Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “parallel performance”

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 865 records · Page 48
All sourcesNASA NTRSDOE OSTI
NASA NTRS2001

Hiding the Disk and Network Latency of Out-of-Core Visualization

This paper describes an algorithm that improves the performance of application-controlled demand paging for out-of-core visualization by hiding the latency of reading data from both local disks or disks on remote servers. The performance improvements come from better overlapping the computation with the page reading process, and by performing multiple page reads in parallel. The paper includes measurements that show that the new multithreaded paging algorithm decreases the time needed to compute visualizations by one third when using one processor and reading data from local disk. The time needed when using one processor and reading data from remote disk decreased by two thirds. Visualization runs using data from remote disk actually ran faster than ones using data from local disk because the remote runs were able to make use of the remote server's high performance disk array.

Ellsworth, David↗
NASA NTRS1987

Algorithms and programming tools for image processing on the MPP:3

This is the third and final report on the work done for NASA Grant 5-403 on Algorithms and Programming Tools for Image Processing on the MPP:3. All the work done for this grant is summarized in the introduction. Work done since August 1986 is reported in detail. Research for this grant falls under the following headings: (1) fundamental algorithms for the MPP; (2) programming utilities for the MPP; (3) the Parallel Pascal Development System; and (4) performance analysis. In this report, the results of two efforts are reported: region growing, and performance analysis of important characteristic algorithms. In each case, timing results from MPP implementations are included. A paper is included in which parallel algorithms for region growing on the MPP is discussed. These algorithms permit different sized regions to be merged in parallel. Details on the implementation and peformance of several important MPP algorithms are given. These include a number of standard permutations, the FFT, convolution, arbitrary data mappings, image warping, and pyramid operations, all of which have been implemented on the MPP. The permutation and image warping functions have been included in the standard development system library.

Reeves, Anthony P.↗
NASA NTRS1990

Supercomputing systems - A projection to 2000

Advances in computer architecture, computer science, computational methods, and constituent technologies are expected to lead to significant advances in the performance of scientific supercomputing system capabilities over the next decade. By the year 2000, single 1-in-sq dies are projected to incorporate four processors, each of which would be operating faster than 750 million instructions per second (MIPS) for a total on-chip processing performance in excess of 2000 MIPS. Scalable parallel processors can be expected to contain thousands of such multiple processor chips. In general, semiconductor performance advances appear to change about one order of magnitude every five years. Rotating magnetic memory and communications technology are not advancing as rapidly, with the result that the allocation of functions within the system configurations fo future supercomputer systems will require important changes. Availability of massively parallel heterogeneous processing capabilities should be a catalyst leading to new approaches for applications.

Lundstrom, S. F.↗
NASA NTRS1994

Implementation of a fully-balanced periodic tridiagonal solver on a parallel distributed memory architecture

While parallel computers offer significant computational performance, it is generally necessary to evaluate several programming strategies. Two programming strategies for a fairly common problem - a periodic tridiagonal solver - are developed and evaluated. Simple model calculations as well as timing results are presented to evaluate the various strategies. The particular tridiagonal solver evaluated is used in many computational fluid dynamic simulation codes. The feature that makes this algorithm unique is that these simulation codes usually require simultaneous solutions for multiple right-hand-sides (RHS) of the system of equations. Each RHS solutions is independent and thus can be computed in parallel. Thus a Gaussian elimination type algorithm can be used in a parallel computation and the more complicated approaches such as cyclic reduction are not required. The two strategies are a transpose strategy and a distributed solver strategy. For the transpose strategy, the data is moved so that a subset of all the RHS problems is solved on each of the several processors. This usually requires significant data movement between processor memories across a network. The second strategy attempts to have the algorithm allow the data across processor boundaries in a chained manner. This usually requires significantly less data movement. An approach to accomplish this second strategy in a near-perfect load-balanced manner is developed. In addition, an algorithm will be shown to directly transform a sequential Gaussian elimination type algorithm into the parallel chained, load-balanced algorithm.

