SEARCH · Search NASA

Results for “Parallelization efficiency”

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.

SourceResearch connectionFrom yearThrough year

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 829 records · Page 46

All sources NASA NTRS DOE OSTI

NASA NTRS1992

A transient FETI methodology for large-scale parallel implicit computations in structural mechanics

Explicit codes are often used to simulate the nonlinear dynamics of large-scale structural systems, even for low frequency response, because the storage and CPU requirements entailed by the repeated factorizations traditionally found in implicit codes rapidly overwhelm the available computing resources. With the advent of parallel processing, this trend is accelerating because explicit schemes are also easier to parallelize than implicit ones. However, the time step restriction imposed by the Courant stability condition on all explicit schemes cannot yet -- and perhaps will never -- be offset by the speed of parallel hardware. Therefore, it is essential to develop efficient and robust alternatives to direct methods that are also amenable to massively parallel processing because implicit codes using unconditionally stable time-integration algorithms are computationally more efficient when simulating low-frequency dynamics. Here we present a domain decomposition method for implicit schemes that requires significantly less storage than factorization algorithms, that is several times faster than other popular direct and iterative methods, that can be easily implemented on both shared and local memory parallel processors, and that is both computationally and communication-wise efficient. The proposed transient domain decomposition method is an extension of the method of Finite Element Tearing and Interconnecting (FETI) developed by Farhat and Roux for the solution of static problems. Serial and parallel performance results on the CRAY Y-MP/8 and the iPSC-860/128 systems are reported and analyzed for realistic structural dynamics problems. These results establish the superiority of the FETI method over both the serial/parallel conjugate gradient algorithm with diagonal scaling and the serial/parallel direct method, and contrast the computational power of the iPSC-860/128 parallel processor with that of the CRAY Y-MP/8 system.

Farhat, Charbel↗

NASA NTRS1993

A transient FETI methodology for large-scale parallel implicit computations in structural mechanics, part 2

Explicit codes are often used to simulate the nonlinear dynamics of large-scale structural systems, even for low frequency response, because the storage and CPU requirements entailed by the repeated factorizations traditionally found in implicit codes rapidly overwhelm the available computing resources. With the advent of parallel processing, this trend is accelerating because explicit schemes are also easier to parallellize than implicit ones. However, the time step restriction imposed by the Courant stability condition on all explicit schemes cannot yet and perhaps will never be offset by the speed of parallel hardware. Therefore, it is essential to develop efficient and robust alternatives to direct methods that are also amenable to massively parallel processing because implicit codes using unconditionally stable time-integration algorithms are computationally more efficient than explicit codes when simulating low-frequency dynamics. Here we present a domain decomposition method for implicit schemes that requires significantly less storage than factorization algorithms, that is several times faster than other popular direct and iterative methods, that can be easily implemented on both shared and local memory parallel processors, and that is both computationally and communication-wise efficient. The proposed transient domain decomposition method is an extension of the method of Finite Element Tearing and Interconnecting (FETI) developed by Farhat and Roux for the solution of static problems. Serial and parallel performance results on the CRAY Y-MP/8 and the iPSC-860/128 systems are reported and analyzed for realistic structural dynamics problems. These results establish the superiority of the FETI method over both the serial/parallel conjugate gradient algorithm with diagonal scaling and the serial/parallel direct method, and contrast the computational power of the iPSC-860/128 parallel processor with that of the CRAY Y-MP/8 system.

Farhat, Charbel↗

DOE OSTI2015

Lattice Quantum Chromodynamics with Overlap Fermions on GPUs

Lattice quantum chromodynamics (QCD) calculations were one of the first applications to demonstrate the potential of GPUs in the area of high-performance computing; the nature of lattice QCD calculations matches well with the GPU's computational model. This article discusses ways to effectively use GPUs for lattice calculations using the overlap operator, a discretization that preserves chiral symmetry even at nonzero lattice spacing and makes possible lattice QCD simulations in the parameter region relevant to nuclear physics. The author shows that the large memory footprint of these codes requires the use of multiple GPUs in parallel and discusses methods for implementing this operator efficiently: hybrid CPU/GPU memory use for eigensolvers and MPI/OpenMP/CUDA parallelization strategies required to take full advantage of both GPU and CPU resources. He then compares the performance of codes on a GPU cluster and a CPU cluster with similar interconnects, discussing the strong scaling for problem sizes relevant to current lattice QCD simulations.

