Search NASA⌕ Search

SEARCH · Search NASA

Results for “Parallel computers.”

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 1,657 records · Page 92

End-to-End Data Movement Using MPI-IO Over Routed Terabots Infrastructures

Scientific discovery is nowadays driven by large-scale simulations running on massively parallel high-performance computing (HPC) systems. These applications each generate a large amount of data, which then needs to be post-processed for example for data mining or visualization. Unfortunately, the computing platform used for post processing might be different from the one on which the data is initially generated, introducing the challenge of moving large amount of data between computing platforms. This is especially challenging when these two platforms are geographically separated since the data needs to be moved between computing facilities. This is even more critical when scientists tightly couple their domain specific applications with a post processing application. The paper presents a solution for the data transfer between MPI applications using a dedicated wide area network (WAN) terabit infrastructure. The proposed solution is based on parallel access to data files and the Message Passing Interface (MPI) over the Common Communication Infrastructure (CCI) for the data transfer over a routed infrastructure. In the context of this research, the Energy Sciences Network (ESnet) of the U.S. Department of Energy (DOE) is targeted for the transfer of data between DOE national laboratories.

Vallee, Geoffroy R↗

Efficient Parallelization of Irregular Applications on GPU Architectures

With the enlarging computation capacity of general Graphics Processing Units (GPUs), leveraging GPUs to accelerate parallel applications has become a critical topic in academia and industry. However, a wide range of irregular applications with the computation-/memory-intensive nature cannot easily achieve high GPU utilization. The challenges mainly involve the following aspects: first, data dependence leads to coarse-grained kernel and inefficient parallelism; second, heavy GPU memory usage may cause frequent memory evictions and extra overhead of I/O; third, specific computation patterns produce memory redundancies; last, workload balance and data reusability conjunctly benefit the overall performance, but there may exist a dynamic trade-off between them. Targeting these challenges, this dissertation proposes multiple optimizations to accelerate two real-world applications: many-body correlation functions to simulate nuclear physics in a large-scale scientific system; the other is the eALS-based matrix factorization recommendation system. To accelerate the calculations of many-body correlation functions, this dissertation presents three frameworks in GPU memory management and multi-GPU scheduling. Firstly, an optimized systematic GPU memory management framework, MemHC, utilizes a series of new memory reduction designs in GPU memory allocation, CPU/GPU communications, and GPU memory oversubscription. Secondly, an enhanced multi-GPU scheduling framework, MICCO, particularly by taking both data dimension (e.g., data reuse and data eviction) and computation dimension into account. MICCO designs a heuristic scheduling algorithm and a machine learning-based regression model to generate the optimal settings of a proposed new concept to manage the trade-off. Thirdly, a locality-aware multi-GPU scheduling framework. This scheduler leverages pipeline batch generation with a looking-ahead strategy by building local dependency graphs for memory transfer reduction and better data reuse, achieving up to 79.92% memory cost reduction and 1.67x speedup. To parallelize the eALS-based recommendation system, this dissertation proposes an efficient CPU/GPU heterogeneous recommendation system, HEALS. HEALS employs newly designed architecture-adaptive data formats to achieve load balance and good data locality on CPU and GPU. To mitigate the data dependence, HEALS presents a CPU/GPU collaboration model for both task parallelism and data parallelism with multiple kernel computation optimizations. In summary, this dissertation efficiently accelerates two typical irregular applications on GPUs by building four frameworks, including CPU/GPU collaboration, GPU memory management, and multi-GPU scheduling.

Wang, Qihan↗

Automatic Differentiation for Adjoint Stencil Loops

Stencil loops are a common motif in computations including convolutional neural networks, structured-mesh solvers for partial differential equations, and image processing. Stencil loops are easy to parallelise, and their fast execution is aided by compilers, libraries, and domain-specific languages. Reverse-mode automatic differentiation, also known as algorithmic differentiation, autodiff, adjoint differentiation, or back-propagation, is sometimes used to obtain gradients of programs that contain stencil loops. Unfortunately, conventional automatic differentiation results in a memory access pattern that is not stencil-like and not easily parallelisable. In this paper we present a novel combination of automatic differentiation and loop transformations that preserves the structure and memory access pattern of stencil loops, while computing fully consistent derivatives. The generated loops can be parallelised and optimised for performance in the same way and using the same tools as the original computation. We have implemented this new technique in the Python tool PerforAD, which we release with this paper along with test cases derived from seismic imaging and computational fluid dynamics applications.

