Search NASA⌕ Search

SEARCH · Search NASA

Results for “distributed parallelization”

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 1,711 records · Page 95

Isochronous Architecture-Based Voltage-Active Power Droop for Multi-Inverter Systems

This article proposes an isochronous architecture for parallel inverters with only voltage-active power droop (VP-D) control for improving active power sharing as well as plug-and-play of multi-inverter-based distributed energy resources (DERs). The isochronous framework obviates the need for an explicit regulation of the frequency while allowing for sharing of reactive power. The article shows that the detrimental effects of circulating currents between inverters can be addressed in the framework developed. The isochronous architecture is implemented by employing a GPS to disseminate clock timing signals that enable the microgrid to maintain the nominal system frequency. Small signal eigenvalue analysis of a Multi-inverter Microgrid system near the steady-state operating point is presented to evaluate the system stability. Moreover, unlike traditional droop-based methods, the isochronous strategy lends itself to analytical guarantees which are developed in the article. Furthermore, the effect of delays in receiving GPS clock signals by an inverter is experimentally characterized and an inverter isolation mechanism is proposed that disconnects inverters facing large GPS communication delays. Hardware experiments on an 1.2 kVA-prototype and controller-hardware-in-the-loop validation on the CIGRE distribution network are conducted to demonstrate the effectiveness of the proposed architecture towards active and reactive power sharing between inverters with load scenarios.

24 POWER TRANSMISSION AND DISTRIBUTION↗

Multistep inversion workflow for 3D long-offset damped elastic waves in the Fourier domain

We present a new workflow for imaging damped three-dimensional elastic wavefields in the Fourier domain. The workflow employs a multiscale imaging approach, in which offset lengths are laddered, where frequency content and damping of the data are changed cyclically. Thus, the inversion process is launched using short-offset and low-frequency data to recover the long spatial wavelength of the image at a shallow depth. Increasing frequency and offset length leads to the recovery of the fine-scale features of the model at greater depths. For the fixed offset, we employ (in the imaging process) a few discrete frequencies with a set of Laplace damping parameters. The forward problem is solved with a finite-difference frequency-domain method based on a massively parallel iterative solver. The inversion code is based upon the solution of a least squares optimisation problem and is solved using a nonlinear gradient method. It is fully parallelised for distributed memory computational platforms. Our full-waveform inversion workflow is applied to the 3D Marmousi-2 and SEG/EAGE Salt models with long-offset data. The maximum inverted frequencies are 6 Hz for the Marmousi model and 2 Hz for the SEG/EAGE Salt model. The detailed structures are imaged successfully up to the depth approximately equal to one-third of the maximum offset length at a resolution consistent with the inverted frequencies.

3D full waveform inversion↗

Design and evaluation of a HDR skin applicator with flattening filter

The purposes of this study are: (i) to design field flattening filters for the Leipzig applicators of 2 and 3 cm of inner diameter with the source traveling parallel to the applicator contact surface, which are accessories of the microSelectron-HDR afterloader (Nucletron, Veenendaal, The Netherlands). These filters, made of tungsten, aim to flatten the heterogeneous dose distribution obtained with the Leipzig applicators. (ii) To estimate the dose rate distributions for these Leipzig+filter applicators by means of the Monte Carlo (MC) method. (iii) To experimentally verify these distributions for prototypes of these new applicators, and (iv) to obtain the correspondence factors to measure the output of the applicators by the user using an insert into a well chamber. The MC GEANT4 code has been used to design the filters and to obtain the dose rate distributions in liquid water for the two Leipzig+filter applicators. In order to validate this specific application and to guarantee that realistic source-applicator geometry has been considered, an experimental verification procedure was implemented in this study, in accordance with the updated recommendations of the American Association of Physicists in Medicine Task Group No. 43 U1 Report. Thermoluminescent dosimeters, radiochromic film, and a pin-point ionization chamber in a plastic [polymethylmethacrylate (PMMA)] phantom were used to verify the MC results for the two applicators of a microSelectron-HDR afterloader with the mHDR-v2 source. To verify the output of the Leipzig+filter applicators, correspondence factors were deduced for the well chambers HDR100-plus (Standard Imaging, Inc., Middleton, WI) and TM33004 (PTW, Freiburg, Germany) using a specific insert for both applicators. The doses measured in the PMMA phantom agree within experimental uncertainties with the dose obtained by the MC calculations. Percentage depth dose and off-axis profiles were obtained normalized at a depth of 3 mm along the central applicator axis in a cylindrical 20x20 cm water phantom. A table of output factors, normalized to 1 U of source air kerma strength at this depth, is presented. Correspondence factors were obtained for the two well chambers considered. The matrix data obtained in the MC simulation with a grid separation of 0.5 mm has been used to build a data set in a convenient format to model these distributions for routine use with a brachytherapy treatment planning system.

