Search NASA⌕ Search

SEARCH · Search NASA

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.

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

At least 577 records · Page 32

Rapid Calculation of Max-Min Fair Rates for Multi-Commodity Flows in Fat-Tree Networks

Max-min fairness is often used in the performance modeling of interconnection networks. Existing methods to compute max-min fair rates for multi-commodity flows have high complexity and are computationally infeasible for large networks. In this paper, we show that by considering topological features, this problem can be solved efficiently for the fat-tree topology that is widely used in data centers and high performance compute clusters. Several efficient new algorithms are developed for this problem, including a parallel algorithm that can take advantage of multi-core and shared-memory architectures. Using these algorithms, we demonstrate that it is possible to find the max-min fair rate allocation for multi-commodity flows in fat-tree networks that support tens of thousands of nodes. We evaluate the run-time performance of the proposed algorithms and show improvement in orders of magnitude over the previously best known method. Finally, we further demonstrate a new application of max-min fair rate allocation that is only computationally feasible using our new algorithms.

97 MATHEMATICS AND COMPUTING↗

Throughput Analytics of Cloud Networks

A network of virtual machines at cloud server sites connected over virtual IO connections is a flexible, easily deployable, and cost-effective alternative to a physical network infrastructure with dedicated servers connected over leased fiber lines. We study the throughput performance of such a cloud network by collecting measurements over Google Cloud infrastructure spanning multiple continents. To study its ideal performance and impact of packet losses, we utilize its emulation using dedicated servers and connection hardware emulation devices. We compare the measurements over the cloud network to those over its emulation on a testbed. We examine the throughput profiles of both networks as a function of the round trip time and their utilization-concavity coefficients, estimated using measurements for common TCP versions. The throughput profile's concave-convex shape and its coefficient are critical indicators of the network performance, qualitatively and quantitatively, respectively. The results indicate their overall agreement between the production cloud network and its emulation using dedicated connections, and a near optimal throughput performance of the former except for a few under-performing connections. Also, the number of parallel flows is found to be a dominant factor in optimizing the throughput across various conditions and TCP versions.

Phanekham, Derek↗

Optimal environmental and economic performance trade-offs for fifth generation district heating and cooling network topologies with waste heat recovery

Network topology greatly influences both the economic and environmental performance of fifth generation district heating and cooling (5GDHC) systems. In this study the optimal trade-offs between the environmental and economic performance of 5GDHC network topologies for a five-building district with waste heat recovery were explored. A life cycle assessment method was used to calculate the total life cycle CO 2 emissions (LCCO2) associated with the installation and operation of various network topologies. Twelve months of empirical data from a data center cooling system were analyzed to assess its suitability for integration into a 5GDHC system. The most suitable method for utilizing this waste heat was selected based on the ambient loop warm pipe setpoint, waste heat temperature, and district energy system configuration. A multi-objective optimization algorithm was used to select the 5GDHC network topology that provided the optimal trade-off between LCCO2 and life cycle cost (LCC). A trade-off parameter was employed to weigh the importance of each objective in the selection process. The results showed waste heat from the data center was suitable for integration into the 5GDHC system due to its availability and consistent temperature profile. When return temperatures of 25 °C or higher were available from the liquid-cooled system, direct pre-heating of the ambient loop warm pipe was found to be the most effective waste heat integration method. The selection of the network topology that provided the optimal trade-off between LCCO2 and LCC (optimal trade-off topology) was highly dependent on factors such as fuel prices, CO 2 prices, electricity CO 2 emissions factors, availability of waste heat, embodied CO 2 emissions associated with network installation and network infrastructure costs. Optimal trade-off topologies produced substantial LCCO2 reductions relative to corresponding LCC increases. LCCO2 reduction to LCC increase ratios from 5.78 to 117.79 were identified with CO 2 offset costs ranging from 4.77 to 60.08 ($/tCO 2 e).

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗

A Workflow-based Intelligent Network Data Movement Advisor with End-to-end Performance Optimization

Next-generation eScience applications often generate large amounts of simulation, experimental, or observational data that must be shared and managed by collaborative organizations. Advanced networking technologies and services have been rapidly developed and deployed to facilitate such massive data transfer. However, these technologies and services have not been fully utilized mainly because their use typically requires significant domain knowledge and in many cases application users are even not aware of their existence. By leveraging the functionalities of an existing Network-Aware Data Movement Advisor (NADMA) utility, we propose a new Workflow-based Intelligent Network Data Movement Advisor (WINDMA) with end-to-end performance optimization for this DOE funded project. This WINDMA system integrates three major components: resource discovery, data movement, and status monitoring, and supports the sharing of common data movement workflows through account and database management. This system provides a web interface and interacts with existing data/space management and discovery services such as Storage Resource Management, transport methods such as GridFTP and GlobusOnline, and network resource provisioning brokers such as ION and OSCARS. We demonstrate the efficacy of the proposed transport-support workflow system in several use cases based on its implementation and deployment in DOE wide-area networks.