Computer Science↗

NASA NTRS1996

Parallel NPARC: Implementation and Performance

Version 3 of the NPARC Navier-Stokes code includes support for large-grain (block level) parallelism using explicit message passing between a heterogeneous collection of computers. This capability has the potential for significant performance gains, depending upon the block data distribution. The parallel implementation uses a master/worker arrangement of processes. The master process assigns blocks to workers, controls worker actions, and provides remote file access for the workers. The processes communicate via explicit message passing using an interface library which provides portability to a number of message passing libraries, such as PVM (Parallel Virtual Machine). A Bourne shell script is used to simplify the task of selecting hosts, starting processes, retrieving remote files, and terminating a computation. This script also provides a simple form of fault tolerance. An analysis of the computational performance of NPARC is presented, using data sets from an F/A-18 inlet study and a Rocket Based Combined Cycle Engine analysis. Parallel speedup and overall computational efficiency were obtained for various NPARC run parameters on a cluster of IBM RS6000 workstations. The data show that although NPARC performance compares favorably with the estimated potential parallelism, typical data sets used with previous versions of NPARC will often need to be reblocked for optimum parallel performance. In one of the cases studied, reblocking increased peak parallel speedup from 3.2 to 11.8.

Townsend, S. E.↗

NASA NTRS1989

Case studies in serial and parallel simulation

The design of a ground combat simulation in a time-driven and in an event-driven style is discussed. The differences between time-driven and event-driven simulation are illustrated, and performance results for the two systems executing both serially and in parallel are presented. It is noted that the event-driven style produces a more efficient system both serially and in parallel, primarily because it minimizes the number of messages produced and thereby reduces the amount of computation.

Wieland, Frederick↗

NASA NTRS1989

Experiences with hypercube operating system instrumentation

The difficulties in conceptualizing the interactions among a large number of processors make it difficult both to identify the sources of inefficiencies and to determine how a parallel program could be made more efficient. This paper describes an instrumentation system that can trace the execution of distributed memory parallel programs by recording the occurrence of parallel program events. The resulting event traces can be used to compile summary statistics that provide a global view of program performance. In addition, visualization tools permit the graphic display of event traces. Visual presentation of performance data is particularly useful, indeed, necessary for large-scale parallel computers; the enormous volume of performance data mandates visual display.

Reed, Daniel A.↗

NASA NTRS1997

Massively Parallel and Scalable Implicit Time Integration Algorithms for Structural Dynamics

