Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Data parallelism”

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 73 records · Page 4
All sourcesNASA NTRSDOE OSTI
NASA NTRS1987

The use of transputers in processing telemetry data

Parallelism will be an essential ingredient of high performance systems of the future. The Inmos transputer is a high performance single-chip computer whose architecture facilitates the construction of parallel processing systems. Occam is a high level language developed for use with the Inmos transputer. This paper describes a project to evaluate the feasibility of using the transputer to implement real time processing of telemetry data.

Delgado, Hugo M., Jr.↗
NASA NTRS1997

Opus: A Coordination Language for Multidisciplinary Applications

Data parallel languages, such as High Performance fortran, can be successfully applied to a wide range of numerical applications. However, many advanced scientific and engineering applications are multidisciplinary and heterogeneous in nature, and thus do not fit well into the data parallel paradigm. In this paper we present Opus, a language designed to fill this gap. The central concept of Opus is a mechanism called ShareD Abstractions (SDA). An SDA can be used as a computation server, i.e., a locus of computational activity, or as a data repository for sharing data between asynchronous tasks. SDAs can be internally data parallel, providing support for the integration of data and task parallelism as well as nested task parallelism. They can thus be used to express multidisciplinary applications in a natural and efficient way. In this paper we describe the features of the language through a series of examples and give an overview of the runtime support required to implement these concepts in parallel and distributed environments.

Chapman, Barbara↗
DOE OSTI2026

User Manual - HydraGNN v5.0: Distributed Implementation of Multi-Tasking Graph Neural Networks

This document serves as the user manual for HydraGNN v5.0, a scalable graph neural network (GNN) architecture for simultaneous prediction of multiple target properties using multi-task learning (MTL). This version of HydraGNN has been developed primarily to support the development, training, and deployment of predictive graph-based deep learning (DL) models for atomistic materials modeling. HydraGNN is templated over 13 message-passing policies, including invariant models (GIN, PNA, PNAPlus, GAT, MFC, CGCNN, SAGE, SchNet, DimeNet) and equivariant models (EGNN, PNAEq, PAINN, MACE), and supports distributed training via distributed data parallelism (DDP), DeepSpeed, and Fully Sharded Data Parallelism (FSDP) on leadership-class supercomputers. Although HydraGNN can be applied to problems beyond atomistic materials modeling, its current use is confined to homogeneous graphs. Additional capabilities include machine-learned interatomic potentials with energy-conserving forces, General, Powerful, and Scalable Graph Transformer (GraphGPS) global attention, periodic boundary conditions, hyperparameter optimization, mixed-precision training, and uncertainty quantification.

97 MATHEMATICS AND COMPUTING↗
NASA NTRS1995

Implementation and Characterization of Three-Dimensional Particle-in-Cell Codes on Multiple-Instruction-Multiple-Data Massively Parallel Supercomputers

A three-dimensional electrostatic particle-in-cell (PIC) plasma simulation code has been developed on coarse-grain distributed-memory massively parallel computers with message passing communications. Our implementation is the generalization to three-dimensions of the general concurrent particle-in-cell (GCPIC) algorithm. In the GCPIC algorithm, the particle computation is divided among the processors using a domain decomposition of the simulation domain. In a three-dimensional simulation, the domain can be partitioned into one-, two-, or three-dimensional subdomains ("slabs," "rods," or "cubes") and we investigate the efficiency of the parallel implementation of the push for all three choices. The present implementation runs on the Intel Touchstone Delta machine at Caltech; a multiple-instruction-multiple-data (MIMD) parallel computer with 512 nodes. We find that the parallel efficiency of the push is very high, with the ratio of communication to computation time in the range 0.3%-10.0%. The highest efficiency (> 99%) occurs for a large, scaled problem with 64(sup 3) particles per processing node (approximately 134 million particles of 512 nodes) which has a push time of about 250 ns per particle per time step. We have also developed expressions for the timing of the code which are a function of both code parameters (number of grid points, particles, etc.) and machine-dependent parameters (effective FLOP rate, and the effective interprocessor bandwidths for the communication of particles and grid points). These expressions can be used to estimate the performance of scaled problems--including those with inhomogeneous plasmas--to other parallel machines once the machine-dependent parameters are known.

