Engineering Papers⌕ Search

SEARCH · Engineering Papers

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 991 records · Page 55

Accelerating DCA++ (Dynamical Cluster Approximation) Scientific Application on the Summit supercomputer

Optimizing scientific applications on today’s accelerator-based high performance computing systems can be challenging, especially when multiple GPUs and CPUs with heterogeneous memories and persistent non-volatile memories are present. An example is Summit, an accelerator-based system at the Oak Ridge Leadership Computing Facility (OLCF) that is rated as the world’s fastest supercomputer to-date. New strategies are thus needed to expose the parallelism in legacy applications, while being amenable to efficient mapping to the underlying architecture.In this paper we discuss our experiences and strategies to port a scientific application, DCA++, to Summit. DCA++ is a highperformance research application that solves quantum manybody problems with a cutting edge quantum cluster algorithm, the dynamical cluster approximation.Our strategies aim to synergize the strengths of the different programming models in the code. These include: (a) streamlining the interactions between the CPU threads and the GPUs, (b) implementing computing kernels on the GPUs and decreasing CPU-GPU memory transfers, (c) allowing asynchronous GPU communications, and (d) increasing compute intensity by combining linear algebraic operations.Full-scale production runs using all 4600 Summit nodes attained a peak performance of 73.5 PFLOPS with a mixed precision implementation.We observed a perfect strong and weak scaling for the quantum Monte Carlo solver in DCA++, while encountering about 2× input/output (I/O) and MPI communication overhead on the time-to-solution for the full machine run. Our hardware agnostic optimizations are designed to alleviate the communication and I/O challenges observed, while improving the compute intensity and obtaining optimal performance on a complex, hybrid architecture like Summit.

Balduzzi, Giovanni↗

Drekar v.2.0

Drekar is an application code that solves partial differential equations for fluids that can be optionally coupled to electromagnetics. Drekar solves low-mach compressible and incompressible computational fluid dynamics (CFD), compressible and incompressible resistive magnetohydrodynamics (MHD), and multiple species plasmas interacting with electromagnetic fields. Drekar discretization technology includes continuous and discontinuous finite element formulations, stabilized finite element formulations, mixed integration finite element bases (nodal, edge, face, volume) and an initial arbitrary Lagrangian Eulerian (ALE) capability. Drekar contains the implementation of the discretized physics and leverages the open source Trilinos project for both parallel solver capabilities and general finite element discretization tools. The code will be released open source under a BSD license. The code is used for fundamental research for simulation of fluids and plasmas on high performance computing environments.

Seefeldt, Ben↗

Classical eikonal from Magnus expansion

In a classical scattering problem, the classical eikonal is defined as the generator of the canonical transformation that maps in-states to out-states. It can be regarded as the classical limit of the log of the quantum S-matrix. In a classical analog of the Born approximation in quantum mechanics, the classical eikonal admits an expansion in oriented tree graphs, where oriented edges denote retarded/advanced worldline propagators. The Magnus expansion, which takes the log of a time-ordered exponential integral, offers an efficient method to compute the coefficients of the tree graphs to all orders. We exploit a Hopf algebra structure behind the Magnus expansion to develop a fast algorithm which can compute the tree coefficients up to the 12th order (over half a million trees) in less than an hour. In a relativistic setting, our methods can be applied to the post-Minkowskian (PM) expansion for gravitational binaries in the worldline formalism. We demonstrate the methods by computing the 3PM eikonal and find agreement with previous results based on amplitude methods. Importantly, the Magnus expansion yields a finite eikonal, while the naïve eikonal based on the time-symmetric propagator is infrared-divergent from 3PM on.

Black Holes↗

Adsorption of Potassium on the MoS2(100) Surface: A First-Principles Investigation