Automatic Differentiation↗

Large-scale atomistic calculations of clusters in intense x-ray pulses

Here, we present the methodology of our recently developed Monte-Carlo/ Molecular-Dynamics method for studying the fundamental ultrafast dynamics induced by high-fluence, high-intensity x-ray free electron laser (XFEL) pulses in clusters. The quantum nature of the initiating ionization process is accounted for by a Monte Carlo method to calculate probabilities of electronic transitions, including photo absorption, inner-shell relaxation, photon scattering, electron collision and recombination dynamics, and thus track the transient electronic configurations explicitly. The freed electrons and ions are followed by classical particle trajectories using a molecular dynamics algorithm. These calculations reveal the surprising role of electron-ion recombination processes that lead to the development of nonuniform spatial charge density profiles in x-ray excited clusters over femtosecond timescales. In the high-intensity limit, it is important to include the recombination dynamics in the calculated scattering response even for a 2- fs pulse. We also demonstrate that our numerical codes and algorithms can make e!cient use of the computational power of massively parallel supercomputers to investigate the intense-field dynamics in systems with increasing complexity and size at the ultrafast timescale and in non-linear x-ray interaction regimes. In particular, picosecond trajectories of XFEL clusters with attosecond time resolution containing millions of particles can be e!ciently computed on upwards of 262,144 processes.

74 ATOMIC AND MOLECULAR PHYSICS↗

Reducing the bulk of the bulk synchronous parallel model

For over two decades the dominant means for enabling portable performance of computational science and engineering applications on parallel processing architectures has been the bulk-synchronous parallel programming (BSP) model. Code developers, motivated by performance considerations to minimize the number of messages transmitted, have typically pursued a strategy of aggregating message data into fewer, larger messages. Emerging and future high-performance architectures, especially those seen as targeting Exascale capabilities, provide motivation and capabilities for revisiting this approach. In this paper we explore alternative configurations within the context of a large-scale complex multi-physics application and a proxy that represents its behavior, presenting results that demonstrate some important advantages as the number of processors increases in scale.

97 MATHEMATICS AND COMPUTING↗

psst: An open-source power system simulation toolbox in Python

This paper is an overview of Power System Simulation Toolbox (psst). psst is an open-source Python application for the simulation and analysis of power system models. psst simulates the wholesale market operation by solving a DC Optimal Power Flow (DCOPF), Security Constrained Unit Commitment (SCUC) and a Security Constrained Economic Dispatch (SCED). psst also includes models for the various entities in a power system such as Generator Companies (GenCos), Load Serving Entities (LSEs) and an Independent System Operator (ISO). psst features an open modular object oriented architecture that will make it useful for researchers to customize, expand, experiment beyond solving traditional problems. psst also includes a web based Graphical User Interface (GUI) that allows for user friendly interaction and for implementation on remote High Performance Computing (HPCs) clusters for parallelized operations. This paper also provides an illustrative application of psst and benchmarks with standard IEEE test cases to show the advanced features and the performance of toolbox.

97 MATHEMATICS AND COMPUTING↗

Multiscale Mathematics for Biomass Conversion to Renewable Hydrogen

Our two key accomplishments in the first three years were towards the development of, (1) a mathematically rigorous and at the same time computationally flexible framework for parallelization of Kinetic Monte Carlo methods, and its implementation on GPUs, and (2) spatial multilevel coarse-graining methods for Monte Carlo sampling and molecular simulation. A common underlying theme in both these lines of our work is the development of numerical methods which are at the same time both computationally efficient and reliable, the latter in the sense that they provide controlled-error approximations for coarse observables of the simulated molecular systems. Finally, our key accomplishment in the last year of the grant is that we started developing (3) pathwise information theory-based and goal-oriented sensitivity analysis and parameter identification methods for complex high-dimensional dynamics and in particular of nonequilibrium extended (high-dimensional) systems. We discuss these three research directions in some detail below, along with the related publications.

