Search NASA⌕ Search

SEARCH · Search NASA

Results for “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 379 records · Page 21

Parallel Algorithms and Patterns

This is a powerpoint presentation on parallel algorithms and patterns. A parallel algorithm is a well-defined, step-by-step computational procedure that emphasizes concurrency to solve a problem. Examples of problems include: Sorting, searching, optimization, matrix operations. A parallel pattern is a computational step in a sequence of independent, potentially concurrent operations that occurs in diverse scenarios with some frequency. Examples are: Reductions, prefix scans, ghost cell updates. We only touch on parallel patterns in this presentation. It really deserves its own detailed discussion which Gabe Rockefeller would like to develop.

97 MATHEMATICS AND COMPUTING↗

Parallel Algorithms for the Exascale Era

New parallel algorithms are needed to reach the Exascale level of parallelism with millions of cores. We look at some of the research developed by students in projects at LANL. The research blends ideas from the early days of computing while weaving in the fresh approach brought by students new to the field of high performance computing. We look at reproducibility of global sums and why it is important to parallel computing. Next we look at how the concept of hashing has led to the development of more scalable algorithms suitable for next-generation parallel computers. Nearly all of this work has been done by undergraduates and published in leading scientific journals.

97 MATHEMATICS AND COMPUTING↗

Efficient Extraction of Regional Subsets from Massive Climate Datasets using Parallel IO

The size of datasets produced by current climate models is increasing rapidly to the scale of petabytes. To handle data at this scale parallel analysis tools are required, however the majority of climate analysis software remains at the scale of workstations. Further, many climate analysis tools adequately process regularly gridded data but lack sufficient features when handling unstructured grids. This paper presents a data-parallel subsetter capable of correctly handling unstructured grids while scaling to over 2000 cores. The approach is based on the partitioned global address space (PGAS) parallel programming model and one-sided communication. The paper demonstrates that IO remains the single greatest bottleneck for this domain of applications and that parallel analysis of climate data succeeds in practice.

54 ENVIRONMENTAL SCIENCES↗

Reducing Concurrency Bottlenecks in Parallel I/O Workloads

To enable high performance parallel checkpointing we introduced the Parallel Log Structured File System (PLFS). PLFS is middleware interposed on the file system stack to transform concurrent writing of one application file into many non-concurrently written component files. The promising effectiveness of PLFS makes it important to examine its performance for workloads other than checkpoint capture, notably the different ways that state snapshots may be later read, to make the case for using PLFS in the Exascale I/O stack. Reading a PLFS file involved reading each of its component files. In this paper we identify performance limitations on broader workloads in an early version of PLFS, specifically the need to build and distribute an index for the overall file, and the pressure on the underlying parallel file system's metadata server, and show how PLFS's decomposed components architecture can be exploited to alleviate bottlenecks in the underlying parallel file system.

COMPUTER ARCHITECTURE↗

Computing contingency statistics in parallel.

Statistical analysis is typically used to reduce the dimensionality of and infer meaning from data. A key challenge of any statistical analysis package aimed at large-scale, distributed data is to address the orthogonal issues of parallel scalability and numerical stability. Many statistical techniques, e.g., descriptive statistics or principal component analysis, are based on moments and co-moments and, using robust online update formulas, can be computed in an embarrassingly parallel manner, amenable to a map-reduce style implementation. In this paper we focus on contingency tables, through which numerous derived statistics such as joint and marginal probability, point-wise mutual information, information entropy, and {chi}{sup 2} independence statistics can be directly obtained. However, contingency tables can become large as data size increases, requiring a correspondingly large amount of communication between processors. This potential increase in communication prevents optimal parallel speedup and is the main difference with moment-based statistics where the amount of inter-processor communication is independent of data size. Here we present the design trade-offs which we made to implement the computation of contingency tables in parallel.We also study the parallel speedup and scalability properties of our open source implementation. In particular, we observe optimal speed-up and scalability when the contingency statistics are used in their appropriate context, namely, when the data input is not quasi-diffuse.