Farhat, Charbel↗

DOE OSTI2015

Secure Storage Architectures

The purpose of this report is to clarify the challenges associated with storage for secure enclaves. The major focus areas for the report are: - review of relevant parallel filesystem technologies to identify assets and gaps; - review of filesystem isolation/protection mechanisms, to include native filesystem capabilities and auxiliary/layered techniques; - definition of storage architectures that can be used for customizable compute enclaves (i.e., clarification of use-cases that must be supported for shared storage scenarios); - investigate vendor products related to secure storage. This study provides technical details on the storage and filesystem used for HPC with particular attention on elements that contribute to creating secure storage. We outline the pieces for a a shared storage architecture that balances protection and performance by leveraging the isolation capabilities available in filesystems and virtualization technologies to maintain the integrity of the data. Key Points: There are a few existing and in-progress protection features in Lustre related to secure storage, which are discussed in (Chapter 3.1). These include authentication capabilities like GSSAPI/Kerberos and the in-progress work for GSSAPI/Host-keys. The GPFS filesystem provides native support for encryption, which is not directly available in Lustre. Additionally, GPFS includes authentication/authorization mechanisms for inter-cluster sharing of filesystems (Chapter 3.2). The limitations of key importance for secure storage/filesystems are: (i) restricting sub-tree mounts for parallel filesystem (which is not directly supported in Lustre or GPFS), and (ii) segregation of hosts on the storage network and practical complications with dynamic additions to the storage network, e.g., LNET. A challenge for VM based use cases will be to provide efficient IO forwarding of the parallel filessytem from the host to the guest (VM). There are promising options like para-virtualized filesystems to help with this issue, which are a particular instances of the more general challenge of efficient host/guest IO that is the focus of interfaces like virtio. A collection of bridging technologies have been identified in Chapter 4, which can be helpful to overcome the limitations and challenges of supporting efficient storage for secure enclaves. The synthesis of native filesystem security mechanisms and bridging technologies led to an isolation-centric storage architecture that is proposed in Chapter 5, which leverages isolation mechanisms from different layers to facilitate secure storage for an enclave. Recommendations: The following highlights recommendations from the investigations done thus far. - The Lustre filesystem offers excellent performance but does not support some security related features, e.g., encryption, that are included in GPFS. If encryption is of paramount importance, then GPFS may be a more suitable choice. - There are several possible Lustre related enhancements that may provide functionality of use for secure-enclaves. However, since these features are not currently integrated, the use of Lustre as a secure storage system may require more direct involvement (support). (*The network that connects the storage subsystem and users, e.g., Lustre s LNET.) - The use of OpenStack with GPFS will be more streamlined than with Lustre, as there are available drivers for GPFS. - The Manilla project offers Filesystem as a Service for OpenStack and is worth further investigation. Manilla has some support for GPFS. - The proposed Lustre enhancement of Dynamic-LNET should be further investigated to provide more dynamic changes to the storage network which could be used to isolate hosts and their tenants. - The Linux namespaces offer a good solution for creating efficient restrictions to shared HPC filesystems. However, we still need to conduct a thorough round of storage/filesystem benchmarks. - Vendor products should be more closely reviewed, possibly to include evaluation of performance/protection of select products. (Note, we are investigation the option of evaluating equipment from Seagate/Xyratex.) Outline: The remainder of this report is structured as follows: - Section 1: Describes the growing importance of secure storage architectures and highlights some challenges for HPC. - Section 2: Provides background information on HPC storage architectures, relevant supporting technologies for secure storage and details on OpenStack components related to storage. Note, that background material on HPC storage architectures in this chapter can be skipped if the reader is already familiar with Lustre and GPFS. - Section 3: A review of protection mechanisms in two HPC filesystems; details about available isolation, authentication/authorization and performance capabilities are discussed. - Section 4: Describe technologies that can be used to bridge gaps in HPC storage and filesystems to facilitate...

Aderholdt, Ferrol↗

DOE OSTI2019

Parallel beamlet dose calculation via beamlet contexts in a distributed multi‐GPU framework

Purpose Dose calculation is one of the most computationally intensive, yet essential tasks in the treatment planning process. With the recent interest in automatic beam orientation and arc trajectory optimization techniques, there is a great need for more efficient model‐based dose calculation algorithms that can accommodate hundreds to thousands of beam candidates at once. Foundational work has shown the translation of dose calculation algorithms to graphical processing units ( GPU s), lending to remarkable gains in processing efficiency. But these methods provide parallelization of dose for only a single beamlet, serializing the calculation of multiple beamlets and under‐utilizing the potential of modern GPU s. In this paper, the authors propose a framework enabling parallel computation of many beamlet doses using a novel beamlet context transformation and further embed this approach in a scalable network of multi‐ GPU computational nodes. Methods The proposed context‐based transformation separates beamlet‐local density and TERMA into distinct beamlet contexts that independently provide sufficient data for beamlet dose calculation. Beamlet contexts are arranged in a composite context array with dosimetric isolation, and the context array is subjected to a GPU collapsed‐cone convolution superposition procedure, producing the set of beamlet‐specific dose distributions in a single pass. Dose from each context is converted to a sparse representation for efficient storage and retrieval during treatment plan optimization. The context radius is a new parameter permitting flexibility between the speed and fidelity of the dose calculation process. A distributed manager‐worker architecture is constructed around the context‐based GPU dose calculation approach supporting an arbitrary number of worker nodes and resident GPU s. Phantom experiments were executed to verify the accuracy of the context‐based approach compared to Monte Carlo and a reference CPU ‐ CCCS implementation for single beamlets and broad beams composed by addition of beamlets. Dose for representative 4 π beam sets was calculated in lung and prostate cases to compare its efficiency with that of an existing beamlet‐sequential GPU ‐ CCCS implementation. Code profiling was also performed to evaluate the scalability of the framework across many networked GPU s. Results The dosimetric accuracy of the context‐based method displays <1.35% and 2.35% average error from the existing serialized CPU ‐ CCCS algorithm and Monte Carlo simulation for beamlet‐specific PDD s in water and slab phantoms, respectively. The context‐based method demonstrates substantial speedup of up to two orders of magnitude over the beamlet‐sequential GPU ‐ CCCS method in the tested configurations. The context‐based framework demonstrates near linear scaling in the number of distributed compute nodes and GPU s employed, indicating that it is flexible enough to meet the performance requirements of most users by simply increasing the hardware utilization. Conclusions The context‐based approach demonstrates a new expectation of performance for beamlet‐based dose calculation methods. This approach has been successful in accelerating the dose calculation process for very large‐scale treatment planning problems ‐ such as automatic 4 π IMRT beam orientation and VMAT arc trajectory selection, with hundreds of thousands of beamlets ‐ in clinically feasible timeframes. The flexibility of this framework makes it as a strong candidate for use in a variety of other very large‐scale treatment planning tasks and clinical workflows.

