SEARCH · Engineering Papers

Results for “network performance”

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 1,657 records · Page 92

All sources NASA NTRS DOE OSTI

NASA NTRS1992

Membership generation using multilayer neural network

There has been intensive research in neural network applications to pattern recognition problems. Particularly, the back-propagation network has attracted many researchers because of its outstanding performance in pattern recognition applications. In this section, we describe a new method to generate membership functions from training data using a multilayer neural network. The basic idea behind the approach is as follows. The output values of a sigmoid activation function of a neuron bear remarkable resemblance to membership values. Therefore, we can regard the sigmoid activation values as the membership values in fuzzy set theory. Thus, in order to generate class membership values, we first train a suitable multilayer network using a training algorithm such as the back-propagation algorithm. After the training procedure converges, the resulting network can be treated as a membership generation network, where the inputs are feature values and the outputs are membership values in the different classes. This method allows fairly complex membership functions to be generated because the network is highly nonlinear in general. Also, it is to be noted that the membership functions are generated from a classification point of view. For pattern recognition applications, this is highly desirable, although the membership values may not be indicative of the degree of typicality of a feature value in a particular class.

Kim, Jaeseok↗

DOE OSTI2020

Broadband Characterization and Circuit Model Development of Transmission-Scale Transformers

This report describes broadband measurements of transmission-scale transformers typical in the electric power grid. This work was performed as part of the EMP Resilient Grid LDRD project at Sandia National Laboratories to generate circuit models that can be used for high-altitude electromagnetic pulse (HEMP) coupling simulations and response predictions. The objective of the work was to obtain characterization data of substation yard equipment across a frequency range relevant to HEMP. Vector network analyzer measurements up to 100 MHz were performed on two power transformers at ABB-Hitachi and a single ITEC potential transformer. Custom cable breakouts were designed to interface with the transformer terminals and provide ground connections to the chassis at the base of the transformer bushings. The three-phase terminals of the power transformers were measured as a common mode impedance using a parallel resistive splitter, and the single-phase terminals of the potential transformer were measured directly. A vector fitting algorithm was used to empirically fit circuit models to the resulting two-port networks and input impedances of the measured objects. Simplified circuit representations of the input impedances were also generated to assess the degree of precision needed for high-altitude electromagnetic pulse response predictions, which were performed in Sandia's XYCE circuit simulator platform. HEMP coupling simulations using the transformer models showed significant reduction in the voltage peak and broadening in the pulse width seen at the power transformer compared to the traveling wave voltage. This indicated the importance of the load condition when defining the coupled insult in an electric power substation. Simplified circuit models showed a similar voltage at the transformer with a smoothed waveform. The presence of potential transformers in the simulation did not significantly change the simulated voltage at the power transformer. Single-port input impedance models were also developed to define load conditions when transfer characteristics were not necessary.

24 POWER TRANSMISSION AND DISTRIBUTION↗

DOE OSTI2021

SYMBIOSYS: A Methodology for Performance Analysis of Composable HPC Data Services

Microservices are a powerful new way of building, customizing, and deploying distributed services owing to their flexibility and maintainability. Several large-scale distributed platforms have emerged to serve the growing needs of data-centric workloads and services in commercial computing. Concurrently, high-performance computing (HPC) systems and software are rapidly evolving to meet the demands of diversified applications and heterogeneity. The interplay of hardware factors, software configuration parameters, and the flexibility offered with a microservice architecture makes it nontrivial to estimate the optimal service instantiation for a given application workload. Further, this problem is exacerbated when considering that these services operate in a dynamic and heterogeneous HPC environment. An optimally integrated service can be vastly more performant than a haphazardly integrated one. Existing performance tools for HPC either fail to understand the request-response model of communication inherent to microservices or they operate within a narrow scope, limiting the insight that can be gleaned from employing them in isolation. We propose a methodology for integrated performance analysis of HPC microservices frameworks and applications called SYMBIOSYS. We describe its design and implementation within the context of the Mochi framework. This integration is achieved by combining distributed callpath profiling and tracing with a performance data exchange strategy that collects fine-grained, low-level metrics from the RPC communication library and network layers. The result is a portable, low-overhead performance analysis setup that provides a holistic profile of the dependencies among microservices and how they interact with the Mochi RPC software stack. Using HEPnOS, a production-quality Mochi data service, we demonstrate the low-overhead operation of SYMBIOSYS at scale and use it to identify the root causes of poorly performing service configurations.

