Search NASA⌕ Search

SEARCH · Search NASA

Results for “distributed memory”

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 757 records · Page 42

Loss circulation prevention in geothermal drilling by shape memory polymer

Geothermal formations are naturally fractured with large fracture openings and networks. This can often lead to frequent drilling fluid loss events which is a major contributor to the cost and non-productive time in geothermal drilling. Development of smarter technologies and methods to tackle problem lost circulation is vital for geothermal drilling cost reduction required for geothermal energy to be recognized as a competitive alternative energy source. Mitigation of this problem and other drilling problems such as stuck pipe can minimize overall project cost. In this report, a thermoset shape memory polymer that can be activated by formation natural heat was assessed to seal near wellbore fractures in geothermal wells. The performance of the shape memory polymer (SMP) was evaluated using artificial fractures created in aluminum discs and cylindrical granite cores. Rheology and particle size distribution were considered. A novel testing setup was built for this work, which allows testing of sealing efficiency under dynamic conditions at high temperature. Analysis showed that the SMP has efficiently succeeded in forming a strong plug inside the fractures and stopped fluid loss at high sealing pressure. This smart loss circulation material can expand within the fractures to reduces non-drilling time and strengthen the wellbore in high-temperature drilling operations.

58 GEOSCIENCES↗

An Accurate, Error-Tolerant, and Energy-Efficient Neural Network Inference Engine Based on SONOS Analog Memory

In this work, we demonstrate SONOS (silicon-oxide-nitrideoxide- silicon) analog memory arrays that are optimized for neural network inference. The devices are fabricated in a 40nm process and operated in the subthreshold regime for in-memory matrix multiplication. Subthreshold operation enables low conductances to be implemented with low error, which matches the typical weight distribution of neural networks, which is heavily skewed toward near-zero values. This leads to high accuracy in the presence of programming errors and process variations. We simulate the end-to-end neural network inference accuracy, accounting for the measured programming error, read noise, and retention loss in a fabricated SONOS array. Evaluated on the ImageNet dataset using ResNet50, the accuracy using a SONOS system is within 2.16% of floating-point accuracy without any retraining. The unique error properties and high On/Off ratio of the SONOS device allow scaling to large arrays without bit slicing, and enable an inference architecture that achieves 20 TOPS/W on ResNet50, a >10× gain in energy efficiency over state-of-the-art digital and analog inference accelerators.

97 MATHEMATICS AND COMPUTING↗

Advanced Data Structures for Improved Cyber Resilience and Awareness in Untrusted Environments: LDRD Report

This report summarizes the work performed under the project "Advanced Data Structures for Improved Cyber Resilience and Awareness in Untrusted Environments." The goal of the project was to design, analyze, and test new data structures for cybersecurity applications. We had two major thrusts: 1) using new/improved write-optimized data structures and/or algorithms to better man- age and analyze high-speed massive-volume cyberstreams, and 2) adding security features to data structures at minimum cost. Write optimization allows data structures to better use secondary memory to store and search a larger amount of useful information. Secondary memory is large compared to main memory, but data movement to and from secondary memory must be carefully managed to run quickly enough to keep up with fast streams. The first thrust included managing cyberstreams in parallel, both multi-threaded and distributed, and improving the benchmarking infrastructure for testing new streaming data structures. We considered both (near) real-time discovery of particular patterns, and improved logging for improved forensics. We considered two kinds of security-feature problem. The first was high-performance history- independent external-memory data structures. These provide certain protections to data if a disk is stolen. We also prove some trade-offs between speed and security in this setting. The second data-security problem is more secure data look-up in secret-shared data bases. This report summarizes the project's major accomplishments, with the background to under- stand these accomplishments. It gathers the abstracts and references for the six refereed publications that have appeared as part of this work. We summarize several accomplishments that will be submitted for publication. We then archive one piece of partial work that is not likely to be published in the near future: validation of history-independent data structure implementations.

