Search NASA⌕ Search

SEARCH · Search NASA

Results for “Time 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 883 records · Page 49
All sourcesNASA NTRSDOE OSTI
DOE OSTI2017

GPU-accelerated Tersoff potentials for massively parallel Molecular Dynamics simulations

The Tersoff potential is one of the empirical many-body potentials that has been widely used in simulation studies at atomic scales. Unlike pair-wise potentials, the Tersoff potential involves three-body terms, which require much more arithmetic operations and data dependency. In this contribution, we have implemented the GPU-accelerated version of several variants of the Tersoff potential for LAMMPS, an open-source massively parallel Molecular Dynamics code. Compared to the existing MPI implementation in LAMMPS, the GPU implementation exhibits a better scalability and offers a speedup of 2.2X when run on 1000 compute nodes on the Titan supercomputer. On a single node, the speedup ranges from 2.0 to 8.0 times, depending on the number of atoms per GPU and hardware configurations. The most notable features of our GPU-accelerated version include its design for MPI/accelerator heterogeneous parallelism, its compatibility with other functionalities in LAMMPS, its ability to give deterministic results and to support both NVIDIA CUDA- and OpenCL-enabled accelerators. Our implementation is now part of the GPU package in LAMMPS and accessible for public use.

Computer Science↗
DOE OSTI2025

Computational Algorithms for Unit Commitment with AC Power Flows (Final Report)

Security-constrained unit commitment (SCUC) is a key component in power system operations. When AC power flow constraints are considered in the SCUC model (AC-SCUC), the problem becomes extremely difficult due to its discrete and non-convex nature, as described in “Grid Optimization Competition Challenge 3 Problem Formulation (GOCC)”. There are four main challenges: (i) Discrete decisions regarding unit online/offline status and start-up/shut-down procedures for every single unit. The number of discrete decision variables increases considerably when a system integrates multiple generators; (ii) Configuration-based combined-cycle formulations, and multi-commodity models that include ramping products, spin/non-spin products, and regulation up/down products. The combined-cycle units introduce additional discrete decision variables and auxiliary service products further complicate the model by connecting multi-commodity products’ continuous and discrete variables; (iii) SCUC models with AC power flow constraints are far more complex due to massive bilinear terms in the large-scale nonlinear power balance equations. The nonlinear power balance equations are further complicated by the discrete step control variables of shunts; (iv) N − 1 contingency analysis. The size of the model increases linearly with the number of contingencies considered, greatly increasing the size of the optimization model. Accordingly, there is an emergent need to develop a robust algorithm capable of deriving a high-quality solution in a short time and passing through contingency tests simultaneously. In this project, we explore innovative techniques to address this challenging problem by integrating advanced polyhedral theory, approximation methods, relaxation strategies, decomposition techniques, and parallel computing. Each technique approaches the problem from a different perspective, leveraging its specific strengths to tackle distinct challenges. Each individual method has demonstrated its effectiveness in the PI’s previous research. Their integration is expected to significantly reduce the computational time required to solve the proposed complex problem. Successful completion of this project has the potential to transform the industry by enhancing optimization solvers capable of handling large-scale day-ahead energy market clearing models within strict time constraints, while incorporating AC power flow constraints. This advancement will lead to reduced overall generation costs and, consequently, increased social welfare.

