Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “distributed parallelization”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

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

At least 847 records · Page 47
All sourcesNASA NTRSDOE OSTI
NASA NTRS1986

Stability analysis of double-peaked proton distribution functions in the solar wind

Twelve proton velocity distribution functions as measured for different solar wind speeds by the plasma analyzer at the Helios spacecraft are modeled analytically by a superposition of two bi-Maxwellians. The stability of these distributions is investigated by solving the linear dispersion equation for parallel propagation of both left- and right-hand polarized modes. Due to the temperature anisotropy of the main proton component the presence of the resonant left-hand ion cyclotron instability is found to be typical for high-speed streams and is independent of an ion beam. The results indicate that the ratio of the beam density to total particle density and the beam drift velocity are generally close to the threshold values necessary to drive a highly unstable nonresonant right-hand mode. This instability is found for one proton distribution and is a consequence of the beam particles. It is concluded that both the ion cyclotron instability and the ion beam instability act as a regulating mechanism on the shape of solar wind double humped proton distribution functions.

Leubner, Manfred P.↗
NASA NTRS1998

Performance Analysis of an Actor-Based Distributed Simulation

Object-oriented design of simulation programs appears to be very attractive because of the natural association of components in the simulated system with objects. There is great potential in distributing the simulation across several computers for the purpose of parallel computation and its consequent handling of larger problems in less elapsed time. One approach to such a design is to use "actors", that is, active objects with their own thread of control. Because these objects execute concurrently, communication is via messages. This is in contrast to an object-oriented design using passive objects where communication between objects is via method calls (direct calls when they are in the same address space and remote procedure calls when they are in different address spaces or different machines). This paper describes a performance analysis program for the evaluation of a design for distributed simulations based upon actors.

Schoeffler, James D.↗
NASA NTRS1999

Evidence of Component Merging Equatorward of the Cusp

The Polar spacecraft passed through a region near the dayside magnetopause on May 29, 1996, at a geocentric distance of approx. 8 R (sub E) and high, northern magnetic latitudes. The interplanetary magnetic field (IMF) was northward during the pass. Data from the Thermal Ion Dynamics Experiment revealed the existence of low-speed (approx. 50 km s (exp-1)) ion D-shaped distributions mixed with cold ions (approx. 2 eV) over a period of 2.5 hours. These ions were traveling parallel to the magnetic field toward the Northern Hemisphere ionosphere and were convecting primarily eastward. The D-shaped distributions are distinct from a convecting Maxwellian and, along with the magnetic field direction, are taken as evidence that the spacecraft was inside the magnetosphere and not in the magnetosheath. Furthermore, the absence of ions in the antiparallel direction is taken as evidence that low-shear merging was occurring at a location southward of the spacecraft and equatorward of the Southern Hemisphere cusp. The cold ions were of ionospheric origin, with initially slow field-aligned speeds, which were accelerated upon reflection from the magnetopause. These observations provide significant new evidence consistent with component magnetic merging sites equatorward of the cusp for northward IMF.

Chandler, M. O.↗
DOE OSTI2023

Evaluation of pre-training large language models on leadership-class supercomputers

Large language models (LLMs) have arisen rapidly to the center stage of artificial intelligence as the foundation models applicable to many downstream learning tasks. However, how to effectively build, train, and serve such models for many high-stake and first-principle-based scientific use cases are both of great interests and of great challenges. Moreover, pre-training LLMs with billions or even trillions of parameters can be prohibitively expensive not just for academic institutions, but also for well-funded industrial and government labs. Furthermore, the energy cost and the environmental impact of developing LLMs must be kept in mind. Here, in this work, we conduct a first-of-its-kind performance analysis to understand the time and energy cost of pre-training LLMs on the Department of Energy (DOE)’s leadership-class supercomputers. Employing state-of-the-art distributed training techniques, we evaluate the computational performance of various parallelization approaches at scale for a range of model sizes, and establish a projection model for the cost of full training. Our findings provide baseline results, best practices, and heuristics for pre-training such large models that should be valuable to HPC community at large. We also offer insights and optimization strategies for using the first exascale computing system, Frontier, to train models of the size of GPT-3 and beyond.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2025

Parallel derivative-free optimization for simulation-based design of behind-the-meter energy systems