97 MATHEMATICS AND COMPUTING↗

High-Performance Computer Modeling of the Cosmos-Iridium Collision

This paper describes the application of a new, integrated modeling and simulation framework, encompassing the space situational awareness (SSA) enterprise, to the recent Cosmos-Iridium collision. This framework is based on a flexible, scalable architecture to enable efficient simulation of the current SSA enterprise, and to accommodate future advancements in SSA systems. In particular, the code is designed to take advantage of massively parallel, high-performance computer systems available, for example, at Lawrence Livermore National Laboratory. We will describe the application of this framework to the recent collision of the Cosmos and Iridium satellites, including (1) detailed hydrodynamic modeling of the satellite collision and resulting debris generation, (2) orbital propagation of the simulated debris and analysis of the increased risk to other satellites (3) calculation of the radar and optical signatures of the simulated debris and modeling of debris detection with space surveillance radar and optical systems (4) determination of simulated debris orbits from modeled space surveillance observations and analysis of the resulting orbital accuracy, (5) comparison of these modeling and simulation results with Space Surveillance Network observations. We will also discuss the use of this integrated modeling and simulation framework to analyze the risks and consequences of future satellite collisions and to assess strategies for mitigating or avoiding future incidents, including the addition of new sensor systems, used in conjunction with the Space Surveillance Network, for improving space situational awareness.

42 ENGINEERING↗

Simulating the Effect of Fracture Connectivity on Repository Performance with GDSA Framework - 18589

This work examines the possible migration of radionuclides from a deep geologic nuclear waste repository sited in fractured crystalline host rock. The key safety concern is the potential effect on waste isolation of the inter-connectivity of the fracture network, which is primarily established by the temporal evolution of the temperature and stress fields at the time of original rock intrusion and cooling. Two end members are considered in these simulations, one with a high degree of connectivity to the biosphere, such that advective transport through the fracture and fault network controls radionuclide migration, and the other with a low degree of connectivity, such that slow diffusive transport through the crystalline rock matrix is the controlling process for radionuclide migration to the biosphere. Both end members have several areally extensive, high-transmissivity deformation zones, whose distance to the repository is an important factor for waste isolation capability. Uncertainties in fracture properties (transmissivity, orientation, radius) and fracture distribution also give rise to uncertainty in waste isolation capability (i.e., the overall connectivity of the repository with the surface biosphere). These types of natural system uncertainties are generally present in any geologic site-characterization program and thus represent an important factor in assessing repository performance in hard rock environments. The waste considered in these simulations are spent fuel rod assemblies from the U.S. commercial nuclear reactor fleet. Thermal output of the waste must be considered in the simulations and can influence the rate and timing of waste package failure, waste form degradation, and fluid flux due to thermal expansion around the repository horizon. To achieve a high-fidelity representation of radionuclide transport in fractures and rock matrix, combined with thermal energy transport and fluid flow in fractures and matrix, the mathematical model is solved numerically in a parallel high-performance computing (HPC) environment on a finite volume unstructured grid consisting of approximately 5 million cells, using Geologic Disposal Safety Assessment (GDSA) Framework (https://pa.sandia.gov), an open-source performance assessment tool for deep underground disposal of nuclear waste. GDSA Framework uses PFLOTRAN to solve the balance equations on a three-dimensional grid with heterogeneous properties, using multiple processors in a parallel configuration based on domain decomposition. The fracture networks in these simulations are originally generated as discrete fracture networks (DFNs), which are sets of two-dimensional planes distributed in a three-dimensional domain. The method used in GDSA Framework maps the stochastically generated DFN to an equivalent continuous porous medium (ECPM) domain that allows for the simulation of coupled heat flow, fluid flow, and radionuclide transport, including heat conduction through the matrix of the fractured rock, which is a process not easily modeled in a DFN representation. Computational efficiency is also greatly enhanced using the ECPM method, allowing for a realistic representation and analysis of uncertainties in a multi-realization performance assessment of a deep geologic repository. The effect of fracture connectivity on the waste isolation safety function, as brought to light by these GDSA Framework simulations, points to the importance of including a realistic representation of uncertainties in fracture properties and distribution (effectively, uncertainty in spatial heterogeneity) in repository safety assessment simulations. (authors)