97 MATHEMATICS AND COMPUTING↗

Efficient generation of sum-of-products representations of high-dimensional potential energy surfaces based on multimode expansions

The transformation of multi-dimensional potential energy surfaces (PESs) from a grid-based multimode representation to an analytical one is a standard procedure in quantum chemical programs. Within the framework of linear least squares fitting, a simple and highly efficient algorithm is presented, which relies on a direct product representation of the PES and a repeated use of Kronecker products. It shows the same scalings in computational cost and memory requirements as the POTFIT approach. In comparison to customary linear least squares fitting algorithms, this corresponds to a speed-up and memory saving by several orders of magnitude. Different fitting bases are tested, namely, polynomials, B-splines, and distributed Gaussians. Benchmark calculations are provided for the PESs of a set of small molecules.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Regional Medium-Term Hourly Electricity Demand Forecasting Based on LSTM

This paper aims to forecast high-resolution (hourly) aggregated load for a certain region in the medium term (a few days to over a year). One region is defined as some places with similar climate characteristics because the climate influences people's daily lifestyles and hence the electric usage. We decom- pose the electric usage records into two parts: base load and seasonal load. Considering both temperature and time factors, different deep learning methods are adopted to characterize them. The first goal of our approach is to predict the peak load which is critical for power system planning. Furthermore, our proposed forecast method can provide the depiction of the hourly load profile to provide customized load curves for high- level real-time applications. The proposed method is tested on real-world historical data collected by CAISO, BPA, and PACW. The experimental results show that trained by three years of data, our method could reduce the prediction error for one-year lead hourly load below 5% MAPE, and predict the occurrence of the peak load for next year in CAISO with an error within three days. Furthermore, as a byproduct, an interesting observation on the impact of COVID-19 on human life was made and discussed based on these case studies.

deep learning↗

F and Cl detection limits in secondary ion mass spectrometry measurements of Si and SiO{sub 2} samples

Usually, F and Cl negative secondary ion currents forming background signals determine detection limits for these elements in secondary ion mass spectrometry (SIMS) measurements. Technique involving changing the primary-ion current was used to determine that the ''memory effect'' is the main source of the background signal for these elements. It is shown that energy distributions of F and Cl ions forming background signals are distinctly narrower than energy distributions of these ions reflecting element bulk concentration in a Si wafer. Using this fact, the F and Cl detection limits in Si were improved by energy filtering of secondary ions. For experimental conditions used, the filtering of the secondary ion energy in a region around 60-70 eV improves the F and Cl detection limits by a factor of more than 6. The effect of an electron beam used for charge neutralization of insulator samples in a negative SIMS mode on the F and Cl detection limits was also investigated. It is shown that the simultaneous use of the primary ion and electron beams causes a synergistic increase of the background signals. It was determined that both F and Cl detection limits for the SiO{sub 2} sample are almost the same as for the Si sample when the electron and ion beams are applied simultaneously. The possible mechanism of this effect is discussed.

36 MATERIALS SCIENCE↗

PLATO (Parallel Load Assignment Tool): A Parallel Workload Partitioner for Particle-based Methods

A key first step in many large-scale high-performance scientific computing applications is the distribution of the computational domain across all participating processors of a parallel computing platform such as a shared memory machine, cluster or a tightly-coupled supercomputer. The primary goal of such a partitioning step is to ensure that the computational work performed by each process remains load-balanced throughout the execution of the parallel application. In this project, a general-purpose parallel partitioning tool, code-named PLATO, was designed and implemented. PLATO assumes that the workload is a set of points in an n-dimensional metric space. These points are abstractions of computational units that may represent point-like computational units such as celestial objects in astrophysics, atoms in materials sciences and molecules in biology/chemistry, or may represent more abstract computational units such as grid points in finite difference methods, grid elements in finite element methods or n-dimensional points in the feature space of a host of other applications. PLATO supports several key and useful capabilities, namely: (a) multiple load partitioning strategies, (b) support for arbitrary user-defined processor topology, (c) pre-fetching data from neighboring processors based on user-defined parameters, and (d) support for arbitrary number and types of attributes associated with each point in the computational domain. PLATO is implemented in the C programming language using Message Passing Interface (MPI) for inter-processor communications and its parallel performance tested on up to 48 processors of a shared memory machine. This report includes a detailed performance study of PLATO using parallel speedup and iso-efficiency as the two main metrics of parallel performance. In addition, a user manual and DOXYGEN generated code documentation have been included for completeness.

97 MATHEMATICS AND COMPUTING↗

Regional Medium-Term Hourly Electricity Demand Forecasting Based on LSTM: Preprint

This paper aims to forecast high-resolution (hourly) aggregated load for a certain region in the medium term (a few days to over a year). One region is defined as some places with similar climate characteristics because the climate influences people's daily lifestyles and hence the electric usage. We decompose the electric usage records into two parts: base load and seasonal load. Considering both temperature and time factors, different deep-learning methods are adopted to characterize them. The first goal of our approach is to predict the peak load which is critical for power system planning. Furthermore, our proposed forecast method can provide the depiction of the hourly load profile to provide customized load curves for high-level real-time applications. The proposed method is tested on real-world historical data collected by CAISO, BPA, and PACW. The experimental results show that trained by three years of data, our method could reduce the prediction error for a one-year lead hourly load below 5% MAPE, and predict the occurrence of the peak load for next year in CAISO with an error within three days. Furthermore, as a byproduct, an interesting observation on the impact of COVID-19 on human life was made and discussed based on these case studies.