97 MATHEMATICS AND COMPUTING↗

Topology-Dependent Performance of Free-Space Photonic Quantum Networks Under Noise

Photonic quantum communication enables secure and high-fidelity information transfer beyond classical limits, with direct relevance to emerging quantum networks operating in free-space environments. While physical-layer models of depolarizing noise, Gamma–Gamma turbulence statistics, entanglement swapping, and decoy-state QKD security bounds are individually well established, prior work typically treats these components in isolation or under fixed network assumptions. In this work, we develop a unified topology-aware analytical framework that simultaneously integrates free-space optical link budgets, turbulence-induced visibility degradation, depolarizing qubit noise, multi-hop entanglement cascade dynamics, teleportation fidelity thresholds, CHSH nonlocality certification, and asymptotic decoy-state secret key rate bounds across star, mesh, and ring graph structures. Rather than introducing new physical channel models, we demonstrate that identical physical links exhibit fundamentally different end-to-end performance once embedded within different network topologies. Mesh architectures minimize visibility cascade through hop-count reduction but incur quadratic hardware scaling. Star topologies minimize link count but concentrate noise and synchronization overhead at the hub. Ring configurations offer linear hardware scaling with multiplicative fidelity degradation. The results establish topology as a first-order design parameter in near-term free-space quantum networks operating without full quantum repeater infrastructures. While motivated by distributed multi-agent architectures, the framework applies broadly to terrestrial, airborne, and satellite-assisted photonic quantum communication systems.

QKD↗

Predicting Future Hourly Residential Electrical Consumption: A Machine Learning Case Study

Whole building input models for energy simulation programs are frequently created in order to evaluate specific energy savings potentials. They are also often utilized to maximize cost-effective retrofits for existing buildings as well as to estimate the impact of policy changes toward meeting energy savings goals. Traditional energy modeling suffers from several factors, including the large number of inputs required to characterize the building, the specificity required to accurately model building materials and components, simplifying assumptions made by underlying simulation algorithms, and the gap between the as-designed and as-built building. Prior works have attempted to mitigate these concerns by using sensor-based machine learning approaches to model energy consumption. However, a majority of these prior works focus only on commercial buildings. The works that focus on modeling residential buildings primarily predict monthly electrical consumption, while commercial models predict hourly consumption. This means there is not a clear indicator of which techniques best model residential consumption, since these methods are only evaluated using low-resolution data. We address this issue by testing seven different machine learning algorithms on a unique residential data set, which contains 140 different sensors measurements, collected every 15 minutes. In addition, we validate each learner's correctness on the ASHRAE Great Energy Prediction Shootout, using the original competition metrics. Our validation results confirm existing conclusions that Neural Network-based methods perform best on commercial buildings. However, the results from testing our residential data set show that Feed Forward Neural Networks, Support Vector Regression (SVR), and Linear Regression methods perform poorly, and that Hierarchical Mixture of Experts (HME) with Least Squares Support Vector Machines (LS-SVM) performs best - a technique not previously applied to this domain.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗

An image-driven machine learning approach to kinetic modeling of a discontinuous precipitation reaction

Micrograph quantification is an essential component of several materials science studies. Machine learning methods, in particular convolutional neural networks, have previously demonstrated performance in image recognition tasks across several disciplines (e.g. materials science, medical imaging, facial recognition). Here, we apply these well-established methods to develop an approach to microstructure quantification for kinetic modeling of a discontinuous precipitation reaction in a case study on the uranium-molybdenum system. Prediction of material processing history based on image data (classification), calculation of area fraction of phases present in the micrographs (segmentation), and kinetic modeling from segmentation results were performed. Results indicate that convolutional neural networks represent microstructure image data well, and segmentation using the k-means clustering algorithm yields results that agree well with manually annotated images. Classification accuracies of original and segmented images are both 94% for a 5-class classification problem. Kinetic modeling results agree well with previously reported data using manual thresholding. The image quantification and kinetic modeling approach developed and presented here aims to reduce researcher bias introduced into the characterization process, and allows for leveraging information in limited image data sets.

