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 1,405 records · Page 78

Programming Strategies for Irregular Algorithms on the Emu Chick

The Emu Chick prototype implements migratory memory-side processing in a novel hardware system. Rather than transferring large amounts of data across the system interconnect, the Emu Chick moves lightweight thread contexts to near-memory cores before the beginning of each remote memory read. Previous work has characterized the performance of the Chick prototype in terms of memory bandwidth and programming differences from more typical, non-migratory platforms, but there has not yet been an analysis of algorithms on this system. This work evaluates irregular algorithms that could benefit from the lightweight, memory-side processing of the Chick and demonstrates techniques and optimization strategies for achieving performance in sparse matrix-vector multiply operation (SpMV), breadth-first search (BFS), and graph alignment across up to eight distributed nodes encompassing 64 nodelets in the Chick system. We also define and justify relative metrics to compare prototype FPGA-based hardware with established ASIC architectures. The Chick currently supports up to 68x scaling for graph alignment, 80 MTEPS for BFS on balanced graphs, and 50% of measured STREAM bandwidth for SpMV.

Hein, Eric↗

Three-Dimensional Simulations of Electron Beams Focused by Periodic Permanent Magnets

A fully three-dimensional (3D) model of an electron beam focused by a periodic permanent magnet (PPM) stack has been developed. First, the simulation code MAFIA was used to model a PPM stack using the magnetostatic solver. The exact geometry of the magnetic focusing structure was modeled; thus, no approximations were made regarding the off-axis fields. The fields from the static solver were loaded into the 3D particle-in-cell (PIC) solver of MAFIA where fully 3D behavior of the beam was simulated in the magnetic focusing field. The PIC solver computes the time-integration of electromagnetic fields simultaneously with the time integration of the equations of motion of charged particles that move under the influence of those fields. Fields caused by those moving charges are also taken into account; thus, effects like space charge and magnetic forces between particles are fully simulated. The electron beam is simulated by a number of macro-particles. These macro-particles represent a given charge Q amounting to that of several million electrons in order to conserve computational time and memory. Particle motion is unrestricted, so particle trajectories can cross paths and move in three dimensions under the influence of 3D electric and magnetic fields. Correspondingly, there is no limit on the initial current density distribution of the electron beam, nor its density distribution at any time during the simulation. Simulation results including beam current density, percent ripple and percent transmission will be presented, and the effects current, magnetic focusing strength and thermal velocities have on beam behavior will be demonstrated using 3D movies showing the evolution of beam characteristics in time and space. Unlike typical beam optics models, this 3D model allows simulation of asymmetric designs such as non- circularly symmetric electrostatic or magnetic focusing as well as the inclusion of input/output couplers.

Kory, Carol L.↗

Navier-Stokes Computations on Commodity Computers

In this paper we discuss and demonstrate the feasibility of solving high-fidelity, nonlinear computational fluid dynamics (CFD) problems of practical interest on commodity machines, namely Pentium Pro PC's. Such calculations have now become possible due to the progress in computational power and memory of the off-the-shelf commodity computers, along with the growth in bandwidth and communication speeds of networks. A widely used CFD code known as TLNS3D, which was developed originally on large shared memory computers was selected for this effort. This code has recently been ported to massively parallel processor (MPP) type machines, where natural partitioning along grid blocks is adopted in which one or more blocks are distributed to each of the available processors. In this paper, a similar approach is adapted to port this code to a cluster of Pentium Pro computers. The message passing among the processors is accomplished through the use of standard message passing interface (MPI) libraries. Scaling studies indicate fairly high level of parallelism on such clusters of commodity machines, thus making solutions to Navier-Stokes equations for practical problems more affordable.