deep learning↗

Velocity-Space Hybridization of Direct Simulation Monte Carlo and a Quasi-Particle Boltzmann Solver

This paper presents a new method for modeling rarefied gas flows based on hybridization of direct simulation Monte Carlo (DSMC) and discrete velocity method (DVM)-based quasi-particle representations of the velocity distribution function. It is aimed at improving the resolution of the tails of the distribution function (compared with DSMC) and computational efficiency (compared with DVM). Details of the method, such as the collision algorithm and the particle merging scheme, are discussed. The hybrid approach is applied to the study of noise in a Maxwellian distribution, computation of electron-impact ionization rate coefficient, as well as numerical simulation of a supersonic Couette flow. The hybrid-based solver is compared with pure DSMC and DVM approaches in terms of accuracy, computational speed, and memory use. It is shown that such a hybrid approach can provide a lower computational cost than a pure DVM approach, while being able to retain accuracy in modeling high-velocity tails of the distribution function. For problems where trace species have a significant impact on the flow physics, the proposed method is shown to be capable of providing better computational efficiency and accuracy compared with standard fixed-weight DSMC.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

FPGA Acceleration of GCN in Light of the Symmetry of Graph Adjacency Matrix

Graph Convolutional Neural Networks (GCNs) are widely used to process large-scale graph data. Different from deep neural networks (DNNs), GCNs are sparse, irregular, and unstructured, posing unique challenges to hardware acceleration with regular processing elements (PEs). In particular, the adjacency matrix of a GCN is extremely sparse, leading to frequent but irregular memory access, low spatial/temporal data locality and poor data reuse. Furthermore, a realistic graph usually consists of unstructured data (e.g., unbalanced distributions), creating significantly different processing times and imbalanced workload for each node in GCN acceleration. To overcome these challenges, we propose an end-to-end hardware-software co-design to accelerate GCNs on resource-constrained FPGAs with the features including: (1) A custom dataflow that leverages symmetry along the diagonal of the adjacency matrix to accelerate feature aggregation for undirected graphs. We utilize either the upper or the lower triangular matrix of the adjacency matrix to perform aggregation in GCN to improve data reuse. (2) Unified compute cores for both aggregation and transform phases, with full support to the symmetry-based dataflow. These cores can be dynamically reconfigured to the systolic mode for transformation or as individual accumulators for aggregation in GCN processing. (3) Preprocessing of the graph in software to rearrange the edges and features to match the custom dataflow. This step improves the regularity in memory access and data reuse in the aggregation phase. Moreover, we quantize the GCN precision from FP32 to INT8 to reduce the memory footprint without losing the inference accuracy. We implement our accelerator design in Intel Stratix10 MX FPGA board with HBM2, and demonstrate 1.3x-110.5x improvement in end-to-end GCN latency as compared to the state-of the-art FPGA implementations, on the graph datasets of Cora, Pubmed, Citeseer and Reddit.

Nair, Gopikrishnan R.↗

Long–short-term memory encoder–decoder with regularized hidden dynamics for fault detection in industrial processes

The ability of recurrent neural networks (RNN) to model nonlinear dynamics of high dimensional process data has enabled data-driven RNN-based fault detection algorithms. Previous studies have focused on detecting faults by identifying the discrepancies in data distribution between the faulty and normal data, as reflected in prediction errors generated by RNN models. However, in industrial processes, variations in data distribution can also result from changes in normal control setpoints and compensatory control adjustments in response to disturbances, making it hard to differentiate between normal and faulty conditions. This paper proposes a fault detection method utilizing a long short-term memory (LSTM) encoder–decoder structure with regularized hidden dynamics and reversible instance normalization (RevIN) to compactly represent high-dimensional measurements for effective monitoring. During training, the hidden states of the model are regularized to form a low-dimensional latent space representation of the original multivariate time series data. As a result, the prediction errors of the latent states can be used to monitor the abnormal dynamic variations, while the reconstruction errors of the measured variables are used to monitor the abnormal static variations. Furthermore, the proposed indices can reflect operating conditions, even when the distribution of test data changes, which helps distinguish faults from normal adjustments and disturbances that controllers can settle. Here, data from numerical simulation and the Tennessee Eastman process are used to illustrate the effectiveness of the proposed fault detection method.

