Search NASA⌕ Search

SEARCH · Search NASA

Results for “Parallel optimization”

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 1,369 records · Page 76

High-Throughput Discovery of Robust Metal-Organic Frameworks for CO 2 Capture (Final Report)

The goal of the project is to develop and utilize high-throughput instrumentation for the discovery of robust, cost-effective metal-organic frameworks for the efficient capture of CO 2 from power plant flue gas. To this end, we have used highly-parallel and automated synthesize methods to accelerate the discovery of new MOF structures and their optimized synthesis and activation routes. To be able to rapidly screen these new materials for their potential CO 2 capture performance, we have developed high-throughput structural characterization instruments and gas adsorption instruments. To guide the materials choice and design, we have also developed computational tools to analyze the life cycle and sorbent performance of existing and hypothetical materials.

36 MATERIALS SCIENCE↗

Understanding the Machine Learning Needs of ECP Applications

In order to support the codesign needs of ECP applications in current and future hardware in the area of machine learning, the ExaLearn team at Sandia studied the different machine learning use cases in three different ECP applications. This report is a summary of the needs of the three applications. The Sandia ExaLearn team will develop a proxy application representative of ECP application needs, specifically the ExaSky and EXAALT ECP projects. The proxy application will allow us to demonstrate performance portable kernels within machine learning codes. Furthermore, current training scalability of machine learning networks in these applications is negatively affected by large batch sizes. Training throughput of the network will increase as batch size increases, but network accuracy and generalization worsens. The proxy application will contain hybrid model- and data-parallelism to improve training efficiency while maintaining network accuracy. The proxy application will also target optimizing 3D convolutional layers, specific to scientific machine learning, which have not been as thoroughly explored by industry.

97 MATHEMATICS AND COMPUTING↗

Optimizing the Accelerated Recursive Doubling Algorithm for Block Tridiagonal Systems of Equations

The need to solve block tridiagonal systems with hundreds or thousands of right-hand sides for the same block tridiagonal matrix is common in a variety of disciplines. To meet this need, the Accelerated Recursive Doubling Algorithm was developed. After a right-hand side independent phase, the algorithm allows for the quick, online calculation of solutions for different right-hand sides. In this work, we present methods to optimize the Accelerated Recursive Doubling Algorithm in memory usage and computation time in a hybrid parallelization model. The right-hand side independent phase of the naïve implementation takes ≥ 11/3 the amount of memory required to store the tridiagonal matrix, while our implementation reduces the fraction to ≈ 5/3 . The right-hand side dependent phase of the naïve implementation takes ≥ 6 times the amount of memory required to store the right-hand side, while our implementation reduces the fraction to ≈ 3. The computation time for the independent phase is reduced to ≈ 2/3 times that of the naïve implementation, while the computation time for the dependent phase is reduced to ≈ 5/9 . With increasing numbers of shared-memory threads q on every distributed processing element, we have O(q) theoretical speedup.

97 MATHEMATICS AND COMPUTING↗

Multi-Node Program Fuzzing on High Performance Computing Resources

Significant effort is placed on tuning the internal parameters of fuzzers to explore the state space, measured as coverage, of binaries. In this work, we investigate the effects of the external environment on the resulting coverage after fuzzing two binaries with AFL for 24 hours. Parameters such as scaling to multiple nodes, node saturation, and parallel file system type on HPC resources are controlled in order to maximize coverage. It will be shown that employing a parallel file system such as IBM's General Parallel File System offers an advantage for fuzzing operations, since it contains enhancements for performance optimization. When combined with scaling to two and four nodes, while simultaneously restricting the number of coordinated AFL tasks per node on the low end (10-50% of available physical cores), coverage may be enhanced within a shorter period of time. Thus, controlling the external environment is a useful effort.

97 MATHEMATICS AND COMPUTING↗

Parallel computing for probabilistic fatigue analysis

This paper presents the results of Phase I research to investigate the most effective parallel processing software strategies and hardware configurations for probabilistic structural analysis. We investigate the efficiency of both shared and distributed-memory architectures via a probabilistic fatigue life analysis problem. We also present a parallel programming approach, the virtual shared-memory paradigm, that is applicable across both types of hardware. Using this approach, problems can be solved on a variety of parallel configurations, including networks of single or multiprocessor workstations. We conclude that it is possible to effectively parallelize probabilistic fatigue analysis codes; however, special strategies will be needed to achieve large-scale parallelism to keep large number of processors busy and to treat problems with the large memory requirements encountered in practice. We also conclude that distributed-memory architecture is preferable to shared-memory for achieving large scale parallelism; however, in the future, the currently emerging hybrid-memory architectures will likely be optimal.

