Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “parallel, scaling”

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 1,675 records · Page 93

PROTON HEATING BY NONLINEAR FIELD-ALIGNED ALFVEN WAVES IN SOLAR CORONAL HOLES

Field-aligned Alfven waves are often viewed as a source of the proton heating that accelerates the fast solar wind. However, the energy that they can inject into the protons in the limit of cyclotron-resonant quasi-linear diffusion is insufficient to account for the observed acceleration. To test the validity of this limit in coronal holes, nonlinear Alfven waves are modeled using a hybrid code. It is found that the nonlinearity is particularly strong when the intensity of antisunward-propagating waves is comparable to that of sunward waves. The sunward waves can be generated by the proton distribution as it evolves with the heliocentric distance. The ponderomotive force and beat interaction are identified as the most important nonlinear effects. The nonlinearity of the field-aligned Alfven waves produces density fluctuations. In the simulations, the amplitude of the density fluctuations was kept within the observed constraints from the interplanetary scintillation measurements in the corona. In this case, the characteristic time of the proton heating is almost 2 orders of magnitude smaller than the solar wind expansion time. Therefore, it can contribute to the energization of the solar wind on the global scale. The nonlinear wave damping operating alone cannot be responsible for the energization because it only causes particle diffusion parallel to the magnetic field. However, it can relax the limitation on the perpendicular diffusion imposed by the cyclotron resonance condition. The nonlinear damping combined with the linear one can then inject the additional thermal energy needed to accelerate the solar wind.

99 GENERAL AND MISCELLANEOUS↗

Scaling Linear Algebra Kernels using Remote Memory Access

This paper describes the scalability of linear algebra kernels based on remote memory access approach. The current approach differs from the other linear algebra algorithms by the explicit use of shared memory and remote memory access (RMA) communication rather than message passing. It is suitable for clusters and scalable shared memory systems. The experimental results on large scale systems (Linux-Infiniband cluster, Cray XT) demonstrate consistent performance advantages over ScaLAPACK suite, the leading implementation of parallel linear algebra algorithms used today. For example, on a Cray XT4 for a matrix size of 102400, our RMA-based matrix multiplication achieved over 55 teraflops while ScaLAPACK’s pdgemm measured close to 42 teraflops on 10000 processes.

99 GENERAL AND MISCELLANEOUS↗

The SprayList: a scalable relaxed priority queue

High-performance concurrent priority queues are essential for applications such as task scheduling and discrete event simulation. Unfortunately, even the best performing implementations do not scale past a number of threads in the single digits. This is because of the sequential bottleneck in accessing the elements at the head of the queue in order to perform a DeleteMin operation. In this paper, we present the SprayList, a scalable priority queue with relaxed ordering semantics. Starting from a non-blocking SkipList, the main innovation behind our design is that the DeleteMin operations avoid a sequential bottleneck by ``spraying'' themselves onto the head of the SkipList list in a coordinated fashion. The spraying is implemented using a carefully designed random walk, so that DeleteMin returns an element among the first O(p log^3 p) in the list, with high probability, where p is the number of threads. We prove that the running time of a DeleteMin operation is O(log^3 p), with high probability, independent of the size of the list. Our experiments show that the relaxed semantics allow the data structure to scale for high thread counts, comparable to a classic unordered SkipList. As a result, we observe that, for reasonably parallel workloads, the scalability benefits of relaxation considerably outweigh the additional work due to out-of-order execution.

97 MATHEMATICS AND COMPUTING↗

ExaTN

A challenging scientific problem that will benefit from exascale computing is predictive simulation of two- and three-dimensional quantum many-body Hamiltonians in condensed matter physics, together with the related strong entanglement problems in quantum chemistry and quantum computing. Efficient numerical simulations will require new insights into wavefunction compression algorithms that scale on extremely heterogeneous HPC architectures. To address this need, we propose to develop a scalable math library of parallel multi-linear algebraic numerical primitives for hierarchical tensor compression. Our library will subsequently be used for building an efficient, performance-portable HPC framework for simulating strongly correlated quantum systems. In particular, our will will be the first to deliver a massively parallel implementation of the multiscale entanglement renormalization ansatz (MERA), a hierarchical tensor decomposition quantum scheme capable of expressing local expectation values of strongly entangled higher-dimensional quantum systems efficiently. To address the extreme complexity and high computational cost of such schemes, we will automate their parallel implementation by delivering an integrated software framework, ExaTN (Exascale Tensor Networks), that will process hierarchical tensor representations on exascale HPC systems via an asymchronous task-based programming model. As a specific application, our framework will provide scientists with a capability of simulating strongly entangled quantum systems, thus positioning ORNL as a strategic leader in modeling of strongly correlated materials.