42 ENGINEERING↗

A new hybrid-Lagrangian numerical scheme for gyrokinetic simulation of tokamak edge plasma

In order to enable kinetic simulation of non-thermal edge plasmas at a reduced computational cost, a new hybrid-Lagrangian δf scheme has been developed that utilizes the phase space grid in addition to the usual marker particles, taking advantage of the computational strengths from both sides. The new scheme splits the particle distribution function of a kinetic equation into two parts. Marker particles contain the fast space-time varying, δf, part of the distribution function and the coarse-grained phase-space grid contains the slow space-time varying part. The coarse-grained phase-space grid reduces the memory-requirement and the computing cost, while the marker particles provide scalable computing ability for the fine-grained physics. Weights of the marker particles are determined by a direct weight evolution equation instead of the differential form weight evolution equations that the conventional delta-f schemes use. The particle weight can be slowly transferred to the phase space grid, thereby reducing the growth of the particle weights. The non-Lagrangian part of the kinetic equation – e.g., collision operation, ionization, charge exchange, heat-source, radiative cooling, and others – can be operated directly on the phase space grid. Deviation of the particle distribution function on the velocity grid from a Maxwellian distribution function – driven by ionization, charge exchange and wall loss – is allowed to be arbitrarily large. The numerical scheme is implemented in the gyrokinetic particle code XGC1, which specializes in simulating the tokamak edge plasma that crosses the magnetic separatrix and is in contact with the material wall.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Kelvin Probe Force Microscopy Imaging of Plasticity in Hydrogenated Perovskite Nickelate Multilevel Neuromorphic Devices

Ion drift in nanoscale electronically inhomogeneous semiconductors is among the most important mechanisms being studied for designing neuromorphic computing hardware. However, nondestructive imaging of the ion drift in operando devices directly responsible for multiresistance states and synaptic memory represents a formidable challenge. Here, we present Kelvin probe force microscopy imaging of hydrogen-doped perovskite nickelate device channels subject to high-speed electric field pulses to directly visualize proton distribution by monitoring surface potential changes spatially, which is also supported with finite element-based electric field distribution studies. First-principles calculations provide mechanistic insights into the origin of surface potential changes as a function of hydrogen donor doping that serves as the contrast mechanism. We demonstrate 128 (7-bit) nonvolatile conductance levels in such devices relevant to in-memory computing applications. The synaptic plasticity measurements are implemented in spiking neural networks and show promising results for classification (SciKit Learn’s Iris and Wine data sets) and control (OpenAI’s CartPole-v1 and BipedalWalker-v3) simulation tasks.

Kelvin probe force microscopy↗

Crystal Phase Distribution and Ferroelectricity in Ultrathin HfO 2 –ZrO 2 Bilayers

(Hf,Zr)O 2 ultrathin films are used as ferroelectric layers in emerging digital logic and nonvolatile memory devices. The ferroelectric properties of (Hf,Zr)O 2 can be improved by interface engineering, such as the formation of nanolaminates with distinct HfO 2 and ZrO 2 layers. Herein, the ferroelectric performance of HfO 2 –ZrO 2 ultrathin bilayer devices is shown to depend on the stacking order of HfO 2 and ZrO 2 , which affects the quantity of the noncentrosymmetric orthorhombic Pca 2 1 crystal phase. By combining X‐ray diffraction with a novel extended X‐ray absorption fine structure (EXAFS) analysis procedure, the orthorhombic, tetragonal, and monoclinic phase fractions are quantified for bilayers composed of 3 nm HfO 2 and 3 nm ZrO 2 . A significantly larger orthorhombic ZrO 2 phase fraction is found when ZrO 2 has an unconstrained surface during annealing, whereas the presence of a ZrO 2 interface with the substrate results in a substantial tetragonal ZrO 2 phase fraction and a 2.4× smaller remanent polarization. HfO 2 is found to be less susceptible than ZrO 2 to crystal phase templating. The methods presented herein enable mechanistic studies of ferroelectric wake‐up, fatigue, and processing effects in (Hf,Zr)O 2 films, accelerating the development of electronic devices that rely on ferroelectric oxides.