12 MANAGEMENT OF RADIOACTIVE AND NON-RADIOACTIVE W↗

An accelerated threshold-based back-projection algorithm for Compton camera image reconstruction

Purpose: Compton camera imaging (CCI) systems are currently under investigation for radiotherapy dose reconstruction and verification. The ability of such a system to provide real-time images during dose delivery will be limited by the computational speed of the image reconstruction algorithm. In this work, the authors present a fast and simple method by which to generate an initial back-projected image from acquired CCI data, suitable for use in a filtered back-projection algorithm or as a starting point for iterative reconstruction algorithms, and compare its performance to the current state of the art. Methods: Each detector event in a CCI system describes a conical surface that includes the true point of origin of the detected photon. Numerical image reconstruction algorithms require, as a first step, the back-projection of each of these conical surfaces into an image space. The algorithm presented here first generates a solution matrix for each slice of the image space by solving the intersection of the conical surface with the image plane. Each element of the solution matrix is proportional to the distance of the corresponding voxel from the true intersection curve. A threshold function was developed to extract those pixels sufficiently close to the true intersection to generate a binary intersection curve. This process is repeated for each image plane for each CCI detector event, resulting in a three-dimensional back-projection image. The performance of this algorithm was tested against a marching algorithm known for speed and accuracy. Results: The threshold-based algorithm was found to be approximately four times faster than the current state of the art with minimal deficit to image quality, arising from the fact that a generically applicable threshold function cannot provide perfect results in all situations. The algorithm fails to extract a complete intersection curve in image slices near the detector surface for detector event cones having axes nearly parallel to the image plane. This effect decreases the sum of the image, thereby also affecting the mean, standard deviation, and SNR of the image. All back-projected events associated with a simulated point source intersected the voxel containing the source and the FWHM of the back-projected image was similar to that obtained from the marching method. Conclusions: The slight deficit to image quality observed with the threshold-based back-projection algorithm described here is outweighed by the 75% reduction in computation time. The implementation of this method requires the development of an optimum threshold function, which determines the overall accuracy of the method. This makes the algorithm well-suited to applications involving the reconstruction of many large images, where the time invested in threshold development is offset by the decreased image reconstruction time. Implemented in a parallel-computing environment, the threshold-based algorithm has the potential to provide real-time dose verification for radiation therapy.

60 APPLIED LIFE SCIENCES↗

Enhancing scalability of a matrix-free eigensolver for studying many-body localization

We propose several techniques to enhance the parallel scalability of a matrix-free eigensolver designed for studying many-body localization (MBL) of quantum spin chain models with nearest-neighbor interactions and on-site disorder. This type of problem is computationally challenging because the dimension of the associated Hamiltonian matrix grows exponentially with respect to the number of spins L, and we need to average over different realizations of the random disorder to obtain relevant statistical behavior. For each disorder realization, we need to compute eigenvalues from different regions of the spectrum and their corresponding eigenvectors. In previous work, the interior eigenstates for a single eigenvalue problem are computed via the shift-and-invert Lanczos algorithm. Due to the extremely high memory footprint of the LU factorizations, this technique is not well suited for large L’s. For example, we need thousands of compute nodes on modern high performance computing infrastructures to go beyond L = 24. The matrix-free approach does not suffer from this memory bottleneck, however, its scalability is limited by a computation and communication load imbalance. To reduce this imbalance and to significantly enhance the scalability of the matrix-free eigensolver, we reorder the matrix and leverage the consistent space runtime, CSPACER. We also show its efficiency in managing irregular communication patterns at scale compared to optimized MPI non-blocking two-sided and one-sided RMA implementation variants. This effort enables us to study MBL for spin chains with a larger number of spins. The efficiency and effectiveness of the proposed algorithm is demonstrated by computing eigenstates on a massively parallel many-core high performance computer.