Neph, Ryan↗

DOE OSTI2013

Steal Tree: Low-Overhead Tracing of Work Stealing Schedulers

Work stealing is a popular approach to scheduling task-parallel programs. The flexibility inherent in work stealing when dealing with load imbalance results in seemingly irregular computation structures, complicating the study of its runtime behavior. In this paper, we present an approach to efficiently trace async-finish parallel programs scheduled using work stealing. We identify key properties that allow us to trace the execution of tasks with low time and space overheads. We also study the usefulness of the proposed schemes in supporting algorithms for data-race detection and retentive steal- ing presented in the literature. We demonstrate that the perturbation due to tracing is within the variation in the execution time with 99% confidence and the traces are concise, amounting to a few tens of kilobytes per thread in most cases. We also demonstrate that the traces enable significant reductions in the cost of detecting data races and result in low, stable space overheads in supporting retentive stealing for async-finish programs.

async-finish parallelism↗

DOE OSTI2018

ARRC: A random ray neutron transport code for nuclear reactor simulation

Not Available

Tramm, John R.↗

DOE OSTI2021

Multigrid reduction in time with Richardson extrapolation

The advent of exascale computing will leave many users with access to more computational resources than they can simultaneously use, e.g., billion-way parallelism. In particular, this is true for time-dependent simulations that limit parallelism to the spatial domain. One method to add parallelism in time to existing simulation codes and thus take advantage of ever larger compute resources is Multigrid Reduction in Time (MGRIT). The goal is to achieve a smaller time-to-solution through parallelism in time. In this paper, MGRIT is enhanced with Richardson extrapolation in a cost-efficient way to produce a parallel-in-time method with improved accuracy. Overall, this leads to a large improvement in the accuracy per computational cost of MGRIT.

97 MATHEMATICS AND COMPUTING↗

NASA NTRS2001

Application of a Scalable, Parallel, Unstructured-Grid-Based Navier-Stokes Solver

A parallel version of an unstructured-grid based Navier-Stokes solver, USM3Dns, previously developed for efficient operation on a variety of parallel computers, has been enhanced to incorporate upgrades made to the serial version. The resultant parallel code has been extensively tested on a variety of problems of aerospace interest and on two sets of parallel computers to understand and document its characteristics. An innovative grid renumbering construct and use of non-blocking communication are shown to produce superlinear computing performance. Preliminary results from parallelization of a recently introduced "porous surface" boundary condition are also presented.

Parikh, Paresh↗

DOE OSTI2018

Producing a glycosylating Escherichia coli cell factory: The placement of the bacterial oligosaccharyl transferase pglB onto the genome

Highlights: • Expressing pglB from the genome creates a “plug and play” E. coli strain for homogeneous recombinant glycoprotein production. • This engineered strain can increase glycosylation efficiency and overall glycoprotein titres. • Parallel Reaction Monitoring can simultaneously confirm glycan addition and quantify absolute glycoprotein titres. Although Escherichia coli has been engineered to perform N-glycosylation of recombinant proteins, an optimal glycosylating strain has not been created. By inserting a codon optimised Campylobacter oligosaccharyltransferase onto the E. coli chromosome, we created a glycoprotein platform strain, where the target glycoprotein, sugar synthesis and glycosyltransferase enzymes, can be inserted using expression vectors to produce the desired homogenous glycoform. To assess the functionality and glycoprotein producing capacity of the chromosomally based OST, a combined Western blot and parallel reaction monitoring mass spectrometry approach was applied, with absolute quantification of glycoprotein. We demonstrated that chromosomal oligosaccharyltransferase remained functional and facilitated N-glycosylation. Although the engineered strain produced less total recombinant protein, the glycosylation efficiency increased by 85%, and total glycoprotein production was enhanced by 17%.

