Search NASA⌕ Search

SEARCH · Search NASA

Results for “Edge Computing”

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 541 records · Page 30

Large-Scale NASA Science Applications on the Columbia Supercluster

Columbia, NASA's newest 61 teraflops supercomputer that became operational late last year, is a highly integrated Altix cluster of 10,240 processors, and was named to honor the crew of the Space Shuttle lost in early 2003. Constructed in just four months, Columbia increased NASA's computing capability ten-fold, and revitalized the Agency's high-end computing efforts. Significant cutting-edge science and engineering simulations in the areas of space and Earth sciences, as well as aeronautics and space operations, are already occurring on this largest operational Linux supercomputer, demonstrating its capacity and capability to accelerate NASA's space exploration vision. The presentation will describe how an integrated environment consisting not only of next-generation systems, but also modeling and simulation, high-speed networking, parallel performance optimization, and advanced data analysis and visualization, is being used to reduce design cycle time, accelerate scientific discovery, conduct parametric analysis of multiple scenarios, and enhance safety during the life cycle of NASA missions. The talk will conclude by discussing how NAS partnered with various NASA centers, other government agencies, computer industry, and academia, to create a national resource in large-scale modeling and simulation.

Brooks, Walter↗

SU-E-J-248: Contributions of Tumor and Stroma Phenotyping in Computer-Aided Diagnosis

Purpose: To gain insight into the role of parenchyma stroma in the characterization of breast tumors by incorporating computerized mammographic parenchyma assessment into breast CADx in the task of distinguishing between malignant and benign lesions. Methods: This study was performed on 182 biopsy-proven breast mass lesions, including 76 benign and 106 malignant lesions. For each full-field digital mammogram (FFDM) case, our quantitative imaging analysis was performed on both the tumor and a region-of-interest (ROI) from the normal contralateral breast. The lesion characterization includes automatic lesion segmentation and feature extraction. Radiographic texture analysis (RTA) was applied on the normal ROIs to assess the mammographic parenchymal patterns of these contralateral normal breasts. Classification performance of both individual computer extracted features and the output from a Bayesian artificial neural network (BANN) were evaluated with a leave-one-lesion-out method using receiver operating characteristic (ROC) analysis with area under the curve (AUC) as the figure of merit. Results: Lesion characterization included computer-extracted phenotypes of spiculation, size, shape, and margin. For parenchymal pattern characterization, five texture features were selected, including power law beta, contrast, and edge gradient. Merging of these computer-selected features using BANN classifiers yielded AUC values of 0.79 (SE=0.03) and 0.67 (SE=0.04) in the task of distinguishing between malignant and benign lesions using only tumor phenotypes and texture features from the contralateral breasts, respectively. Incorporation of tumor phenotypes with parenchyma texture features into the BANN yielded improved classification performance with an AUC value of 0.83 (SE=0.03) in the task of differentiating malignant from benign lesions. Conclusion: Combining computerized tumor and parenchyma phenotyping was found to significantly improve breast cancer diagnostic accuracy highlighting the need to consider both tumor and stroma in decision making. Funding: University of Chicago Dean Bridge Fund, NCI U24-CA143848-05, P50-CA58223 Breast SPORE program, and Breast Cancer Research Foundation. COI: MLG is a stockholder in R2 technology/Hologic and receives royalties from Hologic, GE Medical Systems, MEDIAN Technologies, Riverain Medical, Mitsubishi, and Toshiba. MLG is a cofounder and stockholder in Quantitative Insights.

60 APPLIED LIFE SCIENCES↗

Method for concurrent execution of primitive operations by dynamically assigning operations based upon computational marked graph and availability of data

Computationally complex primitive operations of an algorithm are executed concurrently in a plurality of functional units under the control of an assignment manager. The algorithm is preferably defined as a computationally marked graph contianing data status edges (paths) corresponding to each of the data flow edges. The assignment manager assigns primitive operations to the functional units and monitors completion of the primitive operations to determine data availability using the computational marked graph of the algorithm. All data accessing of the primitive operations is performed by the functional units independently of the assignment manager.