McCaskey, AlexanderJ↗

Exploring Machine Learning Techniques For Dynamic Modeling on Future Exascale Systems

Future exascale systems must be optimized for both power and performance at scale in order to achieve DOE’s goal of a sustained petaflop within 20 Megawatts by 2022 [1]. Massive parallelism of the future systems combined with complex memory hierarchies will form a barrier to efficient application and architecture design. These challenges are exacerbated with emerging complex architectures such as GPGPUs and Intel Xeon Phi as parallelism increases orders of magnitude and system power consumption can easily triple or quadruple. Therefore, we need techniques that can reduce the search space for optimization, isolate power-performance bottlenecks, identify root causes for software/hardware inefficiency, and effectively direct runtime scheduling.

Song, Shuaiwen↗

Design and Construction of a High-Resolution Hodoscope for the GlueX Experiment with High-Statistics Analysis of the p0, ¿, and ¿1 Photoproduction Cross Sections from the RadPhi Experiment

Differential cross sections for forward-angle photoproduction of p0, ¿, and ¿ 1 pseudoscalar mesons were measured using data from the RadPhi experiment conducted in Hall B at Jef ferson Lab. RadPhi utilized a tagged bremsstrahlung photon beam incident on a stationary 9Be target, with a detector system configured to trigger on a recoil proton in coincidence with multiple neutral showers in the calorimeter. Events were reconstructed and subjected to kinematic constraints, with background suppressed via sideband subtraction guided by Monte Carlo modeling of background contributions. Cross sections were extracted over the photon energy range 4.4– 5.4 GeV and binned in invariant momentum transfer t, providing measurements from one of the first high-statistics experiments of forward ¿ and ¿1 pro duction from a nuclear target at these energies. Acceptance corrections were applied using a detailed GEANT-based simulation of the detector geometry and response. The resulting cross sections are consistent with 2020 CLAS results, when scaled by the number of protons in beryllium, and show broad agreement with other data and theoretical models. In parallel, a high-resolution photon tagger detector, the Tagger Microscope (TAGM), was designed, constructed, and commissioned for the GlueX experiment in Hall D at Jefferson Lab. The TAGM was developed to provide high-rate tagging capability in the coherent bremsstrahlung peak by detecting post-bremsstrahlung electrons across a one GeV range along the focal plane of the tagging spectrometer. The detector consists of a 5ˆ102 array of 2ˆ2 mm2 square BCF-20 plastic scintillating fibers thermally fused to BCF-98 light guide fibers optically coupled to silicon photomultipliers. These fibers are mounted in a precision machined framework enabling fine positional adjustments to maintain precise alignment with post-bremsstrahlung electron trajectories, while ensuring mechanical rigidity, thermal stability, optical isolation, minimal inactive area, and radiation shielding for electronics. The construction effort involved extensive testing of fiber quality, light transmission, thermal fusing, radiation hardness, and defect analysis using SEM and EDX techniques. Following its installation and commissioning, the TAGM became a critical component of the GlueX beamline, enabling high-rate tagging essential for studies of hybrid mesons and gluonic ex citations.

McIntyre, James [Univ. of Connecticut, Storrs, CT ↗

Silo & HDF5 I/O Scaling Improvements on BG/P Systems

Silo and HDF5 are I/O libraries used by many codes important to the LLNL's Weapons and Complex Integration (WCI) mission. In the past year, modest adjustments and tuning of Silo, HDF5 and the I/O configuration of the BG/P platform, Dawn, were undertaken. A key goal of this work was to improve I/O performance without requiring any changes in the application codes themselves. In particular, the application codes have been allowed to continue to use a simplified yet highly flexible I/O paradigm known as 'Poor Man's Parallel I/O', where scalability is achieved through concurrent, serial I/O to multiple files. The results demonstrate substantial performance gains (better than 50x in many cases) at large scale (greater than 64,000 MPI tasks). They describe key enhancements made to Silo, HDF5 and the I/O configuration of our BG/P platform and present very favorable results from scalability studies over a wide range of operating scenarios.