Lyster, P. M.↗
NASA NTRS1994

High Performance Programming Using Explicit Shared Memory Model on Cray T3D1

The Cray T3D system is the first-phase system in Cray Research, Inc.'s (CRI) three-phase massively parallel processing (MPP) program. This system features a heterogeneous architecture that closely couples DEC's Alpha microprocessors and CRI's parallel-vector technology, i.e., the Cray Y-MP and Cray C90. An overview of the Cray T3D hardware and available programming models is presented. Under Cray Research adaptive Fortran (CRAFT) model four programming methods (data parallel, work sharing, message-passing using PVM, and explicit shared memory model) are available to the users. However, at this time data parallel and work sharing programming models are not available to the user community. The differences between standard PVM and CRI's PVM are highlighted with performance measurements such as latencies and communication bandwidths. We have found that the performance of neither standard PVM nor CRI s PVM exploits the hardware capabilities of the T3D. The reasons for the bad performance of PVM as a native message-passing library are presented. This is illustrated by the performance of NAS Parallel Benchmarks (NPB) programmed in explicit shared memory model on Cray T3D. In general, the performance of standard PVM is about 4 to 5 times less than obtained by using explicit shared memory model. This degradation in performance is also seen on CM-5 where the performance of applications using native message-passing library CMMD on CM-5 is also about 4 to 5 times less than using data parallel methods. The issues involved (such as barriers, synchronization, invalidating data cache, aligning data cache etc.) while programming in explicit shared memory model are discussed. Comparative performance of NPB using explicit shared memory programming model on the Cray T3D and other highly parallel systems such as the TMC CM-5, Intel Paragon, Cray C90, IBM-SP1, etc. is presented.

Simon, Horst D.↗
NASA NTRS2012

Physics Mining of Multi-Source Data Sets

Powerful new parallel data mining algorithms can produce diagnostic and prognostic numerical models and analyses from observational data. These techniques yield higher-resolution measures than ever before of environmental parameters by fusing synoptic imagery and time-series measurements. These techniques are general and relevant to observational data, including raster, vector, and scalar, and can be applied in all Earth- and environmental science domains. Because they can be highly automated and are parallel, they scale to large spatial domains and are well suited to change and gap detection. This makes it possible to analyze spatial and temporal gaps in information, and facilitates within-mission replanning to optimize the allocation of observational resources. The basis of the innovation is the extension of a recently developed set of algorithms packaged into MineTool to multi-variate time-series data. MineTool is unique in that it automates the various steps of the data mining process, thus making it amenable to autonomous analysis of large data sets. Unlike techniques such as Artificial Neural Nets, which yield a blackbox solution, MineTool's outcome is always an analytical model in parametric form that expresses the output in terms of the input variables. This has the advantage that the derived equation can then be used to gain insight into the physical relevance and relative importance of the parameters and coefficients in the model. This is referred to as physics-mining of data. The capabilities of MineTool are extended to include both supervised and unsupervised algorithms, handle multi-type data sets, and parallelize it.