Vatsa, Veer N.↗

Representation of Serendipitous Scientific Data

A computer program defines and implements an innovative kind of data structure than can be used for representing information derived from serendipitous discoveries made via collection of scientific data on long exploratory spacecraft missions. Data structures capable of collecting any kind of data can easily be implemented in advance, but the task of designing a fixed and efficient data structure suitable for processing raw data into useful information and taking advantage of serendipitous scientific discovery is becoming increasingly difficult as missions go deeper into space. The present software eases the task by enabling definition of arbitrarily complex data structures that can adapt at run time as raw data are transformed into other types of information. This software runs on a variety of computers, and can be distributed in either source code or binary code form. It must be run in conjunction with any one of a number of Lisp compilers that are available commercially or as shareware. It has no specific memory requirements and depends upon the other software with which it is used. This program is implemented as a library that is called by, and becomes folded into, the other software with which it is used.

James, Mark↗

Registration and fusion of large-scale melt pool temperature and morphology monitoring data demonstrated for surface topography prediction in LPBF

In-situ monitoring technologies for laser powder bed fusion (LPBF) additive manufacturing often face one key challenge, extracting the ultrafast melt pool (MP) signatures for understanding the localized part properties. Further, the spatial information of each monitored MP signature is essential for correlating the MP – part property. This spatial information is often unavailable especially from commercial LPBF printers. Many MP monitoring methods have been reported and utilized. However, very few of these have the MP’s spatial information. To overcome this challenge, in this work we report a method for spatially registering the key MP signatures (MP intensity, temperature, and area) to the monitored print parts. The MP signatures are obtained from our coaxial high-speed single-camera based two-wavelength imaging pyrometry (STWIP) system and the MP spatial information is obtained from an off-axis camera system. A machine learning aided image analysis method is employed to retrieve the spatial distribution of MPs within the corresponding part’s coordinates system. Then, the MP signature maps (MPSMs) are reconstructed by mapping the STWIP measured MP signatures to the registered MP coordinates. Further, a long short-term memory (LSTM) neural network is developed for estimating the layer surface topography from the registered MPSMs. The obtained results indicate that the layer surface topography can be more accurately estimated by using MP temperature signature rather than MP intensity and/or area signatures as in common practice. Finally, our developed methods for MP monitoring, registration, and MP-surface topography prediction offer advanced capabilities for the online detection of process anomalies and part defects.

36 MATERIALS SCIENCE↗

Optimizing checkpoint data placement with guaranteed burst buffer endurance in large-scale hierarchical storage systems

Non-volatile devices, such as SSDs, will be an integral part of the deepening storage hierarchy on large-scale HPC systems. These devices can be on the compute nodes as part of a distributed burst buffer service or they can be external. Wherever they are located in the hierarchy, one critical design issue is the SSD endurance under the write-heavy workloads, such as the checkpoint I/O for scientific applications. For these environments, it is widely assumed that checkpoint operations can occur once every 60 min and for each checkpoint step as much as half of the system memory can be written out. Unfortunately, for large-scale HPC applications, the burst buffer SSDs can be worn out much more quickly given the extensive amount of data written at every checkpoint step. One possible solution is to control the amount of data written by reducing the checkpoint frequency. However, a direct effect caused by reduced checkpoint frequency is the increased vulnerability window of system failures and therefore potentially wasted computation time, especially for large-scale compute jobs.

Computer Science↗

Electric Potential and Electric Field Imaging with Dynamic Applications & Extensions

The technology and methods for remote quantitative imaging of electrostatic potentials and electrostatic fields in and around objects and in free space is presented. Electric field imaging (EFI) technology may be applied to characterize intrinsic or existing electric potentials and electric fields, or an externally generated electrostatic field made be used for volumes to be inspected with EFI. The baseline sensor technology (e-Sensor) and its construction, optional electric field generation (quasi-static generator), and current e- Sensor enhancements (ephemeral e-Sensor) are discussed. Critical design elements of current linear and real-time two-dimensional (2D) measurement systems are highlighted, and the development of a three dimensional (3D) EFI system is presented. Demonstrations for structural, electronic, human, and memory applications are shown. Recent work demonstrates that phonons may be used to create and annihilate electric dipoles within structures. Phonon induced dipoles are ephemeral and their polarization, strength, and location may be quantitatively characterized by EFI providing a new subsurface Phonon-EFI imaging technology. Results from real-time imaging of combustion and ion flow, and their measurement complications, will be discussed. Extensions to environment, Space and subterranean applications will be presented, and initial results for quantitative characterizing material properties are shown. A wearable EFI system has been developed by using fundamental EFI concepts. These new EFI capabilities are demonstrated to characterize electric charge distribution creating a new field of study embracing areas of interest including electrostatic discharge (ESD) mitigation, manufacturing quality control, crime scene forensics, design and materials selection for advanced sensors, combustion science, on-orbit space potential, container inspection, remote characterization of electronic circuits and level of activation, dielectric morphology of structures, tether integrity, organic molecular memory, atmospheric science, weather prediction, earth quake prediction, and medical diagnostic and treatment efficacy applications such as cardiac polarization wave propagation and electromyography imaging.