microservices↗

DOE OSTI2023

Neural network-based control of an ultrafast laser

With the recent advances in machine learning (ML) and data science (DS), the control, modeling, and analysis of these complex systems continues to improve. In this work, we report on the optimization of the intensity of a femtosecond laser using feedforward neural networks (FFNN) that model the input–output relationships of the data. The input parameters of the system were optimized to achieve the required performance of the femtosecond laser. We propose a neural network-based control system to model the relationship between the spectral amplitude and phase of the input laser pulse at the amplifier input and the shape of the output pulse. Low-jitter laser parameter inputs and the resulting laser pulse duration were modeled, and the resulting correlation between the input and output data was used to optimize the laser pulse. Here, we demonstrate improved processing and laser control performance.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

DOE OSTI2024

Scaling neural simulations in STACS

Abstract As modern neuroscience tools acquire more details about the brain, the need to move towards biological-scale neural simulations continues to grow. However, effective simulations at scale remain a challenge. Beyond just the tooling required to enable parallel execution, there is also the unique structure of the synaptic interconnectivity, which is globally sparse but has relatively high connection density and non-local interactions per neuron. There are also various practicalities to consider in high performance computing applications, such as the need for serializing neural networks to support potentially long-running simulations that require checkpoint-restart. Although acceleration on neuromorphic hardware is also a possibility, development in this space can be difficult as hardware support tends to vary between platforms and software support for larger scale models also tends to be limited. In this paper, we focus our attention on Simulation Tool for Asynchronous Cortical Streams (STACS), a spiking neural network simulator that leverages the Charm++ parallel programming framework, with the goal of supporting biological-scale simulations as well as interoperability between platforms. Central to these goals is the implementation of scalable data structures suitable for efficiently distributing a network across parallel partitions. Here, we discuss a straightforward extension of a parallel data format with a history of use in graph partitioners, which also serves as a portable intermediate representation for different neuromorphic backends. We perform scaling studies on the Summit supercomputer, examining the capabilities of STACS in terms of network build and storage, partitioning, and execution. We highlight how a suitably partitioned, spatially dependent synaptic structure introduces a communication workload well-suited to the multicast communication supported by Charm++. We evaluate the strong and weak scaling behavior for networks on the order of millions of neurons and billions of synapses, and show that STACS achieves competitive levels of parallel efficiency.

59 BASIC BIOLOGICAL SCIENCES↗

NASA NTRS2012

NASA Integrated Space Communications Network

The NASA Integrated Network for Space Communications and Navigation (SCaN) has been in the definition phase since 2010. It is intended to integrate NASA s three existing network elements, i.e., the Space Network, Near Earth Network, and Deep Space Network, into a single network. In addition to the technical merits, the primary purpose of the Integrated Network is to achieve a level of operating cost efficiency significantly higher than it is today. Salient features of the Integrated Network include (a) a central system element that performs service management functions and user mission interfaces for service requests; (b) a set of common service execution equipment deployed at the all stations that provides return, forward, and radiometric data processing and delivery capabilities; (c) the network monitor and control operations for the entire integrated network are conducted remotely and centrally at a prime-shift site and rotating among three sites globally (a follow-the-sun approach); (d) the common network monitor and control software deployed at all three network elements that supports the follow-the-sun operations.

Tai, Wallace↗

NASA NTRS2012

NASA Integrated Space Communications Network

The NASA Integrated Network for Space Communications and Navigation (SCaN) has been in the definition phase since 2010. It is intended to integrate NASA’s three existing network elements, i.e., the Space Network, Near Earth Network, and Deep Space Network, into a single network. In addition to the technical merits, the primary purpose of the Integrated Network is to achieve a level of operating cost efficiency significantly higher than it is today. Salient features of the Integrated Network include (a) a central system element that performs service management functions and user mission interfaces for service requests; (b) a set of common service execution equipment deployed at the all stations that provides return, forward, and radiometric data processing and delivery capabilities; (c) the network monitor and control operations for the entire integrated network are conducted remotely and centrally at a prime-shift site and rotating among three sites globally (a follow-the-sun approach); (d) the common network monitor and control software deployed at all three network elements that supports the follow-the-sun operations.