29 ENERGY PLANNING, POLICY, AND ECONOMY↗
DOE OSTI2017

On the Evolution of the Central Density of Quiescent Galaxies

We investigate the origin of the evolution of the population-averaged central stellar mass density (Σ{sub 1}) of quiescent galaxies (QGs) by probing the relation between stellar age and Σ{sub 1} at z ∼ 0. We use the Zurich ENvironmental Study (ZENS), which is a survey of galaxy groups with a large fraction of satellite galaxies. QGs shape a narrow locus in the Σ{sub 1}– M {sub ⋆} plane, which we refer to as Σ{sub 1} ridgeline. Colors of ( B − I ) and ( I − J ) are used to divide QGs into three age categories: young (<2 Gyr), intermediate (2–4 Gyr), and old (>4 Gyr). At fixed stellar mass, old QGs on the Σ{sub 1} ridgeline have higher Σ{sub 1} than young QGs. This shows that galaxies landing on the Σ{sub 1} ridgeline at later epochs arrive with lower Σ{sub 1}, which drives the zeropoint of the ridgeline down with time. We compare the present-day zeropoint of the oldest population at z = 0 with the zeropoint of the quiescent population 4 Gyr back in time, at z = 0.37. These zeropoints are identical, showing that the intrinsic evolution of individual galaxies after they arrive on the Σ{sub 1} ridgeline must be negligible, or must evolve parallel to the ridgeline during this interval. The observed evolution of the global zeropoint of 0.07 dex over the last 4 Gyr is thus largely due to the continuous addition of newly quenched galaxies with lower Σ{sub 1} at later times (“progenitor bias”). While these results refer to the satellite-rich ZENS sample as a whole, our work suggests a similar age–Σ{sub 1} trend for central galaxies.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2015

Data consistency conditions for truncated fanbeam and parallel projections

Purpose: In image reconstruction from projections, data consistency conditions (DCCs) are mathematical relationships that express the overlap of information between ideal projections. DCCs have been incorporated in image reconstruction procedures for positron emission tomography, single photon emission computed tomography, and x-ray computed tomography (CT). Building on published fanbeam DCCs for nontruncated projections along a line, the authors recently announced new DCCs that can be applied to truncated parallel projections in classical (two-dimensional) image reconstruction. These DCCs take the form of polynomial expressions for a weighted backprojection of the projections. The purpose of this work was to present the new DCCs for truncated parallel projections, to extend these conditions to truncated fanbeam projections on a circular trajectory, to verify the conditions with numerical examples, and to present a model of how DCCs could be applied with a toy problem in patient motion estimation with truncated projections. Methods: A mathematical derivation of the new parallel DCCs was performed by substituting the underlying imaging equation into the mathematical expression for the weighted backprojection and demonstrating the resulting polynomial form. This DCC result was extended to fanbeam projections by a substitution of parallel to fanbeam variables. Ideal fanbeam projections of a simple mathematical phantom were simulated and the DCCs for these projections were evaluated by fitting polynomials to the weighted backprojection. For the motion estimation problem, a parametrized motion was simulated using a dynamic version of the mathematical phantom, and both noiseless and noisy fanbeam projections were simulated for a full circular trajectory. The fanbeam DCCs were applied to extract the motion parameters, which allowed the motion contamination to be removed from the projections. A reconstruction was performed from the corrected projections. Results: The mathematical derivation revealed the anticipated polynomial behavior. The conversion to fanbeam variables led to a straight-forward weighted fanbeam backprojection which yielded the same function and therefore the same polynomial behavior as occurred in the parallel case. Plots of the numerically calculated DCCs showed polynomial behavior visually indistinguishable from the fitted polynomials. For the motion estimation problem, the motion parameters were satisfactorily recovered and ten times more accurately for the noise-free case. The reconstructed images showed that only a faint trace of the motion blur was still visible after correction from the noisy motion-contaminated projections. Conclusions: New DCCs have been established for fanbeam and parallel projections, and these conditions have been validated using numerical experiments with truncated projections. It has been shown how these DCCs could be applied to extract parameters of unwanted physical effects in tomographic imaging, even with truncated projections.

62 RADIOLOGY AND NUCLEAR MEDICINE↗
DOE OSTI2025

Robustness of Deep Learning Classification to Adversarial Input on GPUs: Asynchronous Parallel Accumulation Is a Source of Vulnerability