99 GENERAL AND MISCELLANEOUS↗

Discrete Event Modeling and Massively Parallel Execution of Epidemic Outbreak Phenomena

In complex phenomena such as epidemiological outbreaks, the intensity of inherent feedback effects and the significant role of transients in the dynamics make simulation the only effective method for proactive, reactive or post-facto analysis. The spatial scale, runtime speed, and behavioral detail needed in detailed simulations of epidemic outbreaks make it necessary to use large-scale parallel processing. Here, an optimistic parallel execution of a new discrete event formulation of a reaction-diffusion simulation model of epidemic propagation is presented to facilitate in dramatically increasing the fidelity and speed by which epidemiological simulations can be performed. Rollback support needed during optimistic parallel execution is achieved by combining reverse computation with a small amount of incremental state saving. Parallel speedup of over 5,500 and other runtime performance metrics of the system are observed with weak-scaling execution on a small (8,192-core) Blue Gene / P system, while scalability with a weak-scaling speedup of over 10,000 is demonstrated on 65,536 cores of a large Cray XT5 system. Scenarios representing large population sizes exceeding several hundreds of millions of individuals in the largest cases are successfully exercised to verify model scalability.

99 GENERAL AND MISCELLANEOUS↗

Singular value decomposition utilizing parallel algorithms on graphical processors

One of the current challenges in underwater acoustic array signal processing is the detection of quiet targets in the presence of noise. In order to enable robust detection, one of the key processing steps requires data and replica whitening. This, in turn, involves the eigen-decomposition of the sample spectral matrix, Cx = 1/K xKX(k)XH(k) where X(k) denotes a single frequency snapshot with an element for each element of the array. By employing the singular value decomposition (SVD) method, the eigenvectors and eigenvalues can be determined directly from the data without computing the sample covariance matrix, reducing the computational requirements for a given level of accuracy (van Trees, Optimum Array Processing). (Recall that the SVD of a complex matrix A involves determining V, , and U such that A = U VH where U and V are orthonormal and is a positive, real, diagonal matrix containing the singular values of A. U and V are the eigenvectors of AAH and AHA, respectively, while the singular values are the square roots of the eigenvalues of AAH.) Because it is desirable to be able to compute these quantities in real time, an efficient technique for computing the SVD is vital. In addition, emerging multicore processors like graphical processing units (GPUs) are bringing parallel processing capabilities to an ever increasing number of users. Since the computational tasks involved in array signal processing are well suited for parallelization, it is expected that these computations will be implemented using GPUs as soon as users have the necessary computational tools available to them. Thus, it is important to have an SVD algorithm that is suitable for these processors. This work explores the effectiveness of two different parallel SVD implementations on an NVIDIA Tesla C2050 GPU (14 multiprocessors, 32 cores per multiprocessor, 1.15 GHz clock - peed). The first algorithm is based on a two-step algorithm which bidiagonalizes the matrix using Householder transformations, and then diagonalizes the intermediate bidiagonal matrix through implicit QR shifts. This is similar to that implemented for real matrices by Lahabar and Narayanan ("Singular Value Decomposition on GPU using CUDA", IEEE International Parallel Distributed Processing Symposium 2009). The implementation is done in a hybrid manner, with the bidiagonalization stage done using the GPU while the diagonalization stage is done using the CPU, with the GPU used to update the U and V matrices. The second algorithm is based on a one-sided Jacobi scheme utilizing a sequence of pair-wise column orthogonalizations such that A is replaced by AV until the resulting matrix is sufficiently orthogonal (that is, equal to U ). V is obtained from the sequence of orthogonalizations, while can be found from the square root of the diagonal elements of AH A and, once is known, U can be found from column scaling the resulting matrix. These implementations utilize CUDA Fortran and NVIDIA's CUB LAS library. The primary goal of this study is to quantify the comparative performance of these two techniques against themselves and other standard implementations (for example, MATLAB). Considering that there is significant overhead associated with transferring data to the GPU and with synchronization between the GPU and the host CPU, it is also important to understand when it is worthwhile to use the GPU in terms of the matrix size and number of concurrent SVDs to be calculated.

99 GENERAL AND MISCELLANEOUS↗

Data-Parallel Halo Finder Operator in PISTON

PISTON is a portable framework which supports the development of visualization and analysis operators using a platform-independent, data-parallel programming model. Operators such as isosurface, cut-surface and threshold have been implemented in this framework, with the exact same operator code achieving good parallel performance on different architectures. An important analysis operator in cosmology is the halo finder. A halo is a cluster of particles and is considered a common feature of interest found in cosmology data. As the number of cosmological simulations carried out in the recent past has increased, the resultant data of these simulations and the required analysis tasks have increased as well. As a consequence, there is a need to develop scalable and efficient tools to carry out the needed analysis. Therefore, we are currently implementing a halo finder operator using PISTON. Researchers have developed a wide variety of techniques to identify halos in raw particle data. The most basic algorithm is the friend-of-friends (FOF) halo finder, where the particles are clustered based on two parameters: linking length and halo size. In a FOF halo finder, all particles which lie within the linking length are considered as one halo and the halos are filtered based on the halo size parameter. A naive implementation of a FOF halo finder compares each and every particle pair, requiring O(n{sup 2}) operations. Our data-parallel halo finder operator uses a balanced k-d tree to reduce this number of operations in the average case, and implements the algorithm using only the data-parallel primitives in order to achieve portability and performance.