Bhasin, Kul↗

DOE OSTI2023

Pan-cancer evaluation of clinical value of mitotic network activity index (MNAI) and its predictive value for immunotherapy

Increased mitotic activity is associated with the genesis and aggressiveness of many cancers. To assess the clinical value of mitotic activity as prognostic biomarker, we performed a pan-cancer study on the mitotic network activity index (MNAI) constructed based on 54-gene mitotic apparatus network. Our pan-cancer assessment on TCGA (33 tumor types, 10,061 patients) and validation on other publicly available cohorts (23 tumor types, 9,209 patients) confirmed the significant association of MNAI with overall survival, progression-free survival, and other prognostic endpoints in multiple cancer types, including lower-grade gliomas (LGG), breast invasive carcinoma (BRCA), as well as many others. We also showed significant association between MNAI and genetic instability, which provides a biological explanation of its prognostic impact at pan-cancer landscape. Our association analysis revealed that patients with high MNAI benefitted more from anti-PD-1 and Anti-CTLA-4 treatment. In addition, we demonstrated that multimodal integration of MNAI and the AI-empowered Cellular Morphometric Subtypes (CMS) significantly improved the predictive power of prognosis compared to using MNAI and CMS alone. Our results suggest that MNAI can be used as a potential prognostic biomarker for different tumor types toward different clinical endpoints, and multimodal integration of MNAI and CMS exceeds individual biomarker for precision prognosis.

60 APPLIED LIFE SCIENCES↗

DOE OSTI2022

Learning high-dimensional parametric maps via reduced basis adaptive residual networks

We propose a scalable framework for the learning of high-dimensional parametric maps via adaptively constructed residual network (ResNet) maps between reduced bases of the inputs and outputs. When just few training data are available, it is beneficial to have a compact parametrization in order to ameliorate the ill-posedness of the neural network training problem. By linearly restricting high-dimensional maps to informed reduced bases of the inputs, one can compress high-dimensional maps in a constructive way that can be used to detect appropriate basis ranks, equipped with rigorous error estimates. A scalable neural network learning framework is thus to learn the nonlinear compressed reduced basis mapping. Unlike the reduced basis construction, however, neural network constructions are not guaranteed to reduce errors by adding representation power, making it difficult to achieve good practical performance. Inspired by recent approximation theory that connects ResNets to sequential minimizing flows, we present an adaptive ResNet construction algorithm. This algorithm allows for depth-wise enrichment of the neural network approximation, in a manner that can achieve good practical performance by first training a shallow network and then adapting. We prove universal approximation of the associated neural network class for $L^2_v$ functions on compact sets. Our overall framework allows for constructive means to detect appropriate breadth and depth, and related compact parametrizations of neural networks, significantly reducing the need for architectural hyperparameter tuning. Numerical experiments for parametric PDE problems and a 3D CFD wing design optimization parametric map demonstrate that the proposed methodology can achieve remarkably high accuracy for limited training data, and outperformed other neural network strategies we compared against.

42 ENGINEERING↗

DOE OSTI2024

Nonlinear encoding in diffractive information processing using linear optical materials

Nonlinear encoding of optical information can be achieved using various forms of data representation. Here, we analyze the performances of different nonlinear information encoding strategies that can be employed in diffractive optical processors based on linear materials and shed light on their utility and performance gaps compared to the state-of-the-art digital deep neural networks. For a comprehensive evaluation, we used different datasets to compare the statistical inference performance of simpler-to-implement nonlinear encoding strategies that involve, e.g., phase encoding, against data repetition-based nonlinear encoding strategies. We show that data repetition within a diffractive volume (e.g., through an optical cavity or cascaded introduction of the input data) causes the loss of the universal linear transformation capability of a diffractive optical processor. Therefore, data repetition-based diffractive blocks cannot provide optical analogs to fully connected or convolutional layers commonly employed in digital neural networks. However, they can still be effectively trained for specific inference tasks and achieve enhanced accuracy, benefiting from the nonlinear encoding of the input information. Our results also reveal that phase encoding of input information without data repetition provides a simpler nonlinear encoding strategy with comparable statistical inference accuracy to data repetition-based diffractive processors. Our analyses and conclusions would be of broad interest to explore the push-pull relationship between linear material-based diffractive optical systems and nonlinear encoding strategies in visual information processors.

