Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “computational complexity”

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 1,189 records · Page 66

Modeling Analog Tile-Based Accelerators Using SST

Analog computing has been widely proposed to improve the energy efficiency of multiple important workloads including neural network operations, and other linear algebra kernels. To properly evaluate analog computing and explore more complex workloads such as systems consisting of multiple analog data paths, system level simulations are required. Moreover, prior work on system architectures for analog computing often rely on custom simulators creating signficant additional design effort and complicating comparisons between different systems. To remedy these issues, this report describes the design and implementation of a flexible tile-based analog accelerator element for the Structural Simulation Toolkit (SST). The element focuses on heavily on the tile controller—an often neglected aspect of prior work—that is sufficiently versatile to simulate a wide range of different tile operations including neural network layers, signal processing kernels, and generic linear algebra operations without major constraints. The tile model also interoperates with existing SST memory and network models to reduce the overall development load and enable future simulation of heterogeneous systems with both conventional digital logic and analog compute tiles. Finally, both the tile and array models are designed to easily support future extensions as new analog operations and applications that can benefit from analog computing are developed.

97 MATHEMATICS AND COMPUTING↗

Information geometry of Gaussian channels

We define a local Riemannian metric tensor in the manifold of Gaussian channels and the distance that it induces. We adopt an information-geometric approach and define a metric derived from the Bures-Fisher metric for quantum states. The resulting metric inherits several desirable properties from the Bures-Fisher metric and is operationally motivated by distinguishability considerations: It serves as an upper bound to the attainable quantum Fisher information for the channel parameters using Gaussian states, under generic constraints on the physically available resources. Our approach naturally includes the use of entangled Gaussian probe states. We prove that the metric enjoys some desirable properties like stability and covariance. As a by-product, we also obtain some general results in Gaussian channel estimation that are the continuous-variable analogs of previously known results in finite dimensions. We prove that optimal probe states are always pure and bounded in the number of ancillary modes, even in the presence of constraints on the reduced state input in the channel. This has experimental and computational implications. It limits the complexity of optimal experimental setups for channel estimation and reduces the computational requirements for the evaluation of the metric: Indeed, we construct a converging algorithm for its computation. We provide explicit formulas for computing the multiparametric quantum Fisher information for dissipative channels probed with arbitrary Gaussian states and provide the optimal observables for the estimation of the channel parameters (e.g., bath couplings, squeezing, and temperature).

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Global Sensitivity Measures from Given Data

Simulation models support managers in the solution of complex problems. International agencies recommend uncertainty and global sensitivity methods as best practice in the audit, validation and application of scientific codes. However, numerical complexity, especially in the presence of a high number of factors, induces analysts to employ less informative but numerically cheaper methods. This work introduces a design for estimating global sensitivity indices from given data (including simulation input–output data), at the minimum computational cost. We address the problem starting with a statistic based on the L1-norm. A formal definition of the estimators is provided and corresponding consistency theorems are proved. The determination of confidence intervals through a bias-reducing bootstrap estimator is investigated. The strategy is applied in the identification of the key drivers of uncertainty for the complex computer code developed at the National Aeronautics and Space Administration (NASA) assessing the risk of lunar space missions. We also introduce a symmetry result that enables the estimation of global sensitivity measures to datasets produced outside a conventional input–output functional framework.

99 GENERAL AND MISCELLANEOUS↗

Performance Analysis of an Optimization Algorithm for Metamaterial Design on the Integrated High-Performance Computing and Quantum Systems

Optimizing metamaterials with complex geometries is a big challenge. Although an active learning algorithm, combining machine learning (ML), quantum computing, and optical simulation, has emerged as an efficient optimization tool, it still faces difficulties in optimizing complex structures that have potentially high performance. In this work, we comprehensively analyze the performance of an optimization algorithm for metamaterial design on the integrated HPC and quantum systems. We demonstrate significant time advantages through message-passing interface (MPI) parallelization on the high-performance computing (HPC) system showing approximately 54% faster ML tasks and 67 times faster optical simulation against serial workloads. Furthermore, we analyze the performance of a quantum algorithm designed for optimization, which runs with various quantum simulators on a local computer or HPC-quantum system. Results showcase ~24 times speedup when executing the optimization algorithm on the HPC-quantum hybrid system. This study paves a way to optimize complex metamaterials using the integrated HPC-quantum system.

