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 1,621 records · Page 90

The OMPS Limb Profiler Instrument: Two-Dimensional Retrieval Algorithm

The upcoming Ozone Mapper and Profiler Suite (OMPS), which will be launched on the NPOESS Preparatory Project (NPP) platform in early 2011, will continue monitoring the global distribution of the Earth's middle atmosphere ozone and aerosol. OMPS is composed of three instruments, namely the Total Column Mapper (heritage: TOMS, OMI), the Nadir Profiler (heritage: SBUV) and the Limb Profiler (heritage: SOLSE/LORE, OSIRIS, SCIAMACHY, SAGE III). The ultimate goal of the mission is to better understand and quantify the rate of stratospheric ozone recovery. The focus of the paper will be on the Limb Profiler (LP) instrument. The LP instrument will measure the Earth's limb radiance (which is due to the scattering of solar photons by air molecules, aerosol and Earth surface) in the ultra-violet (UV), visible and near infrared, from 285 to 1000 nm. The LP simultaneously images the whole vertical extent of the Earth's limb through three vertical slits, each covering a vertical tangent height range of 100 km and each horizontally spaced by 250 km in the cross-track direction. Measurements are made every 19 seconds along the orbit track, which corresponds to a distance of about 150km. Several data analysis tools are presently being constructed and tested to retrieve ozone and aerosol vertical distribution from limb radiance measurements. The primary NASA algorithm is based on earlier algorithms developed for the SOLSE/LORE and SAGE III limb scatter missions. All the existing retrieval algorithms rely on a spherical symmetry assumption for the atmosphere structure. While this assumption is reasonable in most of the stratosphere, it is no longer valid in regions of prime scientific interest, such as polar vortex and UTLS regions. The paper will describe a two-dimensional retrieval algorithm whereby the ozone distribution is simultaneously retrieved vertically and horizontally for a whole orbit. The retrieval code relies on (1) a forward 2D Radiative Transfer code (to model limb radiances within a non-uniform atmosphere and evaluate 2D analytical partial derivatives) and (2) an optimal estimator inversion routine. The algorithm uses the typically sparse nature of the kernel matrices as well as fast matrix inversion techniques to allow for fast inversion of limb data with efficient memory management (as was done for MIPAS data processing). While the method has so far only been developed in the context of Single Scatter, the paper will show how the CPU intensive Multiple Scatter modeling can be implemented using parallel CPU processing. Initial results will be presented in terms of retrieved ozone profiles and code performance.

Rault, Didier F.↗

Distributed parallel messaging for multiprocessor systems

A method and apparatus for distributed parallel messaging in a parallel computing system. The apparatus includes, at each node of a multiprocessor network, multiple injection messaging engine units and reception messaging engine units, each implementing a DMA engine and each supporting both multiple packet injection into and multiple reception from a network, in parallel. The reception side of the messaging unit (MU) includes a switch interface enabling writing of data of a packet received from the network to the memory system. The transmission side of the messaging unit, includes switch interface for reading from the memory system when injecting packets into the network.

97 MATHEMATICS AND COMPUTING↗

Structural and electronic characteristics of amorphous Ge 8 Sb 2 Te 11

GeTe-rich phase-change materials can be utilized in rewriteable optical memory due to the large contrast in reflectivity between amorphous and crystalline phases. Here we explored the structure and electronic properties of amorphous Ge 8 Sb 2 Te 11 using ab initio molecular dynamics simulations. The results indicate that amorphous Ge 8 Sb 2 Te 11 is dominantly composed of Ge-, Sb- and Te-centered octahedrons with distortions, while 30.4 % of Ge-centered clusters are in the form of tetrahedrons which are randomly distributed. The 5-fold rings possess a large proportion, and the Gesingle bondTe and Sbsingle bondTe bonds present larger formation energies than other bonds, leading to the ABAB bonding sequence (A: Ge and Sb, B: Te). The lone pair electrons locating at the opposite direction of bonds possess a large fraction of 14.8 %, which may enhance the distortions of local clusters. Importantly, these structural properties lead to the high stability of amorphous Ge 8 Sb 2 Te 11 and thus long data retention in the information storage.