Helly, John↗
DOE OSTI2024

Pretraining Billion-Scale Geospatial Foundational Models on Frontier

As AI workloads increase in scope, generalization capability becomes challenging for small task-specific models and their demand for large amounts of labeled training samples increases. On the contrary, Foundation Models (FMs) are trained with internet-scale unlabeled data via self-supervised learning and have been shown to adapt to various tasks with minimal fine-tuning. Although large FMs have demonstrated significant impact in natural language processing and computer vision, efforts toward FMs for geospatial applications have been restricted to smaller size models, as pretraining larger models requires very large computing resources equipped with state-of-the-art hardware accelerators. Current satellite constellations collect 100+TBs of data a day, resulting in images that are billions of pixels and multimodal in nature. Such geospatial data poses unique challenges opening up new opportunities to develop FMs. We investigate billion scale FMs and HPC training profiles for geospatial applications by pretraining on publicly available data. We studied from end-to-end the performance and impact in the solution by scaling the model size. Our larger 3B parameter size model achieves up to 30% improvement in top1 scene classification accuracy when comparing a 100M parameter model. Moreover, we detail performance experiments on the Frontier supercomputer, America's first exascale system, where we study different model and data parallel approaches using PyTorch's Fully Sharded Data Parallel library. Specifically, we study variants of the Vision Transformer architecture (ViT), conducting performance analysis for ViT models with size up to 15B parameters. By discussing throughput and performance bottlenecks under different parallelism configurations, we offer insights on how to leverage such leadership-class HPC resources when developing large models for geospatial imagery applications.

Tsaris, Aristeidis (aris)↗
DOE OSTI2021

Asynchronous and Load-Balanced Union-Find for Distributed and Parallel Scientific Data Visualization and Analysis

We present a novel distributed union-find algorithm that features asynchronous parallelism and k-d tree based load balancing for scalable visualization and analysis of scientific data. Applications of union-find include level set extraction and critical point tracking, but distributed union-find can suffer from high synchronization costs and imbalanced workloads across parallel processes. In this study, we prove that global synchronizations in existing distributed union-find can be eliminated without changing final results, allowing overlapped communications and computations for scalable processing. We also use a k-d tree decomposition to redistribute inputs, in order to improve workload balancing. We benchmark the scalability of our algorithm with up to 1,024 processes using both synthetic and application data. Here, we demonstrate the use of our algorithm in critical point tracking and super-level set extraction with high-speed imaging experiments and fusion plasma simulations, respectively.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2024

Quantum-parallel vectorized data encodings and computations on trapped-ion and transmon QPUs

Compact data representations in quantum systems are crucial for the development of quantum algorithms for data analysis. In this study, we present two innovative data encoding techniques, known as QCrank and QBArt, which exhibit significant quantum parallelism via uniformly controlled rotation gates. The QCrank method encodes a series of real-valued data as rotations on data qubits, resulting in increased storage capacity. On the other hand, QBArt directly incorporates a binary representation of the data within the computational basis, requiring fewer quantum measurements and enabling well-established arithmetic operations on binary data. We showcase various applications of the proposed encoding methods for various data types. Notably, we demonstrate quantum algorithms for tasks such as DNA pattern matching, Hamming weight computation, complex value conjugation, and the retrieval of a binary image with 384 pixels, all executed on the Quantinuum trapped-ion QPU. Furthermore, we employ several cloud-accessible QPUs, including those from IBMQ and IonQ, to conduct supplementary benchmarking experiments.

97 MATHEMATICS AND COMPUTING↗
NASA NTRS1993

Efficient partitioning and assignment on programs for multiprocessor execution

The general problem studied is that of segmenting or partitioning programs for distribution across a multiprocessor system. Efficient partitioning and the assignment of program elements are of great importance since the time consumed in this overhead activity may easily dominate the computation, effectively eliminating any gains made by the use of the parallelism. In this study, the partitioning of sequentially structured programs (written in FORTRAN) is evaluated. Heuristics, developed for similar applications are examined. Finally, a model for queueing networks with finite queues is developed which may be used to analyze multiprocessor system architectures with a shared memory approach to the problem of partitioning. The properties of sequentially written programs form obstacles to large scale (at the procedure or subroutine level) parallelization. Data dependencies of even the minutest nature, reflecting the sequential development of the program, severely limit parallelism. The design of heuristic algorithms is tied to the experience gained in the parallel splitting. Parallelism obtained through the physical separation of data has seen some success, especially at the data element level. Data parallelism on a grander scale requires models that accurately reflect the effects of blocking caused by finite queues. A model for the approximation of the performance of finite queueing networks is developed. This model makes use of the decomposition approach combined with the efficiency of product form solutions.