Periodic density functional theory calculations were performed to investigate the interaction that potassium with the Mo and S edges of the MoS2(100) surface. Both neutral and cationic (+1) charged potassium-promoted systems at different sulfur coverages were considered. Our calculations indicate that the potassium atom readily donates its single 4s valence electron to the MoS2 structure for the neutral potassium-promoted system, and the neutral and cationic potassium-promoted systems demonstrate a similar adsorption behavior. Moreover, potassium changes the magnetic properties known to occur at the metallic edge surface, which have implications for electron spin dependent surface characterization methods (i.e., electron spin/paramagnetic spectroscopy). Potassium in both the neutral and cationic systems tends to maximize its interactions with the available sulfur atoms at the edge surface, preferring sites over four-fold S hollows on fully sulfided Mo and S edges and over the interstitial gap where two to four edge surface S atoms are available for coordination. As the potassium coverage increases, the adsorption energy per potassium atom, surface work function, and transfer of the K 4s electron to the MoS2(100) surface decreases, which is in line with an increased metallization of the potassium adlayer. The potassium adlayer tends to form chains along the interstitial with K-K distances ~1 Å, which is notably less than those of bulk bcc K metal (4.61 Å). Density of states for the potassium-saturated surface suggests enhanced involvement of broad K 3d states beginning just above the Fermi level. Potassium-promotion of MoS2(100) has implications for alcohol catalysis: increasing the surface basicity by increasing the electron charge of the surface, providing hydrogenation-promoting CO site, blocking edge surface that dissociate CO and lead to methanation, and limiting H2 dissociative adsorption to the edge surface and possibly inhibiting the H2 dissociative adsorption via s character electron repulsion. This research was performed in part using the Molecular Science Computing Facility in the William R. Wiley Environmental Molecular Sciences Laboratory, a U.S. Department of Energy (DOE) national scientific user facility located at the Pacific Northwest National Laboratory (PNNL). PNNL is operated by Battelle for DOE.

10 SYNTHETIC FUELS↗

Calculations of two-fluid linear response to non-axisymmetric fields in tokamaks

The zero-frequency linear plasma response to static applied non-axisymmetric fields is calculated using a resistive two-fluid model in diverted, toroidal geometry. Within this model, the effects on the plasma response of resistivity, rotation, differential ion and electron velocity, and dissipation are explored. Rotation is generally found to inhibit the formation of islands in the plasma, in qualitative agreement with theoretical results. When two-fluid effects are included, it is found that the penetration of the non-axisymmetric fields is generally greatest when the part of the electron rotation perpendicular to the equilibrium magnetic field is small at the mode-rational surface. Strong rotation shear in the edge is found to enhance the plasma response there. The entire plasma, including the separatrix and scrape-off layer, is included in the computational domain.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

O3BNN-R: An Out-Of-Order Architecture for High-Performance and Regularized BNN Inference

Binarized Neural Networks (BNN) have drawn tremendous attention due to significantly reduced computational complexity and memory demand. They have especially shown great potential in cost- and power-restricted domains, such as IoT and smart edge-devices, where reaching a certain accuracy bar is often sufficient, and real-time is highly desired.In this work, we demonstrate that the highly-condensed BNN model can be shrunk significantly further by dynamically pruning irregular redundant edges. Based on two new observations on BNN-specific properties, an out-of-order (OoO) architecture – O3BNN-R, can curtail edge evaluation in cases where the binary output of a neuron can be determined early. Similar to Instruction-Level-Parallelism(ILP), these fine-grained, irregular, runtime pruning opportunities are traditionally presumed to be difficult to exploit. In order to increase the pruning opportunities, we also optimize the training process by adding 2 regularization items in the loss function (1) for pooling pruning and (2) for threshold pruning. We evaluate our design on an FPGA platform using three well-known networks, including VggNet-16, AlexNet for ImageNet, and a VGG-like network for Cifar-10.

Geng, Tong↗

SU-E-J-133: Autosegmentation of Linac CBCT: Improved Accuracy Via Penalized Likelihood Reconstruction