36 MATERIALS SCIENCE↗

Wavelet-Smoothed Interpolation of Masked Scientific Data for JPEG 2000 Compression [Slides]

How should we manage scientific data with 'holes'? Some applications, like JPEG 2000, expect logically rectangular data, but some sources, like the Parallel Ocean Program (POP), generate data that isn't defined on certain subsets. We refer to grid points that lack well-defined, scientifically meaningful sample values as 'masked' samples. Wavelet-smoothing is a highly scalable interpolation scheme for regions with complex boundaries on logically rectangular grids. Computation is based on forward/inverse discrete wavelet transforms, so runtime complexity and memory scale linearly with respect to sample count. Efficient state-of-the-art minimal realizations yield small constants (O(10)) for arithmetic complexity scaling, and in-situ implementation techniques make optimal use of memory. Implementation in two dimensions using tensor product filter banks is straighsorward and should generalize routinely to higher dimensions. No hand-tuning required when the interpolation mask changes, making the method aeractive for problems with time-varying masks. Well-suited for interpolating undefined samples prior to JPEG 2000 encoding. The method outperforms global mean interpolation, as judged by both SNR rate-distortion performance and low-rate artifact mitigation, for data distributions whose histograms do not take the form of sharply peaked, symmetric, unimodal probability density functions. These performance advantages can hold even for data whose distribution differs only moderately from the peaked unimodal case, as demonstrated by POP salinity data. The interpolation method is very general and is not tied to any particular class of applications, could be used for more generic smooth interpolation.

42 ENGINEERING↗

SMALE: Enhancing Scalability of Machine Learning Algorithms on Extreme-Scale Computing Platforms

Deployment and execution of machine learning tasks on extreme-scale computing platforms face several significant technical challenges: 1) High computing cost incurred by dense networks – The computing workload of deep networks with densely-connected topology increases rapidly with the network size, imposing a non-scalable computing model of extreme-scale computing platforms; 2) Non-optimized workload distribution – Many advanced deep learning algorithms, e.g., sparsification and irregular net-work topology, produce very unbalanced workload distribution on extreme-scale computing platforms. The computation efficiency is greatly hindered by the incurred data and computation redundancies as well as long tails of the node with extensive workload; 3) Constraints in data movement and I/O bottle-neck – Inter-node data movement in extreme-scale computing platforms are associated with high energy and latency costs, and subject to the constraints of I/O bandwidth; and 4) Generalization of algorithm realization and acceleration on computing platforms – The large varieties of machine learning algorithms and structures of extreme-scale computing platforms make the derivation of a generalized algorithm realization and acceleration method very challenging, which, however, is the requirement by domain scientists and interested users. We call the above challenges Smale’s Problems in Machine Learning and Understanding for High-Performance Computing Scientific Discovery. The objective of our three-year research project is to develop a holistic innovation set at structure, assembly, and acceleration layers of machine learning algorithms to address the above challenges in algorithm deployment and execution. Three tasks are particularly performed, including: At the algorithm structure level, we investigate the techniques that can structurally sparsify on the topology of deep networks for computing workload reduction. We also study clustering and pruning techniques that can optimize the workload distributions over the extreme-scale computing platforms; At the algorithm assembly level, we derive a unified learning framework for unsupervised transfer learning and dynamic growing capabilities. Novel training methods are also exploited to enhance the training efficiency of the proposed framework; At the algorithm acceleration level, we will develop a series of techniques that can accelerate the computation of sparse matrix operations, which are one of the core executions in deep learning and optimize memory access of the concerned platforms. Our proposed techniques attack the fundamental problems in machine learning algorithms running on extreme-scale computing platforms by vertically integrating the solutions at three closely entangled layers, paving the long-term scaling path of machine learning applications under DOE context. Three tasks corresponding to the above respective research orientations are performed during the three-year project period with our collaborators at ORNL. The outcome of the proposed project is anticipated to form a holistic solution set of novel algorithms and network topologies, efficient training techniques, and fast acceleration methods to promote the computing scalability of the machine learning applications of particular interest to DOE.