Stoughton, John W.↗

Distilling particle knowledge for fast reconstruction at high-energy physics experiments

Knowledge distillation is a form of model compression that allows artificial neural networks of different sizes to learn from one another. Its main application is the compactification of large deep neural networks to free up computational resources, in particular on edge devices. In this article, we consider proton-proton collisions at the High-Luminosity Large Hadron Collider (HL-LHC) and demonstrate a successful knowledge transfer from an event-level graph neural network (GNN) to a particle-level small deep neural network (DNN). Our algorithm, DistillNet, is a DNN that is trained to learn about the provenance of particles, as provided by the soft labels that are the GNN outputs, to predict whether or not a particle originates from the primary interaction vertex. The results indicate that for this problem, which is one of the main challenges at the HL-LHC, there is minimal loss during the transfer of knowledge to the small student network, while improving significantly the computational resource needs compared to the teacher. This is demonstrated for the distilled student network on a CPU, as well as for a quantized and pruned student network deployed on a field programmable gate array. Our study proves that knowledge transfer between networks of different complexity can be used for fast artificial intelligence (AI) in high-energy physics that improves the expressiveness of observables over non-AI-based reconstruction algorithms. Such an approach can become essential at the HL-LHC experiments, e.g. to comply with the resource budget of their trigger stages.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

The effects of Reynolds number, rotor incidence angle, and surface roughness on the heat transfer distribution in a large-scale turbine rotor passage

A combined experimental and computational program was conducted to examine the heat transfer distribution in a turbine rotor passage geometrically similiar to the Space Shuttle Main Engine (SSME) High Pressure Fuel Turbopump (HPFTP). Heat transfer was measured and computed for both the full-span suction and pressure surfaces of the rotor airfoil as well as for the hub endwall surface. The primary objective of the program was to provide a benchmark-quality data base for the assessment of rotor passage heat transfer computational procedures. The experimental portion of the study was conducted in a large-scale, ambient temperature, rotating turbine model. Heat transfer data were obtained using thermocouple and liquid-crystal techniques to measure temperature distributions on the thin, electrically-heated skin of the rotor passage model. Test data were obtained for various combinations of Reynolds number, rotor incidence angle and model surface roughness. The data are reported in the form of contour maps of Stanton number. These heat distribution maps revealed numerous local effects produced by the three-dimensional flows within the rotor passage. Of particular importance were regions of local enhancement produced on the airfoil suction surface by the main-passage and tip-leakage vortices and on the hub endwall by the leading-edge horseshoe vortex system. The computational portion consisted of the application of a well-posed parabolized Navier-Stokes analysis to the calculation of the three-dimensional viscous flow through ducts simulating the a gas turbine passage. These cases include a 90 deg turning duct, a gas turbine cascade simulating a stator passage, and a gas turbine rotor passage including Coriolis forces. The calculated results were evaluated using experimental data of the three-dimensional velocity fields, wall static pressures, and wall heat transfer on the suction surface of the turbine airfoil and on the end wall. Particular attention was paid to an accurate modeling of the passage vortex and to the development of the wall boundary layers including crossflow.

Blair, Michael F.↗

Seismic Waveform Inversion Capability on Resource-Constrained Edge Devices

Seismic full wave inversion (FWI) is a widely used non-linear seismic imaging method used to reconstruct subsurface velocity images, however it is time consuming, has high computational cost and depend heavily on human interaction. Recently, deep learning has accelerated it’s use in several data-driven techniques, however most deep learning techniques suffer from overfitting and stability issues. In this work, we propose an edge computing-based data-driven inversion technique based on supervised deep convolutional neural network to accurately reconstruct the subsurface velocities. Deep learning based data-driven technique depends mostly on bulk data training. In this work, we train our deep convolutional neural network (DCN) (UNet and InversionNet) on the raw seismic data and their corresponding velocity models during the training phase to learn the non-linear mapping between the seismic data and velocity models. The trained network is then used to estimate the velocity models from new input seismic data during the prediction phase. The prediction phase is performed on a resource-constrained edge device such as Raspberry Pi. Raspberry Pi provides real-time and on-device computational power to execute the inference process. In addition, we demonstrate robustness of our models to perform inversion in the presence on noise by performing both noise-aware and no-noise training and feeding the resulting trained models with noise at different signal-to-noise (SNR) ratio values. We make great efforts to achieve very feasible inference times on the Raspberry Pi for both models. Specifically, the inference times per prediction for UNet and InversionNet models on Raspberry Pi were 22 and 4 s respectively whilst inference times for both models on the GPU were 2 and 18 s which are very comparable. Finally, we have designed a user-friendly interactive graphical user interface (GUI) to automate the model execution and inversion process on the Raspberry Pi.