In this work, the integrated design and dispatch of behind-the-meter or distributed resources (e.g. stationary battery storage and solar PV generation) is considered. A simulation-based framework is employed, generating high-fidelity results with closed-loop predictive control at a fine resolution, at the expense of high computational cost (several minutes to a few hours per design point). To address this challenge, parallel derivative-free design methods are considered. Four methods are compared, including state-of-the-art surrogate-based methods (Radial-Basis Functions and Gaussian processes) and sampling strategies, an evolutionary-based method, and a simple sequential grid refinement method. As a case study, two types of design problem with increasing complexity are considered, namely, the design of behind-the-meter resources (three design variables) and the inclusion of grid capacity (four design variables). The second yields a constrained design problem for which violations can only be determined after solving the computationally expensive simulation. For the three-dimensional case, all methods present a good performance, achieving a solution within 1% of the optimum after the first iteration, with the sequential grid refinement exhibiting the fastest convergence and achieving the best final objective value. This indicates that the parallel evaluation of multiple sampling points may be more important than the choice of method for small decision spaces. For the four-dimensional constrained case, the Genetic Algorithm presents the best tradeoff between performance and computational effort, while the rough objective function terrain generated by constraint violation penalties reduces the performance of surrogate-based methods. Contour plots with flat regions indicate flexibility in the optimal design and highlight the importance of characterizing the solution space.

24 POWER TRANSMISSION AND DISTRIBUTION↗
DOE OSTI2026

Mode Multiplexing for Scalable Cavity-Enhanced Operations in Neutral-Atom Arrays

Neutral-atom arrays provide a versatile platform for quantum information processing. However, in large-scale arrays, efficient photon collection remains a bottleneck for key tasks such as fast, nondestructive qubit readout and remote entanglement distribution. We propose a cavity-based approach that enables fast, parallel operations over many atoms using multiple modes of a single optical cavity. By selectively shifting the relevant atomic transitions, each atom can be coupled to a distinct cavity mode, allowing independent simultaneous processing. We present practical system designs that support cavity-mode multiplexing with up to 50 modes, enabling rapid mid-circuit syndrome extraction and significantly enhancing entanglement distribution rates between remote atom arrays. This approach offers a scalable solution to core challenges in neutral-atom arrays, advancing the development of practical quantum technologies.

Aqua, Ziv [Massachusetts Institute of Technology (↗
DOE OSTI2024

Parallel Runtime Interface for Fortran (PRIF): A Multi-Image Solution for LLVM Flang

Fortran compilers that provide support for Fortran’s native parallel features often do so with a runtime library that depends on details of both the compiler implementation and the communication library, while others provide limited or no support at all. This paper introduces a new generalized interface that is both compiler- and runtime-library-agnostic, providing flexibility while fully supporting all of Fortran’s parallel features. The Parallel Runtime Interface for Fortran (PRIF) was developed to be portable across shared- and distributed-memory systems, with varying operating systems, toolchains and architectures. It achieves this by defining a set of Fortran procedures corresponding to each of the parallel features defined in the Fortran standard that may be invoked by a Fortran compiler and implemented by a runtime library. PRIF aims to be used as the solution for LLVM Flang to provide parallel Fortran support. This paper also briefly describes our PRIF prototype implementation: Caffeine.

Bonachea, Dan↗
NASA NTRS1993

Distributed computing feasibility in a non-dedicated homogeneous distributed system

The low cost and availability of clusters of workstations have lead researchers to re-explore distributed computing using independent workstations. This approach may provide better cost/performance than tightly coupled multiprocessors. In practice, this approach often utilizes wasted cycles to run parallel jobs. The feasibility of such a non-dedicated parallel processing environment assuming workstation processes have preemptive priority over parallel tasks is addressed. An analytical model is developed to predict parallel job response times. Our model provides insight into how significantly workstation owner interference degrades parallel program performance. A new term task ratio, which relates the parallel task demand to the mean service demand of nonparallel workstation processes, is introduced. It was proposed that task ratio is a useful metric for determining how large the demand of a parallel applications must be in order to make efficient use of a non-dedicated distributed system.

Leutenegger, Scott T.↗
DOE OSTI2024

Parallel computing for power system climate resiliency: Solving a large-scale stochastic capacity expansion problem with mpi-sppy

Here we propose a nodal stochastic generation and transmission expansion planning model that incorporates the output from high-resolution global climate models through load and generation availability scenarios. We implement our model in Pyomo and perform computational studies on a realistically-sized test case of the California electric grid in a high performance computing environment. We propose model reformulations and algorithm tuning to efficiently solve this large problem using a variant of the Progressive Hedging Algorithm. We utilize the parallelization capabilities and overall versatility of mpi-sppy, exploiting its hub-and-spoke architecture to concurrently obtain inner and outer bounds on an optimal expansion plan. Initial results show that instances with 360 representative days on a system with over 8,000 buses can be solved to within 5% of optimality in under 4 h of wall clock time, a first step towards solving a large-scale power system expansion planning problem across a wide range of climate-informed operational scenarios.