97 MATHEMATICS AND COMPUTING↗

Fault tolerance using self-checking building-block computers

The paper attempts to define and characterize a set of VLSI (very large scale integration) building-block circuits which can be used to combine existing microprocessors and memories into a wide variety of fault-tolerant computing systems. Such VLSI circuits would transform fault-tolerant computing into an off-the-shelf technology and enable its routine use for new applications. The self-checking computer module (SCCM) is the basic component out of which fault-tolerant computer systems are constructed. Several fault-tolerant configurations of SCCM are discussed, including the standby redundant uniprocessor, the voted/hybrid uniprocessor, and the distributed computer network.

Rennels, D. A.↗

High-speed pitch angle sorter

A high-speed method was developed to compress the two-dimensional angular distribution of space particles gathered by space plasma instrumentation into the angle distribution, where the pitch angle is polar angle with respect to the ambient magnetic field. The pitch angle sorter can handle rates of up to 2 MHz and it is designed to accommodate high angular resolution plasma analyzers that are placed on a rotating spacecraft. This compression is achieved by relying on digitally encoded lookup tables to eliminate all arithmetic operations while applying the high symmetry of this compression to reduce the amount of digital memory.

Keller, John W.↗

Advanced Unstructured Grid Generation for Complex Aerodynamic Applications

A new approach for distribution of grid points on the surface and in the volume has been developed and implemented in the NASA unstructured grid generation code VGRID. In addition to the point and line sources of prior work, the new approach utilizes surface and volume sources for automatic curvature-based grid sizing and convenient point distribution in the volume. A new exponential growth function produces smoother and more efficient grids and provides superior control over distribution of grid points in the field. All types of sources support anisotropic grid stretching which not only improves the grid economy but also provides more accurate solutions for certain aerodynamic applications. The new approach does not require a three-dimensional background grid as in the previous methods. Instead, it makes use of an efficient bounding-box auxiliary medium for storing grid parameters defined by surface sources. The new approach is less memory-intensive and more efficient computationally. The grids generated with the new method either eliminate the need for adaptive grid refinement for certain class of problems or provide high quality initial grids that would enhance the performance of many adaptation methods.

Pirzadeh, Shahyar Z.↗

Advanced Unstructured Grid Generation for Complex Aerodynamic Applications

A new approach for distribution of grid points on the surface and in the volume has been developed. In addition to the point and line sources of prior work, the new approach utilizes surface and volume sources for automatic curvature-based grid sizing and convenient point distribution in the volume. A new exponential growth function produces smoother and more efficient grids and provides superior control over distribution of grid points in the field. All types of sources support anisotropic grid stretching which not only improves the grid economy but also provides more accurate solutions for certain aerodynamic applications. The new approach does not require a three-dimensional background grid as in the previous methods. Instead, it makes use of an efficient bounding-box auxiliary medium for storing grid parameters defined by surface sources. The new approach is less memory-intensive and more efficient computationally. The grids generated with the new method either eliminate the need for adaptive grid refinement for certain class of problems or provide high quality initial grids that would enhance the performance of many adaptation methods.

Pirzadeh, Shahyar↗

Quicksilver

Quicksilver is proxy application for the Monte Carlo Transport code, Mercury. The purpose of the Quicksilver mini-app is to create a freely distributable code base that represents the key elements of the Mercury workload by solving a simplified time-dependent neutron transport problem. As an open source code, Quicksilver will be hosted on the CZ git/stash repo alongside other mini-apps in the ACS so-design suite. It will be used by vendors and other partners for co-design activities such as identifying root causes of Mercury performance bottlenecks on future architectures (i.e., memory bandwidth, memory latency, flop intensity) and developing Quicksilver will be used within LLNL as a nimble platform to evaluate programming models, algorithms, etc. that might be useful in Mercury. Quicksilver may also become a replacement for MCB in future procurement benchmarks, as well as a way to engage vendors in the Trinity and Sierra Centers of Excellence as well as other co-design efforts.