Standley, Hilda M.↗
NASA NTRS1978

Preliminary design for a standard 10 sup 7 bit Solid State Memory (SSM)

A modular concept with three separate modules roughly separating bubble domain technology, control logic technology, and power supply technology was employed. These modules were respectively the standard memory module (SMM), the data control unit (DCU), and power supply module (PSM). The storage medium was provided by bubble domain chips organized into memory cells. These cells and the circuitry for parallel data access to the cells make up the SMM. The DCU provides a flexible serial data interface to the SMM. The PSM provides adequate power to enable one DCU and one SMM to operate simultaneously at the maximum data rate. The SSM was designed to handle asynchronous data rates from dc to 1.024 Mbs with a bit error rate less than 1 error in 10 to the eight power bits. Two versions of the SSM, a serial data memory and a dual parallel data memory were specified using the standard modules. The SSM specification includes requirements for radiation hardness, temperature and mechanical environments, dc magnetic field emission and susceptibility, electromagnetic compatibility, and reliability.

Hayes, P. J.↗
DOE OSTI2021

Prediction of DIII-D Pedestal Structure From Externally Controllable Parameters

The sharp increase of pressure at the edge of a high confinement mode (H-mode) plasma, the pedestal, strongly impacts overall plasma performance. Predicting the pedestal is a necessity to control and optimize tokamak operations. Here, an experimental data-driven machine learning (ML) approach is presented that predicts the pedestal heights and widths of electron density (n e ) and electron temperature (T e ) profiles as well as the separatrix ne from externally controllable parameters such as the plasma shape, heating method and power, and gas puff rate and integrated gas puff. The OMFIT framework was used with DIII-D data to efficiently, robustly, and automatically build a database of pedestal parameters to train machine learning models. Database creation was enabled by the search engine tool for DIII-D data, TokSearch, which parallelizes data fetching, enabling fast searches through basic signals of thousands of DIII-D shots and selection of relevant time intervals. Principal Component Analysis (PCA) separated the database into three clusters that represent classes of plasma shapes that are regularly used in DIII-D. The most important parameters for setting the pedestal structure were plasma current (I p ), toroidal magnetic field (B Φ ), neutral beam heating power (P NBI ) and shaping quantities. The Deep Jointly Informed Neural Networks (DJINN) algorithm was applied to identify suitable neural network (NN) architectures that appropriately capture the features of the pedestal database. Separate NNs were implemented for each pedestal parameter, and ensembling methods were used to improve the prediction accuracy and allowed estimation of the prediction uncertainty. The pedestal predictions of the test dataset lie within the measurement uncertainties of the pedestal parameters. The NN outperformed simple Linear Regression (LR) analysis, indicating non-linear dependencies in the pedestal structure. The presented achievements illustrate a promising path for future research, using feature extraction to infer experimental trends and thereby improve pedestal models as well as deploying NN for a fast pedestal prediction in DIII-D scenario development.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
DOE OSTI2022

Prediction of DIII-D Pedestal Structure from Externally Controllable Parameters

The sharp increase of pressure at the edge of a high confinement mode (H-mode) plasma, the pedestal, strongly impacts overall plasma performance. Predicting the pedestal is a necessity to control and optimize tokamak operations. An experimental data-driven machine learning (ML) approach is presented that predicts the pedestal heights and widths of electron density (ne) and electron temperature (Te) profiles as well as the separatrix ne from externally controllable parameters such as the plasma shape, heating method and power, and gas puff rate and integrated gas puff. The OMFIT framework was used with DIII-D data to efficiently, robustly, and automatically build a database of pedestal parameters to train machine learning models. Database creation was enabled by the search engine tool for DIII-D data, TokSearch, which parallelizes data fetching, enabling fast searches through basic signals of thousands of DIII-D shots and selection of relevant time intervals. Principal Component Analysis (PCA) separated the database into three clusters that represent classes of plasma shapes that are regularly used in DIII-D. The most important parameters for setting the pedestal structure were plasma current (Ip), toroidal magnetic field (Bφ), neutral beam heating power (PNBI) and shaping quantities. The Deep Jointly Informed Neural Networks (DJINN) algorithm was applied to identify suitable neural network (NN) architectures that appropriately capture the features of the pedestal database. Separate NNs were implemented for each pedestal parameter, and ensembling methods were used to improve the prediction accuracy and allowed estimation of the prediction uncertainty. The pedestal predictions of the test dataset lie within the measurement uncertainties of the pedestal parameters. The NN outperformed simple Linear Regression (LR) analysis, indicating non-linear dependencies in the pedestal structure. The presented achievements illustrate a promising path for future research, using feature extraction to infer experimental trends and thereby improve pedestal models as well as deploying NN for a fast pedestal prediction in DIII-D scenario development.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
DOE OSTI2020

