Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Scalable Algorithms”

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 685 records · Page 38

Benchmarking the Performance of Neuromorphic and Spiking Neural Network Simulators

Software simulators play a critical role in the development of new algorithms and system architectures in any field of engineering. Neuromorphic computing, which has shown potential in building brain-inspired energy-efficient hardware, suffers a slow-down in the development cycle due to a lack of flexible and easy-to-use simulators of either neuromorphic hardware itself or of spiking neural networks (SNNs), the type of neural network computation executed on most neuromorphic systems. While there are several openly available neuromorphic or SNN simulation packages developed by a variety of research groups, they have mostly targeted computational neuroscience simulations, and only a few have targeted small-scale machine learning tasks with SNNs. Evaluations or comparisons of these simulators have often targeted computational neuroscience-style workloads. In this work, we seek to evaluate the performance of several publicly available SNN simulators with respect to non-computational neuroscience workloads, in terms of speed, flexibility, and scalability. We evaluate the performance of the NEST, Brian2, Brian2GeNN, BindsNET and Nengo packages under a common front-end neuromorphic framework. Our evaluation tasks include a variety of different network architectures and workload types to mimic the computation common in different algorithms, including feed-forward network inference, genetic algorithms, and reservoir computing. We also study the scalability of each of these simulators when running on different computing hardware, from single core CPU workstations to multi-node supercomputers. Our results show that the BindsNET simulator has the best speed and scalability for most of the SNN workloads (sparse, dense, and layered SNN architectures) on a single core CPU. However, when comparing the simulators leveraging the GPU capabilities, Brian2GeNN outperforms the others for these workloads in terms of scalability. NEST performs the best for small sparse networks and is also the most flexible simulator in terms of reconfiguration capability NEST shows a speedup of at least 2x compared to the other packages when running evolutionary algorithms for SNNs. The multi-node and multi-thread capabilities of NEST show at least 2x speedup compared to the rest of the simulators (single core CPU or GPU based simulators) for large and sparse networks. We conclude our work by providing a set of recommendations on the suitability of employing these simulators for different tasks and scales of operations. We also present the characteristics for a future generic ideal SNN simulator for different neuromorphic computing workloads.

97 MATHEMATICS AND COMPUTING↗

Incremental Interval Assignment (IIA) for Scalable Mesh Preparation

Interval Assignment (IA) means selecting the number of mesh edges for each CAD curve. IIA is a discrete algorithm over integers. A priority queue iteratively selects compatible sets of intervals to increase in lock-step by integers. In contrast, the current capability in Cubit is floating-point Linear Programming with Branch-and-Bound for integerization (BBIA).

97 MATHEMATICS AND COMPUTING↗

Development and Evaluation of Algorithms to Improve Small- and Medium-Size Commercial Building Operations

Small- and medium-sized (<100,000 sf) commercial buildings (SMBs) represent over 95% of the U.S. commercial building stock and consume over 60% of total site energy consumption. Many of these buildings use rudimentary controls that are mostly manual, with limited scheduling capability, no monitoring or failure management. Therefore, many of these buildings are operated inefficiently and consume excess energy. SMBs typically utilize packaged rooftop units (RTUs) that are controlled by an individual thermostat. There is increased urgency to improve the operating efficiency of existing commercial building stock in the U.S. for many reasons, chief among them is to mitigate the climate change impacts. Studies have shown that managing set points and schedules of the RTUs will result in up to 20% energy and cost savings. Another problem associated with RTUs is short-cycling, where an RTU goes through ON and OFF cycles too frequently. Excessive cycling can lead to excessive wear and lead to premature failure of the compressor or its components. The short cycling can result in a significantly decreased average efficiency (up to 10%), even if there are no physical failures in the equipment. Also, SMBs use a time-of-day scheduling is to start the RTUs before the building will be occupied and shut it off when unoccupied. Ensuring correct use of the zone set points and eliminating frequent cycling of RTUs thereby leading to persistent building operations can significantly increase the operational efficiency of the SMBs. A growing trend is to use low-cost control infrastructure that can enable scalable and cost-effective intelligent building operations. The work reported in this report describes three algorithms for detecting the zone set point temperature, RTU cycling rate and occupancy schedule detection that can be deployed on the low-cost infrastructure. These algorithms only require the zone temperature data for detection. The algorithms have been tested and validated using field data from a number of RTUs from six buildings in different climate locations. Overall, the algorithms were successful in detecting the set points and ON/OFF cycles accurately using the peak detection technique and occupancy schedule using symbolic aggregate approximation technique. The report describes the three algorithms, results from testing the algorithms using field data, how the algorithms can be used to improve SMBs efficiency, and presents related conclusions.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗

Locally purified maximally mixed states at scale: Entanglement pruning and symmetries

Locally Purified Density Operators (LPDOs) are state-of-the-art tensor network ansatze candidates that efficiently represent mixed quantum states at scale. However, given their non-uniqueness, their representational complexity is generally sub-optimal in practical computations. Here, in this work we perform a comprehensive numerical and analytical analysis and resolve this issue in the experimentally relevant limit where noise depolarizes the density operator into a maximally mixed state. To resolve the sub-optimality issue, we analyze two numerical tools, one analytic method, and detail the relations between them. The numerical tools used are fidelity-preserving truncations and isometric gauge transformations leveraging Riemannian optimizations over entropic objective functions. In addition, by invoking the injectivity and symmetry constraints of the maximally mixed LPDO, we also present analytical closed-form expressions for the disentangler and discuss their relation to numerical optimizers. Further, away from the maximally mixed state, our simulations highlight how the truncation threshold smoothly interpolate, as a function of depolarization, between established matrix product results and our new results. Our work shows how, by minimizing the resources required to represent key states of practical interest in experiment, the efficiency of tensor network algorithms can be substantially increased. This paves the path for uncovering tensor network’s fundamental scalability limits and latent potential in representing the wide locus of mixed quantum states that are accessible on near-term quantum devices.

Gangapuram, Amit Jamadagni [Oak Ridge National Lab↗

Advanced Quantum Poisson Solver in the NISQ era

The Poisson equation has many applications across the broad areas of science and engineering. Most quantum algorithms for the Poisson solver presented so far, either suffer from lack of accuracy and/or are limited to very small sizes of the problem, and thus have no practical usage. Here we present an advanced quantum algorithm for solving the Poisson equation with high accuracy and dynamically tunable problem size. After converting the Poisson equation to the linear systems through the finite difference method, we adopt the Harrow-Hassidim-Lloyd (HHL) algorithm as the basic framework. Particularly, in this work we present an advanced circuit that ensures the accuracy of the solution by implementing non-truncated eigenvalues through eigenvalue amplification as well as by increasing the accuracy of the controlled rotation angular coefficients, which are the critical factors in the HHL algorithm. We show that our algorithm not only increases the accuracy of the solutions, but also composes more practical and scalable circuits by dynamically controlling problem size in the NISQ devices. We present both simulated and experimental results, and discuss the sources of errors. Finally, we conclude that overall results on the quantum hardware are dominated by the error in the CNOT gates.

Robson, Walter↗

Decentralized Collaborative Learning with Probabilistic Data Protection

We discuss future directions of Blockchain as a collaborative value co-creation platform, in which network participants can gain extra insights that cannot be accessed when disconnected from the others. As such, we propose a decentralized machine learning framework that is carefully designed to respect the values of democracy, diversity, and privacy. Specifically, we propose a federated multi-task learning framework that integrates a privacy-preserving dynamic consensus algorithm. We show that a specific network topology called the expander graph dramatically improves the scalability of global consensus building. We conclude the paper by making some remarks on open problems.

Ide, Tsuyoshi↗

Scalable Causal Graph Learning through a Deep Neural Network

Learning the causal graph in a complex system is crucial for knowledge discovery and decision making, yet it remains a challenging problem because of the unknown nonlinear interaction among system components. Most of the existing methods either rely on predefined kernel or data distribution, or they focus simply on the causality between a single target and the remaining system. This work presents a deep neural network for scalable causal graph learning (SCGL) through low-rank approximation. The SCGL model can explore nonlinearity on both temporal and intervariable relationships without any predefined kernel or distribution assumptions. Through low-rank approximation, the noise influence is reduced, and better accuracy and high scalability are achieved. Experiments using synthetic and real-world datasets show that our SCGL algorithm outperforms existing state-of-the-art methods for causal graph learning.

97 MATHEMATICS AND COMPUTING↗

Parallel auto-correlative statistics with VTK.

This report summarizes existing statistical engines in VTK and presents both the serial and parallel auto-correlative statistics engines. It is a sequel to [PT08, BPRT09b, PT09, BPT09, PT10] which studied the parallel descriptive, correlative, multi-correlative, principal component analysis, contingency, k-means, and order statistics engines. The ease of use of the new parallel auto-correlative statistics engine is illustrated by the means of C++ code snippets and algorithm verification is provided. This report justifies the design of the statistics engines with parallel scalability in mind, and provides scalability and speed-up analysis results for the autocorrelative statistics engine.