36 MATERIALS SCIENCE↗

Livermore Big Artificial Neural Network Toolkit

LBANN is a toolkit that is designed to train artificial neural networks efficiently on high performance computing architectures. It is optimized to take advantages of key High Performance Computing features to accelerate neural network training. Specifically it is optimized for low-latency, high bandwidth interconnects, node-local NVRAM, node-local GPU accelerators, and high bandwidth parallel file systems. It is built on top of the open source Elemental distributed-memory dense and spars-direct linear algebra and optimization library that is released under the BSD license. The algorithms contained within LBANN are drawn from the academic literature and implemented to work within a distributed-memory framework.

Essen, BrianVan↗

ASCR@40: Four Decades of Department of Energy Leadership in Advanced Scientific Computing Research

Throughout its long history, the Office of Advanced Scientific Computing Research (ASCR) has built the critical technologies to ensure U.S. leadership in energy science and national security. It has made its parent agency, the Department of Energy (DOE) and its Office of Science, the world’s recognized leader in computational science. ASCR’s stated mission is “to discover, develop, and deploy computational and networking capabilities to analyze, model, simulate, and predict complex phenomena important to the DOE.” To accomplish this goal, ASCR oversees a large complex of computing and networking facilities and is responsible for procuring, deploying and operating high-performance computing (HPC), networking and storage resources; conducting basic research in mathematics and computer science; developing and sustaining a large body of software; and collaborating with other Office of Science programs, academia and industry. ASCR’s computational science leadership has a long history, predating even DOE’s inception. Applied mathematics and advanced computing were both elements of the Manhattan Project’s Theoretical Division. In the 1950s, DOE’s predecessor, the Atomic Energy Commission, created a mathematics program to develop and apply digital computing by supporting researchers at universities and AEC laboratories. Several organizational and name changes later, this program would grow and become ASCR.

97 MATHEMATICS AND COMPUTING↗

The Demonstration of Power over Ethernet (PoE) Technologies in Commercial and Institutional Buildings

Power over Ethernet (PoE) technologies are currently ubiquitous in offices in the form of phones, access points, and IP security cameras. They are systems that are familiar to and managed by on-site IT staff. As low voltage DC-power capabilities have increased with the advancement of IT network switches within offices, PoE devices are now being integrated across additional building systems, including lighting, computers, and HVAC controls. The connection of PoE devices to network switches offers the opportunity for energy monitoring and management to be performed through IT network management software. Data and results from technology validations of PoE equipment remain very limited. This lack of information hinders expansion of PoE as an energy management strategy as its benefits and challenges have been insufficiently quantified in real-world scenarios. This final report presents the findings of a DOE/State of Minnesota-funded project that demonstrated and assessed the energy and non-energy benefits of Power over Ethernet (PoE) technologies at six sites: two offices, three schools, and a hotel. The goals of the demonstrations were to: (1) assess the receptiveness of commercial and institutional markets to these new technologies, (2) demonstrate the feasibility of PoE technologies within standard design/construction practices and commercial codes, (3) assess the energy and cost savings opportunities of these technologies, (4) provide energy management opportunities where not typically available, and (5) formulate approaches to spur adoption of these technologies in the marketplace.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗

Livermore Big Artificial Neural Network Toolkit

LBANN is a toolkit that is designed to train artificial neural networks efficiently on high performance computing architectures. It is optimized to take advantages of key High Performance Computing features to accelerate neural network training. Specifically it is optimized for low-latency, high bandwidth interconnects, node-local NVRAM, node-local GPU accelerators, and high bandwidth parallel file systems. It is built on top of the open source Elemental distributed-memory dense and spars-direct linear algebra and optimization library that is released under the BSD license. The algorithms contained within LBANN are drawn from the academic literature and implemented to work within a distributed-memory framework.

Essen, Brian Van↗

An Evaluation of the Effect of Network Cost Optimization for Leadership Class Supercomputers