Evaluation of Graph Analytics Frameworks Using the GAP Benchmark Suite

The analysis of connected data is an increasingly important application in high-performance computing. Such analyses can reveal fraudulent patterns in financial transactions, optimize telecommunications networks, predict information flow in social networks, etc. However, the landscape of graph analytics is highly diverse. Graph algorithms stress processor architectures differently, and no one graph can represent all topologies. Consequently, no single approach or framework is expected to be optimal for all graph analytics problems. To help make sense of this diverse landscape, we evaluated four approaches to graph analytics: GraphBLAS, Galois, BGL17, GraphIt; and compare them against hand-tuned implementations that take advantage of hardware features on our test platform. Graph- BLAS formulates graph analytics as sparse linear algebra. Galois provides syntactic constructs for data parallelism over irregular data structures. BGL17 is a generic C++ template library for implementing graph algorithms. GraphIt provides a domain- specific language to describe and optimize graph algorithms. We use the GAP Benchmark Suite to establish baseline performance and guide the side-by-side evaluation of each framework. GAP consists of 30 tests: six graph analytics algorithms (breadth- first search, single-source shortest path, PageRank, betweenness centrality, connected components, and triangle counting) run on five graphs, each with different topological characteristics (e.g., high diameter, skewed degree distribution, high average degree). High-performance reference implementations are included for each benchmark algorithm. Because a graph can be loaded into memory a number of ways (e.g., flat file on disk, compressed sparse format, data frames, retrieved from SQL or NoSQL databases), our evaluation focused on computational performance rather than I/O. Our results show the relative strengths of each framework.

Graph algorithms, Benchmarking, shared-memory prog↗
NASA NTRS1983

Segmentation of remotely sensed data using parallel region growing

The improved spatial resolution of the new earth resources satellites will increase the need for effective utilization of spatial information in machine processing of remotely sensed data. One promising technique is scene segmentation by region growing. Region growing can use spatial information in two ways: only spatially adjacent regions merge together, and merging criteria can be based on region-wide spatial features. A simple region growing approach is described in which the similarity criterion is based on region mean and variance (a simple spatial feature). An effective way to implement region growing for remote sensing is as an iterative parallel process on a large parallel processor. A straightforward parallel pixel-based implementation of the algorithm is explored and its efficiency is compared with sequential pixel-based, sequential region-based, and parallel region-based implementations. Experimental results from on aircraft scanner data set are presented, as is a discussioon of proposed improvements to the segmentation algorithm.

Tilton, J. C.↗
NASA NTRS2021

Revisiting Topographic Horizons in the Era of Big Data and Parallel Computing

Widely used to calculate illumination geometry forestimates of solar and emitted longwave radiation, and forcorrecting remotely sensed data for topographic effects, digitalelevation models (DEMs) are now extensive globally at 10–30-mspatial resolution and locally at spatial resolutions down to afew centimeters. Globally, regionally, or locally, elevation datasetshave many grid points. Many software packages calculate gradi-ents over every grid cell or point, but in the mountains, shadingby nearby terrain must also be assessed. Terrain may obscure aslope that would otherwise face the Sun. Four decadesago, a fastmethod to calculate topographic horizons at every point in anelevation grid required computations related only linearly to thesize of the grid, but grids now have so many points that parallelcomputing still provides an advantage. Exploiting parallelismover terrain grids can use alternative strategies: among columnsof a rotated grid, or simultaneously at multiple rotation angles,or on different tiles of a grid. On a multi-processor machine, theimprovement in computing time approaches 2/3 the number ofprocessors deployed,