62 RADIOLOGY AND NUCLEAR MEDICINE↗

Distributed Cross-Channel Hierarchical Aggregation for Foundation Models

Vision-based scientific foundation models hold significant promise for advancing scientific discovery and innovation. This potential stems from their ability to aggregate images from diverse sources—such as varying physical groundings or data acquisition systems—and to learn spatio-temporal correlations using transformer architectures. However, tokenizing and aggregating images can be compute-intensive, a challenge not fully addressed by current distributed methods. In this work, we introduce the Distributed Cross-Channel Hierarchical Aggregation (D-CHAG) approach designed for datasets with a large number of channels across image modalities. Our method is compatible with any model-parallel strategy and any type of vision transformer architecture, significantly improving computational efficiency. We evaluated D-CHAG on hyperspectral imaging and weather forecasting tasks. When integrated with tensor parallelism and model sharding, our approach achieved up to a 75% reduction in memory usage and more than doubled sustained throughput on up to 1,024 AMD GPUs on the Frontier Supercomputer.

Tsaris, Aristeidis (aris) [ORNL] (ORCID:0000000277↗

DistGANS- Distributed Generative Adversarial Neural Networks

DistGANs is a Python package to perform distributed training of conditional generative adversarial neural networks for multi-class labeled image data. DistGANs partitionins the training data according to data labels, and enhances scalability by performing a parallel training where multiple generators are concurrently trained, each one of them focusing on a single data label.

Lupo Pasini, Massimiliano [Oak Ridge National Lab.↗

A Path to Operating System and Runtime Support for Extreme Scale Tools

In this project, we cast distributed resource access as operations on files in a global name space and developed a common, scalable solution for group operations on distributed processes and files. The resulting solution enables tool and middleware developers to quickly create new scalable software or easily improve the scalability of existing software. The cornerstone of the project was the design of a new programming idiom called group file operations that eliminates iterative behavior when a single process must apply the same set of file operations to a group of related files. To demonstrate our novel and scalable ideas for group file operations and global name space composition, we developed a group file system called TBON-FS that leverages a tree-based overlay network (TBON), specifically MRNet, for logarithmic communication and distributed data aggregation. We also developed proc++, a new synthetic file system co-designed for use in scalable group file operations. Over the course of the project, we evaluated the utility and performance of group file operations, global name space composition, TBON-FS, and proc++ in three case studies. The first study focused on the ease in using group file operations and TBON-FS to quickly develop several new scalable tools for distributed system administration and monitoring. The second study evaluated the integration of group file operation and TBON-FS within the Ganglia Distributed Monitoring System to improve its scalability for clusters. The final study involved the integration of group file operations, TBON-FS, and proc++ within TotalView, the widely-used parallel debugger. For this project, the work of the Oak Ridge National Laboratory (ORNL) team occurred primarily in two directions: bringing the MRNet tree-based overlay network (TBON) implementation to the Cray XT platform, and investigating techniques for predicting the performance of MRNet topologies on such systems. Rogue Wave Software (RWS), formerly TotalView Technologies Inc., worked with the University ofWisconsin (UW) team on the design and prototyping of a scalable version of the TotalView debugger. RWS assisted UW with their "proc++" design effort. RWS assisted UW with strategy for integrating proc++ into TotalView.