Eidson, T. M.↗
DOE OSTI2025

Asynchronous-many-task systems: Challenges and opportunities - Scaling an AMR astrophysics code on exascale machines using Kokkos and HPX

Dynamic and adaptive mesh refinement is pivotal in high-resolution, multi-physics, multi-model simulations, necessitating precise physics resolution in localized areas across expansive domains. Today’s supercomputers’ extreme heterogeneity presents a significant challenge for dynamically adaptive codes, highlighting the importance of achieving performance portability at scale. Our research focuses on astrophysical simulations, particularly stellar mergers, to elucidate early universe dynamics. Here, we present Octo-Tiger, leveraging Kokkos, HPX, and SIMD for portable performance at scale in complex, massively parallel adaptive multi-physics simulations. Octo-Tiger supports diverse processors, accelerators, and network backends. Experiments demonstrate exceptional scalability across several heterogeneous supercomputers including Perlmutter, Frontier, and Fugaku, encompassing major GPU architectures and x86, ARM, and RISC-V CPUs. Parallel efficiency of 47.59% (110,080 cores and 6880 hybrid A100 GPUs) on a full-system run on Perlmutter (26% HPCG peak performance) and 51.37% (using 32,768 cores and 2048 MI250X) on Frontier are achieved.

97 MATHEMATICS AND COMPUTING↗
NASA NTRS1993

Automatic array alignment in data-parallel programs

FORTRAN 90 and other data-parallel languages express parallelism in the form of operations on data aggregates such as arrays. Misalignment of the operands of an array operation can reduce program performance on a distributed-memory parallel machine by requiring nonlocal data accesses. Determining array alignments that reduce communication is therefore a key issue in compiling such languages. We present a framework for the automatic determination of array alignments in array-based, data-parallel languages. Our language model handles array sectioning, reductions, spreads, transpositions, and masked operations. We decompose alignment functions into three constituents: axis, stride, and offset. For each of these subproblems, we show how to solve the alignment problem for a basic block of code, possibly containing common subexpressions. Alignments are generated for all array objects in the code, both named program variables and intermediate results. We assign computation to processors by virtue of explicit alignment of all temporaries; the resulting work assignment is in general better than that provided by the 'owner-computes' rule. Finally, we present some ideas for dealing with control flow, replication, and dynamic alignments that depend on loop induction variables.

Chatterjee, Siddhartha↗
NASA NTRSDate not supplied

Multiscale and Multifidelity Modeling of a 3D Woven Composite Thermal Protection System

Complex three-dimensional (3D) woven composites have been considered by multiple NASA projects in recent years as a means of offering improved mechanical and thermal performance over traditional laminated composite systems. Parallel efforts have focused on developing simulation capabilities for these systems, which have traditionally and heavily relied on experimental testing to evaluate composite performance. One system is the Heatshield for Extreme Entry Environment Technology (HEEET), which is being considered for the thermal protection system on reentry spacecraft. Optical microscopy was used to characterize the blended carbon and phenolic fiber tows. A section of HEEET insulation layer was imaged with high-resolution micro-computed tomography (microCT) and segmented to separate individual tows, porous matrix, and voids. These data were used to develop multiscale thermomechanical computational models within the NASA Multiscale Analysis Tool (NASMAT). Two NASMAT modeling approaches were considered to capture the details of the 3D woven architecture: a coarse model appropriate for inclusion in multiscale structural analyses and a high-fidelity model created by downsampling the microCT data. Both elastic and thermal properties were computed and compared. The feasibility and challenges associated with modeling complex, hybrid 3D woven composites were also addressed.