79 ASTRONOMY AND ASTROPHYSICS↗

penORNL: a parallel Monte Carlo photon and electron transport package using PENELOPE

The parallel Monte Carlo photon and electron transport code package penORNL was developed at Oak Ridge National Laboratory to enable advanced scanning electron microscope (SEM) simulations on high-performance computing systems. This paper discusses the implementations, capabilities and parallel performance of the new code package. penORNL uses PENELOPE for its physics calculations and provides all available PENELOPE features to the users, as well as some new features including source definitions specifically developed for SEM simulations, a pulse-height tally capability for detailed simulations of gamma and x-ray detectors, and a modified interaction forcing mechanism to enable accurate energy deposition calculations. The parallel performance of penORNL was extensively tested with several model problems, and very good linear parallel scaling was observed with up to 512 processors. penORNL, along with its new features, will be available for SEM simulations upon completion of the new pulse-height tally implementation.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗

Development Of A Parallel Performance Model For The THOR Neutral Particle Transport Code

The THOR neutral particle transport code enables simulation of complex geometries for various problems from reactor simulations to nuclear non-proliferation. It is undergoing a thorough V&V requiring computational efficiency. This has motivated various improvements including angular parallelization, outer iteration acceleration, and development of peripheral tools. For guiding future improvements to the code’s efficiency, better characterization of its parallel performance is useful. A parallel performance model (PPM) can be used to evaluate the benefits of modifications and to identify performance bottlenecks. Using INL’s Falcon HPC, the PPM development incorporates an evaluation of network communication behavior over heterogeneous links and a functional characterization of the per-cell/angle/group runtime of each major code component. After evaluating several possible sources of variability, this resulted in a communication model and a parallel portion model. The former’s accuracy is bounded by the variability of communication on Falcon while the latter has an error on the order of 1%.

97 MATHEMATICS AND COMPUTING↗

A Reduced-Order Aggregated Model for Parallel Inverter Systems with Virtual Oscillator Control: Preprint

This paper introduces a reduced-order aggregated model for parallel-connected inverters controlled with virtual oscillator control (VOC). The premise of VOC is to modulate inverter dynamics to emulate those of nonlinear oscillators with the goal of realizing a stable ac microgrid in the absence of communication, synchronous generation, or a stiff grid. To obtain a reduced-order model for a system of parallel-connected inverters with VOC, we first formulate a set of scaling laws that describe how the controller and filter parameters of a given inverter depend on its voltage and power rating. Subsequently, we show that N parallel inverters which adhere to this scaling law can be modeled with the same structure and hence the same computational burden of the model of a single inverter. The proposed aggregate model is experimentally validated on a system of three parallel inverters with heterogeneous power ratings.

24 POWER TRANSMISSION AND DISTRIBUTION↗

New Parallel computing framework for radiation transport codes

A new parallel computing framework has been developed to use with general-purpose radiation transport codes. The framework was implemented as a C++ module that uses MPI for message passing. The module is significantly independent of radiation transport codes it can be used with, and is connected to the codes by means of a number of interface functions. The framework was integrated with the MARS15 code, and an effort is under way to deploy it in PHITS. Besides the parallel computing functionality, the framework offers a checkpoint facility that allows restarting calculations with a saved checkpoint file. The checkpoint facility can be used in single process calculations as well as in the parallel regime. Several checkpoint files can be merged into one thus combining results of several calculations. The framework also corrects some of the known problems with the scheduling and load balancing found in the original implementations of the parallel computing functionality in MARS15 and PHITS. The framework can be used efficiently on homogeneous systems and networks of workstations, where the interference from the other users is possible.