Sues, Robert H.↗

Global Design Optimization for Fluid Machinery Applications

Recent experiences in utilizing the global optimization methodology, based on polynomial and neural network techniques for fluid machinery design are summarized. Global optimization methods can utilize the information collected from various sources and by different tools. These methods offer multi-criterion optimization, handle the existence of multiple design points and trade-offs via insight into the entire design space can easily perform tasks in parallel, and are often effective in filtering the noise intrinsic to numerical and experimental data. Another advantage is that these methods do not need to calculate the sensitivity of each design variable locally. However, a successful application of the global optimization method needs to address issues related to data requirements with an increase in the number of design variables and methods for predicting the model performance. Examples of applications selected from rocket propulsion components including a supersonic turbine and an injector element and a turbulent flow diffuser are used to illustrate the usefulness of the global optimization method.

Shyy, Wei↗

Computational Analyses of Propulsion Aeroacoustics for Mixed Flow Nozzle Pylon Installation at Takeoff

A CFD analyses is presented for a set of baseline and noise suppression mixed flow nozzles with and without a pylon installation. The five model configurations are as follows; a baselinecore/fan dual-stream nozzle with an external plug, a chevron mixer nozzle with a peak on the symmetry plane with external plug, both of the above nozzles with an installed bifurcatingpylon and lastly a clocked chevron mixer nozzle such that a trough is aligned with the center of the pylon. The fluid flow is simulated by solving the asymptotically steady, compressible, Reynolds-averaged Navier-Stokes equations using an implicit, up-wind, flux-difference splitting finite volume scheme and standard two equation k-epsilon turbulence model with a linear stress representation. All computations are performed using the multiblock, parallel, structuredcode PAB3D. Results indicate that the clocked chevron with pylon case achieves the most optimal levels of average and peak turbulence kinetic energy and vorticity and therefore is expected to be the quietest of the five configurations tested. Further study is required to refine expressions which are indicative of noise and mate these with rigorous noise prediction models.

Massey, Steven J.↗

Line-plane broadcasting in a data communications network of a parallel computer

Methods, apparatus, and products are disclosed for line-plane broadcasting in a data communications network of a parallel computer, the parallel computer comprising a plurality of compute nodes connected together through the network, the network optimized for point to point data communications and characterized by at least a first dimension, a second dimension, and a third dimension, that include: initiating, by a broadcasting compute node, a broadcast operation, including sending a message to all of the compute nodes along an axis of the first dimension for the network; sending, by each compute node along the axis of the first dimension, the message to all of the compute nodes along an axis of the second dimension for the network; and sending, by each compute node along the axis of the second dimension, the message to all of the compute nodes along an axis of the third dimension for the network.

Archer, Charles J↗

Line-plane broadcasting in a data communications network of a parallel computer

Methods, apparatus, and products are disclosed for line-plane broadcasting in a data communications network of a parallel computer, the parallel computer comprising a plurality of compute nodes connected together through the network, the network optimized for point to point data communications and characterized by at least a first dimension, a second dimension, and a third dimension, that include: initiating, by a broadcasting compute node, a broadcast operation, including sending a message to all of the compute nodes along an axis of the first dimension for the network; sending, by each compute node along the axis of the first dimension, the message to all of the compute nodes along an axis of the second dimension for the network; and sending, by each compute node along the axis of the second dimension, the message to all of the compute nodes along an axis of the third dimension for the network.

97 MATHEMATICS AND COMPUTING↗

Integrated Network Decompositions and Dynamic Programming for Graph Optimization (INDDGO)

The INDDGO software package offers a set of tools for finding exact solutions to graph optimization problems via tree decompositions and dynamic programming algorithms. Currently the framework offers serial and parallel (distributed memory) algorithms for finding tree decompositions and solving the maximum weighted independent set problem. The parallel dynamic programming algorithm is implemented on top of the MADNESS task-based runtime.