97 MATHEMATICS AND COMPUTING↗

Transport Studies and Modeling in PEM Fuel Cells

This project’s aim was to develop fuel cell components (i.e. membranes, gas-diffusion media (GDM), bipolar plates and flow fields) that possess specific properties (i.e. water transport and conductivity). A computational fluid dynamics model was developed to elucidate the effect of certain parameters on these specific properties. Ultimately, the model will be used to determine sensitivity of fuel cell performance to component properties to determine limiting components and to guide research. We have successfully reached our objectives and achieved most of the milestones of this project. We have designed and synthesized a variety of hydrocarbon block polymer membranes with lower equivalent weight, structure, chemistry, phase separation and process conditions. These membranes provide a broad selection with optimized water transport properties. We have also designed and constructed a variety of devices that are capable of accurately measuring the water transport properties (water uptake, water diffusivity and electro-osmatic drag) of these membranes. These transport properties are correlated to the membranes’ structures derived from X-ray and microscopy techniques to determine the structure-property relationship. We successfully integrated hydrocarbon membrane MEAs with a current distribution board (CBD) to study the impact of hydrocarbon membrane on water transport in fuel cells. We have designed and fabricated various GDM with varying substrate, diffusivity and micro-porous layers (MPL) and characterized their pore structure, tortuosity and hydrophobicity. We have derived a universal chart (MacMullin number as function of wet proofing and porosity) that can be used to characterize various GDM. The abovementioned GDMs have been evaluated in operating fuel cells; their performance is correlated to various pore structure, tortuosity and hydrophobicity of the GDM. Unfortunately, determining a universal relationship between the MacMullin number and these properties was not achieved. We have simulated fuel cell performance, current distribution and water distribution at various values of the water uptake, membrane diffusivity, and electro-osmotic drag coefficient (EODC) and compared modeling results with segmented-cell data for both serpentine and parallel flow-fields. We have developed iterations of fuel cell flow fields to achieve specific water transport and thermal management targets. This work demonstrated the importance of membrane diffusivity on fuel cell performance, the necessity of a high membrane diffusion coefficient, and the desirability of a low EODC at low levels of relative humidity.

08 HYDROGEN↗

EdgeAI: Machine learning via direct attached accelerator for streaming data processing at high shot rate x-ray free-electron lasers

We present a case for low batch-size inference with the potential for adaptive training of a lean encoder model. We do so in the context of a paradigmatic example of machine learning as applied in data acquisition at high data velocity scientific user facilities such as the Linac Coherent Light Source-II x-ray Free-Electron Laser. We discuss how a low-latency inference model operating at the data acquisition edge can capitalize on the naturally stochastic nature of such sources. We simulate the method of attosecond angular streaking to produce representative results whereby simulated input data reproduce high-resolution ground truth probability distributions. By minimizing the mean-squared error between the decoded output of the latent representation and the ground truth distributions, we ensure that the encoding layers and resulting latent representation maintains full fidelity for any downstream task, be it classification or regression. We present throughput results for data-parallel inference of various batch sizes, some with throughput exceeding 100 k images per second. We also show in situ training below 10 s per epoch for the full encoder–decoder model as would be relevant for streaming and adaptive real-time data production at our nation’s scientific light sources.

97 MATHEMATICS AND COMPUTING↗

Assembly and Tracking of Microbial Community Development within a Microwell Array Platform

The development of microbial communities depends on a combination of complex deterministic and stochastic factors that can dramatically alter the spatial distribution and activities of community members. In this study, we have developed a microwell array platform that can be used to rapidly assemble and track thousands of bacterial communities in parallel. This protocol highlights the utility of the platform and describes its use for optically monitoring the development of simple, two-member communities within an ensemble of arrays within the platform. This demonstration uses two mutants of Pseudomonas aeruginosa, part of a series of mutants developed to study Type VI secretion pathogenicity. Chromosomal inserts of either mCherry or GFP genes facilitate the constitutive expression of fluorescent proteins with distinct emission wavelengths that can be used to monitor community member abundance and location within each microwell. This protocol describes a detailed method for assembling mixtures of bacteria into the wells of the array and using time-lapse fluorescence imaging and quantitative image analysis to measure the relative growth of each member population over time. Lastly, the seeding and assembly of the microwell platform, the imaging procedures necessary for the quantitative analysis of microbial communities within the array, and the methods that can be used to reveal interactions between microbial species area all discussed.