42 ENGINEERING↗

DOE OSTI2024

A Full-Stack Exploration of Language-Based Parallelism in Fortran 2023

This poster explores native parallel features in Fortran 2023 through the lens of supporting applications with libraries, compilers, and parallel runtimes. The language revision informally named Fortran 2008 introduced parallelism in the form of Single Program Multiple Data (SPMD) execution with two broad feature sets: (1) loop-level parallelism via do concurrent and (2) a Partitioned Global Address Space (PGAS) comprised of distributed “coarray” data structures. Fortran’s native parallelism has demonstrated high performance [1] and reduced the burden of inserting what sometimes amounts to more directives than code. Several compilers support both feature sets, typically by translating do concurrent into serial do loops annotated by parallel directives and by translating SPMD/PGAS features into direct calls to a communication library. Our research focuses primarily on two questions: (1) can the compiler’s parallel runtime library be developed in the language being compiled (Fortran) and (2) can we define an interface to the runtime that liberates compilers from being hardwired to one runtime and vice versa. We are answering these questions by developing the Parallel Runtime Interface for Fortran (PRIF) [2] and the Co-Array Fortran Framework of Efficient Interfaces to Network Environments (Caffeine) [3]. Caffeine is initially targeting adoption by LLVM Flang, a new open-source Fortran compiler developed by a broad community in industry, academia, and government labs. We are also exploring the use of these features in Inference-Engine, a deep learning library designed to facilitate neural network training and inference for high-performance computing applications written in modern Fortran.

Rasmussen, Katherine↗

DOE OSTI2022

A Framework for Neural Network Inference on FPGA-Centric SmartNICs

FPGA-based SmartNICs offer great potential to significantly improve the performance of high-performance computing and warehouse data processing by tightly coupling support for reconfigurable data-intensive computation with cross-node communication, thereby mitigating the von Neumann bottleneck. Existing work, however, has been generally been limited in that it assumes an accelerator model where kernels are offloaded to SmartNICs, but most control tasks are left to the CPUs. This leads to frequent waiting, inferior performance, and scaling challenges. In this work, we propose a new distributive data-centric computing framework, named FCsN, for reconfigurable SmartNIC-based systems. Through a lightweight task circulation execution model and its implementation architecture, FCsN allows the complete detaching of kernel execution, control logic, system scheduling, and network communication to the SmarNICs. This boosts performance by: (i) avoiding the control dependency with CPUs and (ii) supporting streaming kernel execution and network communication at line rate and in a very fine-grained manner. We demonstrate the efficiency and flexibility of FCsN using various types of neural network applications including graph neural networks; as these last are both irregular and data intensive they offer an especially robust demonstration. Evaluations using commonly-used neural network models and graph datasets show that a system with the support of FCsN can achieve, on average, 144 speedups over the MPI-based standard CPU baselines.

Guo, Anqi↗

DOE OSTI2020

A deep neural network to search for new long-lived particles decaying to jets

A tagging algorithm to identify jets that are significantly displaced from the proton-proton (pp) collision region in the CMS detector at the LHC is presented. Displaced jets can arise from the decays of long-lived particles (LLPs), which are predicted by several theoretical extensions of the standard model. The tagger is a multiclass classifier based on a deep neural network, which is parameterised according to the proper decay length $\mathrm{c}\tau_0$ of the LLP. A novel scheme is defined to reliably label jets from LLP decays for supervised learning. Samples of pp collision data, recorded by the CMS detector at a centre-of-mass energy of 13 TeV, and simulated events are used to train the neural network. Domain adaptation by backward propagation is performed to improve the simulation modelling of the jet class probability distributions observed in pp collision data. The potential performance of the tagger is demonstrated with a search for long-lived gluinos, a manifestation of split supersymmetric models. The tagger provides a rejection factor of 10 000 for jets from standard model processes, while maintaining an LLP jet tagging efficiency of 30-80% for gluinos with 1 mm $\leq$ $c\tau_0$ $\leq$ 10 m. The expected coverage of the parameter space for split supersymmetry is presented.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