Generazio, Ed↗

Structural signature and transition dynamics of Sb 2 Te 3 melt upon fast cooling

Crystalline Sb 2 Te 3 is widely studied due to its important applications in memory materials and topological insulators. The liquid and amorphous structures of this compound as well as the dynamics upon quenching, however, are yet to be fully understood. In this paper, we have systematically studied the dynamical properties and local structure of Sb 2 Te 3 at different temperatures using ab initio molecular dynamics simulations. The calculated structure factors agree well with the experimental results. The atomic number density and mean-squared displacement as a function of temperature clearly indicate three states as the temperature decreases, namely, melt, undercooled liquid and glass state, respectively. By analyzing the chemical environments and bond-angle distribution functions, we demonstrate that the most probable short-range motifs in the Sb 2 Te 3 system are defective octahedrons, and they are connected with each other via four-fold rings. Finally, this interesting structural feature may be responsible for the high fragility and easy phase transition upon glass forming that is applied in memory devices.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Quantum key distribution over probabilistic quantum repeaters

A feasible route toward implementing long-distance quantum key distribution (QKD) systems relies on probabilistic schemes for entanglement distribution and swapping as proposed in the work of Duan, Lukin, Cirac, and Zoller (DLCZ) [Nature (London) 414, 413 (2001)]. Here, we calculate the conditional throughput and fidelity of entanglement for DLCZ quantum repeaters by accounting for the DLCZ self-purification property in the presence of multiple excitations in the ensemble memories as well as loss and other sources of inefficiency in the channel and measurement modules. We then use our results to find the generation rate of secure key bits for QKD systems that rely on DLCZ quantum repeaters. We compare the key generation rate per logical memory employed in the two cases with and without a repeater node. We find the crossover distance beyond which the repeater system outperforms the nonrepeater one. That provides us with the optimum internode distancing in quantum repeater systems. We also find the optimal excitation probability at which the QKD rate peaks. Such an optimum probability, in most regimes of interest, is insensitive to the total distance.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

NSTX-U advances in real-time deterministic PCIe-based internode communication

Distributing control mechanisms across modern commercial off the shelf (COTS) GNU/Linux systems often introduces difficult to mitigate non-deterministic behavior. Existing methods to address this problem involve non-real-time technologies such as Remote Direct Memory Access (RDMA) over Infiniband or custom Ethernet solutions, such as RDMA over Converged Ethernet (RoCE) or userspace Ethernet drivers, that trade determinism for ease of use or lower cost. The National Spherical Torus Experiment Upgrade (NSTX-U) is pursuing a new design that allows direct communication between heterogeneous systems with scalable, microsecond latency with 1 μs of jitter on that latency, outside of the constant transmission delay at the physical layer. The future design of the NSTX-U Real-time Communication System will utilize direct PCIe-to-PCIe communication with kernel support tuned for low overhead, allowing two (or more, through a switch) real-time (RT) systems to communicate and share resources as one larger entity. As a result, this greatly increases the processing capability of the primary Plasma Control System (PCS), turning previously insurmountable computational challenges into a more manageable divide and conquer parallel task.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Frog: Asynchronous Graph Processing on GPU with Hybrid Coloring Model