Source record↗

A complete implementation of the conjugate gradient algorithm on a reconfigurable supercomputer

The conjugate gradient is a prominent iterative method for solving systems of sparse linear equations. Large-scale scientific applications often utilize a conjugate gradient solver at their computational core. In this paper we present a field programmable gate array (FPGA) based implementation of a double precision, non-preconditioned, conjugate gradient solver for fmite-element or finite-difference methods. OUf work utilizes the SRC Computers, Inc. MAPStation hardware platform along with the 'Carte' software programming environment to ease the programming workload when working with the hybrid (CPUIFPGA) environment. The implementation is designed to handle large sparse matrices of up to order N x N where N <= 116,394, with up to 7 non-zero, 64-bit elements per sparse row. This implementation utilizes an optimized sparse matrix-vector multiply operation which is critical for obtaining high performance. Direct parallel implementations of loop unrolling and loop fusion are utilized to extract performance from the various vector/matrix operations. Rather than utilize the FPGA devices as function off-load accelerators, our implementation uses the FPGAs to implement the core conjugate gradient algorithm. Measured run-time performance data is presented comparing the FPGA implementation to a software-only version showing that the FPGA can outperform processors running up to 30x the clock rate. In conclusion we take a look at the new SRC-7 system and estimate the performance of this algorithm on that architecture.

97 MATHEMATICS AND COMPUTING↗

I/O Bottleneck Detection and Tuning: Connecting the Dots using Interactive Log Analysis

Using parallel file systems efficiently is a tricky problem due to inter-dependencies among multiple layers of I/O software, including high-level I/O libraries (HDF5, netCDF, etc.), MPI-IO, POSIX, and file systems (GPFS, Lustre, etc.). Profiling tools such as Darshan collect traces to help understand the I/O performance behavior. However, there are significant gaps in analyzing the collected traces and then applying tuning options offered by various layers of I/O software. Seeking to connect the dots between I/O bottleneck detection and tuning, we propose DXT Explorer, an interactive log analysis tool. In this paper, we present a case study using our interactive log analysis tool to identify and apply various I/O optimizations. We report an evaluation of performance improvement achieved for four I/O kernels extracted from science applications.

Bez, Jean Luca↗

Interlocked optimization and fast gradient algorithm for a seismic inverse problem

Highlights: {yields} A 2D extension of the 1D nonlinear inversion of well-seismic data is given. {yields} Appropriate regularization yields a well-determined large scale inverse problem. {yields} An interlocked optimization loop acts as an efficient preconditioner. {yields} The adjoint state method is used to compute the misfit function gradient. {yields} Domain decomposition method yields an efficient parallel implementation. - Abstract: We give a nonlinear inverse method for seismic data recorded in a well from sources at several offsets from the borehole in a 2D acoustic framework. Given the velocity field, approximate values of the impedance are recovered. This is a 2D extension of the 1D inversion of vertical seismic profiles . The inverse problem generates a large scale undetermined ill-conditioned problem. Appropriate regularization terms render the problem well-determined. An interlocked optimization algorithm yields an efficient preconditioning. A gradient algorithm based on the adjoint state method and domain decomposition gives a fast parallel numerical method. For a realistic test case, convergence is attained in an acceptable time with 128 processors.

97 MATHEMATICS AND COMPUTING↗

Nonlinear Dynamic Model-Based Multiobjective Sensor Network Design Algorithm for a Plant with an Estimator-Based Control System

Here, a novel sensor network design (SND) algorithm is developed for maximizing process efficiency while minimizing sensor network cost for a nonlinear dynamic process with an estimator-based control system. The multiobjective optimization problem is solved following a lexicographic approach where the process efficiency is maximized first followed by minimization of the sensor network cost. The partial net present value, which combines the capital cost due to the sensor network and the operating cost due to deviation from the optimal efficiency, is proposed as an alternative objective. The unscented Kalman filter is considered as the nonlinear estimator. The large-scale combinatorial optimization problem is solved using a genetic algorithm. The developed SND algorithm is applied to an acid gas removal (AGR) unit as part of an integrated gasification combined cycle (IGCC) power plant with CO 2 capture. Due to the computational expense, a reduced order nonlinear model of the AGR process is identified and parallel computation is performed during implementation.

