Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “algorithms scalability”

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 1,225 records · Page 68

Probabilistic Neural Computing with Stochastic Devices

Abstract The brain has effectively proven a powerful inspiration for the development of computing architectures in which processing is tightly integrated with memory, communication is event‐driven, and analog computation can be performed at scale. These neuromorphic systems increasingly show an ability to improve the efficiency and speed of scientific computing and artificial intelligence applications. Herein, it is proposed that the brain's ubiquitous stochasticity represents an additional source of inspiration for expanding the reach of neuromorphic computing to probabilistic applications. To date, many efforts exploring probabilistic computing have focused primarily on one scale of the microelectronics stack, such as implementing probabilistic algorithms on deterministic hardware or developing probabilistic devices and circuits with the expectation that they will be leveraged by eventual probabilistic architectures. A co‐design vision is described by which large numbers of devices, such as magnetic tunnel junctions and tunnel diodes, can be operated in a stochastic regime and incorporated into a scalable neuromorphic architecture that can impact a number of probabilistic computing applications, such as Monte Carlo simulations and Bayesian neural networks. Finally, a framework is presented to categorize increasingly advanced hardware‐based probabilistic computing technologies.

Misra, Shashank↗

Unorthodox parallelization for Bayesian quantum state estimation

Quantum state tomography (QST) allows for the reconstruction of quantum states through measurements and some inference technique under the assumption of repeated state preparations. Bayesian inference provides a promising platform to achieve both efficient QST and accurate uncertainty quantification, yet is generally plagued by the computational limitations associated with long Markov chains. In this work, we present a novel Bayesian QST approach that leverages modern distributed parallel computer architectures to efficiently sample a D-dimensional Hilbert space. Using a parallelized preconditioned Crank–Nicholson Metropolis–Hastings algorithm, we demonstrate our approach on simulated data and experimental results from IBM Quantum systems up to four qubits, showing significant speedups through parallelization. Although highly unorthodox in pooling independent Markov chains, our method proves remarkably practical, with validation ex post facto via diagnostics like the intrachain autocorrelation time. We conclude by discussing scalability to higher-dimensional systems, offering a path toward efficient and accurate Bayesian characterization of large quantum systems.

Bayesian inference↗

Data-Driven Multi-agent Deep Reinforcement Learning for Distribution System Decentralized Voltage Control with High Penetration of PVs

This paper proposes a novel model-free/data-driven centralized training and decentralized execution multi-agent deep reinforcement learning (MADRL) framework for distribution system voltage control with high penetration of PVs. The proposed MADRL can coordinate both the real and reactive power control of PVs with existing static var compensators and battery storage systems. Unlike the existing DRL-based voltage control methods, our proposed method does not rely on a system model during both the training and execution stages. This is achieved by developing a new interaction scheme between the surrogate modeling of the original system and the multi-agent soft actor critic (MASAC) MADRL algorithm. In particular, the sparse pseudo-Gaussian process with a few-shots of measurements is utilized to construct the surrogate model of the original environment, i.e., power flow model. This is a data-driven process and no model parameters are needed. Furthermore, the MASAC enabled MADRL allows to achieve better scalability by dividing the original system into different voltage control regions with the aid of real and reactive power sensitivities to voltage, where each region is treated as an agent. This also serves as the foundation for the centralized training and decentralized execution, thus significantly reducing the communication requirements as only local measurements are required for control. Comparative results with other alternatives on the IEEE 123-nodes and 342-nodes systems demonstrate the superiority of the proposed method.

14 SOLAR ENERGY↗

Quantum computing in power systems

Electric power systems provide the backbone of modern industrial societies. Enabling scalable grid analytics is the keystone to successfully operating large transmission and distribution systems. However, today's power systems are suffering from ever-increasing computational burdens in sustaining the expanding communities and deep integration of renewable energy resources, as well as managing huge volumes of data accordingly. These unprecedented challenges call for transformative analytics to support the resilient operations of power systems. Recently, the explosive growth of quantum computing techniques has ignited new hopes of revolutionizing power system computations. Quantum computing harnesses quantum mechanisms to solve traditionally intractable computational problems, which may lead to ultra-scalable and efficient power grid analytics. This paper reviews the newly emerging application of quantum computing techniques in power systems. We present a comprehensive overview of existing quantum-engineered power analytics from different operation perspectives, including static analysis, transient analysis, stochastic analysis, optimization, stability, and control. We thoroughly discuss the related quantum algorithms, their benefits and limitations, hardware implementations, and recommended practices. We also review the quantum networking techniques to ensure secure communication of power systems in the quantum era. Finally, we discuss challenges and future research directions. This paper will hopefully stimulate increasing attention to the development of quantum-engineered smart grids.

