Among the main challenges in shape optimization is the coupling of Finite Element Method (FEM) codes in a way that facilitates efficient computation of shape derivatives. This is particularly difficult with multi-physics problems involving legacy codes, where the costs of implementing and maintaining shape derivative capabilities are prohibitive. There are two mathematically equivalent approaches to computing the shape derivative: the volume method, and the boundary method. Each has a major drawback: the boundary method is less accurate, while the volume method is more invasive to the FEM code. Prior implementations of shape derivatives at Sandia have been based on the volume method. We introduce the strip method, which computes shape derivatives on a strip adjacent to the boundary. The strip method makes code coupling simple. Like the boundary method, it queries the state and adjoint solutions at quadrature nodes, but requires no knowledge of the FEM code implementations. At the same time, it exhibits the higher accuracy of the volume method. The development of the strip method also offers us the opportunity to share some lessons learned about implementing the volume method and boundary method, to show shape optimization results on problems of interest, and to begin addressing the other main challenges at hand: constraints on optimized shapes, and their interplay with optimization algorithms.
Nearly every protocol used to analyze the performance of quantum information processors is based on an assumption that the errors experienced by the device during logical operations are constant in time and are insensitive to external contexts. This assumption is pervasive, rarely stated, and almost always wrong. Quantum devices that do behave this way are termed "Markovian:' but nearly every system we have ever probed has displayed drift or crosstalk or memory effects they are all non-Markovian. Strong non-Markovianity introduces spurious effects in characterization protocols and violates assumptions of the fault-tolerance threshold theorems. This SAND report details a three year laboratory-directed research and development (LDRD) project entitled, "Diagnosing and Destroying non-Markovian Noise in Quantum Information Processors." This program was initiated to build tools to study non-Markovian dynamics and quantum systems and develop robust methodologies for eliminating it. The program achieved a number of notable successes, including the first statistically rigorous protocol for identifying and characterizing drift in quantum systems, a formalism for modeling memory effects in quantum devices, and the successful suppression of drift in a Sandia trapped-ion quantum processor.
Intuition tells us that a rolling or spinning sphere will eventually stop due to the presence of friction and other dissipative interactions. The resistance to rolling and spinning or twisting torque that stops a sphere also changes the microstructure of a granular packing of frictional spheres by increasing the number of constraints on the degrees of freedom of motion. We perform discrete element modeling simulations to construct sphere packings implementing a range of frictional constraints under a pressure-controlled protocol. Mechanically stable packings are achievable at volume fractions and average coordination numbers as low as 0.53 and 2.5, respectively, when the particles experience high resistance to sliding, rolling, and twisting. Only when the particle model includes rolling and twisting friction were experimental volume fractions reproduced.
The present disclosure provides a method and device for filtering sensor data. Signals from an array of sensor pixels are received and checked for changes in pixel values. Motion is detected based on the changes in pixel values, and motion output signals are transmitted to a processing station. If the sum of correlated changes in pixel values across a predetermined field of view exceeds a predetermined value, indicating sensor jitter, the motion output signals are suppressed. If a sum of motion values within a defined subsection of the field of view exceeds a predetermined threshold, indicating the presence of a large object of no interest, the motion output signals are suppressed for that subsection.
The present disclosure provides a method and device for filtering sensor data. Signals from an array of sensor pixels are received and checked for changes in pixel values. Motion is detected based on the changes in pixel values, and motion output signals are transmitted to a processing station. If the sum of correlated changes in pixel values across a predetermined field of view exceeds a predetermined value, indicating sensor jitter, the motion output signals are suppressed. If a sum of motion values within a defined subsection of the field of view exceeds a predetermined threshold, indicating the presence of a large object of no interest, the motion output signals are suppressed for that subsection.
The Computer Science Research Institute (CSRI) brings university faculty and students to Sandia for focused collaborative research on Department of Energy (DOE) computer and computational science problems. The institute provides an opportunity for university researchers to learn about problems in computer and computational science at DOE laboratories. Participants conduct leading-edge research, interact with scientists and engineers at the laboratories, and help transfer results of their research to programs at the labs. Some specific CSRI research interest areas are: scalable solvers, optimization, adaptivity and mesh refinement, graph-based, discrete, and combinatorial algorithms, uncertainty estimation, mesh generation, dynamic load-balancing, virus and other malicious-code defense, visualization, scalable cluster computers and heterogeneous computers, data-intensive computing, environments for scalable computing, parallel input/output, advanced architectures, and theoretical computer science. The CSRI Summer Program includes the organization of a weekly seminar series and the publication of this summer proceedings.