Pebay, Philippe Pierre↗

Minimal Energy Routing of a Leader and a Wingmate with Periodic Connectivity

We consider a route planning problem in which two unmanned vehicles are required to complete a set of tasks present at distinct locations, referred to as targets, with minimum energy consumption. The mission environment is hazardous, and to ensure a safe operation, the UVs are required to communicate with each other at every target they visit. The problem objective is to determine the allocation of the tasks to the UVs and plan tours for the UVs to visit the targets such that the weighted sum of the distances traveled by the UVs and the distances traveled by the communicating signals between them is minimized. We formulate this problem as an Integer program and show that naively solving the problem using commercially available off-the-shelf solvers is insufficient in determining scalable solutions efficiently. To address this computational challenge, we develop an approximation and a heuristic algorithm, and employ them to compute high-quality solutions to a special case of the problem where equal weights are assigned to the distances traveled by the vehicles and the communicating signals. For this special case, we show that the approximation algorithm has a fixed approximation ratio of 3.75. We also develop lower bounds to the optimal cost of the problem to evaluate the performance of these algorithms on large-scale instances. We demonstrate the performance of these algorithms on 500 randomly generated instances with the number of targets ranging from 6 to 100, and show that the algorithms provide high-quality solutions to the problem swiftly; the average computation time of the algorithmic solutions is within a fraction of a second for instances with at most 100 targets. Finally, we show that the approximation ratio has a variable ratio for the weighted case of the problem. Specifically, if ρ denotes the ratio of the weights assigned to the distances representing the communication and travel costs, the algorithm has an a posteriori ratio of $3 + \frac{3ρ}{4}$ when ρ ≥ 1, and $\frac{3}{ρ}$ + $\frac{3}{4}$ when ρ ≤ 1.

42 ENGINEERING↗

Integrated Transmission-Distribution Multi-Period Switching for Wildfire Risk Mitigation: Improving Speed and Scalability with Distributed Optimization: Preprint

With increasingly severe wildfire conditions driven by climate change, utilities must manage the risk of wildfire ignitions from electric power lines. During "public safety power shutoff'" events, utilities de-energize power lines to reduce wildfire ignition risk, which may result in load shedding. Distributed energy resources provide flexibility that can help support the system to reduce load shedding when lines are de-energized. We investigate a coordinated transmission-distribution optimization problem that balances wildfire risk mitigation and load shedding. We model distribution systems that include battery energy storage systems which may support loads when transmission lines are de-energized. This multi-period integrated transmission-distribution optimal switching problem jointly optimizes line switching decisions, the generators' setpoints, load shedding, and the batteries' states of charge, resulting in significant computational challenges. To improve scalability, we decompose the problem over both space and time and apply a distributed optimization algorithm. Using a large-scale synthetic California test case with realistic distribution models and real wildfire risk data, we show that distributed optimization can solve large-scale multi-period switching problems that are otherwise intractable for centralized solvers. We also discuss challenges and future directions for improving the distributed algorithm's convergence performance as the number of time periods increases.

24 POWER TRANSMISSION AND DISTRIBUTION↗

The Impact of a Fault Tolerant MPI on Scalable Systems Services and Applications

Exascale targeted scientific applications must be prepared for a highly concurrent computing environment where failure will be a regular event during execution. Natural and algorithm-based fault tolerance (ABFT) techniques can often manage failures more efficiently than traditional checkpoint/restart techniques alone. Central to many petascale applications is an MPI standard that lacks support for ABFT. The Run-Through Stabilization (RTS) proposal, under consideration for MPI 3, allows an application to continue execution when processes fail. The requirements of scalable, fault tolerant MPI implementations and applications will stress the capabilities of many system services. System services must evolve to efficiently support such applications and libraries in the presence of system component failures. This paper discusses how the RTS proposal impacts system services, highlighting specific requirements. Early experimentation results from Cray systems at ORNL using prototype MPI and runtime implementations are presented. Additionally, this paper outlines fault tolerance techniques targeted at leadership class applications.