24 POWER TRANSMISSION AND DISTRIBUTION↗

Scalable deep learning for watershed model calibration

Watershed models such as the Soil and Water Assessment Tool (SWAT) consist of high-dimensional physical and empirical parameters. These parameters often need to be estimated/calibrated through inverse modeling to produce reliable predictions on hydrological fluxes and states. Existing parameter estimation methods can be time consuming, inefficient, and computationally expensive for high-dimensional problems. In this paper, we present an accurate and robust method to calibrate the SWAT model (i.e., 20 parameters) using scalable deep learning (DL). We developed inverse models based on convolutional neural networks (CNN) to assimilate observed streamflow data and estimate the SWAT model parameters. Scalable hyperparameter tuning is performed using high-performance computing resources to identify the top 50 optimal neural network architectures. We used ensemble SWAT simulations to train, validate, and test the CNN models. We estimated the parameters of the SWAT model using observed streamflow data and assessed the impact of measurement errors on SWAT model calibration. We tested and validated the proposed scalable DL methodology on the American River Watershed, located in the Pacific Northwest-based Yakima River basin. Our results show that the CNN-based calibration is better than two popular parameter estimation methods (i.e., the generalized likelihood uncertainty estimation [GLUE] and the dynamically dimensioned search [DDS], which is a global optimization algorithm). For the set of parameters that are sensitive to the observations, our proposed method yields narrower ranges than the GLUE method but broader ranges than values produced using the DDS method within the sampling range even under high relative observational errors. The SWAT model calibration performance using the CNNs, GLUE, and DDS methods are compared using R 2 and a set of efficiency metrics, including Nash-Sutcliffe, logarithmic Nash-Sutcliffe, Kling-Gupta, modified Kling-Gupta, and non-parametric Kling-Gupta scores, computed on the observed and simulated watershed responses. The best CNN-based calibrated set has scores of 0.71, 0.75, 0.85, 0.85, 0.86, and 0.91. The best DDS-based calibrated set has scores of 0.62, 0.69, 0.8, 0.77, 0.79, and 0.82. The best GLUE-based calibrated set has scores of 0.56, 0.58, 0.71, 0.7, 0.71, and 0.8. The scores above show that the CNN-based calibration leads to more accurate low and high streamflow predictions than the GLUE and DDS sets. Our research demonstrates that the proposed method has high potential to improve our current practice in calibrating large-scale integrated hydrologic models.

54 ENVIRONMENTAL SCIENCES↗

Coupled Models and Parallel Simulations for Three-Dimensional Full-Stokes Ice Sheet Modeling

A three-dimensional full-Stokes computational model is considered for determining the dynamics, temperature, and thickness of ice sheets. The governing thermomechanical equations consist of the three-dimensional full-Stokes system with nonlinear rheology for the momentum, an advective-diffusion energy equation for temperature evolution, and a mass conservation equation for icethickness changes. Here, we discuss the variable resolution meshes, the finite element discretizations, and the parallel algorithms employed by the model components. The solvers are integrated through a well-designed coupler for the exchange of parametric data between components. The discretization utilizes high-quality, variable-resolution centroidal Voronoi Delaunay triangulation meshing and existing parallel solvers. We demonstrate the gridding technology, discretization schemes, and the efficiency and scalability of the parallel solvers through computational experiments using both simplified geometries arising from benchmark test problems and a realistic Greenland ice sheet geometry.