In this paper, we consider the optimal control of semilinear elliptic PDEs with random inputs. These problems are often nonconvex, infinite-dimensional stochastic optimization problems for which we employ risk measures to quantify the implicit uncertainty in the objective function. In contrast to previous works in uncertainty quantification and stochastic optimization, we provide a rigorous mathematical analysis demonstrating higher solution regularity (in stochastic state space), continuity and differentiability of the control-to-state map, and existence, regularity and continuity properties of the control-to-adjoint map. Our proofs make use of existing techniques from PDE-constrained optimization as well as concepts from the theory of measurable multifunctions. We illustrate our theoretical results with two numerical examples motivated by the optimal doping of semiconductor devices.
GPUs are now a fundamental accelerator for many high-performance computing applications. They are viewed by many as a technology facilitator for the surge in fields like machine learning and Convolutional Neural Networks. To deliver the best performance on a GPU, we need to create monitoring tools to ensure that we optimize the code to get the most performance and efficiency out of a GPU. Since NVIDIA GPUs are currently the most commonly implemented in HPC applications and systems, NVIDIA tools are the solution for performance monitoring. The Light-Weight Distributed Metric System (LDMS) at Sandia is an infrastructure widely adopted for large-scale systems and application monitoring. Sandia has developed CPU application monitoring capability within LDMS. Therefore, we chose to develop a GPU monitoring capability within the same framework. In this report, we discuss the current limitations in the NVIDIA monitoring tools, how we overcame such limitations, and present an overview of the tool we built to monitor GPU performance in LDMS and its capabilities. Also, we discuss our current validation results. Most of the performance counter results are the same in both vendor tools and our tool when using LDMS to collect these results. Furthermore, our tool provides these statistics during the entire runtime of the tool as a time series and not just aggregate statistics at the end of the application run. This allows the user to see the progress of the behavior of the applications during their lifetime.
Analog hardware accelerators, which perform computation within a dense memory array, have the potential to overcome the major bottlenecks faced by digital hardware for data-heavy workloads such as deep learning. Exploiting the intrinsic computational advantages of memory arrays, however, has proven to be challenging principally due to the overhead imposed by the peripheral circuitry and due to the non-ideal properties of memory devices that play the role of the synapse. We review the existing implementations of these accelerators for deep supervised learning, organizing our discussion around the different levels of the accelerator design hierarchy, with an emphasis on circuits and architecture. We explore and consolidate the various approaches that have been proposed to address the critical challenges faced by analog accelerators, for both neural network inference and training, and highlight the key design trade-offs underlying these techniques.
Incorporating experimental data is essential for increasing the credibility of simulation-aided decision making and design. This paper presents a method which uses a computational model to guide the optimal acquisition of experimental data to produce data-informed predictions of quantities of interest (QoI). Many strategies for optimal experimental design (OED) select data that maximize some utility that measures the reduction in uncertainty of uncertain model parameters, for example the expected information gain between prior and posterior distributions of these parameters. In this paper, we seek to maximize the expected information gained from the push-forward of an initial (prior) density to the push-forward of the updated (posterior) density through the parameter-to-prediction map. The formulation presented is based upon the solution of a specific class of stochastic inverse problems which seeks a probability density that is consistent with the model and the data in the sense that the push-forward of this density through the parameter-to-observable map matches a target density on the observable data. While this stochastic inverse problem forms the mathematical basis for our approach, we develop a one-step algorithm, focused on push-forward probability measures, that leverages inference-for-prediction to bypass constructing the solution to the stochastic inverse problem. A number of numerical results are presented to demonstrate the utility of this optimal experimental design for prediction and facilitate comparison of our approach with traditional OED.