Kim, Seongmin↗

Computationally Efficient Multiscale Neural Networks Applied to Fluid Flow in Complex 3D Porous Media

Abstract The permeability of complex porous materials is of interest to many engineering disciplines. This quantity can be obtained via direct flow simulation, which provides the most accurate results, but is very computationally expensive. In particular, the simulation convergence time scales poorly as the simulation domains become less porous or more heterogeneous. Semi-analytical models that rely on averaged structural properties (i.e., porosity and tortuosity) have been proposed, but these features only partly summarize the domain, resulting in limited applicability. On the other hand, data-driven machine learning approaches have shown great promise for building more general models by virtue of accounting for the spatial arrangement of the domains’ solid boundaries. However, prior approaches building on the convolutional neural network (ConvNet) literature concerning 2D image recognition problems do not scale well to the large 3D domains required to obtain a representative elementary volume (REV). As such, most prior work focused on homogeneous samples, where a small REV entails that the global nature of fluid flow could be mostly neglected, and accordingly, the memory bottleneck of addressing 3D domains with ConvNets was side-stepped. Therefore, important geometries such as fractures and vuggy domains could not be modeled properly. In this work, we address this limitation with a general multiscale deep learning model that is able to learn from porous media simulation data. By using a coupled set of neural networks that view the domain on different scales, we enable the evaluation of large ( $$>512^3$$ > 512 3 ) images in approximately one second on a single graphics processing unit. This model architecture opens up the possibility of modeling domain sizes that would not be feasible using traditional direct simulation tools on a desktop computer. We validate our method with a laminar fluid flow case using vuggy samples and fractures. As a result of viewing the entire domain at once, our model is able to perform accurate prediction on domains exhibiting a large degree of heterogeneity. We expect the methodology to be applicable to many other transport problems where complex geometries play a central role.

36 MATERIALS SCIENCE↗

Ammonia Oxidation by Abstraction of Three Hydrogen Atoms from a Mo–NH 3 Complex

We report ammonia oxidation by homolytic cleavage of all three H atoms from a Mo৶ 15 NH 3 complex using the 2,4,6-tri-tert-butylphenoxyl radical to afford a Mo-alkylimido (Mo৶ 15 NR) complex (R = 2,4,6-tri-t-butylcyclohexa-2,5-dien-1-one). Reductive cleavage of Mo৶15NR generates a terminal Mo৶N nitride, and a [Mo৶ 15 NH] + complex is formed by protonation. Computational analysis describes the energetic profile for the stepwise removal of three H atoms from the Mo৶ 15 NH 3 complex and the formation of Mo৶ 15 NR. Acknowledgment. This work was supported as part of the Center for Molecular Electrocatalysis, an Energy Frontier Re-search Center funded by the U.S. Department of Energy (U.S. DOE), Office of Science, Office of Basic Energy Sciences. EPR and mass spectrometry experiments were performed using EMSL, a national scientific user facility sponsored by the DOE’s Office of Biological and Environmental Research and located at PNNL. The authors thank Dr. Eric D. Walter and Dr. Rosalie Chu for assistance in performing EPR and mass spectroscopy analysis, respectively. Computational resources provided by the National Energy Re-search Scientific Computing Center (NERSC) at Lawrence Berkeley National Laboratory. Pacific North-west National Laboratory is operated by Battelle for the U.S. DOE.

Molybdenum, ammonia, phosphines, pendant amines, h↗

Process Image Analysis using Big Data, Machine Learning, and Computer Vision

Within the DOE complex, 3013 canisters are used to store Pu-bearing waste. Wall penetration due to corrosion, specifically stress corrosion cracking, is considered to be the most likely cause of failure over the 50-year lifetime of the canisters. This project has 2 objectives: The first is the development of machine learning algorithms to identify the presence of corrosion from a very large set of images generated by laser confocal microscope scanning of 3013 canister used to store Pu oxides. This portion of the LDRD constitutes image analysis of a metal surface for the presence of corrosion. The image processing algorithms developed for this project will provide a suitable basis for analysis of other types of corrosion data, which can be produced in vast quantities using modern devices. The second component of the LDRD consists of the development of machine learning algorithms that obtain molecular mechanics force-fields from ab-initio Density Functional Theory (DFT) calculations for corrosive attack by chlorides on 304L or 316L stainless steel. The goal of this latter component is to determine means for mitigating corrosion on a fundamental level, including coatings, welding methods, metal composition, etc. Force-field modeling is necessary for this endeavor because the incipience and progression of corrosion is the governed by molecular structures, grain boundaries, dislocations and surface structures represented by large numbers of atoms. Although DFT calculations are extremely adept at describing molecular scale processes, they are computationally expensive making them ill-suited for calculations of more than several hundred atoms, especially for the repeated applications inherent in material design. Fortunately, force-field methods provide an avenue for viable molecular-scale calculations involving the numbers of atoms involved in corrosion processes. The accuracy of the force-field calculations depends strongly on the accuracy of the force-field model, which is extremely difficult and time-consuming to derive from either data or ab-initio calculations. The use of machine learning algorithms has the potential to make the calculation of force-fields much more efficient but has not been explored significantly for corrosion processes. If shown to be viable, the technique would have wide -ranging impact on design of corrosion resistant materials and on the mitigation of corrosion in existing process systems.