99 GENERAL AND MISCELLANEOUS↗

The development of laser-plasma interaction program LAP3D on thousands of processors

Modeling laser-plasma interaction (LPI) processes in real-size experiments scale is recognized as a challenging task. For explorering the influence of various instabilities in LPI processes, a three-dimensional laser and plasma code (LAP3D) has been developed, which includes filamentation, stimulated Brillouin backscattering (SBS), stimulated Raman backscattering (SRS), non-local heat transport and plasmas flow computation modules. In this program, a second-order upwind scheme is applied to solve the plasma equations which are represented by an Euler fluid model. Operator splitting method is used for solving the equations of the light wave propagation, where the Fast Fourier translation (FFT) is applied to compute the diffraction operator and the coordinate translations is used to solve the acoustic wave equation. The coupled terms of the different physics processes are computed by the second-order interpolations algorithm. In order to simulate the LPI processes in massively parallel computers well, several parallel techniques are used, such as the coupled parallel algorithm of FFT and fluid numerical computation, the load balance algorithm, and the data transfer algorithm. Now the phenomena of filamentation, SBS and SRS have been studied in low-density plasma successfully with LAP3D. Scalability of the program is demonstrated with a parallel efficiency above 50% on about ten thousand of processors.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

HPDR: High-Performance Portable Scientific Data Reduction Framework

The rapid growth in scientific data generation is outpacing advancements in computing systems necessary for efficient storage, transfer, and analysis, particularly in the context of exascale computing. With the deployment of first-generation exascale computing systems and next-generation experimental facilities, this gap is widening and necessitates effective data reduction techniques to manage enormous data volumes. Over the past decade, various data reduction methods, including lossless compression, error-controlled lossy compression, and data refactoring, have been developed to accelerate I/O in scientific workflows. Despite significant reductions in data volume, these methods introduce considerable computational overhead, which can become the new bottleneck in data processing. To mitigate this, GPU-accelerated data reduction algorithms have been introduced. However, challenges remain in their integration into exascale workflows, including limited portability across different GPU architectures, substantial memory transfer overhead, and reduced scalability on dense multi-GPU systems. To address these challenges, we propose HPDR, a high-performance and portable data reduction framework. HPDR is designed to enable the execution of state-of-the-art reduction algorithms across diverse processor architectures while reducing memory transfer overhead to 2.3 % of the original, resulting in up to 3.5× faster throughput compared to existing solutions. It also achieves up to 96% of the theoretical speedup in multi-GPU settings. In addition, evaluations on accelerating I/O operations at scale up to 1,024 nodes of the Frontier supercomputer demonstrate that HPDR can achieve up to 103 TB/s reduction throughput, providing up to 4× acceleration in parallel I/O performance compared to existing data reduction routines. This work highlights the potential of HPDR to significantly enhance data reduction efficiency in exascale computing environments.

Chen, Jieyang [University of Oregon]↗

Algebraic multigrid domain and range decomposition (AMG-DD / AMG-RD)*

In modern large-scale supercomputing applications, algebraic multigrid (AMG) is a leading choice for solving matrix equations. However, the high cost of communication relative to that of computation is a concern for the scalability of traditional implementations of AMG on emerging architectures. This paper introduces two new algebraic multilevel algorithms, algebraic multigrid domain decomposition (AMG-DD) and algebraic multigrid range decomposition (AMG-RD), that replace traditional AMG V-cycles with a fully overlapping domain decomposition approach. While the methods introduced here are similar in spirit to the geometric methods developed by Brandt and Diskin [Multigrid solvers on decomposed domains, in Domain Decomposition Methods in Science and Engineering, Contemp. Math. 157, AMS, Providence, RI, 1994, pp. 135--155], Mitchell [Electron. Trans. Numer. Anal., 6 (1997), pp. 224--233], and Bank and Holst [SIAM J. Sci. Comput., 22 (2000), pp. 1411--1443], they differ primarily in that they are purely algebraic: AMG-RD and AMG-DD trade communication for computation by forming global composite “grids” based only on the matrix, not the geometry. (As is the usual AMG convention, “grids” here should be taken only in the algebraic sense, regardless of whether or not it corresponds to any geometry.) Another important distinguishing feature of AMG-RD and AMG-DD is their novel residual communication process that enables effective parallel computation on composite grids, avoiding the all-to-all communication costs of the geometric methods. The main purpose of this paper is to study the potential of these two algebraic methods as possible alternatives to existing AMG approaches for future parallel machines. As a result, this paper develops some theoretical properties of these methods and reports on serial numerical tests of their convergence properties over a spectrum of problem parameters.