97 MATHEMATICS AND COMPUTING↗

Fidelity susceptibility and general quench near an anisotropic quantum critical point

We study the scaling behavior of fidelity susceptibility density ({chi}{sub f}) at or close to an anisotropic quantum critical point (AQCP) characterized by two different correlation length exponents {nu}{sub ||} and {nu}{sub perpendicular} along parallel and perpendicular spatial directions, respectively. Our studies show that the response of the system due to a small change in the Hamiltonian near an AQCP is different from that seen near an isotropic QCP. In particular, for a finite system with linear dimension L{sub ||} (L{sub perpendicular}) in the parallel (perpendicular) directions, the maximum value of {chi}{sub f} is found to increase in a power-law fashion with L{sub ||} for small L{sub ||}, with an exponent depending on both {nu}{sub ||} and {nu}{sub perpendicular} and eventually crosses over to a scaling with L{sub perpendicular} for L{sub ||}{sup 1/{nu}{sub ||}} > or approx. L{sub perpendicular}{sup 1/{nu}{sub perpendicular}}. We also propose scaling relations of heat density and defect density generated following a quench starting from an AQCP and connect them to a generalized fidelity susceptibility. These predictions are verified exactly both analytically and numerically taking the example of a Hamiltonian showing a semi-Dirac band-crossing point.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Three-Dimensional Parallel Adaptive Mesh Refinement Simulations of Shock-Driven Turbulent Mixing in Plane and Converging Geometries

This paper presents the use of a dynamically adaptive mesh refinement strategy for the simulations of shock-driven turbulent mixing. Large-eddy simulations are necessary due the high Reynolds number turbulent regime. In this approach, the large scales are simulated directly and small scales at which the viscous dissipation occurs are modeled. A low-numerical centered finite-difference scheme is used in turbulent flow regions while a shock-capturing method is employed to capture shocks. Three-dimensional parallel simulations of the Richtmyer-Meshkov instability performed in plane and converging geometries are described.

42 ENGINEERING↗

Coupled Lattice Boltzmann Modeling Framework for Pore-Scale Fluid Flow and Reactive Transport

In this paper, we propose a modeling framework for pore-scale fluid flow and reactive transport based on a coupled lattice Boltzmann model (LBM). We develop a modeling interface to integrate the LBM modeling code parallel lattice Boltzmann solver and the PHREEQC reaction solver using multiple flow and reaction cell mapping schemes. The major advantage of the proposed workflow is the high modeling flexibility obtained by coupling the geochemical model with the LBM fluid flow model. Consequently, the model is capable of executing one or more complex reactions within desired cells while preserving the high data communication efficiency between the two codes. Meanwhile, the developed mapping mechanism enables the flow, diffusion, and reactions in complex pore-scale geometries. We validate the coupled code in a series of benchmark numerical experiments, including 2D single-phase Poiseuille flow and diffusion, 2D reactive transport with calcite dissolution, as well as surface complexation reactions. The simulation results show good agreement with analytical solutions, experimental data, and multiple other simulation codes. In addition, we design an AI-based optimization workflow and implement it on the surface complexation model to enable increased capacity of the coupled modeling framework. Compared to the manual tuning results proposed in the literature, our workflow demonstrates fast and reliable model optimization results without incorporating pre-existing domain knowledge.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

STUDY OF TWO SUCCESSIVE THREE-RIBBON SOLAR FLARES ON 2012 JULY 6

This Letter reports two rarely observed three-ribbon flares (M1.9 and C9.2) on 2012 July 6 in NOAA AR 11515, which we found using Hα observations of 0.''1 resolution from the New Solar Telescope and Ca II H images from Hinode. The flaring site is characterized by an intriguing ''fish-bone-like'' morphology evidenced by both Hα images and a nonlinear force-free field (NLFFF) extrapolation, where two semi-parallel rows of low-lying, sheared loops connect an elongated, parasitic negative field with the sandwiching positive fields. The NLFFF model also shows that the two rows of loops are asymmetric in height and have opposite twists, and are enveloped by large-scale field lines including open fields. The two flares occurred in succession within half an hour and are located at the two ends of the flaring region. The three ribbons of each flare run parallel to the magnetic polarity inversion line, with the outer two lying in the positive field and the central one in the negative field. Both flares show surge-like flows in Hα apparently toward the remote region, while the C9.2 flare is also accompanied by EUV jets possibly along the open field lines. Interestingly, the 12-25 keV hard X-ray sources of the C9.2 flare first line up with the central ribbon then shift to concentrate on the top of the higher branch of loops. These results are discussed in favor of reconnection along the coronal null line, producing the three flare ribbons and the associated ejections.