NASMAT↗
DOE OSTI2021

Implementing a neural network interatomic model with performance portability for emerging exascale architectures

The two main thrusts of computational science are increasingly accurate predictions and faster calculations; to this end, the zeitgeist in molecular dynamics (MD) simulations is pursuing machine learned and data driven interatomic models, e.g. neural network potentials, and novel hardware architectures, e.g. GPUs. Current implementations of neural network potentials are orders of magnitude slower than traditional interatomic models and while looming exascale computing offers the ability to run large, accurate simulations with these models, achieving portable performance for MD with new and varied exascale hardware requires rethinking traditional algorithms, using novel data structures, and library solutions. We re-implement a neural network interatomic model in CabanaMD, an MD proxy application, built on libraries developed for performance portability. Our implementation shows significantly improved thread scaling in this complex kernel as compared to a current LAMMPS implementation, across both strong and weak scaling. Our single-source solution enables simulations up to 20 million atoms on a single CPU node and 4 million atoms with improved performance on a single GPU. Furthermore, we also explore parallelism and data layout choices (using flexible data structures called AoSoAs) and their effect on performance, seeing up to ~50% and ~5% improvements in performance on a GPU by choosing the right level of parallelism and data layout respectively.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2023

Multiphysics Modeling of Impurity Transport for FNSF Startup Scenario with ERO2.0

This paper focuses on performing a multiphysics study using the ERO2.0 and UEDGE codes for two standard double null configurations for the Fusion Nuclear Science Facility: (a) 100% recycling and (b) 99% recycling. Results show that the main contributor to tungsten erosion along the divertor plates is impurities from the midplane waveguides. In addition, the standard high-recycling case (100% recycling) shows a significantly higher buildup of impurities along the divertor tiles during the startup phase, which can lead to a higher increase of energy loss in the plasma during steady-state operation. Last, for high recycling, anomalous diffusion can dominate over parallel field diffusion. The work performed in this study can be iteratively applied to a full operation scenario with additional physics such as those from neutrals, wall shaping, and additional external fields.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
NASA NTRS2002

Application of Reconfigurable Computing Technology to Multi-KiloHertz Micro-Laser Altimeter (MMLA) Data Processing

The Multi-KiloHertz Micro-Laser Altimeter (MMLA) is an aircraft based instrument developed by NASA Goddard Space Flight Center with several potential spaceflight applications. This presentation describes how reconfigurable computing technology was employed to perform MMLA signal extraction in real-time under realistic operating constraints. The MMLA is a "single-photon-counting" airborne laser altimeter that is used to measure land surface features such as topography and vegetation canopy height. This instrument has to date flown a number of times aboard the NASA P3 aircraft acquiring data at a number of sites in the Mid-Atlantic region. This instrument pulses a relatively low-powered laser at a very high rate (10 kHz) and then measures the time-of-flight of discrete returns from the target surface. The instrument then bins these measurements into a two-dimensional array (vertical height vs. horizontal ground track) and selects the most likely signal path through the array. Return data that does not correspond to the selected signal path are classified as noise returns and are then discarded. The MMLA signal extraction algorithm is very compute intensive in that a score must be computed for every possible path through the two dimensional array in order to select the most likely signal path. Given a typical array size with 50 x 6, up to 33 arrays must be processed per second. And for each of these arrays, roughly 12,000 individual paths must be scored. Furthermore, the number of paths increases exponentially with the horizontal size of the array, and linearly with the vertical size. Yet, increasing the horizontal and vertical sizes of the array offer science advantages such as improved range, resolution, and noise rejection. Due to the volume of return data and the compute intensive signal extraction algorithm, the existing PC-based MMLA data system has been unable to perform signal extraction in real-time unless the array is limited in size to one column, This limits the ability of the MMLA to operate in environments with sparse signal returns and a high number of noise return. However, under an IR&D project, an FPGA-based, reconfigurable computing data system has been developed that has been demonstrated to perform real-time signal extraction under realistic operating constraints. This reconfigurable data system is based on the commercially available Firebird Board from Annapolis Microsystems. This PCI board consists of a Xilinx Virtex 2000E FPGA along with 36 MB of SRAM arranged in five separately addressable banks. This board is housed in a rackmount PC with dual 850MHz Pentium processors running the Windows 2000 operating system. This data system performs all signal extraction in hardware on the Firebird, but also runs the existing "software based" signal extraction in tandem for comparison purposes. Using a relatively small amount of the Virtex XCV2000E resources, the reconfigurable data system has demonstrated to improve performance improvement over the existing software based data system by an order of magnitude. Performance could be further improved by employing parallelism. Ground testing and a preliminary engineering test flight aboard the NASA P3 has been performed, during which the reconfigurable data system has been demonstrated to match the results of the existing data system.