97 MATHEMATICS AND COMPUTING↗

Hijacking a rapid and scalable metagenomic method reveals subgenome dynamics and evolution in polyploid plants

Premise: The genomes of polyploid plants archive the evolutionary events leading to their present forms. However, plant polyploid genomes present numerous hurdles to the genome comparison algorithms for classification of polyploid types and exploring genome dynamics. Methods: Here, the problem of intra- and inter-genome comparison for examining polyploid genomes is reframed as a metagenomic problem, enabling the use of the rapid and scalable MinHashing approach. To determine how types of polyploidy are described by this metagenomic approach, plant genomes were examined from across the polyploid spectrum for both k-mer composition and frequency with a range of k-mer sizes. In this approach, no subgenome-specific k-mers are identified; rather, whole-chromosome k-mer subspaces were utilized. Results: Given chromosome-scale genome assemblies with sufficient subgenome-specific repetitive element content, literature-verified subgenomic and genomic evolutionary relationships were revealed, including distinguishing auto- from allopolyploidy and putative progenitor genome assignment. The sequences responsible were the rapidly evolving landscape of transposable elements. An investigation into the MinHashing parameters revealed that the downsampled k-mer space (genomic signatures) produced excellent approximations of sequence similarity. Furthermore, the clustering approach used for comparison of the genomic signatures is scrutinized to ensure applicability of the metagenomics-based method. Discussion: The easily implementable and highly computationally efficient MinHashing-based sequence comparison strategy enables comparative subgenomics and genomics for large and complex polyploid plant genomes. Such comparisons provide evidence for polyploidy-type subgenomic assignments. In cases where subgenome-specific repeat signal may not be adequate given a chromosomes' global k-mer profile, alternative methods that are more specific but more computationally complex outperform this approach.

59 BASIC BIOLOGICAL SCIENCES↗

A Linear-Time Algorithm and Analysis of Graph Relative Hausdorff Distance

Graph similarity metrics serve far-ranging purposes across many domains in data science. As graph datasets grow in size, scientists need comparative tools that capture meaningful differences, yet are lightweight and scalable. Graph Relative Hausdorff (RH) distance is a promising, recently proposed measure for quantifying degree distribution similarity. In spite of recent interest in RH distance, little is known about its properties. Here, we conduct an algorithmic and analytic study of RH distance. In particular, we provide the first linear-time algorithm for computing RH distance, analyze examples of RH distance between families of graphs, and prove several analytic results concerning the range, density, and extremal behavior of RH distance values.

Relative Hausdorff distance, degree distribution, ↗

Improving scalability of parallel CNN training by adaptively adjusting parameter update frequency

Synchronous SGD with data parallelism, the most popular parallelization strategy for CNN training, suffers from the expensive communication cost of averaging gradients among all workers. The iterative parameter updates of SGD cause frequent communications and it becomes the performance bottleneck. In this paper, we propose a lazy parameter update algorithm that adaptively adjusts the parameter update frequency to address the expensive communication cost issue. Our algorithm accumulates the gradients if the difference of the accumulated gradients and the latest gradients is sufficiently small. Here, the less frequent parameter updates reduce the per-iteration communication cost while maintaining the model accuracy. Our experimental results demonstrate that the lazy update method remarkably improves the scalability while maintaining the model accuracy. For ResNet50 training on ImageNet, the proposed algorithm achieves a significantly higher speedup (739.6 on 2048 Cori KNL nodes) as compared to the vanilla synchronous SGD (276.6) while the model accuracy is almost not affected (<0.2% difference).

97 MATHEMATICS AND COMPUTING↗

Rolling Horizon Based Temporal Decomposition for the Offline Pickup and Delivery Problem with Time Windows

The offline pickup and delivery problem with time windows (PDPTW) is a classical combinatorial optimization problem in the transportation community, which has proven to be very challenging computationally. Due to the complexity of the problem, practical problem instances can be solved only via heuristics, which trade-off solution quality for computational tractability. Among the various heuristics, a common strategy is problem decomposition, that is, the reduction of a large-scale problem into a collection of smaller sub-problems, with spatial and temporal decompositions being two natural approaches. While spatial decomposition has been successful in certain settings, effective temporal decomposition has been challenging due to the difficulty of stitching together the sub-problem solutions across the decomposition boundaries. In this work, we introduce a novel temporal decomposition scheme for solving a class of PDPTWs that have narrow time windows, for which it is able to provide both fast and high-quality solutions. We utilize techniques that have been popularized recently in the context of online dial-a-ride problems along with the general idea of rolling horizon optimization. To the best of our knowledge, this is the first attempt to solve offline PDPTWs using such an approach. To show the performance and scalability of our framework, we use the optimization of paratransit services as a motivating example. Due to the lack of benchmark solvers similar to ours (i.e., temporal decomposition with an online solver), we compare our results with an offline heuristic algorithm using Google OR-Tools. In smaller problem instances (with an average of 129 requests per instance), the baseline approach is as competitive as our framework. However, in larger problem instances (approximately 2,500 requests per instance), our framework is more scalable and can provide good solutions to problem instances of varying degrees of difficulty, while the baseline algorithm often fails to find a feasible solution within comparable compute times.

Kim, Youngseo↗