Manu, Daniel (ORCID:0000000154982677)↗

Genomics for Everyone

Genomics — the genetic mapping and DNA sequencing of sets of genes or the complete genomes of organisms, along with related genome analysis and database work — is emerging as one of the transformative sciences of the 21st century. But current bioinformatics tools are not accessible to most biological researchers. Now, a new computational and web-based tool called EDGE Bioinformatics is working to fulfill the promise of democratizing genomics.

59 BASIC BIOLOGICAL SCIENCES↗

FPGA Computing

The articles in this special section focus on cutting-edge research on topics that relate to field programmable gate arrays (FPGA) in computing.

97 MATHEMATICS AND COMPUTING↗

Application of thin-layer Navier-Stokes equations near maximum lift

The flowfield about a NACA 0012 airfoil at a Mach number of 0.3 and Reynolds number of 1 million is computed through an angle of attack range, up to 18 deg, corresponding to conditions up to and beyond the maximum lift coefficient. Results obtained using the compressible thin-layer Navier-Stokes equations are presented as well as results from the compressible Euler equations with and without a viscous coupling procedure. The applicability of each code is assessed and many thin-layer Navier-Stokes benchmark solutions are obtained which can be used for comparison with other codes intended for use at high angles of attack. Reasonable agreement of the Navier-Stokes code with experiment and the viscous-inviscid interaction code is obtained at moderate angles of attack. An unsteady solution is obtained with the thin-layer Navier-Stokes code at the highest angle of attack considered. The maximum lift coefficient is overpredicted, however, in comparison to experimental data, which is attributed to the presence of a laminar separation bubble near the leading edge not modeled in the computations. Two comparisons with experimental data are also presented at a higher Mach number.

Anderson, W. K.↗

The Effect on the Buckling of Perfect Cylinders of Prebuckling Deformations and Stresses Induced by Edge Support

Large deflection theory is used to compute buckling loads of simply supported initially perfect cylinders under axial compression, external hydrostatic pressure, and combinations of axial compression and internal or external pressure. Important results are obtained by taking into account prebuckling deformations and stresses induced by edge support. For example, the presence of these deformations and stresses can decrease the axial-compression buckling load of an unpressurized perfect cylinder by 50 percent or more.

CYLINDER↗

Analog Systems for Edge Optimization

Over the past decade, analog computing has the subject of substantial research interest providing a path toward improved computational efficiency in the post-Dennard era. Analog matrix vector multiplication (MVM) accelerators provide a popular approach given the ubiquity of MVM operations in numerous applications. However, historically analog computing systems can struggle with applications requiring high precision due to the inherent susceptibility of these systems to analog non-idealities. Therefore, prior work on analog systems has focused either on applications known to be tolerant of limited precision (e.g., neural network inference), or using expensive techniques to emulate high-precision using many analog MVM operations. In this work, we propose an alternative approach. Motivated by recent advances in inexact nonlinear solvers and optimizers, we explore the potential of co-designing optimization algorithms which can take full advantage of the fundamentally inexact analog MVM operations. To enable these co-designed algorithms we also develop a general mathematical theory of the precision and energy efficiency of analog operations, and a new system architecture for tightly-coupled analog and digital computation. Finally, we examine the applicability of analog computing to a wider class of symmetric positive definite systems and find potential in using analog operations as a sparse approximate inverse preconditioner. With these core innovations, this project provides a path toward effectively implementing optimization algorithms on power-constrained autonomous and semi-autonomous systems.