12 MANAGEMENT OF RADIOACTIVE AND NON-RADIOACTIVE W↗

A Visual Comparison of Silent Error Propagation

High-performance computing (HPC) systems play a critical role in facilitating scientific discoveries. Their scale and complexity (e.g., the number of computational units and software stack) continue to grow as new systems are expected to process increasingly more data and reduce computing time. However, with more processing elements, the probability that these systems will experience a random bit-flip error that corrupts a program's output also increases, which is often recognized as silent data corruption. Analyzing the resiliency of HPC applications in extreme-scale computing to silent data corruption is crucial but difficult. An HPC application often contains a large number of computation units that need to be tested, and error propagation caused by error corruption is complex and difficult to interpret. Here, to accommodate this challenge, we propose an interactive visualization system that helps HPC researchers understand the resiliency of HPC applications and compare their error propagation. Our system models an application's error propagation to study a program's resiliency by constructing and visualizing its fault tolerance boundary. Coordinating with multiple interactive designs, our system enables domain experts to efficiently explore the complicated spatial and temporal correlation between error propagations. At the end, the system integrated a nonmonotonic error propagation analysis with an adjustable graph propagation visualization to help domain experts examine the details of error propagation and answer such questions as why an error is mitigated or amplified by program execution.

97 MATHEMATICS AND COMPUTING↗

Scalable Computation of 2D-Minkowski Sum of Arbitrary Non-Convex Domains: Modeling Flexibility in Energy Resources

The flexibility of active ($p$) and reactive power ($q$) consumption in distributed energy resources (DERs) can be represented as a (potentially non-convex) set of points in the $p$-$q$ plane. Modeling of the aggregated flexibility in a heterogeneous ensemble of DERs as a Minkowski sum (M-sum) is computationally intractable even for moderately sized populations. In this article, we propose a scalable method of computing the M-sum of the flexibility domains of a heterogeneous ensemble of DERs, which are allowed to be non-convex, non-compact. In particular, the proposed algorithm computes a guaranteed superset of the true M-sum, with desired accuracy. The worst-case complexity of the algorithm is computed. Special cases are considered, and it is shown that under certain scenarios, it is possible to achieve a complexity that is linear with the size of the ensemble. Numerical examples are provided by computing the aggregated flexibility of different mix of DERs under varying scenarios.

Kundu, Soumya↗

Accelerating Instanton Theory with the Line Integral Nudged Elastic Band Method and Gaussian Process Regression

Quantum tunneling plays a fundamental role in many chemical reactions, particularly proton transfer processes. Ring polymer instanton theory offers a practical framework for computing tunneling rates in complex molecular systems. However, applying the ring polymer instanton method with a potential energy surface generated on-the-fly using electronic structure calculations can be computationally demanding. Here, in this work, we present a new efficient implementation of the ring polymer instanton method by combining the Line Integral Nudged Elastic Band (LI-NEB) approach with Gaussian Process Regression (GPR). We benchmarked this method on prototypical ground-state proton transfer systems, including the benchmark gas-phase hydrogen abstraction reaction H + CH 4 → H 2 + CH 3 , malonaldehyde, and Z-3-amino-propenal (aminopropenal). Our results show that this approach is an order of magnitude faster than traditional instanton algorithms while maintaining excellent agreement with their tunneling rates. This development opens the door to studying proton transfer in larger systems with improved efficiency.