Purpose: To improve the quality of kV X-ray cone beam CT (CBCT) for use in radiotherapy delivery assessment and re-planning by using penalized likelihood (PL) iterative reconstruction and auto-segmentation accuracy of the resulting CBCTs as an image quality metric. Methods: Present filtered backprojection (FBP) CBCT reconstructions can be improved upon by PL reconstruction with image formation models and appropriate regularization constraints. We use two constraints: 1) image smoothing via an edge preserving filter, and 2) a constraint minimizing the differences between the reconstruction and a registered prior image. Reconstructions of prostate therapy CBCTs were computed with constraint 1 alone and with both constraints. The prior images were planning CTs(pCT) deformable-registered to the FBP reconstructions. Anatomy segmentations were done using atlas-based auto-segmentation (Elekta ADMIRE). Results: We observed small but consistent improvements in the Dice similarity coefficients of PL reconstructions over the FBP results, and additional small improvements with the added prior image constraint. For a CBCT with anatomy very similar in appearance to the pCT, we observed these changes in the Dice metric: +2.9% (prostate), +8.6% (rectum), −1.9% (bladder). For a second CBCT with a very different rectum configuration, we observed +0.8% (prostate), +8.9% (rectum), −1.2% (bladder). For a third case with significant lateral truncation of the field of view, we observed: +0.8% (prostate), +8.9% (rectum), −1.2% (bladder). Adding the prior image constraint raised Dice measures by about 1%. Conclusion: Efficient and practical adaptive radiotherapy requires accurate deformable registration and accurate anatomy delineation. We show here small and consistent patterns of improved contour accuracy using PL iterative reconstruction compared with FBP reconstruction. However, the modest extent of these results and the pattern of differences across CBCT cases suggest that significant further development will be required to make CBCT useful to adaptive radiotherapy.

60 APPLIED LIFE SCIENCES↗

cuAlign: Scalable Network Alignment on GPU Accelerators

Given two graphs, the objective of network alignment is to find the best one-to-one mapping of vertices in one graph (??) to vertices in the other (??), such that the number of overlaps is maximized. We say that edges(??, ??) ???and(??', ??') ??? are overlapped if ?? is mapped to ??' and ?? is mapped to??'. Network alignment is an important optimization problem with several applications in bioinformatics, computer vision and ontology matching. Since it is an NP-hard problem, efficient heuristics and scalable implementations are necessary. In this work, we introduce a new framework that combines the concepts of intra-network proximity using vertex embedding,Belief Propagation (BP) and approximate weighted matching, and provides qualitative improvements up to22%over state-of-the-art approaches. We also provide scalable implementations on GPU accelerators, demonstrating up to19×speedup for Belief Propagation and 3× speedup for approximate weighted matching relative to previous multithreaded implementation. A combination of combinatorial and algebraic kernels within the network alignment algorithm poses significant hurdles for parallelization. Load imbalance and irregular DRAM traffic limit achievable performance on GPUs. Our novel approach identifies and exploits unique structural proper-ties of the BP-based algorithm and employs code fusion to reduce data movement between different steps of the algorithm. Using a diverse set of inputs, we demonstrate qualitative improvements of our algorithms, and performance gains of our GPU-accelerated implementation. We believe that our work will enable algorithmic improvements and practical applications of network alignment.

Xiang, Lizhi↗

An Observation-Driven Agent-Based Modeling and Analysis Framework for C. elegans Embryogenesis

With cutting-edge live microscopy and image analysis, biologists can now systematically track individual cells in complex tissues and quantify cellular behavior over extended time windows. Computational approaches that utilize the systematic and quantitative data are needed to understand how cells interact in vivo to give rise to the different cell types and 3D morphology of tissues. An agent-based, minimum descriptive modeling and analysis framework is presented in this paper to study C. elegans embryogenesis. The framework is designed to incorporate the large amounts of experimental observations on cellular behavior and reserve data structures/interfaces that allow regulatory mechanisms to be added as more insights are gained. Observed cellular behaviors are organized into lineage identity, timing and direction of cell division, and path of cell movement. The framework also includes global parameters such as the eggshell and a clock. Division and movement behaviors are driven by statistical models of the observations. Data structures/interfaces are reserved for gene list, cell-cell interaction, cell fate and landscape, and other global parameters until the descriptive model is replaced by a regulatory mechanism. This approach provides a framework to handle the ongoing experiments of single-cell analysis of complex tissues where mechanistic insights lag data collection and need to be validated on complex observations.

Science & Technology - Other Topics↗

Addressing Rising Energy Demand Through Innovation

The U.S. is facing a significant increase in energy demand, driven by AI advancements, the rapid expansion of data centers, manufacturing and industrial growth, and the electrification of transportation and buildings. Buildings alone account for approximately 75% of U.S. electricity consumption and 40% of total energy use. To address these challenges, NLR leverages its state-of-the-art research facilities, advanced energy modeling, hardware-in-the-loop emulation, and real-world demonstrations to provide data-driven insights that de-risk emerging energy solutions, increase efficiency and demand flexibility, optimize grid controls, and identify vulnerabilities to enhance energy security. This presentation will highlight our research ecosystem and its role in supporting a more reliable, affordable, and adaptive energy infrastructure in the face of accelerating demand.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗

Micro-spectroscopic and freezing characterization of ice-nucleating particles collected in the marine boundary layer in the eastern North Atlantic

Abstract. Formation of atmospheric ice plays a crucial role in the microphysical evolution of mixed-phase and cirrus clouds and thus climate. How aerosol particles impact ice crystal formation by acting as ice-nucleating particles (INPs) is a subject of intense research activities. To improve understanding of atmospheric INPs, we examined daytime and nighttime particles collected during the Aerosol and Cloud Experiments in the Eastern North Atlantic (ACE-ENA) field campaign conducted in summer 2017. Collected particles, representative of a remote marine environment, were investigated for their propensity to serve as INPs in the immersion freezing (IMF) and deposition ice nucleation (DIN) modes. The particle population was characterized by chemical imaging techniques such as computer-controlled scanning electron microscopy with energy-dispersive X-ray analysis (CCSEM/EDX) and scanning transmission X-ray microscopy with near-edge X-ray absorption fine-structure spectroscopy (STXM/NEXAFS). Four major particle-type classes were identified where internally mixed inorganic–organic particles make up the majority of the analyzed particles. Following ice nucleation experiments, individual INPs were identified and characterized by SEM/EDX. The identified INP types belong to the major particle-type classes consisting of fresh sea salt with organics or processed sea salt containing dust and sulfur with organics. Ice nucleation experiments show IMF events at temperatures as low as 231 K, including the subsaturated regime. DIN events were observed at lower temperatures of 210 to 231 K. IMF and DIN observations were analyzed with regard to activated INP fraction, ice-nucleation active site (INAS) densities, and a water activity-based immersion freezing model (ABIFM) yielding heterogeneous ice nucleation rate coefficients. Observed IMF and DIN events of ice formation and corresponding derived freezing rates demonstrate that the marine boundary layer aerosol particles can serve as INPs under typical mixed-phase and cirrus cloud conditions. The derived IMF and DIN parameterizations allow for implementation in cloud and climate models to evaluate predictive effects of atmospheric ice crystal formation.

54 ENVIRONMENTAL SCIENCES↗

Ultrasound waveform tomography with spatial and edge regularization

Synthetic-aperture ultrasound tomography systems and methods using scanning arrays and algorithms configured to simultaneously acquire ultrasound transmission and reflection data, and process the data for improved ultrasound tomography imaging, wherein the tomography imaging comprises total-variation regularization, or a modified total variation regularization, particularly with edge-guided or spatially variant regularization.

47 OTHER INSTRUMENTATION↗

High-Performance Computing for Real-Time Grid Analysis and Operation

Power grids worldwide are undergoing an unprecedented transition as a result of grid evolution meeting information revolution. The grid evolution is largely driven by the desire for green energy. Emerging grid technologies such as renewable generation, smart loads, plug-in hybrid vehicles, and distributed generation provide opportunities to generate energy from green sources and to manage energy use for better system efficiency. With utility companies actively deploying these technologies, a high level of penetration of these new technologies is expected in the next 5-10 years, bringing in a level of intermittency, uncertainties, and complexity that the grid did not see nor design for. On the other hand, the information infrastructure in the power grid is being revolutionized with large-scale deployment of sensors and meters in both the transmission and distribution networks. The future grid will have two-way flows of both electrons and information. The challenge is how to take advantage of the information revolution: pull the large amount of data in, process it in real time, and put information out to manage grid evolution. Without addressing this challenge, the opportunities in grid evolution will remain unfulfilled. This transition poses grand challenges in grid modeling, simulation, and information presentation. The computational complexity of underlying power grid modeling and simulation will significantly increase in the next decade due to an increased model size and a decreased time window allowed to compute model solutions. High-performance computing is essential to enable this transition. The essential technical barrier is to vastly increase the computational speed so operation response time can be reduced from minutes to seconds and sub-seconds. The speed at which key functions such as state estimation and contingency analysis are conducted (typically every 3-5 minutes) needs to be dramatically increased so that the analysis of contingencies is both comprehensive and real time. An even bigger challenge is how to incorporate dynamic information into real-time grid operation. Today’s online grid operation is based on a static grid model and can only provide a static snapshot of current system operation status, while dynamic analysis is conducted offline because of low computational efficiency. The offline analysis uses a worst-case scenario to determine transmission limits, resulting in under-utilization of grid assets. This conservative approach does not necessarily lead to reliability. Many times, actual power grid scenarios are not studied, and they will push the grid over the edge and resulting in outages and blackouts. This chapter addresses the HPC needs in power grid analysis and operations. Example applications such as state estimation and contingency analysis are given to demonstrate the value of HPC in power grid applications. Future research directions are suggested for high performance computing applications in power grids to improve the transparency, efficiency, and reliability of power grids.