97 MATHEMATICS AND COMPUTING↗

The energies and structure of X-ray K-absorption edges

Improved screening parameters are employed to compute the K threshold energies E(K) and photoelectric cross sections sigma(K) for the abundant atomic species of the interstellar gas. The E(K)-values for neutral C, N, and O differ from the values currently being employed in X-ray astronomy by several percent. The values are computed in an effective Coulomb field formulation, introducing an 'effective Coulomb' threshold, and convenient formulas are given for the extrapolation of the results to lower energies in the neighborhood of the real threshold energy. Within about 30-200 eV above threshold, the corrected cross sections display structure associated with direct secondary excitation of outer-shell electrons into the continuum. A unified description of this feature of the cross section is introduced, in terms of the secondary threshold energies and core relaxation factors. Both of these factors are computed with the improved screening parameters associated with the one-electron atomic states of the many-electron systems.

Gould, Robert J.↗

Noise Radiation From a Leading-Edge Slat

This paper extends our previous computations of unsteady flow within the slat cove region of a multi-element high-lift airfoil configuration, which showed that both statistical and structural aspects of the experimentally observed unsteady flow behavior can be captured via 3D simulations over a computational domain of narrow spanwise extent. Although such narrow domain simulation can account for the spanwise decorrelation of the slat cove fluctuations, the resulting database cannot be applied towards acoustic predictions of the slat without invoking additional approximations to synthesize the fluctuation field over the rest of the span. This deficiency is partially alleviated in the present work by increasing the spanwise extent of the computational domain from 37.3% of the slat chord to nearly 226% (i.e., 15% of the model span). The simulation database is used to verify consistency with previous computational results and, then, to develop predictions of the far-field noise radiation in conjunction with a frequency-domain Ffowcs-Williams Hawkings solver.

Lockard, David P.↗

Performance Characteristics of the BlueField-2 SmartNIC

High-performance computing (HPC) researchers have long envisioned scenarios where application workflows could be improved through the use of programmable processing elements embedded in the network fabric. Recently, vendors have introduced programmable Smart Network Interface Cards (SmartNICs) that enable computations to be offloaded to the edge of the network. There is great interest in both the HPC and high-performance data analytics (HPDA) communities in understanding the roles these devices may play in the data paths of upcoming systems. This paper focuses on characterizing both the networking and computing aspects of NVIDIA’s new BlueField-2 SmartNIC when used in a 100Gb/s Ethernet environment. For the networking evaluation we conducted multiple transfer experiments between processors located at the host, the SmartNIC, and a remote host. These tests illuminate how much effort is required to saturate the network and help estimate the processing headroom available on the SmartNIC during transfers. For the computing evaluation we used the stress-ng benchmark to compare the BlueField-2 to other servers and place realistic bounds on the types of offload operations that are appropriate for the hardware. Our findings from this work indicate that while the BlueField-2 provides a flexible means of processing data at the network’s edge, great care must be taken to not overwhelm the hardware. While the host can easily saturate the network link, the SmartNIC’s embedded processors may not have enough computing resources to sustain more than half the expected bandwidth when using kernel-space packet processing. From a computational perspective, encryption operations, memory operations under contention, and on-card IPC operations on the SmartNIC perform significantly better than the general-purpose servers used for comparisons in our experiments. Therefore, applications that mainly focus on these operations may be good candidates for offloading to the SmartNIC.

97 MATHEMATICS AND COMPUTING↗

Comparison of Methods for Determining Boundary Layer Edge Conditions for Transition Correlations

Data previously obtained for the X-33 in the NASA Langley Research Center 20-Inch Mach 6 Air Tunnel have been reanalyzed to compare methods for determining boundary layer edge conditions for use in transition correlations. The experimental results were previously obtained utilizing the phosphor thermography technique to monitor the status of the boundary layer downstream of discrete roughness elements via global heat transfer images of the X-33 windward surface. A boundary layer transition correlation was previously developed for this data set using boundary layer edge conditions calculated using an inviscid/integral boundary layer approach. An algorithm was written in the present study to extract boundary layer edge quantities from higher fidelity viscous computational fluid dynamic solutions to develop transition correlations that account for viscous effects on vehicles of arbitrary complexity. The boundary layer transition correlation developed for the X-33 from the viscous solutions are compared to the previous boundary layer transition correlations. It is shown that the boundary layer edge conditions calculated using an inviscid/integral boundary layer approach are significantly different than those extracted from viscous computational fluid dynamic solutions. The present results demonstrate the differences obtained in correlating transition data using different computational methods.