The ability of machine learning (ML) classification models to resist small, targeted input perturbations—known as adversarial attacks—is a key measure of their safety and reliability. We show that floating-point non associativity (FPNA) coupled with asynchronous parallel programming on GPUs is sufficient to result in misclassification, without any perturbation to the input. Additionally, we show that this misclassification is particularly significant for inputs close to the decision boundary and that standard adversarial robustness results may be overestimated up to 4.6 when not considering machine-level details. We first study a linear classifier, before focusing on standard Graph Neural Network (GNN) architectures and datasets used in robustness assessments. We develop a novel black-box attack using Bayesian optimization to discover external workloads that can change the instruction scheduling which bias the output of reductions on GPUs and reliably lead to misclassification. Motivated by these results, we present a new learnable permutation (LP) gradient-based approach to learning floating-point operation orderings that lead to misclassifications. The LP approach provides a worst-case estimate in a computationally efficient manner, avoiding the need to run identical experiments tens of thousands of times over a potentially large set of possible GPU states or architectures. Finally, using instrumentation-based testing, we investigate parallel reduction ordering across different GPU architectures under external background workloads, when utilizing multi-GPU virtualization, and when applying power capping. Our results demonstrate that parallel reduction ordering varies significantly across architectures under the first two conditions, substantially increasing the search space required to fully test the effects of this parallel scheduler-based vulnerability. These results and the methods developed here can help to include machine-level considerations into adversarial robustness assessments, which can make a difference in safety and mission critical applications.

Shanmugavelu, Sanjif [Maxeler Technologies, a Groq↗
DOE OSTI2017

A fast parallel 3D Poisson solver with longitudinal periodic and transverse open boundary conditions for space-charge simulations

A three-dimensional (3D) Poisson solver with longitudinal periodic and transverse open boundary conditions can have important applications in beam physics of particle accelerators. Here, we present a fast efficient method to solve the Poisson equation using a spectral finite-difference method. This method uses a computational domain that contains the charged particle beam only and has a computational complexity of $O(N_u (logN_{mode}))$, where $N_u$ is the total number of unknowns and $N_{mode}$ is the maximum number of longitudinal or azimuthal modes. This saves both the computational time and the memory usage of using an artificial boundary condition in a large extended computational domain. The new 3D Poisson solver is parallelized using a message passing interface (MPI) on multi-processor computers and shows a reasonable parallel performance up to hundreds of processor cores.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2017

Effective Vectorization with OpenMP 4.5

This paper describes how the Single Instruction Multiple Data (SIMD) model and its extensions in OpenMP work, and how these are implemented in different compilers. Modern processors are highly parallel computational machines which often include multiple processors capable of executing several instructions in parallel. Understanding SIMD and executing instructions in parallel allows the processor to achieve higher performance without increasing the power required to run it. SIMD instructions can significantly reduce the runtime of code by executing a single operation on large groups of data. The SIMD model is so integral to the processor s potential performance that, if SIMD is not utilized, less than half of the processor is ever actually used. Unfortunately, using SIMD instructions is a challenge in higher level languages because most programming languages do not have a way to describe them. Most compilers are capable of vectorizing code by using the SIMD instructions, but there are many code features important for SIMD vectorization that the compiler cannot determine at compile time. OpenMP attempts to solve this by extending the C++/C and Fortran programming languages with compiler directives that express SIMD parallelism. OpenMP is used to pass hints to the compiler about the code to be executed in SIMD. This is a key resource for making optimized code, but it does not change whether or not the code can use SIMD operations. However, in many cases critical functions are limited by a poor understanding of how SIMD instructions are actually implemented, as SIMD can be implemented through vector instructions or simultaneous multi-threading (SMT). We have found that it is often the case that code cannot be vectorized, or is vectorized poorly, because the programmer does not have sufficient knowledge of how SIMD instructions work.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2026

