, we investigate the use of neural networks for the prediction of molecular properties, namely the interatomic potential. We use the machine learning package Tensorflow to build a variety of neural networks and compare performance with a popular Fortran package - Atomic Energy Networks (aenet). There are two primary goals for this work: 1) use the wide availability of different optimization techniques in Tensorflow to outperform aenet and 2) use new descriptors that can outperform Behler descriptors.

74 ATOMIC AND MOLECULAR PHYSICS↗

DOE OSTI2021

All Optical Neural Networks for Low Power Edge Computing

We developed a simplistic physics-based model of an all-optical neural network that mimics the encoder part of an autoencoder neural network for image compression. Our approach relies on the generation of a MATLAB-based model for both data compression and decompression and utilizes MATLAB's built-in autoencoder networks in combination with simple propagation of optical fields between layers constituting phase elements via Fourier transform. We optimize the phase elements using the particle swarm optimization technique and using our model, we demonstrate a compression ratio of 25% for 2828-pixel input images containing numeric digits from 0 to 9.

97 MATHEMATICS AND COMPUTING↗

DOE OSTI2022

Integration of Wireless Sensor Networks and Battery-free RFID for Advanced Reactors

To address an important need for Nuclear Power Plants (NPPs) to significantly reduce the amount of the required cables for sensor data communications, this Phase I SBIR effort successfully developed and demonstrated a novel low-cost proof-of-concept prototype of a secure wireless sensor network backbone communications system. This system combines commercially available low power, low cost XBEE wireless communications network and passive (battery-free) Radio Frequency Identification (RFID) systems to report individual sensor data and their location through the containment wall for rapid response to anomalies in nuclear facilities. This WIreless Sensing and Locating (WISLO) system network architecture allows non-intrusive wireless collection of sensor data with the sensor’s accurate location information from inside of the instrumentation or containment area with minimal need for power sources. The multi-node sensor data in the containment area is wirelessly transmitted to outside through the metal reinforced concrete walls without the need for any batteries. The WISLO system could be readily used in current reactor fleet and future advanced reactors, as well as in small modular reactors (SMR). The WISLO system is the first dual frequency battery-free trough the wall communications backbone system developed specifically for use in nuclear power plants. It not only can modernize the sensor monitoring practices in the existing reactor fleet, but also offers a secure, low-cost solution to advanced reactors and SMRs by removing cables and issues related to them such as cable integrity, reporting delays, and installation and maintenance costs. The intent is to improve process safety, reliability, efficiency, and cost effectiveness to the monitoring and maintenance process in current and future plants. As we address the next productization and manufacturing capabilities in the next phase, this system will serve the needs of many commercial and government applications in remote monitoring of sensor data that demand battery-free transmission of sensor data such as Internet-of-Things (IOT).

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

DOE OSTI2022

The Portals 4.3 Network Programming Interface

This report presents a specification for the Portals 4 network programming interface. Portals 4 is intended to allow scalable, high-performance network communication between nodes of a parallel computing system. Portals 4 is well suited to massively parallel processing and embedded systems. Portals 4 represents an adaption of the data movement layer developed for massively parallel processing platforms, such as the 4500-node Intel TeraFLOPS machine. Sandia's Cplant cluster project motivated the development of Version 3.0, which was later extended to Version 3.3 as part of the Cray Red Storm machine and XT line. Version 4 is targeted to the next generation of machines employing advanced network interface architectures that support enhanced offload capabilities.

97 MATHEMATICS AND COMPUTING↗

DOE OSTI2022

Evaluating the Impacts of the Bipartisan Budget Act of 2018 45Q Tax Credit on CCS Network Costs

This study evaluates the impact of the Bipartisan Budget Act of 2018 (BBA)-amended 45Q tax credit on integrated carbon capture and storage (CCS) networks for source types and geologic storage reservoirs common to the north-central U.S. The integrated CCS networks evaluated were chosen to closely replicate those in the forthcoming National Energy Technology Laboratory (NETL) central U.S. CCS cost options study’s Northwest CCS Network Regional Impact Area (Northwest Impact Area).

29 ENERGY PLANNING, POLICY, AND ECONOMY↗