Jeff Dozier↗
NASA NTRS2002

Wireless Temperature-Monitoring System

A relatively inexpensive instrumentation system that includes units that are connected to thermocouples and that are parts of a radio-communication network has been developed to enable monitoring of temperatures at multiple locations. Because there is no need to string wires or cables for communication, the system is well suited for monitoring temperatures at remote locations and for applications in which frequent changes of monitored or monitoring locations are needed. The system can also be adapted to monitoring of slowly varying physical quantities, other than temperature, that can be transduced by solid-state electronic sensors. electronic sensors. The system comprises any number of transmitting units and a single receiving unit. Each transmitting unit includes connections for as many as four external thermocouples, a signal-conditioning module, a control module, and a radio-communication module. The signal-conditioning module acts as an interface between the thermocouples and the rest of the transmitting unit and includes a built-in solid ambient temperature sensor that is in addition to the external thermocouples. The control module is a system-on-chip embedded processor that includes analog-to-digital converters, serial and parallel data ports, and an interface for local connection to an analog meter that is used during installation to verify correct operation. The radio-communication module contains a commercial spread-spectrum transceiver that operates in the 900-MHz industrial, scientific, and medical (ISM) frequency band. This transceiver transmits data to the receiving unit at a rate of 19,200 baud. The receiving unit includes a transceiver like that of a transmitting unit, plus a control module that contains a system-on-chip processor that includes serial data port for output to a computer that runs monitoring and/or control software, a parallel data port for output to a printer, and a seven-segment light-emitting-diode display. Each transmitting unit is battery-powered and can operate for at least seven days continuously while reporting temperatures every half hour. The receiving unit is powered by a wall-mounted transformer source. The receiving unit responds to each transmitting unit and reports the readings of each of the four thermocouples and of the ambient-temperature sensor of the transmitting unit. The end-to-end accuracy of the system is plus or minus 0.2 C over the temperature range from 0 to 100 C. The radio-communication range between the receiving and transmitting units is approximately equal to 0.5 mile (approximately equal to 0.8 km).

Solano, Wanda↗
NASA NTRS2003

Wireless Temperature-Monitoring System

A relatively inexpensive instrumentation system that includes units that are connected to thermocouples and that are parts of a radio-communication network has been developed to enable monitoring of temperatures at multiple locations. Because there is no need to string wires or cables for communication, the system is well suited for monitoring temperatures at remote locations and for applications in which frequent changes of monitored or monitoring locations are needed. The system can also be adapted to monitoring of slowly varying physical quantities, other than temperature, that can be transduced by solid-state electronic sensors. The system comprises any number of transmitting units and a single receiving unit (see figure). Each transmitting unit includes connections for as many as four external thermocouples, a signal-conditioning module, a control module, and a radio-communication module. The signal- conditioning module acts as an interface between the thermocouples and the rest of the transmitting unit and includes a built-in solid ambient-temperature sensor that is in addition to the external thermocouples. The control module is a "system-on-chip" embedded processor that includes analog-to-digital converters, serial and parallel data ports, and an interface for local connection to an analog meter that is used during installation to verify correct operation. The radio-communication module contains a commercial spread-spectrum transceiver that operates in the 900-MHz industrial, scientific, and medical (ISM) frequency band. This transceiver transmits data to the receiving unit at a rate of 19,200 baud. The receiving unit includes a transceiver like that of a transmitting unit, plus a control module that contains a system-on-chip processor that includes serial data port for output to a computer that runs monitoring and/or control software, a parallel data port for output to a printer, and a seven-segment light-emitting-diode display.

Solano, Wanda↗