The high level trigger and express data production at STAR

To meet the demands of the Beam Energy Scan phase-II (BES-II) program, the STAR experiment at the Relativistic Heavy Ion Collider (RHIC) developed a dual real-time framework consisting of a High Level Trigger (HLT) and an Express Data Production system (xProduction). The HLT operates online within the Data Acquisition (DAQ) chain on a dedicated multi-core CPU cluster with the option to offload compute-intensive kernels to Xeon Phi coprocessors. It uses parallelized algorithms, such as the Cellular Automaton (CA) Track Finder, to perform rapid tracking, vertexing, and event filtering. This allows it to select events of interest in real time and provide immediate feedback on detector and beam conditions. In contrast, the xProduction workflow runs concurrently and independently of the DAQ loop. It applies near offline-quality calibration and reconstruction within hours of data collection. The xProduction input is the express data stream, whose content can be enriched by HLT trigger/priority selections under DAQ/HLT resource constraints, and it uses the STAR calibration/conditions framework, incorporating online calibration/QA information when available. This enables early preliminary physics analysis, including the reconstruction of rare signals, such as hyperons and hypernuclei. It also provides collaboration-wide access to analysis-ready datasets. Together, the HLT and xProduction systems form a complementary architecture: the HLT performs online event selection while the xProduction chain delivers high-quality results within a short amount of time. This integrated framework has enabled the prompt reconstruction of the $^5_Λ$ He hypernucleus with high statistical significance and the efficient processing of hundreds of millions of heavy-ion collision events. In conclusion, its demonstrated scalability and robustness establish a model for future high-luminosity experiments requiring both online event filtering and rapid access to analysis-quality data.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗
DOE OSTI2024

Early experiences on the OLCF Frontier system with AthenaPK and Parthenon–Hydro

The Oak Ridge Leadership Computing Facility (OLCF) has been preparing the nation's first exascale system, Frontier, for production and end users. Frontier is based on HPE Cray's new EX architecture and Slingshot interconnect and features 74 cabinets of optimized 3rd Gen AMD EPYC CPUs for HPC and AI and AMD Instinct 250X accelerators. As a part of this preparation, “real-world” user codes have been selected to help assess the functionality, performance, and usability of the system. This article describes early experiences using the system in collaboration with the Hamburg Observatory for two selected codes, which have since been adopted in the OLCF test harness. Experiences discussed include efforts to resolve performance variability and per-cycle slowdowns. Results are shown for a performance portable astrophysical magnetohydronamics code, AthenaPK, and a mini-application stressing the core functionality of a performance portable block-structured adaptive mesh refinement framework, Parthenon-Hydro. Here, these results show good scaling characteristics to the full system. At the largest scale, the Parthenon-Hydro miniapp reaches a total of $1.7$ $\times$ $10^{13}$ zone-cycles/s on 9216 nodes (73,728 logical GPUs) at ≈92% weak scaling parallel efficiency (starting from a single node using a second-order, finite-volume method).

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2017

The in-depth residual strain heterogeneities due to an indentation and a laser shock peening for Ti-6Al-4V titanium alloy

In this paper, heterogeneity of the through-thickness residual strain due to the laser shock peening (LSP) in comparison with that due to the indentation was studied in Ti-6Al-4V alloy samples. The latter is almost a quasi-static process while the former features extremely high strain-rate deformation. The synchrotron based high-energy X-ray diffraction was employed to investigate the through-thickness residual strain distribution. The studied two samples after the two processing procedures share the following features: (i) the pressure affected depths are both ~2 mm and (ii) the largest magnitudes of the compressive residual strains parallel to the surface are ~4,000 με. However, the pit depth for the indentation sample is ~9 times larger than that for the LSP. The position featuring the largest magnitude of the compressive residual strain is in the sub-surface for the indentation while it is in the surface for the LSP. Results of the elastic-visco-plastic finite element simulation for the indentation indicate that the position featuring the maximum accumulative plastic shear as defined in this paper corresponds to the location with the largest magnitude of compressive residual strain. To validate this finding, full width at half maximum (FWHM) of the X-ray diffraction profile, which is proportional to the level of the plastic deformation, is also studied. Finally, it is found that positions with the largest FWHM indeed correspond to the largest magnitude of compressive residual strain for both tests.