24 POWER TRANSMISSION AND DISTRIBUTION↗
NASA NTRS1975

Analysis of liquid-metal-jet impingement cooling in a corner region and for a row of jets

A conformal mapping method was used to analyze liquid-metal-jet impingement heat transfer. The jet flow region and energy equation are transformed to correspond to uniform flow in a parallel plate channel with nonuniform heat addition along a portion of one wall. The exact solution for the wall-temperature distribution was obtained in the transformed channel, and the results are mapped back into the physical plane. Two geometries are analyzed. One is for a single slot jet directed either into an interior corner formed by two flat plates, or over the external sides of the corner; the flat plates are uniformly heated, and the corner can have various included angles. The heat-transfer coefficient at the stagnation point at the apex of the plates is obtained as a function of the corner angle, and temperature distributions are calculated along the heated walls. The second geometry is an infinite row of uniformly spaced parallel slot jets impinging normally against a uniformly heated plate. The heat-transfer behavior is obtained as a function of the spacing between the jets. Results are given for several jet Peclet numbers from 5 to 50.

Siegel, R.↗
NASA NTRS1989

NETRA: A parallel architecture for integrated vision systems 2: Algorithms and performance evaluation

In part 1 architecture of NETRA is presented. A performance evaluation of NETRA using several common vision algorithms is also presented. Performance of algorithms when they are mapped on one cluster is described. It is shown that SIMD, MIMD, and systolic algorithms can be easily mapped onto processor clusters, and almost linear speedups are possible. For some algorithms, analytical performance results are compared with implementation performance results. It is observed that the analysis is very accurate. Performance analysis of parallel algorithms when mapped across clusters is presented. Mappings across clusters illustrate the importance and use of shared as well as distributed memory in achieving high performance. The parameters for evaluation are derived from the characteristics of the parallel algorithms, and these parameters are used to evaluate the alternative communication strategies in NETRA. Furthermore, the effect of communication interference from other processors in the system on the execution of an algorithm is studied. Using the analysis, performance of many algorithms with different characteristics is presented. It is observed that if communication speeds are matched with the computation speeds, good speedups are possible when algorithms are mapped across clusters.

Choudhary, Alok N.↗
NASA NTRS1988

Structure of scintillations in Neptune's occultation shadow

An exceptionally high-quality data set from a Neptune occultation is used here to derive a number of new results about the statistical properties of the fluctuations of the intensity distribution in various parts of Neptune's occultation shadow. An approximate numerical ray-tracing model which successfully accounts for many of the qualitative aspects of the observed intensity fluctuation distribution is introduced. Strong refractive scintillation is simulated by including the effects of 'turbulence' with projected atmospheric properties allowed to vary in both the direction perpendicular and parallel to the limb, and an explicit two-dimensional picture of a typical intensity distribution throughout an occulting planet's shadow is presented. The results confirm the existence of highly anisotropic turbulence.

Hubbard, W. B.↗
NASA NTRS1991

Aeroacoustic and aerodynamic applications of the theory of nonequilibrium thermodynamics

Recent developments in the field of nonequilibrium thermodynamics associated with viscous flows are examined and related to developments to the understanding of specific phenomena in aerodynamics and aeroacoustics. A key element of the nonequilibrium theory is the principle of minimum entropy production rate for steady dissipative processes near equilibrium, and variational calculus is used to apply this principle to several examples of viscous flow. A review of nonequilibrium thermodynamics and its role in fluid motion are presented. Several formulations are presented of the local entropy production rate and the local energy dissipation rate, two quantities that are of central importance to the theory. These expressions and the principle of minimum entropy production rate for steady viscous flows are used to identify parallel-wall channel flow and irrotational flow as having minimally dissipative velocity distributions. Features of irrotational, steady, viscous flow near an airfoil, such as the effect of trailing-edge radius on circulation, are also found to be compatible with the minimum principle. Finally, the minimum principle is used to interpret the stability of infinitesimal and finite amplitude disturbances in an initially laminar, parallel shear flow, with results that are consistent with experiment and linearized hydrodynamic stability theory. These results suggest that a thermodynamic approach may be useful in unifying the understanding of many diverse phenomena in aerodynamics and aeroacoustics.