58 GEOSCIENCES↗

Dynamic Load Balancing for Adaptive Computations on Distributed-Memory Machines

Dynamic load balancing is central to adaptive mesh-based computations on large-scale parallel computers. The principal investigator has investigated various issues on the dynamic load balancing problem under NASA JOVE and JAG rants. The major accomplishments of the project are two graph partitioning algorithms and a load balancing framework. The S-HARP dynamic graph partitioner is known to be the fastest among the known dynamic graph partitioners to date. It can partition a graph of over 100,000 vertices in 0.25 seconds on a 64- processor Cray T3E distributed-memory multiprocessor while maintaining the scalability of over 16-fold speedup. Other known and widely used dynamic graph partitioners take over a second or two while giving low scalability of a few fold speedup on 64 processors. These results have been published in journals and peer-reviewed flagship conferences.

Source record↗

MetaStrider: Architectures for Scalable Memory-centric Reduction of Sparse Data Streams

Reduction is an operation performed on the values of two or more key-value pairs that share the same key. Reduction of sparse data streams finds application in a wide variety of domains such as data and graph analytics, cybersecurity, machine learning, and HPC applications. However, these applications exhibit low locality of reference, rendering traditional architectures and data representations inefficient. This article presents MetaStrider, a significant algorithmic and architectural enhancement to the state-of-the-art, SuperStrider. Furthermore, these enhancements enable a variety of parallel, memory-centric architectures that we propose, resulting in demonstrated performance that scales near-linearly with available memory-level parallelism.

97 MATHEMATICS AND COMPUTING↗

Modeling Mesoscale Processes of Scalable Synthesis Final Scientific/Technical Report

This is the final scientific and technical report for this award. Molecular dynamics (MD) is a powerful tool to study biomolecules at the nanoscale. One of its main limitations however is that such simulations typically require a timestep of 1 femtosecond while reactions and conformational changes typically happen on the time scale of 1 microsecond or more. This discrepancy severely restricts the impact of this method. We have developed a series of algorithms to mitigate this problem and allow predicting long-time scale events using much shorter simulations. The method is based on the concurrent adaptive sampling method. It consists in using probabilistic weights for trajectories and biasing the sampling such that rare but important trajectories, e.g., those that go over energy barriers and lead to relevant conformational changes in the molecule, are observed with much higher frequency that in a standard MD simulation. We will demonstrate below that this approach has been very successful. In collaboration with Jay Grate at PNNL we have applied this novel method to the study the triazine polymers, which can be con gured using different types of side chains in order to adopt different 3-dimensional structures. We have found novel stable con- formations and have analyzed the impact of side chain modifcations on the native folded shape of the polymer. Direct MD simulations would have been untractable for this problem. The method of replica exchange was also used successfully to calculate free energy profiles, in particular for cis-trans torsional angle changes with high energy barriers. Studying molecules at the atomistic scale is essential to design new molecules in chemical engineering, such as the triazine polymers developed by Grate, in material science to study the strength of metals for examples (e.g., dislocation dynamics), or in biochemistry to understand how proteins fold and how they interact with ligands (a key step in designing drugs). Although new hardware makes these simulations increasingly faster (e.g., GPU, Anton from DE Shaw Research), new algorithms are key to making these methods more tractable and increase their impact. At this point, because of limitations in modeling, one of the main techniques for advancing science remains to a large extent numerous and costly experiments. For exam- ple, when designing drug, where multiple variants must be screened, the ability to run tests on a computer rather than experimentally testing with high throughput methods thousands of compounds would represent a signi cant advance with broad benefits for the public at large.

36 MATERIALS SCIENCE↗

GPU acceleration of the Locally Selfconsistent Multiple Scattering code for first principles calculation of the ground state and statistical physics of materials

The Locally Self-consistent Multiple Scattering (LSMS) code solves the first principles Density Functional theory Kohn–Sham equation for a wide range of materials with a special focus on metals, alloys and metallic nano-structures. It has traditionally exhibited near perfect scalability on massively parallel high performance computer architectures. In this paper, we present our efforts to exploit GPUs to accelerate the LSMS code to enable first principles calculations of O(100,000) atoms and statistical physics sampling of finite temperature properties. We reimplement the scattering matrix calculation for GPUs with a block matrix inversion algorithm that only uses accelerator memory. Finally, using the Cray XK7 system Titan at the Oak Ridge Leadership Computing Facility we achieve a sustained performance of 14.5PFlop/s and a speedup of 8.6 compared to the CPU only code.