Liechty, Derek S.↗

Changes in Flat Plate Wake Characteristics Obtained With Decreasing Plate Thickness

The near and very near wake of a flat plate with a circular trailing edge is investigated with data from direct numerical simulations. Computations were performed for four different Reynolds numbers based on plate thickness (D) and at constant plate length. The value of θ/D varies by a factor of approximately 20 in the computations (θ being the boundary layer momentum thickness at the trailing edge). The separating boundary layers are turbulent in all the cases. One objective of the study is to understand the changes in wake characteristics as the plate thickness is reduced (increasing θ/D). Vortex shedding is vigorous in the low θ/D cases with a substantial decrease in shedding intensity in the largest θ/D case (for all practical purposes shedding becomes almost intermittent). Other characteristics that are significantly altered with increasing θ/D are the roll-up of the detached shear layers and the magnitude of fluctuations in shedding period. These effects are explored in depth. The effects of changing θ/D on the distributions of the time-averaged, near-wake velocity statistics are discussed.

Wakes↗

HPC Annual Report 2017

Sandia National Laboratories has a long history of significant contributions to the high performance community and industry. Our innovative computer architectures allowed the United States to become the first to break the teraFLOP barrier—propelling us to the international spotlight. Our advanced simulation and modeling capabilities have been integral in high consequence US operations such as Operation Burnt Frost. Strong partnerships with industry leaders, such as Cray, Inc. and Goodyear, have enabled them to leverage our high performance computing (HPC) capabilities to gain a tremendous competitive edge in the marketplace. As part of our continuing commitment to providing modern computing infrastructure and systems in support of Sandia missions, we made a major investment in expanding Building 725 to serve as the new home of HPC systems at Sandia. Work is expected to be completed in 2018 and will result in a modern facility of approximately 15,000 square feet of computer center space. The facility will be ready to house the newest National Nuclear Security Administration/Advanced Simulation and Computing (NNSA/ASC) Prototype platform being acquired by Sandia, with delivery in late 2019 or early 2020. This new system will enable continuing advances by Sandia science and engineering staff in the areas of operating system R&D, operation cost effectiveness (power and innovative cooling technologies), user environment and application code performance.

97 MATHEMATICS AND COMPUTING↗

Effects of Nose Bluntness on Stability of Hypersonic Boundary Layers over Blunt Cone

Receptivity and stability of hypersonic boundary layers are numerically investigated for boundary layer flows over a 5-degree straight cone at a free-stream Mach number of 6.0. To compute the shock and the interaction of shock with the instability waves, we solve the Navier-Stokes equations in axisymmetric coordinates. The governing equations are solved using the 5th-order accurate weighted essentially non-oscillatory (WENO) scheme for space discretization and using third-order total-variation-diminishing (TVD) Runge-Kutta scheme for time integration. After the mean flow field is computed, disturbances are introduced at the upstream end of the computational domain. Generation of instability waves from leading edge region and receptivity of boundary layer to slow acoustic waves are investigated. Computations are performed for a cone with nose radii of 0.001, 0.05 and 0.10 inches that give Reynolds numbers based on the nose radii ranging from 650 to 130,000. The linear stability results showed that the bluntness has a strong stabilizing effect on the stability of axisymmetric boundary layers. The transition Reynolds number for a cone with the nose Reynolds number of 65,000 is increased by a factor of 1.82 compared to that for a sharp cone. The receptivity coefficient for a sharp cone is about 4.23 and it is very small, approx.10(exp -3), for large bluntness.

Kara, K.↗