GPUs have been increasingly used to accelerate graph processing for complicated computational problems regarding graph theory. Many parallel graph algorithms adopt the asynchronous computing model to accelerate the iterative convergence. Unfortunately, the consistent asynchronous computing requires locking or atomic operations, leading to significant penalties/overheads when implemented on GPUs. As such, coloring algorithm is adopted to separate the vertices with potential updating conflicts, guaranteeing the consistency/correctness of the parallel processing. Common coloring algorithms, however, may suffer from low parallelism because of a large number of colors generally required for processing a large-scale graph with billions of vertices. We propose a light-weight asynchronous processing framework called Frog with a preprocessing/hybrid coloring model. The fundamental idea is based on Pareto principle (or 80-20 rule) about coloring algorithms as we observed through masses of realworld graph coloring cases. We find that a majority of vertices (about 80%) are colored with only a few colors, such that they can be read and updated in a very high degree of parallelism without violating the sequential consistency. Accordingly, our solution separates the processing of the vertices based on the distribution of colors. In this work, we mainly answer three questions: (1) how to partition the vertices in a sparse graph with maximized parallelism, (2) how to process large-scale graphs that cannot fit into GPU memory, and (3) how to reduce the overhead of data transfers on PCIe while processing each partition. Here, we conduct experiments on real-world data (Amazon, DBLP, YouTube, RoadNet-CA, WikiTalk and Twitter) to evaluate our approach and make comparisons with well-known non-preprocessed (such as Totem, Medusa, MapGraph and Gunrock) and preprocessed (Cusha) approaches, by testing four classical algorithms (BFS, PageRank, SSSP and CC). On all the tested applications and datasets, Frog is able to significantly outperform existing GPU-based graph processing systems except Gunrock and MapGraph. MapGraph gets better performance than Frog when running BFS on RoadNet-CA. The comparison between Gunrock and Frog is inconclusive. Frog can outperform Gunrock more than 1.04X when running PageRank and SSSP, while the advantage of Frog is not obvious when running BFS and CC on some datasets especially for RoadNet-CA.

97 MATHEMATICS AND COMPUTING↗

GridLAB-D/SG

GridLAB-D is a new power system simulation tool that provides valuable information to users who design and operate electric power transmission and distribution systems, and to utilities that wish to take advantage of the latest smart grid technology. This special release of GridLAB-D was developed to study the proposed Smart Grid technology that is used by Battelle Memorial Institute in the AEP gridSMART demonstration project in Northeast Columbus, Ohio.

Chassin, DavidP.↗

Machine Learning for Memory Reduction in the Implicit Monte Carlo Simulations of Thermal Radiative Transfer [Slides]

Project Goal: Use parametric Machine Learning methods in order to reduce memory requirements at checkpointing & restarting in the IMC simulations of Thermal Radiative Transfer using: Expectation Maximization and Weighted Gaussian Mixture Model-based approach for `particle-data compression', introduced in Plasma Physics to model Maxwellian particle distributions by Luis Chacon and Guangye Chen; Expectation Maximization with Weighted Hyper-Erlang Model in order to compress isotropic IMC particle data in the frequency domain; and Expectation Maximization and von Mises-Fisher Mixture Model for compression of anisotropic IMC particle data in the angular domain (work-in-progress).

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗

Plasma enhanced multistate storage capability of single ZnO nanowire based memory

Multiple-state storage (MSS) is common for resistive random access memory, but the effects of plasma treatment on the MSS and the switching properties have been scarcely investigated. We have demonstrated a stable four-state storage capability of single zinc oxide nanowire (ZnO NW) treated by argon plasma. The electrical switching is attributed to the electron trapping and detrapping from the oxygen vacancies (V{sub o}s). The MSS relates to the electrical-thermal induced distribution of the V{sub o}s which determines electron transport behavior to show different resistance states. Additionally, programming (set and reset) voltages decrease with plasma treatment due to the thickness modulation of the interface barrier.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Carbon doping in GeTe drives differences in local structure and properties