36 MATERIALS SCIENCE↗

ExaTN: Scalable GPU-Accelerated High-Performance Processing of General Tensor Networks at Exascale

We present ExaTN (Exascale Tensor Networks), a scalable GPU-accelerated C++ library which can express and process tensor networks on shared- as well as distributed-memory high-performance computing platforms, including those equipped with GPU accelerators. Specifically, ExaTN provides the ability to build, transform, and numerically evaluate tensor networks with arbitrary graph structures and complexity. It also provides algorithmic primitives for the optimization of tensor factors inside a given tensor network in order to find an extremum of a chosen tensor network functional, which is one of the key numerical procedures in quantum many-body theory and quantum-inspired machine learning. Numerical primitives exposed by ExaTN provide the foundation for composing rather complex tensor network algorithms. We enumerate multiple application domains which can benefit from the capabilities of our library, including condensed matter physics, quantum chemistry, quantum circuit simulations, as well as quantum and classical machine learning, for some of which we provide preliminary demonstrations and performance benchmarks just to emphasize a broad utility of our library.

97 MATHEMATICS AND COMPUTING↗

A parallel geometric multigrid method for finite elements on octree meshes

In this article, we present a parallel geometric multigrid algorithm for solving variable-coefficient elliptic partial differential equations on the unit box (with Dirichlet or Neumann boundary conditions) using highly nonuniform, octree-based, conforming finite element discretizations. Our octrees are 2:1 balanced, that is, we allow no more than one octree-level difference between octants that share a face, edge, or vertex. We describe a parallel algorithm whose input is an arbitrary 2:1 balanced fine-grid octree and whose output is a set of coarser 2:1 balanced octrees that are used in the multigrid scheme. Also, we derive matrix-free schemes for the discretized finite element operators and the intergrid transfer operations. The overall scheme is second-order accurate for sufficiently smooth right-hand sides and material properties; its complexity for nearly uniform trees is {Omicron}(N/n{sub p} log N/n{sub p}) + {Omicron}(n{sub p} log n{sub p}), where N is the number of octree nodes and n{sub p} is the number of processors. Our implementation uses the Message Passing Interface standard. We present numerical experiments for the Laplace and Navier (linear elasticity) operators that demonstrate the scalability of our method. Our largest run was a highly nonuniform, 8-billion-unknown, elasticity calculation using 32,000 processors on the Teragrid system, 'Ranger,' at the Texas Advanced Computing Center. Our implementation is publically available in the Dendro library, which is built on top of the PETSc library from Argonne National Laboratory.

99 GENERAL AND MISCELLANEOUS↗

Asynchronous distributed-memory task-parallel algorithm for compressible flows on unstructured 3D Eulerian grids

Here, we discuss the implementation of a finite element method, used to numerically solve the Euler equations of compressible flows, using an asynchronous runtime system (RTS). The algorithm is implemented for distributed-memory machines, using stationary unstructured 3D meshes, combining data-, and task-parallelism on top of the Charm++ RTS. Charm++’s execution model is asynchronous by default, allowing arbitrary overlap of computation and communication. Task-parallelism allows scheduling parts of an algorithm independently of, or dependent on, each other. Built-in automatic load balancing enables continuous redistribution of computational load by migration of work units based on real-time CPU load measurement. The RTS also features automatic checkpointing, fault tolerance, resilience against hardware failure, and supports power-, and energy-aware computation. We demonstrate scalability up to 25 x 10 9 cells at $\mathscr{O}$10 4 compute cores and the benefits of automatic load balancing for irregular workloads. The full source code with documentation is available at https://quinoacomputing.org.