43 PARTICLE ACCELERATORS↗

Efficient Quaternary Organic Solar Cells with Parallel‐Alloy Morphology

Abstract Two compatible donors (PBDB‐T and PTB7‐Th) and two miscible acceptors (ITIC and FOIC) are employed to deliver a parallel‐alloy morphology model in non‐fullerene‐based quaternary organic solar cells. PBDB‐T and PTB7‐Th form a parallel link with a slight adjustment of molecular packing into enhanced face‐on crystallites while ITIC disperses into discontinuous FOIC microcrystal regions to form continuous and ordered alloy‐like acceptor phases. Characterization of blend morphology highlights the parallel‐alloy model—enabled by the introduction of PBDB‐T and ITIC, which contributes to improved molecular packing and reduced domain size resulting in efficient charge generation and consistent transport channels. This successful parallel‐alloy quaternary blend morphology demonstrates an enhanced optical absorption, optimized domain size, and nanostructures toward simultaneous improvement in charge transfer and transport. Therefore, a power conversion efficiency of 12.52% is realized for a quaternary device which is 6% higher than the ternary device (PBDB‐T:PTB7‐Th:FOIC) and 12% higher than the binary device (PTB7‐Th:FOIC). Domination of quaternary devices over ternary and binary blends, which is another feasible way to realize highly efficient devices through further investigation of quaternary OSCs, is presented.

Bi, Zhaozhao↗

A Parallel Computing Framework for Finding the Supported Solutions to a Biobjective Network Optimization Problem

Abstract Solving multi‐objective combinatorial optimization problems can quickly become computationally challenging when applied to large networks generated from big data. Present‐day first‐rate Mixed Integer Programming (MIP) solvers have parallelism built‐in to take advantage of multicore architectures; but specialized network optimization algorithms that can often solve graph problems more efficiently than a general MIP solver are typically programmed serially. Thus, these network algorithm implementations do not take full advantage of modern multicore computer capabilities. Many parallel computing languages include a fork/join framework into their concurrency packages that allow for simple conversion of recursive serial algorithms to operate in parallel without having to use complicated low‐level message‐passing interfaces. Fork/join is particularly well suited to divide and conquer algorithms such as Non‐Inferior Set Estimation (NISE) method for computing the supported Pareto front of a multi‐objective optimization problem. This work develops a simple general parallel NISE (pNISE) implementation using the Java 7 concurrency tools, and then offers results from a specific implementation of solving a bi‐objective shortest path problem. The results indicate that this method is capable of sizeable speed‐ups on both small‐scale personal computers as well as large‐scale shared memory supercomputers. Finally, other network problems suitable to this method are discussed, including multi‐objective variants of the minimum spanning tree problem, transportation problem, assignment problem, maximum flow problem, and the minimum‐cost flow problem. Copyright © 2015 John Wiley & Sons, Ltd.

Medrano, Fernando Antonio↗

Robustness of Deep Learning Classification to Adversarial Input on GPUs: Asynchronous Parallel Accumulation Is a Source of Vulnerability

The ability of machine learning (ML) classification models to resist small, targeted input perturbations—known as adversarial attacks—is a key measure of their safety and reliability. We show that floating-point non associativity (FPNA) coupled with asynchronous parallel programming on GPUs is sufficient to result in misclassification, without any perturbation to the input. Additionally, we show that this misclassification is particularly significant for inputs close to the decision boundary and that standard adversarial robustness results may be overestimated up to 4.6 when not considering machine-level details. We first study a linear classifier, before focusing on standard Graph Neural Network (GNN) architectures and datasets used in robustness assessments. We develop a novel black-box attack using Bayesian optimization to discover external workloads that can change the instruction scheduling which bias the output of reductions on GPUs and reliably lead to misclassification. Motivated by these results, we present a new learnable permutation (LP) gradient-based approach to learning floating-point operation orderings that lead to misclassifications. The LP approach provides a worst-case estimate in a computationally efficient manner, avoiding the need to run identical experiments tens of thousands of times over a potentially large set of possible GPU states or architectures. Finally, using instrumentation-based testing, we investigate parallel reduction ordering across different GPU architectures under external background workloads, when utilizing multi-GPU virtualization, and when applying power capping. Our results demonstrate that parallel reduction ordering varies significantly across architectures under the first two conditions, substantially increasing the search space required to fully test the effects of this parallel scheduler-based vulnerability. These results and the methods developed here can help to include machine-level considerations into adversarial robustness assessments, which can make a difference in safety and mission critical applications.