79 ASTRONOMY AND ASTROPHYSICS↗

MEASUREMENT OF INTERMITTENCY OF ANISOTROPIC MAGNETOHYDRODYNAMIC TURBULENCE IN HIGH-SPEED SOLAR WIND

We investigate the intermittency of anisotropic magnetohydrodynamic (MHD) turbulence in high-speed solar wind. Using the data recorded by the Ulysses spacecraft, we apply the Castaing function to model the probability density functions of the fluctuating magnetic field and calculate the magnetic structure functions (SFs) S{sup p} of the order p in the coordinates (r, {Theta}), with r being the length scale and {Theta} the direction of the local mean field. The scaling exponent {zeta}, from S{sup p} (r, {Theta}){proportional_to}r {sup {zeta}(p,{Theta})}, has an anomalous nonlinear dependence on p, implying the intermittent scaling of solar wind turbulence, which has been observed for decades. Furthermore, we study the anisotropy of solar wind turbulence introduced by the strong mean magnetic field. From S{sup p} ({Theta} = 0){proportional_to}S{sup p} ({Theta} = {pi}/2), we obtain r{sub perpendicular{proportional_to}r {sup {alpha}}p||} with {alpha}{sub p} = {zeta}{sub ||}/{zeta}{sub perpendicular} denoting the perpendicular-parallel spatial correlation of the moment of the pth order. For the magnetic field difference {delta}B, we find {alpha}{sub 2} = 1.78 {+-} 0.26, consistent with recent theories and observations. However, when the contribution from the intermittent fluctuations begins to dominate the scaling, {alpha} is not a constant but increases with p, e.g., {alpha}{sub 5} = 1.97 {+-} 0.41 and {alpha}{sub 8} {approx} 2.42 {+-} 0.64. This complication of the perpendicular-parallel spatial correlation due to the intermittency raises new questions for MHD turbulence theory.

79 ASTRONOMY AND ASTROPHYSICS↗

Generalizable coordination of large multiscale workflows: challenges and learnings at scale

The advancement of machine learning techniques and the heterogeneous architectures of most current supercomputers are propelling the demand for large multiscale simulations that can automatically and autonomously couple diverse components and map them to relevant resources to solve complex problems at multiple scales. Nevertheless, despite the recent progress in workflow technologies, current capabilities are limited to coupling two scales. In the first-ever demonstration of using three scales of resolution, we present a scalable and generalizable framework that couples pairs of models using machine learning and in situ feedback. We expand upon the massively parallel Multiscale Machine-Learned Modeling Infrastructure (MuMMI), a recent, award-winning workflow, and generalize the framework beyond its original design. We discuss the challenges and learnings in executing a massive multiscale simulation campaign that utilized over 600,000 node hours on Summit and achieved more than 98% GPU occupancy for more than 83% of the time. We present innovations to enable several orders of magnitude scaling, including simultaneously coordinating 24,000 jobs, and managing several TBs of new data per day and over a billion files in total. Finally, we describe the generalizability of our framework and, with an upcoming open-source release, discuss how the presented framework may be used for new applications.

Bhatia, Harsh↗

COHERENT EVENTS AND SPECTRAL SHAPE AT ION KINETIC SCALES IN THE FAST SOLAR WIND TURBULENCE

In this paper we investigate spectral and phase coherence properties of magnetic fluctuations in the vicinity of the spectral transition from large, magnetohydrodynamic to sub-ion scales using in situ measurements of the Wind spacecraft in a fast stream. For the time interval investigated by Leamon et al. (1998) the phase coherence analysis shows the presence of sporadic quasi-parallel Alfvén ion cyclotron (AIC) waves as well as coherent structures in the form of large-amplitude, quasi-perpendicular Alfvén vortex-like structures and current sheets. These waves and structures importantly contribute to the observed power spectrum of magnetic fluctuations around ion scales; AIC waves contribute to the spectrum in a narrow frequency range whereas the coherent structures contribute to the spectrum over a wide frequency band from the inertial range to the sub-ion frequency range. We conclude that a particular combination of waves and coherent structures determines the spectral shape of the magnetic field spectrum around ion scales. This phenomenon provides a possible explanation for a high variability of the magnetic power spectra around ion scales observed in the solar wind.