Powell, Wesley↗
NASA NTRS1992

Two dimensional PIC simulations of plasma heating by the dissipation of Alfven waves

Two dimensional plasma particle simulations of the evolution of large amplitude circularly polarized Alfven waves propagating parallel to the magnetic field show that the waves decay via both one- and two- dimensional parametric decay instabilities. For parameters studied, one-dimensional processes dominate the simulations, but two-dimensional decay processes, including the recently predicted filamentation instability are also observed. The daughter waves generated by the parametric decay are primarily damped by the ions, leading to ion heating. The parametric decay processes efficiently convert the ordered fluid ion motion in the Alfven wave into ion thermal energy. These processes may be important for the dissipation of Alfven waves in the solar wind, the corona and other space plasma environments. The computations were performed on the Intel Touchstone parallel supercomputer.

Liewer, P. C.↗
NASA NTRS2013

Profiling and Improving I/O Performance of a Large-Scale Climate Scientific Application

Exascale computing systems are soon to emerge, which will pose great challenges on the huge gap between computing and I/O performance. Many large-scale scientific applications play an important role in our daily life. The huge amounts of data generated by such applications require highly parallel and efficient I/O management policies. In this paper, we adopt a mission-critical scientific application, GEOS-5, as a case to profile and analyze the communication and I/O issues that are preventing applications from fully utilizing the underlying parallel storage systems. Through in-detail architectural and experimental characterization, we observe that current legacy I/O schemes incur significant network communication overheads and are unable to fully parallelize the data access, thus degrading applications' I/O performance and scalability. To address these inefficiencies, we redesign its I/O framework along with a set of parallel I/O techniques to achieve high scalability and performance. Evaluation results on the NASA discover cluster show that our optimization of GEOS-5 with ADIOS has led to significant performance improvements compared to the original GEOS-5 implementation.

Exascale↗
NASA NTRS2001

Accelerating Demand Paging for Local and Remote Out-of-Core Visualization

This paper describes a new algorithm that improves the performance of application-controlled demand paging for the out-of-core visualization of data sets that are on either local disks or disks on remote servers. The performance improvements come from better overlapping the computation with the page reading process, and by performing multiple page reads in parallel. The new algorithm can be applied to many different visualization algorithms since application-controlled demand paging is not specific to any visualization algorithm. The paper includes measurements that show that the new multi-threaded paging algorithm decreases the time needed to compute visualizations by one third when using one processor and reading data from local disk. The time needed when using one processor and reading data from remote disk decreased by up to 60%. Visualization runs using data from remote disk ran about as fast as ones using data from local disk because the remote runs were able to make use of the remote server's high performance disk array.

Ellsworth, David↗
DOE OSTI2026

Tula: Optimizing Time, Cost, and Generalization in Distributed Large-Batch Training

Distributed training increases the number of batches processed per iteration either by scaling-out (adding more nodes) or scaling-up (increasing the batch-size). However, the largest configuration does not necessarily yield the best performance. Horizontal scaling introduces additional communication overhead, while vertical scaling is constrained by computation cost and device memory limits. Thus, simply increasing the batch-size leads to diminishing returns: training time and cost decrease initially but eventually plateaus, creating a knee-point in the time/cost vs. batch-size pareto curve. The optimal batch-size therefore depends on the underlying model, data and available compute resources. Large batches also suffer from worse model quality due to the well-known “generalization gap”. In this paper, we present Tula, an online service that automatically optimizes time, cost, and convergence quality for large-batch training of convolutional models. It combines parallel-systems modeling with statistical performance prediction to identify the optimal batchsize. Tula predicts training time and cost within 7.5−14% error across multiple models, and achieves up to 20× overall speedup and improves test accuracy by ≈9% on average over standard large-batch training on various vision tasks, thus successfully mitigating the generalization gap and accelerating training at the same time.