60 APPLIED LIFE SCIENCES↗

DOE OSTI2018

Parallel Nonnegative CP Decomposition of Dense Tensors

The CP tensor decomposition is a low-rank approximation of a tensor. We present a distributed-memory parallel algorithm and implementation of an alternating optimization method for computing a CP decomposition of dense tensors that can enforce nonnegativity of the computed low-rank factors. The principal task is to parallelize the Matricized-Tensor Times Khatri-Rao Product (MTTKRP) bottleneck subcomputation. The algorithm is computation efficient, using dimension trees to avoid redundant computation across MTTKRPs within the alternating method. Our approach is also communication efficient, using a data distribution and parallel algorithm across a multidimensional processor grid that can be tuned to minimize communication. We benchmark our software on synthetic as well as hyperspectral image and neuroscience dynamic functional connectivity data, demonstrating that our algorithm scales well to 100s of nodes (up to 4096 cores) and is faster and more general than the currently available parallel software.

Ballard, Grey↗

NASA NTRS1989

Novel techniques for data decomposition and load balancing for parallel processing of vision systems: Implementation and evaluation using a motion estimation system

Computer vision systems employ a sequence of vision algorithms in which the output of an algorithm is the input of the next algorithm in the sequence. Algorithms that constitute such systems exhibit vastly different computational characteristics, and therefore, require different data decomposition techniques and efficient load balancing techniques for parallel implementation. However, since the input data for a task is produced as the output data of the previous task, this information can be exploited to perform knowledge based data decomposition and load balancing. Presented here are algorithms for a motion estimation system. The motion estimation is based on the point correspondence between the involved images which are a sequence of stereo image pairs. Researchers propose algorithms to obtain point correspondences by matching feature points among stereo image pairs at any two consecutive time instants. Furthermore, the proposed algorithms employ non-iterative procedures, which results in saving considerable amounts of computation time. The system consists of the following steps: (1) extraction of features; (2) stereo match of images in one time instant; (3) time match of images from consecutive time instants; (4) stereo match to compute final unambiguous points; and (5) computation of motion parameters.

Choudhary, Alok Nidhi↗

DOE OSTI2017

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↗

DOE OSTI2022

aphBO-2GP-3B: a budgeted asynchronous parallel multi-acquisition functions for constrained Bayesian optimization on high-performing computing architecture

High-fidelity complex engineering simulations are often predictive, but also computationally expensive and often require substantial computational efforts. The mitigation of computational burden is usually enabled through parallelism in high-performance cluster (HPC) architecture. Optimization problems associated with these applications is a challenging problem due to the high computational cost of the high-fidelity simulations. In this paper, an asynchronous parallel constrained Bayesian optimization method is proposed to efficiently solve the computationally expensive simulation-based optimization problems on the HPC platform, with a budgeted computational resource, where the maximum number of simulations is a constant. The advantage of this method are three-fold. Firstly, the efficiency of the Bayesian optimization is improved, where multiple input locations are evaluated parallel in an asynchronous manner to accelerate the optimization convergence with respect to physical runtime. This efficiency feature is further improved so that when each of the inputs is finished, another input is queried without waiting for the whole batch to complete. Second, the proposed method can handle both known and unknown constraints. Third, the proposed method samples several acquisition functions based on their rewards using a modified GP-Hedge scheme. The proposed framework is termed aphBO-2GP-3B, which means asynchronous parallel hedge Bayesian optimization with two Gaussian processes and three batches. The numerical performance of the proposed framework aphBO-2GP-3B is comprehensively benchmarked using 16 numerical examples, compared against other 6 parallel Bayesian optimization variants and 1 parallel Monte Carlo as a baseline, and demonstrated using two real-world high-fidelity expensive industrial applications. The first engineering application is based on finite element analysis (FEA) and the second one is based on computational fluid dynamics (CFD) simulations.

97 MATHEMATICS AND COMPUTING↗