42 ENGINEERING↗

State Dependent Optimization with Quantum Circuit Cutting

Quantum circuits can be reduced through optimization to better fit the constraints of quantum hardware. One such method, initial-state dependent optimization (ISDO), reduces gate count by leveraging knowledge of the input quantum states. Surprisingly, we found that ISDO is broadly applicable to the downstream circuits produced by circuit cutting. Circuit cutting also requires measuring upstream qubits and has some flexibility of selection observables to do reconstruction. Therefore, we propose a state-dependent optimization (SDO) framework that incorporates ISDO, our newly proposed measure-state dependent optimization (MSDO), and a biased observable selection strategy. Building on the strengths of the SDO framework and recognizing the scalability challenges of circuit cutting, we propose nonseparate circuit cutting-a more flexible approach that enables optimizing gates without fully separating them. We validate our methods on noisy simulations of QAOA, QFT, and BV circuits. Results show that our approach consistently mitigates noise and improves overall circuit performance, demonstrating its promise for enhancing quantum algorithm execution on near-term hardware.

Li, Xinpeng↗

Learning Sequential Distribution System Restoration via Graph-Reinforcement Learning

We report a distribution service restoration algorithm as a fundamental resilient paradigm for system operators provides an optimally coordinated, resilient solution to enhance the restoration performance. The restoration problem is formulated to coordinate distribution generators and controllable switches optimally. A model-based control scheme is usually designed to solve this problem, relying on a precise model and resulting in low scalability. To tackle these limitations, this work proposes a graph-reinforcement learning framework for the restoration problem. We link the power system topology with a graph convolutional network, which captures the complex mechanism of network restoration in power networks and understands the mutual interactions among controllable devices. Latent features over graphical power networks produced by graph convolutional layers are exploited to learn the control policy for network restoration using deep reinforcement learning. The solution scalability is guaranteed by modeling distributed generators as agents in a multi-agent environment and a proper pre-training paradigm. Comparative studies on IEEE 123-node and 8500-node test systems demonstrate the performance of the proposed solution.

24 POWER TRANSMISSION AND DISTRIBUTION↗

Trigger-DAQ and Slow Controls Systems in the Mu2e Experiment

The muon campus program at Fermilab includes the Mu2e experiment that will search for a charged-lepton flavor violating processes where a negative muon converts into an electron in the field of an aluminum nucleus, improving by four orders of magnitude the search sensitivity reached so far. Mu2e's Trigger and Data Acquisition System (TDAQ) uses otsdaq as its solution. Developed at Fermilab, otsdaq uses the artdaq DAQ framework and art analysis framework, under the-hood, for event transfer, filtering, and processing. otsdaq is an online DAQ software suite with a focus on flexibility and scalability, while providing a multi-user, web-based, interface accessible through the Chrome or Firefox web browser. The detector Read Out Controller (ROC), from the tracker and calorimeter, stream out zero-suppressed data continuously to the Data Transfer Controller (DTC). Data is then read over the PCIe bus to a software filter algorithm that selects events which are finally combined with the data flux that comes froma Cosmic Ray Ve to System (CRV). A Detector Control System (DCS) for monitoring, controlling, alarming, and archiving has been developed using the Experimental Physics and Industrial Control System (EPICS) Open Source Platform. The DCS System has also been itegrated into otsdaq. The installation of the TDAQ and the DCS systems in the Mu2e building is planned for 2021-2022, and a prototype has been built at Fermilab's Feynman Computing Center. We report here on the developments and achievements of the integration of Mu2e's DCS system into the online otsdaq software.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

Scalable Static and Dynamic Community Detection Using Grappolo

Graph clustering, popularly known as community detection, is a fundamental kernel for several applications of relevance to the Defense Advanced Research Projects Agency’s (DARPA) Hierarchical Identify Verify Exploit (HIVE) Pro- gram. Clusters or communities represent natural divisions within a network that are densely connected within a cluster and sparsely connected to the rest of the network. The need to compute clustering on large scale data necessitates the development of efficient algorithms that can exploit modern architectures that are fundamentally parallel in nature. How- ever, due to their irregular and inherently sequential nature, many of the current algorithms for community detection are challenging to parallelize. In response to the HIVE Graph Challenge, we present several parallelization heuristics for fast community detection using the Louvain method as the serial template. We implement all the heuristics in a software library called Grappolo. Using the inputs from the HIVE Challenge, we demonstrate superior performance and high quality solutions based on four parallelization heuristics. We use Grappolo on static graphs as the first step towards community detection on streaming graphs.