Horne, W. Clifton↗
DOE OSTI2025

Initial position optimization in molecular dynamics simulations for a Coulomb system

A new algorithm for molecular dynamics (MD) simulations is developed to optimize plasma particle distributions at given initial temperatures. By combining velocity scaling and reassignment, the method effectively eliminates the initial rise and oscillation in temperatures observed with randomly distributed positions. These rises and oscillations are undesired numerical artifacts observed in conventional plasma MD simulations, arising from unoptimized particle positions. The algorithm demonstrates temperature relaxation without initial rises or oscillations, as well as precise flow velocity relaxation, enabling accurate measurement of relaxation times. The code is accelerated using graphics processing units for parallel processing, enhancing the study of plasma dynamics. The proposed method for distributing physically valid particles in MD simulations enables accurate studies of intrinsic collision processes in plasmas, including the dynamics of strongly coupled plasmas, plasma–wave interactions, and transport phenomena in magnetized plasmas. The paper concludes with a discussion of potential applications and future enhancements to the algorithm.

Jo, Jawon (ORCID:0009000924193285)↗
NASA NTRS1987

Mapping between parallel processor structures and programs

This paper reports some ongoing research efforts at Stanford in allocation of parallel processing resources. Both processor structures and program structures have their own characteristics. Resource allocation binds the two structures during program execution. The mapping problem determines what processor structure and program structure may be combined to obtain maximum speedup. Three approaches to this mapping problem are considered. Two important factors, granularity and interaction delay, are also considered. A new hierarchical approach to structure definition is outlined. Effective and efficient tools are necessary for the study of the mapping problem. A fast turn-around simulation environment developed for investigating partition strategies for distributed computations and a computationally efficient method to predict performance of parallel processor structures are described.

Ngai, Tin-Fook↗
NASA NTRS2008

Advanced Precipitation Radar Antenna to Measure Rainfall From Space

To support NASA s planned 20-year mission to provide sustained global precipitation measurement (EOS-9 Global Precipitation Measurement (GPM)), a deployable antenna has been explored with an inflatable thin-membrane structure. This design uses a 5.3 5.3-m inflatable parabolic reflector with the electronically scanned, dual-frequency phased array feeds to provide improved rainfall measurements at 2.0-km horizontal resolution over a cross-track scan range of up to 37 , necessary for resolving intense, isolated storm cells and for reducing the beam-filling and spatial sampling errors. The two matched radar beams at the two frequencies (Ku and Ka bands) will allow unambiguous retrieval of the parameters in raindrop size distribution. The antenna is inflatable, using rigidizable booms, deployable chain-link supports with prescribed curvatures, a smooth, thin-membrane reflecting surface, and an offset feed technique to achieve the precision surface tolerance (0.2 mm RMS) for meeting the low-sidelobe requirement. The cylindrical parabolic offset-feed reflector augmented with two linear phased array feeds achieves dual-frequency shared-aperture with wide-angle beam scanning and very low sidelobe level of -30 dB. Very long Ku and Ka band microstrip feed arrays incorporating a combination of parallel and series power divider lines with cosine-over-pedestal distribution also augment the sidelobe level and beam scan. This design reduces antenna mass and launch vehicle stowage volume. The Ku and Ka band feed arrays are needed to achieve the required cross-track beam scanning. To demonstrate the inflatable cylindrical reflector with two linear polarizations (V and H), and two beam directions (0deg and 30deg), each frequency band has four individual microstrip array designs. The Ku-band array has a total of 166x2 elements and the Ka-band has 166x4 elements with both bands having element spacing about 0.65 lambda(sub 0). The cylindrical reflector with offset linear array feeds reduces the complexity from "NxN" transmit/receive (T/R) modules of a conventional planar-phased array to just "N" T/R modules. The antenna uses T/R modules with electronic phase-shifters for beam steering. The offset reflector does not provide poor cross-polarization like a double- curved offset reflector would, and it allows the wide scan angle in one plane required by the mission. Also, the cylindrical reflector with two linear array feeds provides dual-frequency performance with a single, shared aperture. The aperture comprises a reflective surface with a focal length of 1.89 m and is made from aluminized Kapton film. The reflective surface is of uniform thickness in the range of a few thousandths of an inch and is attached to the chain-link support structure via an adjustable suspension system. The film aperture rolls up, together with the chain-link structure, for launch and can be deployed in space by the deployment of the chain-link structure.

Rahmat-Samii, Yahya↗