Tyagi, Sahil [ORNL] (ORCID:0009000783144745)↗
DOE OSTI2024

VAN-DAMME: GPU-accelerated and symmetry-assisted quantum optimal control of multi-qubit systems

We present an open-source software package, VAN-DAMME (Versatile Approaches to Numerically Design, Accelerate, and Manipulate Magnetic Excitations), for massively-parallelized quantum optimal control (QOC) calculations of multi-qubit systems. To enable large QOC calculations, the VAN-DAMME software package utilizes symmetry-based techniques with custom GPU-enhanced algorithms. This combined approach allows for the simultaneous computation of hundreds of matrix exponential propagators that efficiently leverage the intra-GPU parallelism found in high-performance GPUs. In addition, to maximize the computational efficiency of the VAN-DAMME code, we carried out several extensive tests on data layout, computational complexity, memory requirements, and performance. These extensive analyses allowed us to develop computationally efficient approaches for evaluating complex-valued matrix exponential propagators based on Padé approximants. To assess the computational performance of our GPU-accelerated VAN-DAMME code, we carried out QOC calculations of systems containing 10 - 15 qubits, which showed that our GPU implementation is 18.4× faster than the corresponding CPU implementation. Our GPU-accelerated enhancements allow efficient calculations of multi-qubit systems, which can be used for the efficient implementation of QOC applications across multiple domains.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2025

Spotlight: efficient automated global optimization in rietveld analysis of diffraction data

Performing reliable Rietveld analysis on tens or hundreds of powder diffraction datasets from parametric or time-resolved experiments often poses a bottleneck in extracting meaningful results from the data. While automated analysis of data has recently been demonstrated, high temperature annealing studies, during which phase transformations occur and lattice parameters may change due to repartitioning of elements, are prime examples where automation by a simple phase identification from a database of room temperature structures or automation by sequential refinements is likely to fail. To enable reliable, efficient, automated Rietveld analysis, we present a Python package named Spotlight , building on established Rietveld packages such as MAUD, GSAS , or GSAS-II , which extends the refinement of best fit parameters to a global optimization using an ensemble of optimizers leveraging hierarchical parallel execution on high-performance computing clusters. Spotlight further enables the efficient design of refinement plans through the iterative automated machine-learning of a surrogate for the refinement on which the global optimizations are performed until results from the surrogate converge to the response surface data. We demonstrate Spotlight with the analysis of uranium molybdenum and Ti–6Al–4V datasets, as well as in two open-source tutorials analyzing aluminium oxide and lead sulphate.

36 MATERIALS SCIENCE↗
NASA NTRS1996

Parallelization of Finite Element Analysis Codes Using Heterogeneous Distributed Computing

Performance gains in computer design are quickly consumed as users seek to analyze larger problems to a higher degree of accuracy. Innovative computational methods, such as parallel and distributed computing, seek to multiply the power of existing hardware technology to satisfy the computational demands of large applications. In the early stages of this project, experiments were performed using two large, coarse-grained applications, CSTEM and METCAN. These applications were parallelized on an Intel iPSC/860 hypercube. It was found that the overall speedup was very low, due to large, inherently sequential code segments present in the applications. The overall execution time T(sub par), of the application is dependent on these sequential segments. If these segments make up a significant fraction of the overall code, the application will have a poor speedup measure.

Ozguner, Fusun↗