79 ASTRONOMY AND ASTROPHYSICS↗

Analysis of the impact of parallel magnetic fluctuations on linear gyrokinetic stability in NSTX-U and verification of gyro-fluid models

In this work, we use the CGYRO gyrokinetic code to analyze two L- and one H-mode discharges from the National Spherical Torus Experiment (NSTX) and NSTX-Upgrade (NSTX-U) selected due to their different mix of ion-scale driftwaves, ion temperature gradient (ITG) mode and trapped electron mode (TEM), and electromagnetic instabilities, kinetic ballooning mode (KBM), and micro-tearing mode (MTM) in the plasma core. It is found that the effect of parallel magnetic fluctuations is strongly destabilizing to the unstable KBMs compared to calculations with only perpendicular magnetic fluctuations. Two discharges have a mix of ITG/TEM and MTMs that are predicted to be dominant instability across the plasma radius. The parallel magnetic fluctuations are found to have little effect on the MTM stability but are destabilizing to ITG/TEM modes. To test the validity of the gyro-fluid linear stability codes TGLF and GFS at low aspect ratio, a database of linear growth rates has been created using the CGYRO gyrokinetic code. The database is comprised of various parameter scans around a standardized set of NSTX-U core parameters. It contains a group of electrostatic cases and an electromagnetic group that includes the effects of perpendicular and parallel magnetic fluctuations. Comparing the results from the GFS and TGLF models, we find that GFS exhibits the best agreement with the database of CGYRO linear growth rates. Comparing the model results for the electromagnetic scans shows that GFS captures the effects of parallel magnetic fluctuations accurately, while the TGLF model does not, as it lacks sufficient perpendicular energy resolution.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Improving Uintah's Scalability Through the Use of Portable Kokkos-Based Data Parallel Tasks

The University of Utah’s Carbon Capture Multidisciplinary Simulation Center (CCMSC) is using the Uintah Computational Framework to predict performance of a 1000 MWe ultra-supercritical clean coal boiler. The center aims to utilize the Intel Xeon Phi-based DOE systems, Theta and Aurora, through the Aurora Early Science Program by using the Kokkos C++ library to enable node-level performance portability. This paper describes infrastructure advancements and portability improvements made possible by the integration of Kokkos within Uintah. This integration marks a step towards consolidating Uintah’s MPI+PThreads and MPI+CUDA hybrid parallelism approaches into a single MPI+Kokkos approach. Scalability results are presented that compare serial and data parallel task execution models for a challenging radiative heat transfer calculation, central to the center’s predictive boiler simulations. Here, these results demonstrate both good strong-scaling characteristics to 256 Knights Landing (KNL) processors on the NSF Stampede system, and show the KNL-based calculation to compete with prior GPU-based results for the same calculation.

97 MATHEMATICS AND COMPUTING↗

An Open-Source Parallel EMT Simulation Framework

As the integration level of inverter-based resources (IBRs) increases, ensuring the reliable operation of the bulk power systems requires the use of electromagnetic transient (EMT) simulation tools to identify and mitigate system-wide stability risks. Conducting EMT studies for large-scale, IBR-rich grids, however, is challenging due to the inherent computational bottleneck caused by the underlying high-fidelity models and required small time steps. This paper introduces ParaEMT: an open-source, generic EMT simulation framework designed to accelerate simulations by leveraging advanced parallel computational technologies, such as high-performance computers. This paper presents a comprehensive exposition of ParaEMT, covering its modeling library, simulation strategy, framework structure, operational procedures, and auxiliary features, alongside its extensible parallel computational architecture. Notably, ParaEMT is a publicly accessible and modularized framework written in Python, thereby facilitating future development and the integration of new models and algorithms. The accuracy and efficiency of ParaEMT are demonstrated by rigorous validations via multiple case studies.

electromagnetic transient simulation↗

Reducing communication in algebraic multigrid with multi-step node aware communication

Algebraic multigrid (AMG) is often viewed as a scalable [Formula: see text] solver for sparse linear systems. Yet, AMG lacks parallel scalability due to increasingly large costs associated with communication, both in the initial construction of a multigrid hierarchy and in the iterative solve phase. This work introduces a parallel implementation of AMG that reduces the cost of communication, yielding improved parallel scalability. It is common in Message Passing Interface (MPI), particularly in the MPI-everywhere approach, to arrange inter-process communication, so that communication is transported regardless of the location of the send and receive processes. Performance tests show notable differences in the cost of intra- and internode communication, motivating a restructuring of communication. In this case, the communication schedule takes advantage of the less costly intra-node communication, reducing both the number and the size of internode messages. Node-centric communication extends to the range of components in both the setup and solve phase of AMG, yielding an increase in the weak and strong scaling of the entire method.

Computer Science↗