59 BASIC BIOLOGICAL SCIENCES↗

Acquisition, Sharing and Processing of Large Datasets for Strain Imaging: an Example of an Indented Ni3Al/Mo Composite

The local effects of stress from a mechanical indentation have been studied on a Ni3 Al single crystal containing submicron inclusions of molybdenum fibers. Polychromatic X ray Microscopy (PXM) was used to measure elastic and plastic deformations near the indents. Analysis of freshly acquired massive sets of PXM data has been carried out over the Science Studio network using parallel processing software FOXMAS. This network and the FOXMAS software have greatly improved the efficiency of the data processing task.. The analysis was successfully applied to study lattice orientation distribution and strain tensor components for both the Ni3 Al and the Mo phases, particularly around 8 indents patterned at the longitudinal section of the alloy.

data processing↗

Enabling Recycling of Composites: Understanding the Impacts of Multiple Thermal Processing Cycles

When considering the utilization of recycled short carbon fiber feedstock materials for advanced manufacturing, understanding the material degradation behavior is essential in determining how many times a composite material can be effectively reprocessed and remanufactured. This study characterizes the degradation behavior of short carbon fiber acrylonitrile butadiene (CF-ABS) that has been reprocessed five times with twin screw extrusion. Parallel plate rheology was completed to observe the degradation in complex viscosity of the recycled feedstock materials. Gel permeation chromatography (GPC) was utilized to characterize the changes in molecular weight distribution of the recycled materials as a result of thermal and mechanical degradation during the re-processing steps. Rheological characterization, GPC, and twin-screw processing data help inform the process optimizations required to process the recycled feedstock material. Successful characterization of the degradation behavior of short fiber composite feedstock materials aids in increased understanding of the lifespan of high value carbon fiber composite materials and aids in process optimization of recycled composite materials.

Walker, Roo↗

Accelerating population balance-Monte Carlo simulation for coagulation dynamics from the Markov jump model, stochastic algorithm and GPU parallel computing

This paper proposes a comprehensive framework for accelerating population balance-Monte Carlo (PBMC) simulation of particle coagulation dynamics. By combining Markov jump model, weighted majorant kernel and GPU (graphics processing unit) parallel computing, a significant gain in computational efficiency is achieved. The Markov jump model constructs a coagulation-rule matrix of differentially-weighted simulation particles, so as to capture the time evolution of particle size distribution with low statistical noise over the full size range and as far as possible to reduce the number of time loopings. Here three coagulation rules are highlighted and it is found that constructing appropriate coagulation rule provides a route to attain the compromise between accuracy and cost of PBMC methods. Further, in order to avoid double looping over all simulation particles when considering the two-particle events (typically, particle coagulation), the weighted majorant kernel is introduced to estimate the maximum coagulation rates being used for acceptance–rejection processes by single-looping over all particles, and meanwhile the mean time-step of coagulation event is estimated by summing the coagulation kernels of rejected and accepted particle pairs. The computational load of these fast differentially-weighted PBMC simulations (based on the Markov jump model) is reduced greatly to be proportional to the number of simulation particles in a zero-dimensional system (single cell). Finally, for a spatially inhomogeneous multi-dimensional (multi-cell) simulation, the proposed fast PBMC is performed in each cell, and multiple cells are parallel processed by multi-cores on a GPU that can implement the massively threaded data-parallel tasks to obtain remarkable speedup ratio (comparing with CPU computation, the speedup ratio of GPU parallel computing is as high as 200 in a case of 100 cells with 10 000 simulation particles per cell). These accelerating approaches of PBMC are demonstrated in a physically realistic Brownian coagulation case. The computational accuracy is validated with benchmark solution of discrete-sectional method. The simulation results show that the comprehensive approach can attain very favorable improvement in cost without sacrificing computational accuracy.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Nonsymmetric molecules driven by intense few-cycle laser pulses: Phase and orientation dependence of enhanced ionization

The ionization of nonsymmetric heteronuclear diatomic molecules by intense few-cycle laser pulses linearly polarized along the internuclear axis has been investigated. It is found that enhanced ionization (EI) occurs in nonsymmetric molecules and is accompanied by enhanced excitation (EE). We show that the nonsymmetric distribution of the electron cloud between the two nuclei leads to a strong dependence of EI and EE on the carrier envelope phase of few-cycle pulses, and on the orientation of the molecule parallel or antiparallel with the peak electric field of the pulse. This effect is as strong as the pulse duration is short and disappears with increasing pulse duration. The field ionization model, and mainly the ''energy level crossing'' mechanism, are used to explain these phase effects. The newly proposed energy level crossing mechanism, which is relevant to nonsymmetric molecules, occurs as the driving field moves the dressed ground and excited states closer to each other until their energy levels cross, leading to an enhancement of excitation and ionization. A semiclassical nonadiabatic model derived to interpret the level crossing mechanism also predicts the critical internuclear distance R{sub c} at which EI, EE, and energy crossings occur as a function of charge asymmetry and laser intensity, in good agreement with quantum-mechanical simulations.