Richards, David↗

Calculation and use of pointwise effective cross-sections in the unresolved resonance region - 14605

The pointwise neutron elastic scattering, absorption, fission and total cross-sections in the Unresolved Resonance Region (URR) can be calculated with a code C-007 [1]. The volume of these data is large and estimated to make about 106 numbers for each of the reactions. These data volumes may cause time and memory problems in solving radiation transport by the Monte Carlo method. There are methods to thin the data. These are, for example, the widely used probability tables [2] and subgroups [3]. The paper offers another data thinning approach which was successfully used for the Resolved Resonance Region (RRR) [4,5]. It transforms the pointwise resonance cross-sections into the pointwise monotone ones within some energy intervals with the use of cross-section probability distribution functions P(σ). A similar approach was used to calculate pointwise cross-sections in the URR [6]. (authors)

42 ENGINEERING↗

Evolution of galaxy bias, generalized

Fry (1996) showed that galaxy bias has the tendency to evolve towards unity, i.e. in the long run, the galaxy distribution tends to trace that of matter. Generalizing slightly Fry's reasoning, we show that his conclusion remains valid in theories of modified gravity (or equivalently, complex clustered dark energy). This is not surprising: as long as both galaxies and matter are subject to the same force, dynamics would drive them towards tracing each other. This holds, for instance, in theories where both galaxies and matter move on geodesics. This relaxation of bias towards unity is tempered by cosmic acceleration, however: the bias tends towards unity but does not quite make it, unless the formation bias were close to unity. Our argument is extended in a straightforward manner to the case of a stochastic or nonlinear bias. An important corollary is that dynamical evolution could imprint a scale dependence on the large scale galaxy bias. This is especially pronounced if nonstandard gravity introduces new scales to the problem: the bias at different scales relaxes at different rates, the larger scales generally more slowly and retaining a longer memory of the initial bias. A consistency test of the current (general relativity+uniform dark energy) paradigm is therefore to look for departure from a scale-independent bias on large scales. A simple way is to measure the relative bias of different populations of galaxies which are at different stages of bias relaxation. Lastly, we comment on the possibility of directly testing the Poisson equation on cosmological scales, as opposed to indirectly through the growth factor.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

Demonstration of Full Core Reactor Depletion with MPACT

MPACT is a three-dimensional (3-D) whole core transport code capable of generating sub-pin level power distributions. This feature is accomplished by obtaining the integral transport solutions to the heterogeneous reactor problem in which the actual detailed geometrical configuration of fuel components such as the pellet and cladding is modelled explicitly.. The cross section data needed for the neutron transport calculation are obtained directly from a multigroup microscopic cross section library similar to those used in lattice physics codes. Hence MPACT involves neither a priori homogenization nor group condensation for the core spatial solution. The integral transport solution is obtained by means of the method of characteristics (MOC) that employs discrete ray tracing. Since the direct application of the 3-D MOC capability in MPACT to 3-D core configuration requires considerable amounts of memory and computing time for practical reactor applications, an alternative approximate 3-D solution method was implemented in MPACT based on a 2D-1D approach which employs planar MOC solutions in the framework of the 3-D coarse mesh finite difference (CMFD) formulation. The axial coupling is resolved by one-dimensional (1-D) diffusion solutions and the planar and axial problems are coupled through the transverse leakage. The use of a lower order 1-D solution in the axial direction is justified by the fact that most heterogeneity in the core occurs in the radial direction rather than the axial. However, it is possible to use one of the higher order methods in MPACT such as the 1-D SN kernel if improved solution accuracy is desired. The CMFD formulation which originally served as an efficient nodal formulation is used in MPACT to accelerate the whole core transport calculation. The basic mesh in the CMFD formulation is a pin cell which is much coarser than the flat source regions defined for MOC calculations. The concept of dynamic homogenization of group constants for the pin cells provides the basis of the CMFD formulation in whole core transport calculation. The intra-cell flux distribution determined from the MOC calculation is used to generate the homogenized cell constants while the MOC cell surface averaged currents are used to determine the radial nodal coupling coefficients. This dynamic implementation of the equivalence theory formulation ensures the same transport solution is obtained with CMFD as the one obtained using unaccelerated MOC calculations alone. The ability of MPACT to predict the Hot, Zero Power (HZP) condition of a reactor was reported in CASL-U-2014-0045-000. The purpose of this report is to demonstrate the capability of MPACT to simulate the operation of a reactor at Hot, Full Power (HFP) conditions during normal operation throughout a full operating cycle. The time scale for a reactor burnup cycle is typically one to two years and therefore short time phenomena such as the effect of delayed neutrons can be neglected. Instead the time dependence can be treated quasi-statically using the depletion algorithm presented in CASL-U-2013-0276-001. The full power operation requires thermal hydraulic feedback to be present during the entire simulation which will utilize the work in CASL-U-2014-0051-000. However, due to the significant computational burden of the coupled physics simulation, a simplified thermal hydraulic model is used for the analysis reported here which is principally intended to demonstrate the full core neutronics depletion. In addition to the thermal hydraulic feedback, the effect of Boron 10 depletion should be modeled in cases where major boron additions are not performed for extended periods of time. Each of these capabilities will be described in the following sections.