Contingency Analysis↗

Geospatial-temporal semantic graph representations of trajectories from remote sensing and geolocation data

Various technologies for facilitating analysis of large remote sensing and geolocation datasets to identify features of interest are described herein. A search query can be submitted to a computing system that executes searches over a geospatial temporal semantic (GTS) graph to identify features of interest. The GTS graph comprises nodes corresponding to objects described in the remote sensing and geolocation datasets, and edges that indicate geospatial or temporal relationships between pairs of nodes in the nodes. Trajectory information is encoded in the GTS graph by the inclusion of movable nodes to facilitate searches for features of interest in the datasets relative to moving objects such as vehicles.

58 GEOSCIENCES↗

Energy-efficient Mott activation neuron for full-hardware implementation of neural networks

To circumvent the von Neumann bottleneck, substantial progress has been made towards in-memory computing with synaptic devices. However, compact nanodevices implementing non-linear activation functions are required for efficient full-hardware implementation of deep neural networks. Here, in this work, we present an energy-efficient and compact Mott activation neuron based on vanadium dioxide and its successful integration with a conductive bridge random access memory (CBRAM) crossbar array in hardware. The Mott activation neuron implements the rectified linear unit function in the analogue domain. The neuron devices consume substantially less energy and occupy two orders of magnitude smaller area than those of analogue complementary metal–oxide semiconductor implementations. The LeNet-5 network with Mott activation neurons achieves 98.38% accuracy on the MNIST dataset, close to the ideal software accuracy. We perform large-scale image edge detection using the Mott activation neurons integrated with a CBRAM crossbar array. Our findings provide a solution towards large-scale, highly parallel and energy-efficient in-memory computing systems for neural networks.

electrical and electronic engineering↗

An Efficient Pipeline for Abdomen Segmentation in CT Images

Computed tomography (CT) scans usually include some disadvantages due to the nature of the imaging procedure, and these handicaps prevent accurate abdomen segmentation. Discontinuous abdomen edges, bed section of CT, patient information, closeness between the edges of the abdomen and CT, poor contrast, and a narrow histogram can be regarded as the most important handicaps that occur in abdominal CT scans. Currently, one or more handicaps can arise and prevent technicians obtaining abdomen images through simple segmentation techniques. In other words, CT scans can include the bed section of CT, a patient’s diagnostic information, low-quality abdomen edges, low-level contrast, and narrow histogram, all in one scan. These phenomena constitute a challenge, and an efficient pipeline that is unaffected by handicaps is required. In addition, analysis such as segmentation, feature selection, and classification has meaning for a real-time diagnosis system in cases where the abdomen section is directly used with a specific size. A statistical pipeline is designed in this study that is unaffected by the handicaps mentioned above. Intensity-based approaches, morphological processes, and histogram-based procedures are utilized to design an efficient structure. Performance evaluation is realized in experiments on 58 CT images (16 training, 16 test, and 26 validation) that include the abdomen and one or more disadvantage(s). The first part of the data (16 training images) is used to detect the pipeline’s optimum parameters, while the second and third parts are utilized to evaluate and to confirm the segmentation performance. The segmentation results are presented as the means of six performance metrics. Thus, the proposed method achieves remarkable average rates for training/test/validation of 98.95/99.36/99.57% (jaccard), 99.47/99.67/99.79% (dice), 100/99.91/99.91% (sensitivity), 98.47/99.23/99.85% (specificity), 99.38/99.63/99.87% (classification accuracy), and 98.98/99.45/99.66% (precision). In summary, a statistical pipeline performing the task of abdomen segmentation is achieved that is not affected by the disadvantages, and the most detailed abdomen segmentation study is performed for the use before organ and tumor segmentation, feature extraction, and classification.

62 RADIOLOGY AND NUCLEAR MEDICINE↗