36 MATERIALS SCIENCE↗
DOE OSTI2025

MBX V1.2: Accelerating Data-Driven Many-Body Molecular Dynamics Simulations

The MBX software provides an advanced platform for molecular dynamics simulations, leveraging state-of-the-art MB-pol and MB-nrg data-driven many-body potential energy functions. Developed over the past decade, these potential energy functions integrate physics-based and machine-learned many-body terms trained on electronic structure data calculated at the "gold standard" coupled-cluster level of theory. Recent advancements in MBX have focused on optimizing its performance, resulting in the release of MBX v1.2. While the inherently many-body nature of MB-pol and MB-nrg ensures high accuracy, it poses computational challenges. MBX v1.2 addresses these challenges with significant performance improvements, including enhanced parallelism that fully harnesses the power of modern multicore CPUs. In conclusion, these advancements enable simulations on nanosecond time scales for condensed-phase systems, significantly expanding the scope of high-accuracy, predictive simulations of complex molecular systems powered by data-driven many-body potential energy functions.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2018

DNA Strand Buffers

A buffer reaction actively resists changes to the concentration of a chemical species. Typically, buffering reactions have only been able to regulate the concentration of hydronium (i.e., pH) and other ions. In this work, we develop a new class of buffers that regulate the concentrations of short sequences of DNA (i.e., oligonucleotides). A buffer’s behavior is determined by its set point concentration, capacity to resist disturbances, and response time after a disturbance. We provide simple mathematical formulas for selecting rate constants to tune each of these properties and show how to design DNA sequences and concentrations to implement the desired rate constants. We demonstrate several oligonucleotide buffers that maintain oligonucleotide set point concentrations between 10 and 80 nM in the presence of disturbances of 50 to 500 nM, with response times of less than 10 min to 1.5 h. Multiple buffers can regulate different sequences of DNA in parallel without crosstalk. Oligonucleotide buffers could stabilize and restore reactant concentrations in DNA circuits or in self-assembly processes, allowing such systems to operate reliably for extended durations. These buffers might also be coupled to other reactions to buffer molecules other than DNA. In general, an oligonucleotide buffer can be viewed as a chemical “battery” that maintains the total chemical potential of a buffered species in a closed system.

Biopolymers↗
DOE OSTI2022

Neuromorphic scaling advantages for energy-efficient random walk computations

Neuromorphic computing, which aims to replicate the computational structure and architecture of the brain in synthetic hardware, has typically focused on artificial intelligence applications. What is less explored is whether such brain-inspired hardware can provide value beyond cognitive tasks. Here we show that the high degree of parallelism and configurability of spiking neuromorphic architectures makes them well suited to implement random walks via discrete-time Markov chains. Overall, these random walks are useful in Monte Carlo methods, which represent a fundamental computational tool for solving a wide range of numerical computing tasks. Using IBM’s TrueNorth and Intel’s Loihi neuromorphic computing platforms, we show that our neuromorphic computing algorithm for generating random walk approximations of diffusion offers advantages in energy-efficient computation compared with conventional approaches. We also show that our neuromorphic computing algorithm can be extended to more sophisticated jump-diffusion processes that are useful in a range of applications, including financial economics, particle physics and machine learning.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2018

Measurements of the Prompt Fission Neutron Spectrum at LANSCE: The Chi-Nu Experiment

The goal of the Chi-Nu experiment at the Los Alamos Neutron Science Center is to measure the prompt fission neutron spectra from major actinides using a double time-of-flight method with a pulsed, white incoming neutron source. Fission events are detected with a parallel-plate avalanche counter and outgoing neutrons are detected with either a 6 Li-glass or liquid scintillator detector array for low- or high-energy neutrons, respectively. A detector response matrix for the interaction of neutrons with the experimental environment for neutrons measured with the Chi-Nu 6 Li-glass detector array has been calculated to obtain a full understanding of the measured Chi-Nu data and also to allow for nearly instantaneous production of simulated Chi-Nu data spectra. Prompt fission neutron spectra corresponding to 19 incoming neutron energy ranges from 0.7-20 MeV have been extracted using the ratio-of-ratios method with Chi-Nu 6 Li-glass data on the neutron-induced fission of 235 U.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗
DOE OSTI2022

MECHANISMS OF RADIATION CARCINOGENESIS: WHAT IS REALLY INDUCED?