Shanmugavelu, Sanjif [Maxeler Technologies, a Groq↗

Management of Transjugular Intrahepatic Portosystemic Shunt (TIPS)-associated Refractory Hepatic Encephalopathy by Shunt Reduction Using the Parallel Technique: Outcomes of a Retrospective Case Series

Purpose: To investigate the reproducibility and technical and clinical success of the parallel technique of transjugular intrahepatic portosystemic shunt (TIPS) reduction in the management of refractory hepatic encephalopathy (HE). Materials and Methods: A 10-mm-diameter self-expanding stent graft and a 5-6-mm-diameter balloon-expandable stent were placed in parallel inside the existing TIPS in 8 patients via a dual unilateral transjugular approach. Changes in portosystemic pressure gradient and HE grade were used as primary end points. Results: TIPS reduction was technically successful in all patients. Mean {+-} standard deviation portosystemic pressure gradient before and after shunt reduction was 4.9 {+-} 3.6 mmHg (range, 0-12 mmHg) and 10.5 {+-} 3.9 mmHg (range, 6-18 mmHg). Duration of follow-up was 137 {+-} 117.8 days (range, 18-326 days). Clinical improvement of HE occurred in 5 patients (62.5%) with resolution of HE in 4 patients (50%). Single episodes of recurrent gastrointestinal hemorrhage occurred in 3 patients (37.5%). These were self-limiting in 2 cases and successfully managed in 1 case by correction of coagulopathy and blood transfusion. Two of these patients (25%) died, one each of renal failure and hepatorenal failure. Conclusion: The parallel technique of TIPS reduction is reproducible and has a high technical success rate. A dual unilateral transjugular approach is advantageous when performing this procedure. The parallel technique allows repeat bidirectional TIPS adjustment and may be of significant clinical benefit in the management of refractory HE.

62 RADIOLOGY AND NUCLEAR MEDICINE↗

Parallel application-level behavioral attributes for performance and energy management of high-performance computing systems

Run time variability of parallel applications continues to present significant challenges to their performance and energy efficiency in high-performance computing (HPC) systems. When run times are extended and unpredictable, application developers perceive this as a degradation of system (or subsystem) performance. Extended run times directly contribute to proportionally higher energy consumption, potentially negating efforts by applications, or the HPC system, to optimize energy consumption using low-level control techniques, such as dynamic voltage and frequency scaling (DVFS). Therefore, successful systemic management of application run time performance can result in less wasted energy, or even energy savings. We have been studying run time variability in terms of communication time, from the perspective of the application, focusing on the interconnection network. More recently, our focus has shifted to developing a more complete understanding of the effects of HPC subsystem interactions on parallel applications. In this context, the set of executing applications on the HPC system is treated as a subsystem, along with more traditional subsystems like the communication subsystem, storage subsystem, etc. To gain insight into the run time variability problem, our earlier work developed a framework to emulate parallel applications (PACE) that stresses the communication subsystem. Evaluation of run time sensitivity to network performance of real applications is performed with a tool called PARSE, which uses PACE. In this paper, we propose a model defining application-level behavioral attributes, that collectively describes how applications behave in terms of their run time performance, as functions of their process distribution on the system (spacial locality), and subsystem interactions (communication subsystem degradation). These subsystem interactions are produced when multiple applications execute concurrently on the same HPC system. We also revisit our evaluation framework and tools to demonstrate the flexibility of our application characterization techniques, and the ease with which attributes can be quantified. The validity of the model is demonstrated using our tools with several parallel benchmarks and application fragments. Results suggest that it is possible to articulate application-level behavioral attributes as a tuple of numeric values that describe course-grained performance behavior.

Computer Science↗

Massively parallel phase-field simulations targeting exascale

The interface thickness in the phase-field (PF) method limits its simulation scales. Consequently, large-scale PF simulations become prohibitively expensive for resolving the extremely fine microstructures that typically form during rapid solidification processing. This challenge is significant in predicting microstructure evolution in metal additive manufacturing and has been identified by the United States Department of Energy’s Exascale Computing Project. Here, to address this, we develop a multi-GPU and MPI-based massively parallel simulation code, utilizing state-of-the-art algorithms, software, and libraries, for large-scale three-dimensional (3D) PF simulations. We report the first GPU-parallel PF simulations on Frontier (currently the second TOP500 exascale cluster) and Summit machines, taking dendritic growth as an example problem. We evaluate the parallel performance of our implementation using scaling studies with more than 24 000 GPUs (among the largest known computations to date) and the acceleration performance using large-scale simulations of dendritic growth in 3D. Finally, massively parallel GPUs in these supercomputers enabled the first coupled multiscale simulations of laser melting and subsequent dendritic solidification on the scale of a full melt-pool, demonstrating the feasibility of performing PF simulations with a point total over 2 billion grid points within an acceptable time.

Exascale↗