74 ATOMIC AND MOLECULAR PHYSICS↗

PRIMA-X - Performance Retargeting of Instrumentation, Measurement, and Analysis Technologies for Exascale Computing

The PRIMA-X (Performance Retargeting of Instrumentation, Measurement, and Analysis Technologies for Exascale Computing) project is the successor of the DOE PRIMA (Performance Refactoring of Instrumentation, Measurement, and Analysis Technologies for Petascale Computing) project, which addressed the challenge of creating a core measurement infrastructure that would serve as a common platform for both integrating leading parallel performance systems (notably TAU and Scalasca) and developing next-generation scalable performance tools. The PRIMA-X project shifts the focus away from refactorization of robust performance tools towards a re-targeting of the parallel performance measurement and analysis architecture for extreme scales. The massive concurrency, asynchronous execution dynamics, hardware heterogeneity, and multi-objective prerequisites (performance, power, resilience) that identify exascale systems introduce fundamental constraints on the ability to carry forward existing performance methodologies. In particular, there must be a deemphasis of per-thread observation techniques to significantly reduce the otherwise unsustainable flood of redundant performance data. Instead, it will be necessary to assimilate multi-level resource observations into macroscopic performance views, from which resilient performance metrics can be attributed to the computational features of the application. This requires a scalable framework for node-level and system-wide monitoring and runtime analyses of dynamic performance information. Also, the interest in optimizing parallelism parameters with respect to performance and energy drives the integration of tool capabilities in the exascale environment further. Initially, PRIMA-X was a collaborative project between the University of Oregon (lead institution) and the German Research School for Simulation Sciences (GRS). Because Prof. Wolf, the PI at GRS, accepted a position as full professor at Technische Universitaet Darmstadt (TU Darmstadt) starting February 1st, 2015, the work of GRS was continued at TU Darmstadt. This report covers the work at TU Darmstadt after the transition. The first main accomplishment of TU Darmstadt is the development of different techniques to aggregate performance data on the level of threads. TU Darmstadt evaluated different schemes designed at GRS during the first phase of this project and integrated them into Score-P, a widely used performance measurement framework for HPC applications. The second main accomplishment is a substantial increase of Score-P’s scalability, achieved by improving its internal representation of the underlying system. Third, we extended our performance-modeling tool Extra-P to automatically create models with more than one input parameter (e.g., process count and input size), which was previously impossible. These models allow the performance behavior of an application to be extrapolated to larger machines. Specific applications of the new multi-parameter capability realized in this project include modeling the isoefficiency function of task-based applications and a novel co-design methodology. Further extensions of Extra-P comprise online modeling functionality, automatic refinement of the model search space, and segmented performance modeling. Further contributions include new locality metrics for both shared- and distributed-memory applications, a new way of measuring the impact of inter-application inference on the execution time of an application, and finally an automatic method that selects appropriate OpenMP constructs for the parallelization of sequential code and classifies the affected variables based on their sharing semantics.

97 MATHEMATICS AND COMPUTING↗

Magnetic Reconnection in a Quasi-Parallel Shock: Two-Dimensional Local Particle-in-Cell Simulation

Magnetic reconnection in a quasi-parallel bow shock is investigated with two-dimensional local particle-in-cell simulations. In the shock transition and downstream regions, large amplitude magnetic fluctuations exist, and abundant current sheets form. In some current sheets, reconnection occurs, and ion-scale and electron-scale magnetic islands are generated. In electron-scale island regions, only electron outflow jets are observed, producing a quadrupolar out-of-plane magnetic field pattern, while in ion-scale islands, both ions and electrons are involved and energized in reconnection. Finally, normalized reconnection rates are obtained to be between around 0.1 to 0.2, and particle acceleration signatures are seen in distribution functions.