NASA NTRS1987

Problem size, parallel architecture and optimal speedup

The communication and synchronization overhead inherent in parallel processing can lead to situations where adding processors to the solution method actually increases execution time. Problem type, problem size, and architecture type all affect the optimal number of processors to employ. The numerical solution of an elliptic partial differential equation is examined in order to study the relationship between problem size and architecture. The equation's domain is discretized into n sup 2 grid points which are divided into partitions and mapped onto the individual processor memories. The relationships between grid size, stencil type, partitioning strategy, processor execution time, and communication network type are analytically quantified. In so doing, the optimal number of processors was determined to assign to the solution, and identified (1) the smallest grid size which fully benefits from using all available processors, (2) the leverage on performance given by increasing processor speed or communication network speed, and (3) the suitability of various architectures for large numerical problems.

Nicol, David M.↗

NASA NTRS1988

Problem size, parallel architecture, and optimal speedup

Nicol, David M.↗

DOE OSTI2023

Networked Microgrids for Improved Resilient Operation: A Case Study in Adjuntas Puerto Rico

This paper presents a case study of a network microgrid orchestrator designed to allow coordinated operation of microgrids. Through Control Hardware in the Loop (CHIL), the proposed microgrid orchestrator was validated using, as case study two community-owned microgrids in Adjuntas, Puerto Rico. These two microgrids in Adjuntas were designed to provide affordable and reliable access to electricity to 14 businesses located in the town square. In its current design, these two microgrids operate independently from each other. This paper studies how networking them would increase their resiliency metrics while operating as an island. A distributed optimization is implemented in the microgrid orchestrator due to its superior scalability compared to centralized approach as well as to maintain data privacy. Two operational modes are considered in this study to showcase the advantages of networking microgrids: a) normal island, and b) degraded island. CHIL simulations are performed to validate the proposed microgrid orchestrator. CHIL results show meaningful resiliency improvements obtained by networking microgrids during contingencies such as loss of PV generation.

Ferrari Maglia, Max↗

NASA NTRS1996

Queuing Models of Tertiary Storage

Large scale scientific projects generate and use large amounts of data. For example, the NASA Earth Observation System Data and Information System (EOSDIS) project is expected to archive one petabyte per year of raw satellite data. This data is made automatically available for processing into higher level data products and for dissemination to the scientific community. Such large volumes of data can only be stored in robotic storage libraries (RSL's) for near-line access. A characteristic of RSL's is the use of a robot arm that transfers media between a storage rack and the read/write drives, thus multiplying the capacity of the system. The performance of the RSL's can be a critical limiting factor for the performance of the archive system. However, the many interacting components of an RSL make a performance analysis difficult. In addition, different RSL components can have widely varying performance characteristics. This paper describes our work to develop performance models of an RSL in isolation. Next we show how the RSL model can be incorporated into a queuing network model. We use the models to make some example performance studies of archive systems. The models described in this paper, developed for the NASA EODIS project, are implemented in C with a well defined interface. The source code, accompanying documentation, and also sample JAVA applets are available at: http://www.cis.ufl.edu/~ted/

Johnson, Theodore↗

DOE OSTI2021

Optimizing and Extending the Functionality of EXARL for Scalable Reinforcement Learning [Slides]

The main goal of the Co-Design Summer School 2021 is to provide algorithmic improvements to EXARL framework by improving performance and by adding functionalities. This presentation includes an introduction to reinforcement learning and to EXARL. The researchers expanded the capability of EXARL by including additional agents like (Asynchronized) Advantage Actor Critic (A2C/A3C) and Twin Delayed Deep Deterministic Policy Gradient (TD3). They also explored algorithmic improvements such as v-trace and Prioritized Experience Replay. They found that A2C/A3C performed best with v-trace and outperformed Deep Q-Network (DQN) on both the CartPole game and the ExaBooster scientific environment. Additionally, they found that TD3 performed as good as the existing Deep Deterministic Policy Gradient (DDPG) agent and that adding Prioritized Experience Replay to DDPG accelerated convergence.

97 MATHEMATICS AND COMPUTING↗