Abstract It has been difficult to understand why the relative risk for cancer decreases with an increase in time since an exposure to radiation. It was recently recognized that this decline can be explained by a parallel shift of the age-related cancer mortality curve toward younger ages. In fact, it has been known for many years that mouse survival curves exhibit a parallel shift toward younger ages following an exposure to radiation, but it was not recognized that the mutation induction theory is incompatible with this parallel shift. This is because a parallel shift in the survival curve implies that all the irradiated individuals are affected, but the mutation induction theory assumes that only a fraction of the irradiated individuals is affected following an exposure to radiation. Thus, it seems likely that the target of radiation action, which leads to carcinogenesis, is not restricted to epithelial cells but includes all of the surrounding cells leading to an altered microenvironment. Since it is repeatedly observed that radiation-exposed normal tissues can stimulate transplanted or spontaneously arising tumor cells to grow faster, worsen the malignant phenotypes and finally kill the host earlier than usual, an exposure to radiation seems most likely to cause tissue inflammation, which creates conditions favorable for the growth of spontaneously arising tumor cells. This new concept suggests that it might be possible to attenuate the extent of radiation carcinogenesis by intervening in tissue inflammatory processes.

38 RADIATION CHEMISTRY, RADIOCHEMISTRY, AND NUCLEA↗
DOE OSTI2022

Direct evidence of the dominant role of multiphoton permanent-dipole transitions in strong-field dissociation of NO 2 +

We study laser-induced dissociation of a metastable NO 2+ ion-beam target into N + +O + , focusing on the prominent contribution by molecules breaking parallel to the polarization at high peak laser intensity (~10 15 W/cm 2 ). Our experimental results and time-dependent Schrödinger equation calculations show that, contrary to commonly held intuition that electronic transitions always prevail, the dominant process underlying this highly aligned dissociation is a multiphoton permanent-dipole transition involving only the electronic ground state and leading to its vibrational continuum. Strong-field permanent-dipole transitions should thus be considered generally, as they may play a significant role in other heteronuclear molecules. Moreover, their role should only grow in importance for longer wavelengths, a trending direction in ultrafast laser studies.

74 ATOMIC AND MOLECULAR PHYSICS↗
DOE OSTI2020

Deep Generative Models that Solve PDEs: Distributed Computing for Training Large Data-Free Models

Recent progress in scientific machine learning (SciML) has opened up the possibility of training novel neural network architectures that solve complex partial differential equations (PDEs). Several (nearly data free) approaches have been recently reported that successfully solve PDEs, with examples including deep feed forward networks, generative networks, and deep encoder-decoder networks. However, practical adoption of these approaches is limited by the difficulty in training these models, especially to make predictions at large output resolutions (≥1024×1024). Here we report on a software framework for data parallel distributed deep learning that resolves the twin challenges of training these large SciML models - training in reasonable time as well as distributing the storage requirements. Our framework provides several out of the box functionality including (a) loss integrity independent of number of processes, (b) synchronized batch normalization, and (c) distributed higher-order optimization methods. We show excellent scalability of this framework on both cloud as well as HPC clusters, and report on the interplay between bandwidth, network topology and bare metal vs cloud. We deploy this approach to train generative models of sizes hitherto not possible, showing that neural PDE solvers can be viably trained for practical applications. We also demonstrate that distributed higher-order optimization methods are 2-3× faster than stochastic gradient-based methods and provide minimal convergence drift with higher batch-size.

PDEs↗
DOE OSTI2017

Near real time ETEM streaming video analysis

The Environmental Transmission Electron Microscopy (ETEM) provides a powerful tool to observe the formation and evolution of nano-particles over time. However, ETEM generates extremely large amounts of data at level of 3GB/s, which impossible to be analyzed by manually processing or even by using a single PC. Moreover, the image stream obtained from the ETEM is very noisy. In this project, our goal is automatically analyze the physical characteristics of the nanoparticles. We proposed an approach that detect the nano-particles in each frame, and then track all the nano-particles over time, finally we can analyze the dynamical physical characteristics of the nano-particles, such as merging, absorbing, size and distance change over time. Specifically, our proposed approach detects the nano-particles in each frame independently, which could be highly parallelized. The experimental results show the proposed model could detect and track nano-particles robustly.

97 MATHEMATICS AND COMPUTING↗