58 GEOSCIENCES↗

Multi-Level Load Balancing with an Integrated Runtime Approach

The recent trend of increasing numbers of cores per chip has resulted in vast amounts of on-node parallelism. These high core counts result in hardware variability that introduces imbalance. Applications are also becoming more complex, re-sulting in dynamic load imbalance. Load imbalance of any kind can result in loss of performance and system utilization. We address the challenge of handling both transient and persistent load imbalances while maintaining locality with low overhead. In this paper, we propose an integrated runtime system that combines the Charm++ distributed programming model with concurrent tasks to mitigate load imbalances within and across shared memory address spaces. It utilizes a periodic assignment of work to cores based on load measurement, in combination with user created tasks to handle load imbalance. We integrate OpenMP with Charm++ to enable creation of potential tasks via OpenMP's parallel loop construct. This is also available to MPI applications through the Adaptive MPI implementation. We demonstrate the benefits of our work on three applications. We show improvements of Lassen by 29.6% on Cori and 46.5% on Theta. We also demonstrate the benefits on a Charm++ application, ChaNGa by 25.7% on Theta, as well as an MPI proxy application, Kripke, using Adaptive MPI.

Bak, Seonmyeong↗

Application Characterization at Scale: Lessons learned from developing a distributed Open Community Runtime system for High Performance Computing

Since 2012, the U.S. Department of Energy’s X-Stack program has been developing solutions including runtime systems, programming models, languages, compilers, and tools for the Exascale system software to address crucial performance and power requirements. Fine grain programming models and runtime systems show a great potential to efficiently utilize the underlying hardware. Thus, they are essential to many X-Stack efforts. An abundant amount of small tasks can better utilize the vast parallelism available on current and future machines. Moreover, finer tasks can recover faster and adapt better, due to a decrease in state and control. Nevertheless, current applications have been written to exploit old paradigms (such as Communicating Sequential Processor and Bulk Synchronous Parallel processing). To fully utilize the advantages of these new systems, applications need to be adapted to these new paradigms. As part of the applications’ porting process, in-depth characterization studies, focused on both application characteristics and runtime features, need to take place to fully understand the application performance bottlenecks and how to resolve them. This paper presents a characterization study for a novel high performance runtime system, called the Open Community Runtime, using key HPC kernels as its vehicle. This study has the following contributions: one of the first high performance, fine grain, distributed memory runtime system implementing the OCR standard (version 0.99a); and a characterization study of key HPC kernels in terms of runtime primitives running on both intra and inter node environments. Running on a general purpose cluster, we have found up to 1635x relative speed-up for a parallel tiled Cholesky Kernels on 128 nodes with 16 cores each and a 1864x relative speed-up for a parallel tiled Smith-Waterman kernel on 128 nodes with 30 cores.

Application characterization↗

A parallel and performance portable implementation of a full-field crystal plasticity model

We have developed a parallel implementation of an Elasto-Viscoplastic Fast Fourier Transform-based (EVPFFT) micromechanical solver to enable computationally efficient crystal plasticity modeling for polycrystalline materials. Our primary focus lies in achieving performance portability, allowing a single EVPFFT implementation to run optimally on various homogeneous architectures, including multi-core Central Processing Units (CPUs), as well as on heterogeneous computer architectures comprising multi-core CPUs and Graphics Processing Units (GPUs) from different vendors. To accomplish this goal, we have leveraged MATAR, a C++ software library that simplifies the creation and utilization of multidimensional dense or sparse matrix and array data structures. These data structures are designed to be portable across diverse architectures through the use of Kokkos, a performance-portable library. Additionally, we have employed the Message Passing Interface (MPI) to efficiently distribute the computational workload among processors. The heFFTe (Highly Efficient FFT for Exascale) library is used to facilitate the performance portability of the fast Fourier transforms (FFTs) computation. The computational performance of EVPFFT is evaluated and presented in terms of parallel scalability and simulation runtime on different high-performance computing (HPC) architectures. As a result, the utility of the developed framework to efficiently simulate the micro-mechanical fields in polycrystalline microstructures in engineering applications is discussed.

36 MATERIALS SCIENCE↗