22 GENERAL STUDIES OF NUCLEAR REACTORS↗

CUDA Computation of the Feynman Distribution

In 2006, NVIDIA Corporation introduced the Computer Unified Device Architecture (CUDA) for C, C++, and Fortran programming languages. The CUDA parallel computing platform uses the Graphics Processing Unit (GPU) of NVIDIA graphics card, and it is designed for parallel computations. In these computations, the same algorithms are executed on the data in parallel. The GPU contains many Streaming Multiprocessors (SM) that can independently execute blocks of threads. A thread is a sequence of serial programmed instructions. The SM executes threads in groups of 32 and a group of 32 threads is called warp. The threads within a block can be synchronized and have access to a shared memory that is visible only to them. Conversely, blocks of threads cannot be synchronized and have access only to the global memory of the graphics card. This memory is called device memory and is independent of the host memory controlled by the Central Processing Unit (CPU). A single thread has access to a local memory (register) that is not visible to any other thread. The part of the program that is executed on the graphics card is called kernel. Each thread of the kernel has a unique identification index that is accessible by the in-built variable threadIdx; the latter is a 3-component vector. A threads block can contain up to 1024 threads. Blocks of threads have a unique identification index that is accessible by the in-built variable blockIdx; the latter is a 3-component vector. The in-built variable blockDim contains the number of threads per block; this variable is a 3-component vector. Threads and blocks represent different levels of parallelism; a kernel is executed as a grid of blocks of threads. Figure 1 illustrates a grid of thread blocks for a GPU with 2 SM. A SM can execute multiple blocks of threads in parallel; a block of threads is executed on a single SM. A GPU can execute multiple kernels concurrently. This work has used the Quadro M5000 graphics card with the following technical specifications: - 2048 cores with 861 MHz frequency, - 16 streaming multiprocessors, - 5.2 trillion transistors, - 8.2 Gb device memory, - 64 kb register memory (per thread), - 48 kb shared memory (per block) - 211.6 GB/s memory bandwidth, and - 3.5 GFLOPS floating point performance.

97 MATHEMATICS AND COMPUTING↗

Imaging the structural organization of chemical elements in growth cones of developing hippocampal neurons

Abstract During neurodevelopment, neurons form growth cones, F-actin rich extensions located at the distal end of the neurites. Growth cones allow dendrites and axons to build synaptic connections through a process of neurite guidance whose mechanisms have not been fully elucidated. Calcium is an important element in this process by inducing F-actin reorganization. We hypothesized that other biologically active elements might be involved in the growth cone-mediated neurite guidance mechanisms. We performed super resolution and confocal microscopy of F-actin, followed by synchrotron X-ray fluorescence microscopy of phosphorous, sulfur, chlorine, potassium, calcium, iron and zinc on growth cones from primary rat hippocampal neurons. We identified two main patterns of element organization. First, active growth cones presenting an asymmetric distribution of Ca co-localized with the cytoskeleton protein F-actin. In active growth cones, we found that the distributions of P, S, Cl, K, and Zn are correlated with Ca. This correlation is lost in the second pattern, quiescent growth cones, exhibiting a spread elemental distribution. These results suggest that Ca is not the only element required in the F-actin rich active regions of growth cones. In addition, highly concentrated Fe spots of submicrometer size were observed in calcium-rich areas of active growth cones. These results reveal the need for biological active elements in growth cones during neural development and may help explain why early life deficiencies of elements, such as Fe or Zn, induce learning and memory deficits in children.