McBriarty, Martin E.↗

THE BUILDUP OF A SCALE-FREE PHOTOSPHERIC MAGNETIC NETWORK

We use a global Monte Carlo simulation of the formation of the solar photospheric magnetic network to investigate the origin of the scale invariance characterizing magnetic flux concentrations visible on high-resolution magnetograms. The simulations include spatially and temporally homogeneous injection of small-scale magnetic elements over the whole photosphere, as well as localized episodic injection associated with the emergence and decay of active regions. Network elements form in response to cumulative pairwise aggregation or cancellation of magnetic elements, undergoing a random walk on the sphere and advected on large spatial scales by differential rotation and a poleward meridional flow. The resulting size distribution of simulated network elements is in very good agreement with observational inferences. We find that the fractal index and size distribution of network elements are determined primarily by these post-emergence surface mechanisms, and carry little or no memory of the scales at which magnetic flux is injected in the simulation. Implications for models of dynamo action in the Sun are briefly discussed.

79 ASTRONOMY AND ASTROPHYSICS↗

Big Data Meets HPC Log Analytics: Scalable Approach to Understanding Systems at Extreme Scale

Today's high-performance computing (HPC) systems are heavily instrumented, generating logs containing information about abnormal events, such as critical conditions, faults, errors and failures, system resource utilization, and about the resource usage of user applications. These logs, once fully analyzed and correlated, can produce detailed information about the system health, root causes of failures, and analyze an application's interactions with the system, providing valuable insights to domain scientists and system administrators. However, processing HPC logs requires a deep understanding of hardware and software components at multiple layers of the system stack. Moreover, most log data is unstructured and voluminous, making it more difficult for system users and administrators to manually inspect the data. With rapid increases in the scale and complexity of HPC systems, log data processing is becoming a big data challenge. This paper introduces a HPC log data analytics framework that is based on a distributed NoSQL database technology, which provides scalability and high availability, and the Apache Spark framework for rapid in-memory processing of the log data. The analytics framework enables the extraction of a range of information about the system so that system administrators and end users alike can obtain necessary insights for their specific needs. We describe our experience with using this framework to glean insights from the log data about system behavior from the Titan supercomputer at the Oak Ridge National Laboratory.

Park, Byung↗

A hybrid parallel framework for the cellular Potts model simulations

The Cellular Potts Model (CPM) has been widely used for biological simulations. However, most current implementations are either sequential or approximated, which can't be used for large scale complex 3D simulation. In this paper we present a hybrid parallel framework for CPM simulations. The time-consuming POE solving, cell division, and cell reaction operation are distributed to clusters using the Message Passing Interface (MPI). The Monte Carlo lattice update is parallelized on shared-memory SMP system using OpenMP. Because the Monte Carlo lattice update is much faster than the POE solving and SMP systems are more and more common, this hybrid approach achieves good performance and high accuracy at the same time. Based on the parallel Cellular Potts Model, we studied the avascular tumor growth using a multiscale model. The application and performance analysis show that the hybrid parallel framework is quite efficient. The hybrid parallel CPM can be used for the large scale simulation ({approx}10{sup 8} sites) of complex collective behavior of numerous cells ({approx}10{sup 6}).

97 MATHEMATICS AND COMPUTING↗

Interfractional Variations in the Setup of Pelvic Bony Anatomy and Soft Tissue, and Their Implications on the Delivery of Proton Therapy for Localized Prostate Cancer

Purpose: To quantify daily variations in the anatomy of patients undergoing radiation therapy for prostate carcinoma, to estimate their effect on dose distribution, and to evaluate the effectiveness of current standard planning and setup approaches employed in proton therapy. Methods: We used series of computed tomography data, which included the pretreatment scan, and between 21 and 43 in-room scans acquired on different treatment days, from 10 patients treated with intensity-modulated radiation therapy at Morristown Memorial Hospital. Variations in femur rotation angles, thickness of subcutaneous adipose tissue, and physical depth to the distal surface of the prostate for lateral beam arrangement were recorded. Proton dose distributions were planned with the standard approach. Daily variations in the location of the prescription isodose were evaluated. Results: In all 10 datasets, substantial variation was observed in the lateral tissue thickness (standard deviation of 1.7-3.6 mm for individual patients, variations of >5 mm from the planning computed tomography observed in all series), and femur rotation angle (standard deviation between 1.3{sup o} and 4.8{sup o}, with the maximum excursion exceeding 10{sup o} in 6 of 10 datasets). Shifts in the position of treated volume (98% isodose) were correlated with the variations in the lateral tissue thickness. Conclusions: Analysis suggests that, combined with image-guided setup verification, the range compensator expansion technique prevents loss of dose to target from femur rotation and soft-tissue deformation, in the majority of cases. Anatomic changes coupled with the uncertainties of particle penetration in tissue restrict possibilities for margin reduction in proton therapy of prostate cancer.

62 RADIOLOGY AND NUCLEAR MEDICINE↗