METIS↗

An O(N logN) multi-layer boundary element method for direct computation of sound propagation in shallow water environments

Direct three-dimensional (3D) numerical simulations of acoustic fields in range-dependent shallow water environments remains a challenge due to environmental complexities and large computational cost.We develop an efficient 3D boundary element method (BEM) for shallow water acoustic propagation which utilizes a Pre-corrected Fast Fourier Transform (PFFT) approach to reduce the computational effort from O ( N 2~3 ) to O ( N log N ) where N is the total number of boundary unknowns. To account for inhomogeneous media, the method allows for arbitrary number of coupled multi-layer BEM sub-domains. With O ( N log N ) efficiency and the use of massively parallel high-performance computing platforms, we are able to conduct multi-layer 3D direct simulations of low-mid frequency acoustics over kilometer ranges. Furthermore we perform extensive validations of the method and provide two shallow water waveguide examples benchmarked against theoretical solutions. To illustrate the efficacy and usefulness of the PFFT-BEM method, we perform 3D large-scale direct simulations to assess the performance of two established canonical models: axisymmetric coupled mode model for 3D seamount; and Kirchhoff approximation and perturbation theory for 3D rough surface scattering.

42 ENGINEERING↗

Parallel contingency statistics with Titan.

This report summarizes existing statistical engines in VTK/Titan and presents the recently parallelized contingency statistics engine. It is a sequel to [PT08] and [BPRT09] which studied the parallel descriptive, correlative, multi-correlative, and principal component analysis engines. The ease of use of this new parallel engines is illustrated by the means of C++ code snippets. Furthermore, this report justifies the design of these engines with parallel scalability in mind; however, the very nature of contingency tables prevent this new engine from exhibiting optimal parallel speed-up as the aforementioned engines do. This report therefore discusses the design trade-offs we made and study performance with up to 200 processors.

97 MATHEMATICS AND COMPUTING↗

Scalable Deep Learning via I/O Analysis and Optimization

Scalable deep neural network training has been gaining prominence because of the increasing importance of deep learning in a multitude of scientific and commercial domains. Consequently, a number of researchers have investigated techniques to optimize deep learning systems. Much of the prior work has focused on runtime and algorithmic enhancements to optimize the computation and communication. Despite these enhancements, however, deep learning systems still suffer from scalability limitations, particularly with respect to data I/O. This situation is especially true for training models where the computation can be effectively parallelized, leaving I/O as the major bottleneck. In fact, our analysis shows that I/O can take up to 90% of the total training time. Thus, in this article, we first analyze LMDB, the most widely used I/O subsystem of deep learning frameworks, to understand the causes of this I/O inefficiency. Based on our analysis, we propose LMDBIO—an optimized I/O plugin for scalable deep learning. LMDBIO includes six novel optimizations that together address the various shortcomings in existing I/O for deep learning. Lastly, our experimental results show that LMDBIO significantly outperforms LMDB in all cases and improves overall application performance by up to 65-fold on a 9,216-core system.