Lexicographic optimization↗

Parallel variable selection for effective performance prediction

Large data analysis problems often involve a large number of variables, and the corresponding analysis algorithms may examine all variable combinations to find the optimal solution. For example, to model the time required to complete a scientific workflow, we need to consider the impact of dozens of parameters. To reduce the model building time and reduce the likelihood of overfitting, we look to variable selection methods to identify the critical variables for the performance model. In this work, we create a combination of variable selection and performance prediction methods that is as effective as the exhaustive search approach when the exhaustive search could be completed in a reasonable amount of time. To handle the cases where the exhaustive search is too time consuming, we develop the parallelized variable selection algorithm. Additionally, we develop a parallel grouping mechanism that further reduces the variable selection time by 70%.As a case study, we exercise the variable selection technique with the performance measurement data from the Palomar Transient Factory (PTF) workflow. The application scientists have determined that about 50 variables and parameters are important to the performance of the workflows. Our tests show that the Sequential Backward Selection algorithm is able to approximate the optimal subset relatively quickly. By reducing the number of variables used to build the model from 50 to 4, we are able to maintain the prediction quality while reducing the model building time by a factor of 6. Using the parallelization and grouping techniques we developed in this work, the variable selection process was reduced from over 18 hours to 15 minutes while ending up with the same variable subset.

Wang, J↗

Investigation of Benchmark $k$ eff Sensitivity and Uncertainty for 239 Pu fission in Specific Energy Ranges

Nuclear data at intermediate energies (from 1 to 100s of keV) are evaluated based on scarce differential data and theory unable to capture physics’ expected structure. There is also a lack of integral data. This is a known deficiency and is challenging to address. Calculated effective multiplication factor, k eff , values for intermediate energy experiments are ~25× further from experiment than for fast energies and are often well outside the experimental uncertainties. The goal of the PARADIGM (PARallel Approach of Differential and InteGral Measurements) project is to significantly re duce the uncertainties of intermediate energy nuclear data for 239 Pu. To this end, PARADIGM simultaneously optimizes experiments at both the Los Alamos Neutron Science Center (LANSCE) and National Criticality Experiments Research Center (NCERC). The combined set of data will inform new intermediate-energy nuclear data. By execution of differential and integral experiments, establishment of new theory, and undertaking nuclear data evaluation in parallel, the timeline to deliver improved nuclear data to users will be reduced significantly that is to three years. For the PARADIGM project, it was decided to optimize an integral experiment for two neutron energy ranges, within the full intermediate energy range. The low energy range goes from 1 to 30 keV, while the higher energy range goes from 30 to 600 keV. This work focuses on nuclear data sensitivities and uncertainties for 239 Pu fission for existing experiments in the International Criticality Safety Benchmark Evaluation Project (ICSBEP). When designing new experiments, it is important to understand what benchmarks currently exist. For a more traditional experiment design (in which a specific application model(s) exists), comparisons would be made between the application model(s) and existing benchmarks. For PARADIGM, there is no specific application model, but instead the specific nuclear data reaction and energy ranges of interest can be explored for existing benchmarks.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗

Parallel Implementation of the Recursive Approximation of an Unsupervised Hierarchical Segmentation Algorithm

The hierarchical image segmentation algorithm (referred to as HSEG) is a hybrid of hierarchical step-wise optimization (HSWO) and constrained spectral clustering that produces a hierarchical set of image segmentations. HSWO is an iterative approach to region grooving segmentation in which the optimal image segmentation is found at N(sub R) regions, given a segmentation at N(sub R+1) regions. HSEG's addition of constrained spectral clustering makes it a computationally intensive algorithm, for all but, the smallest of images. To counteract this, a computationally efficient recursive approximation of HSEG (called RHSEG) has been devised. Further improvements in processing speed are obtained through a parallel implementation of RHSEG. This chapter describes this parallel implementation and demonstrates its computational efficiency on a Landsat Thematic Mapper test scene.

Tilton, James C.↗

Scaling Semantic Graph Databases in Size and Performance

In this paper we present SGEM, a full software system for accelerating large-scale semantic graph databases on commodity clusters. Unlike current approaches, SGEM addresses semantic graph databases by only employing graph methods at all the levels of the stack. On one hand, this allows exploiting the space efficiency of graph data structures and the inherent parallelism of graph algorithms. These features adapt well to the increasing system memory and core counts of modern commodity clusters. On the other hand, however, these systems are optimized for regular computation and batched data transfers, while graph methods usually are irregular and generate fine-grained data accesses with poor spatial and temporal locality. Our framework comprises a SPARQL to data parallel C compiler, a library of parallel graph methods and a custom, multithreaded runtime system. We introduce our stack, motivate its advantages with respect to other solutions and show how we solved the challenges posed by irregular behaviors. We present the result of our software stack on the Berlin SPARQL benchmarks with datasets up to 10 billion triples (a triple corresponds to a graph edge), demonstrating scaling in dataset size and in performance as more nodes are added to the cluster.

Aggregation↗