60 APPLIED LIFE SCIENCES↗

EmSense: A High-Resolution Emulated Sensor for Experiments with the Smart Grid and Distributed Ledger Technology

This work involves the development of a device - EmSense (“Emulated Sensor”) - that emulates a high-resolution sensor for a power grid. The device collects raw current and voltage sensor data which derive from ORNL's signature library. This library is a dataset that ORNL curates from many different sources that include power systems from various utilities. The EmSense packages the data from the library in the form of IEC 61850 Sampled Value (SV) packets and then broadcasts these SV packets on the network. In another mode, EmSense can generate artificial sinusoidal data that appears as waveforms for voltage and current signals. EmSense has an internal algorithm for determining the period of a signal based on the data so that the period can be specified as a variable in the IEC 61850 packets. The purpose of EmSense is to allow for experimentation with the Dark Net Infrastructure where a variety of power line sensors must be represented along with their typical communication traffic. The EmSense device was developed in coordination with the software for receiving and processing the packets in the Distributed Ledger Technology (DLT) framework of the DarkNet Project. This receiving software must have a methodology for dealing with information of high velocity, variety, and volume. Experimenting with EmSense facilitates the development of such software. The results showed that the DLT framework and the trust-anchoring approach managed to process a large flow of traffic even with up to six instances of EmSense device broadcasting data. This was achieved without overfilling packet queues in the memory of the actual hardware of the DLT devices or causing the Central Processing Unit (CPU) of the hardware to be overwhelmed. The DLTs were also able to store the data in a compact and useful form for later analysis and archival purposes.

Werth, Aaron↗

Monitoring of Liquid Metal Reactor Heater Zones with Recurrent Neural Network Learning of Temperature Time Series

Advanced high-temperature fluid reactors (ARs), such as sodium fast reactors (SFRs) and molten salt cooled reactors (MSCRs) utilize high-temperature fluids at ambient pressure. To melt the fluid during reactor startup and prevent fluid freezing during cooldown, the thermal–hydraulic systems of such ARs include heater zones consisting of specific heaters with controllers, temperature sensors, and thermal insulation. The failure of heater zones due to insulation material degradation or improper installation, resulting in parasitic heat losses, can lead to fluid freezing. The detection of faults using a heat-transfer model is difficult because of a lack of knowledge of the experimental details. Data-driven machine learning of heater zone temperature time series offers a viable alternative. In this study, we benchmarked the performance of recurrent neural networks (RNNs) in an analysis of heat-up transient temperature time series of heater zones installed on a liquid sodium vessel. The RNN models include long short-term memory (LSTM) and gated recurrent unit (GRU) networks, as well as their bi-directional variants, BiLSTM and BiGRU. Anomalous temperature points were designated using a percentile-based threshold applied to residual fluctuations in the detrended temperature time series. Additionally, the impact of the exponentially weighted moving average (EWMA) method on detection accuracy was examined. The RNN models’ performance was assessed using precision, recall, and F 1 score metrics. Results demonstrated that RNN models effectively detect anomalies in temperature time series with the best models for each heater zone achieving F 1 scores of over 93%. To explain the variations in RNN model performance across different heater zones, we used Kullback–Leibler (KL) divergence to quantify the relative entropy between training and testing data, and the Detrended Fluctuation Analysis (DFA) to assess long-range temporal correlations. For datasets with strong long-range correlations and minimal relative entropy between training and testing data, GRU is the best-performing model. When the data exhibits weaker long-term correlations and a significant relative entropy between training and testing distributions, BiGRU shows the best performance. For the data sets with intermediate values of both KL divergence and DFA, the best performance is obtained with LSTM and BiLSTM, respectively.

gated recurrent unit↗

Hybrid zero-voltage switching (ZVS) control for power inverters

A power inverter combination includes a half-bridge power inverter including first and second semiconductor power switches receiving input power having an intermediate node therebetween providing an inductor current through an inductor. A controller includes input comparison circuitry receiving the inductor current having outputs coupled to first inputs of pulse width modulation (PWM) generation circuitry, and a predictive control block having an output coupled to second inputs of the PWM generation circuitry. The predictive control block is coupled to receive a measure of Vin and an output voltage at a grid connection point. A memory stores a current control algorithm configured for resetting a PWM period for a switching signal applied to control nodes of the first and second power switch whenever the inductor current reaches a predetermined upper limit or a predetermined lower limit.

24 POWER TRANSMISSION AND DISTRIBUTION↗