Dragonfly-based networks are an extensively deployed network topology in large-scale high-performance computing due to their cost-effectiveness and efficiency. The US will soon have three Exascale supercomputers for leadership class workloads deployed using dragonfly networks. Compared to indirect networks of similar scale, the dragonfly network has considerably reduced cable lengths, cable counts, and switch counts, resulting in significant network cost savings for a given system size, however, these cost reductions result in reduced global minimal paths and more challenging routing. Additionally, large scale dragonfly networks often require a taper at the global link level, resulting in less bisection bandwidth than is achievable in other traditional non-blocking topologies of equivalent scale. While dragonfly networks have been extensively studied, they have yet to be fully evaluated in an extreme scale (i.e., exascale) system that targets capability workloads. In this paper, we present the results of the first large scale evaluation of a dragonfly network on an exascale system (Frontier) and compare its behavior to a similar scale fat-tree network on a previous generation TOP500 system (Summit). This evaluation aims to determine the effect of network cost optimizations by measuring a tapered topology’s impact on capability workloads. Our evaluation is based on a collection of synthetic microbenchmarks, mini-apps, and full scale applications. It compares the scaling efficiencies of each benchmark between the dragonfly-based Frontier and the fat-tree-based Summit systems. Our results show that a dragonfly network is $\sim \mathbf{3 0 \%}$ more cost efficient than a fat-tree topology, which amortizes to $\sim 3 \%$ of an exascale system cost. Furthermore, while tapered dragonfly networks impose significant tradeoffs, the impacts are not as broad as initially thought and are mostly seen in applications with global communication patterns, particularly all-to-all (e.g., FFT-based algorithms), but also local communication patterns (e.g., nearest-neighbor algorithms) that are sensitive to network performance variability.

Khan, Awais↗

Benchmarking a Tunable Quantum Neural Network on Trapped-Ion and Superconducting Hardware

We implement a quantum generalization of a neural network on trapped-ion and IBM superconducting quantum computers to classify MNIST images, a common benchmark in computer vision. The network feedforward involves qubit rotations whose angles depend on the results of measurements in the previous layer. The network is trained via simulation, but inference is performed experimentally on quantum hardware. The classical-to-quantum correspondence is controlled by an interpolation parameter, $a$, which is zero in the classical limit. Increasing $a$ introduces quantum uncertainty into the measurements, which is shown to improve network performance at moderate values of the interpolation parameter. We then focus on particular images that fail to be classified by a classical neural network but are detected correctly in the quantum network. For such borderline cases, we observe strong deviations from the simulated behavior. We attribute this to physical noise, which causes the output to fluctuate between nearby minima of the classification energy landscape. Such strong sensitivity to physical noise is absent for clear images. We further benchmark physical noise by inserting additional single-qubit and two-qubit gate pairs into the neural network circuits. Our work provides a springboard toward more complex quantum neural networks on current devices: while the approach is rooted in standard classical machine learning, scaling up such networks may prove classically non-simulable and could offer a route to near-term quantum advantage.

FOS: Physical sciences↗

A neural-network based estimator to search for primordial non-Gaussianity in Planck CMB maps

We present an upgraded combined estimator, based on Minkowski Functionals and Neural Networks, with excellent performance in detecting primordial non-Gaussianity in simulated maps that also contain a weighted mixture of Galactic contaminations, besides real pixel's noise from Planck cosmic microwave background radiation data. We rigorously test the efficiency of our estimator considering several plausible scenarios for residual non-Gaussianities in the foreground-cleaned Planck maps, with the intuition to optimize the training procedure of the Neural Network to discriminate between contaminations with primordial and secondary non-Gaussian signatures. We look for constraints of primordial local non-Gaussianity at large angular scales in the foreground-cleaned Planck maps. For the SMICA map we found f{sub NL} = 33 ± 23, at 1σ confidence level, in excellent agreement with the WMAP-9yr and Planck results. In addition, for the other three Planck maps we obtain similar constraints with values in the interval f{sub NL} element of [33, 41], concomitant with the fact that these maps manifest distinct features in reported analyses, like having different pixel's noise intensities.

79 ASTRONOMY AND ASTROPHYSICS↗

Doing the impossible: Why neural networks can be trained at all

As deep neural networks grow in size, from thousands to millions to billions of weights, the performance of those networks becomes limited by our ability to accurately train them. A common naive question arises: if we have a system with billions of degrees of freedom, don't we also need billions of samples to train it? Of course, the success of deep learning indicates that reliable models can be learned with reasonable amounts of data. Similar questions arise in protein folding, spin glasses and biological neural networks. With effectively infinite potential folding/spin/wiring configurations, how does the system find the precise arrangement that leads to useful and robust results? Simple sampling of the possible configurations until an optimal one is reached is not a viable option even if one waited for the age of the universe. On the contrary, there appears to be a mechanism in the above phenomena that forces them to achieve configurations that live on a low-dimensional manifold, avoiding the curse of dimensionality. In the current work we use the concept of mutual information between successive layers of a deep neural network to elucidate this mechanism and suggest possible ways of exploiting it to accelerate training.

deep learning, training, curse of dimensionality, ↗