chemical physics↗

Fast Parallel Tensor Times Same Vector for Hypergraphs

Hypergraphs are a popular paradigm to rep- resent complex real-world networks exhibiting multi-way relationships of varying sizes. Mining centrality in hyper- graphs via symmetric adjacency tensors has only recently become computationally feasible for large and complex datasets. To enable scalable computation of these and related hypergraph analytics, here we focus on the Sparse Symmetric Tensor Times Same Vector (S3TTVC) oper- ation. We introduce the Compound Compressed Sparse Symmetric (CCSS) format, an extension of the compact CSS format for hypergraphs of varying hyperedge sizes and present a shared-memory parallel algorithm to compute S3TTVC. We experimentally show S3TTVC computation using the CCSS format achieves better performance than the naive baseline, and is subsequently more performant for hypergraph H-eigenvector centrality.

Shivakumar, Shruti↗

Tris(carbene)borates; alternatives to cyclopentadienyls in organolanthanide chemistry

The chemistry of the tris-carbene anion phenyltris(3-alkyl-imidazoline-2-yliden-1-yl)borate, [C3 Me ] - ligand, is initiated for f-block metal cations. Neutral, molecular complexes of the form Ln(C3) 2 I are formed for cerium(III), while a separated ion pair [Ln(C3) 2 ]I forms for ytterbium(III). DFT/QTAIM computational analyses of the complexes and related tridentate tris(pyrazolyl)borate (Tp) – supported analogs demonstrates the anticipated strength of the σ donation and confirms greater covalency in the metal–carbon bonds of the [C3 Me ] - complexes in comparison with those in the Tp Me,Me complexes. The DFT calculations demonstrate the crucial role of THF solvent in accurately reproducing the contrasting molecular and ion-pair geometries observed experimentally for the Ce and Yb complexes.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Regional heritability advanced complex trait analysis for GPU and traditional parallel architectures

Motivation: Quantification of the contribution of genetic variation to phenotypic variation for complex traits becomes increasingly computationally demanding with increasing numbers of single-nucleotide polymorphisms and individuals. To meet the challenges in making feasible large-scale studies, we present the REgional heritability advanced complex trait analysis software. Adapted from advanced complex trait analysis (and, in turn, genome-wide complex trait analysis), it is tailored to exploit the parallelism present in modern traditional and graphics processing unit (GPU)-accelerated machines, from workstations to supercomputers. Results: We adapt the genetic relationship matrix estimation algorithm to remove limitations on memory, allowing the analysis of large datasets. Further, we build on this to develop a version of the code able to efficiently exploit GPU-accelerated systems for both the genetic relationship matrix and REstricted maximum likelihood (REML) parts of the analysis, offering substantial speedup over the traditional central processing unit version. We develop the ability to analyze multiple small regions of the genome across multiple compute nodes in parallel, following the ‘regional heritability’ approach. We demonstrate the new software using 1024 GPUs in parallel on one of the world’s fastest supercomputers. Availability: The code is freely available at http://www.epcc.ed.ac.uk/software-products

59 BASIC BIOLOGICAL SCIENCES↗

A Mountaintop View Requires Minimal Sorting: A Faster Contour Tree Algorithm

Consider a scalar field f : M → R, where M is a triangulated simplicial mesh in R d . A level set, or contour, at value v is a connected component of f –1 (v). As v is changed, these contours change topology, merge into each other, or split. Contour trees are concise representations of f that track this contour behavior. The vertices of these trees are the critical points of f, where the gradient is zero. The edges represent changes in the topology of contours. It is a fundamental data structure in data analysis and visualization, and there is significant previous work (both theoretical and practical) on algorithms for constructing contour trees. Suppose M has n vertices, N facets, and t critical points. A classic result of Carr, Snoeyink, and Axen (2000) gives an algorithm that takes O(n log n+Nα(N)) time (where α(·) is the inverse Ackermann function). A further improvement to O(t log t + N) time was given by Chiang et al. All these algorithms involve a global sort of the critical points, a significant computational bottleneck. Unfortunately, lower bounds of Ω(t log t) also exist. We present the first algorithm that can avoid the global sort and has a refined time complexity that depends on the contour tree structure. Intuitively, if the tree is short and fat, we get significant improvements in running time. For a partition of the contour tree into a set of descending paths, P, our algorithm runs in O($\Sigma$ pϵP |p| log |p| + tα(t) + N). This is at most O(t log D + N), where D is the diameter of the contour tree. Moreover, it is O(tα(t) + N) for balanced trees, a significant improvement over the previous complexity. Our algorithm requires numerous ideas: partitioning the contour tree into join and split trees, a local growing procedure to iteratively build contour trees, and the use of heavy path decompositions for the time complexity analysis. There is a crucial use of a family of binomial heaps to maintain priorities, ensuring that any comparison made is between comparable nodes of the contour tree. We also prove lower bounds showing that the $\Sigma$ pϵP |p| log |p| complexity is inherent to computing contour trees.