97 MATHEMATICS AND COMPUTING↗

Adding Data Management Services to Parallel File Systems

The objective of this project, called DAMASC for “Data Management in Scientific Computing”, is to coalesce data management with parallel file system management to present a declarative interface to scientists for managing, querying, and analyzing extremely large data sets efficiently and predictably. Managing extremely large data sets is a key challenge of exascale computing. The overhead, energy, and cost of moving massive volumes of data demand designs where computation is close to storage. In current architectures, compute/analysis clusters access data in a physically separate parallel file system and largely leave it scientist to reduce data movement. Over the past decades the high-end computing community has adopted middleware with multiple layers of abstractions and specialized file formats such as NetCDF-4 and HDF5. These abstractions provide a limited set of high-level data processing functions, but have inherent functionality and performance limitations: middleware that provides access to the highly structured contents of scientific data files stored in the (unstructured) file systems can only optimize to the extent that file system interfaces permit; the highly structured formats of these files often impedes native file system performance optimizations. We are developing Damasc, an enhanced high-performance file system with native rich data management services. Damasc will enable efficient queries and updates over files stored in their native byte-stream format while retaining the inherent performance of file system data storage via declarative queries and updates over views of underlying files. Damasc has four key benefits for the development of data-intensive scientific code: (1) applications can use important data-management services, such as declarative queries, views, and provenance tracking, that are currently available only within database systems; (2) the use of these services becomes easier, as they are provided within a familiar file-based ecosystem; (3) common optimizations, e.g., indexing and caching, are readily supported across several file formats, avoiding effort duplication; and (4) performance improves significantly, as data processing is integrated more tightly with data storage. Our key contributions are: SciHadoop which explores changes to MapReduce assumption by taking advantage of semantics of structured data while preserving MapReduce’s failure and resource management; DataMods which extends common abstractions of parallel file systems so they become programmable such that they can be extended to natively support a variety of data models and can be hooked into emerging distributed runtimes such as Stanford’s Legion; and Miso which combines Hadoop and relational data warehousing to minimize time to insight, taking into account the overhead of ingesting data into data warehousing.

96 KNOWLEDGE MANAGEMENT AND PRESERVATION↗

Computable and Operationally Meaningful Multipartite Entanglement Measures

Multipartite entanglement is an essential resource for quantum communication, quantum computing, quantum sensing, and quantum networks. The utility of a quantum state |ψ$\rangle$ for these applications is often directly related to the degree or type of entanglement present in |ψ$\rangle$. Therefore, efficiently quantifying and characterizing multipartite entanglement is of paramount importance. Here, in this work, we introduce a family of multipartite entanglement measures, called concentratable entanglements. Several well-known entanglement measures are recovered as special cases of our family of measures, and hence we provide a general framework for quantifying multipartite entanglement. We prove that the entire family does not increase, on average, under local operations and classical communications. We also provide an operational meaning for these measures in terms of probabilistic concentration of entanglement into Bell pairs. Finally, we show that these quantities can be efficiently estimated on a quantum computer by implementing a parallelized SWAP test, opening up a research direction for measuring multipartite entanglement on quantum devices.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Management and Storage of Scientific Data

Scientific discoveries rely heavily on efficient access, search, and management of massive data sets. Data management technologies have, for decades, provided foundational capabilities for scientific computing. Just as storage, input/output (I/O), and data management have been fundamental to simulation-based science for many years, so too are capable data-management technologies key to the success of today’s scientific workflows utilizing data intensive and machine learning (ML) techniques. The Department of Energy, Office of Science, Advanced Scientific Computing Research (ASCR) program has invested broadly in data-management research focused on high-performance computing (HPC) systems, from parallel file systems that store data to application software that makes these systems more productive. Still, advances in technology combined with growing diversity of supported science strongly motivate continued investment in this area. In January 2022, ASCR convened a workshop to identify priority research directions in the area of data management for high-performance and scientific computing. Attendees were challenged to identify promising approaches that would support the breadth of the DOE mission, including the explosion of artificial intelligence (AI) uses and the growing needs of experimental and observational science. Technological and science drivers were identified and considered as they relate to key aspects of data management such as interfaces, architectural design, and FAIR principles (Findable, Accessible, Interoperable, and Reusable). The thoughts of the workshop participants were distilled into a set of four priority research directions with the potential for high impact on DOE science. These research directions are summarized in the following pages.

96 KNOWLEDGE MANAGEMENT AND PRESERVATION↗

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↗