Community Detection↗

A novel digital pulse processing architecture for nuclear instrumentation

The field of nuclear instrumentation covers a wide range of applications, including counting, spectrometry, pulse shape discrimination and multi-channel coincidence. These applications are the topic of many researches, new algorithms and implementations are constantly proposed thanks to advances in digital signal processing. However, these improvements are not yet implemented in instrumentation devices. This is especially true for neutron-gamma discrimination applications which traditionally use charge comparison method while literature proposes other algorithms based on frequency domain or wavelet theory which show better performances. Another example is pileups which are generally rejected while pileup correction algorithms also exist. These processes are traditionally performed offline due to two issues. The first is the Poissonian characteristic of the signal, composed of random arrival pulses which requires to current architectures to work in data flow. The second is the real-time requirement, which implies losing pulses when the pulse rate is too high. Despite the possibility of treating the pulses independently from each other, current architectures paralyze the acquisition of the signal during the processing of a pulse. This loss is called dead-time. These two issues have led current architectures to use dedicated solutions based on re-configurable components like Field Programmable Gate Arrays (FPGAs) to overcome the need of performance necessary to deal with dead-time. However, dedicated hardware algorithm implementations on re-configurable technologies are complex and time-consuming. For all these reasons, a programmable Digital pulse Processing (DPP) architecture in a high level language such as Cor C++ which can reduce dead-time would be worthwhile for nuclear instrumentation. This would reduce prototyping and test duration by reducing the level of hardware expertise to implement new algorithms. However, today's programmable solutions do not meet the need of performance to operate online and not allow scaling with the increase in the number of measurement channel. That is why an innovative DPP architecture is proposed in this paper. This architecture is able to overcome dead-time while being programmable and is flexible with the number of measurement channel. Proposed architecture is based on an innovative execution model for pulse processing applications which can be summarized as follow. The signal is not composed of pulses only, consequently, pulses processing does not have to operate on the entire signal. Therefore, the first step of our proposal is pulse extraction by the use of dedicated components named pulse extractors. The triggering step can be achieved after the analog-to-digital conversion without any signal shaping or filtering stages. Pileup detection and accurate pulse time stamping are done at this stage. Any application downstream this step can work on adaptive variable-sized array of samples simplifying pulse processing methods. Then, once the data flow is broken, it is possible to distribute pulses on Functional Units (FUs) which perform processing. As the date of each pulse is known, they can be processed individually out-of-order to provide the results. To manage the pulses distribution, a scheduler and an interconnection network are used. pulses are distributed on the first FU which is not busy without congesting the interconnection network. For this reason, the process duration does not result anymore in dead-time if there are enough FUs. FUs are designed to be standalone and to comprises at least a programmable general purpose processor (ARM, Microblaze) allowing the implementation of complex algorithms without any modification of the hardware. An acquisition chain is composed of a succession of algorithms which lead to organize our FUs as a software macro-pipeline, A simple approach consists in assigning one algorithm per FU. Consequently, the global latency becomes the worst latency of algorithms execution on FU. Moreover, as algorithms are executed locally - i.e. on a FU - this approach limits shared memory requirement. To handle multichannel, we propose FUs sharing, this approach maximize the chance to find a non-busy FU to process an incoming pulse. This is possible since each channel receive random event independently, the pulse extractors associated to them do not necessarily need to access simultaneously to all Computing resources at the same time to distribute their pulses. The major contribution of this paper is the proposition of an execution model and its associated hardware programmable architecture for digital pulse processing that can handle multiple acquisition channels while maintaining the scalability thanks to the use of shared resources. This execution model and associated architecture are validated by simulation of a cycle accurate architecture SystemC model. Proposed architecture shows promising results in terms of scalability while maintaining zero dead-time. This work also permit the sizing of hardware resources requirement required for a predefined set of applications. Future work will focus on the interconnection network and a scheduling policy that can exploit the variable-length of pulses. Then, the hardware implementation of this architecture will be performed and tested for a representative set of application.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