97 MATHEMATICS AND COMPUTING↗

Stephen Jay Kline on systems, or physics, complex systems, and the gap between.

At the end of his life, Stephen Jay Kline, longtime professor of mechanical engineering at Stanford University, completed a book on how to address complex systems. The title of the book is 'Conceptual Foundations of Multi-Disciplinary Thinking' (1995), but the topic of the book is systems. Kline first establishes certain limits that are characteristic of our conscious minds. Kline then establishes a complexity measure for systems and uses that complexity measure to develop a hierarchy of systems. Kline then argues that our minds, due to their characteristic limitations, are unable to model the complex systems in that hierarchy. Computers are of no help to us here. Our attempts at modeling these complex systems are based on the way we successfully model some simple systems, in particular, 'inert, naturally-occurring' objects and processes, such as what is the focus of physics. But complex systems overwhelm such attempts. As a result, the best we can do in working with these complex systems is to use a heuristic, what Kline calls the 'Guideline for Complex Systems.' Kline documents the problems that have developed due to 'oversimple' system models and from the inappropriate application of a system model from one domain to another. One prominent such problem is the Procrustean attempt to make the disciplines that deal with complex systems be 'physics-like.' Physics deals with simple systems, not complex ones, using Kline's complexity measure. The models that physics has developed are inappropriate for complex systems. Kline documents a number of the wasteful and dangerous fallacies of this type.

42 ENGINEERING↗

Computer-Aided Construction of Chemical Kinetic Models

The combustion chemistry of even simple fuels can be extremely complex, involving hundreds or thousands of kinetically significant species. The most reasonable way to deal with this complexity is to use a computer not only to numerically solve the kinetic model, but also to construct the kinetic model in the first place. Because these large models contain so many numerical parameters (e.g. rate coefficients, thermochemistry) one never has sufficient data to uniquely determine them all experimentally. Instead one must work in “predictive” mode, using theoretical rather than experimental values for many of the numbers in the model, and as appropriate refining the most sensitive numbers through experiments. Predictive chemical kinetics is exactly what is needed for computer-aided design of combustion systems based on proposed alternative fuels, particularly for early assessment of the value and viability of proposed new fuels before those fuels are commercially available. This project was aimed at making accurate predictive chemical kinetics practical; this is a challenging goal which requires a range of science advances. The project spanned a wide range from quantum chemical calculations on individual molecules and elementary-step reactions, through the development of improved rate/thermo calculation procedures, the creation of algorithms and software for constructing and solving kinetic simulations, the invention of methods for model-reduction while maintaining error control, and finally comparisons with experiment. Many of the parameters in the models were derived from quantum chemistry calculations, and the models were compared with experimental data measured in our lab or in collaboration with others.

97 MATHEMATICS AND COMPUTING↗

Compute in‐Memory with Non‐Volatile Elements for Neural Networks: A Review from a Co‐Design Perspective

Abstract Deep learning has become ubiquitous, touching daily lives across the globe. Today, traditional computer architectures are stressed to their limits in efficiently executing the growing complexity of data and models. Compute‐in‐memory (CIM) can potentially play an important role in developing efficient hardware solutions that reduce data movement from compute‐unit to memory, known as the von Neumann bottleneck. At its heart is a cross‐bar architecture with nodal non‐volatile‐memory elements that performs an analog multiply‐and‐accumulate operation, enabling the matrix‐vector‐multiplications repeatedly used in all neural network workloads. The memory materials can significantly influence final system‐level characteristics and chip performance, including speed, power, and classification accuracy. With an over‐arching co‐design viewpoint, this review assesses the use of cross‐bar based CIM for neural networks, connecting the material properties and the associated design constraints and demands to application, architecture, and performance. Both digital and analog memory are considered, assessing the status for training and inference, and providing metrics for the collective set of properties non‐volatile memory materials will need to demonstrate for a successful CIM technology.

36 MATERIALS SCIENCE↗