Advances in low-power, energy-efficient information storage and computing require understanding and controlling the atomic and nanoscale structures of functional materials, such as phase-change materials. Phase-change memory technology enables nonvolatile, low-power memory in devices by storing information through reversible changes in a phase-change material's atomic structure (i.e., transformations between amorphous and crystalline phases) that have corresponding changes in properties, including electronic resistivity and optical reflectivity. Here, we apply complementary X-ray absorption spectroscopy and X-ray pair distribution function analyses to experimentally identify the local- and medium-range atomic structure differences of GeTe and C-doped GeTe thin films. Upon controlled heating, composition- and temperature-dependent atomic structure evolution in GeTe and C-doped GeTe films shows differences in bonding behavior and local structure that directly influence crystallization onset temperature. We find that the introduction of C interrupts Ge–Ge bonds in amorphous GeTe, altering the as-deposited structure to be more similar to the distorted rocksalt structure of crystalline α–GeTe. The change alters the response of the amorphous atomic structure to heating and also lowers the crystallization onset temperature, from 230 °C in GeTe to 220 °C in the C-doped film. The combined insights from both X-ray techniques provide understanding of structural transformations that enables the development and optimization of next-generation memory and computing materials.

36 MATERIALS SCIENCE↗

Estimation of stochastic volatility with long memory for index prices of FTSE Bursa Malaysia KLCI

In recent years, modeling in long memory properties or fractionally integrated processes in stochastic volatility has been applied in the financial time series. A time series with structural breaks can generate a strong persistence in the autocorrelation function, which is an observed behaviour of a long memory process. This paper considers the structural break of data in order to determine true long memory time series data. Unlike usual short memory models for log volatility, the fractional Ornstein-Uhlenbeck process is neither a Markovian process nor can it be easily transformed into a Markovian process. This makes the likelihood evaluation and parameter estimation for the long memory stochastic volatility (LMSV) model challenging tasks. The drift and volatility parameters of the fractional Ornstein-Unlenbeck model are estimated separately using the least square estimator (lse) and quadratic generalized variations (qgv) method respectively. Finally, the empirical distribution of unobserved volatility is estimated using the particle filtering with sequential important sampling-resampling (SIR) method. The mean square error (MSE) between the estimated and empirical volatility indicates that the performance of the model towards the index prices of FTSE Bursa Malaysia KLCI is fairly well.

97 MATHEMATICS AND COMPUTING↗

Particle simulation of plasmas on the massively parallel processor

Particle simulations, in which collective phenomena in plasmas are studied by following the self consistent motions of many discrete particles, involve several highly repetitive sets of calculations that are readily adaptable to SIMD parallel processing. A fully electromagnetic, relativistic plasma simulation for the massively parallel processor is described. The particle motions are followed in 2 1/2 dimensions on a 128 x 128 grid, with periodic boundary conditions. The two dimensional simulation space is mapped directly onto the processor network; a Fast Fourier Transform is used to solve the field equations. Particle data are stored according to an Eulerian scheme, i.e., the information associated with each particle is moved from one local memory to another as the particle moves across the spatial grid. The method is applied to the study of the nonlinear development of the whistler instability in a magnetospheric plasma model, with an anisotropic electron temperature. The wave distribution function is included as a new diagnostic to allow simulation results to be compared with satellite observations.

Gledhill, I. M. A.↗

Rheoencephalographic and electroencephalographic measures of cognitive workload: analytical procedures

This investigation demonstrates the feasibility of mental workload assessment by rheoencephalographic (REG) and multichannel electroencephalographic (EEG) monitoring. During the performance of this research, unique testing, analytical and display procedures were developed for REG and EEG monitoring that extend the current state of the art and provide valuable tools for the study of cerebral circulatory and neural activity during cognition. REG records are analyzed to provide indices of the right and left hemisphere hemodynamic changes that take place during each test sequence. The EEG data are modeled using regression techniques and mathematically transformed to provide energy-density distributions of the scalp electrostatic field. These procedures permit concurrent REG/EEG cognitive testing not possible with current techniques. The introduction of a system for recording and analysis of cognitive REG/EEG test sequences facilitates the study of learning and memory disorders, dementia and other encephalopathies.

Arousal/physiology↗