Analog quantum simulation is an approach for studying physical systems that might otherwise be computationally intractable to simulate on classical high-performance computing (HPC) systems. The key idea behind analog quantum simulation is the realization of a physical system with a low-energy effective Hamiltonian that is the same as the low-energy effective Hamiltonian of some target system to be studied. Purpose-built nanoelectronic devices are a natural candidate for implementing the analog quantum simulation of strongly correlated materials that are otherwise challenging to study using classical HPC systems. However, realizing devices that are sufficiently large to study the properties of a non-trivial material system (e.g., those described by a Fermi-Hubbard model) will eventually require the fabrication, control, and measurement of at least 0(10) quantum dots, or other engineered quantum impurities. As a step toward large-scale analog or digital quantum simulation platforms based on nanoelectronic devices, we propose a new approach to analog quantum simulation that makes use of the large Hilbert space dimension of the electronic baths that are used to adjust the occupancy of one or a few engineered quantum impurities. This approach to analog quantum simulation allows us to study a wide array of quantum impurity models. We can further augment the computational power of such an approach by combining it with a classical computer to facilitate dynamical mean-field theory (DMFT) calculations. DMFT replaces the solution of a lattice impurity problem with the solution of a family of localized impurity problems with bath couplings that are adjusted to satisfy a self-consistency condition between the two models. In DMFT, the computationally challenging task is the high-accuracy solution of an instance of a quantum impurity model that is determined self-consistently in coordination with a mean-field calculation. We propose using one or a few engineered quantum impurities with adjustable couplings to baths to realize an analog quantum coprocessor that effects the solution of such a model through measurements of a physical quantum impurity, operating in coordination with a classical computer to achieve a self-consistent solution to a DMFT calculation. We focus on implementation details relevant to a number of technologies for which Sandia has design, fabrication, and measurement expertise. The primary technical advances outlined in this report concern the development of a supporting modeling capability. As with all analog quantum simulation platforms, the successful design and operation of individual devices depends critically on one's ability to predict the effective low-energy Hamiltonian governing its dynamics Our project has made this possible and lays the foundation for future experimental implementations.
The interplay of a rapidly changing climate and infectious disease occurrence is emerging as a critical topic, requiring investigation of possible direct, as well as indirect, connections between disease processes and climate-related variation and phenomena. First, we introduce and overview three infectious disease exemplars (dengue, influenza, valley fever) representing different transmission classes (insect-vectored, human-to-human, environmentally-transmitted) to illuminate the complex and significant interplay between climate disease processes, as well as to motivate discussion of how Sandia can transform the field, and change our understanding of climate-driven infectious disease spread. We also review state-of-the-art epidemiological and climate modeling approaches, together with data analytics and machine learning methods, potentially relevant to climate and infectious disease studies. We synthesize the modeling and disease exemplars information, suggesting initial avenues for research and development (R&D) in this area, and propose potential sponsors for this work. Whether directly or indirectly, it is certain that a rapidly changing climate will alter global disease burden. The trajectory of climate change is an important control on this burden, from local, to regional and global scales. The efforts proposed herein respond to the National Research Councils call for the creation of a multidisciplinary institute that would address critical aspects of these interlocking, cascading crises.
Deep convolutional neural networks (DCNNs) currently provide state-of-the-art performance on image classification and object detection tasks, and there are many global security mission areas where such models could be extremely useful. Crucially, the success of these models is driven in large part by the widespread availability of high-quality open source data sets such as Image Net, Common Objects in Context (COCO), and KITTI, which contain millions of images with thousands of unique labels. However, global security relevant objects-of-interest can be difficult to obtain: relevant events are low frequency and high consequence; the content of relevant images is sensitive; and adversaries and proliferators seek to obscure their activities. For these cases where exemplar data is hard to come-by, even fine-tuning an existing model with available data can be effectively impossible. Recent work demonstrated that models can be trained using a combination of real-world and synthetic images generated from 3D representations; that such models can exceed the performance of models trained using real-world data alone; and that the generated images need not be perfectly realistic (Tremblay, et al., 2018). However, this approach still required hundreds to thousands of real-world images for training and fine tuning, which for sparse, global security-relevant datasets can be an unrealistic hurdle. In this research, we validate the performance and behavior of DCNN models as we drive the number of real-world images used for training object detection tasks down to a minimal set. We perform multiple experiments to identify the best approach to train DCNNs from an extremely small set of real-world images. In doing so, we: Develop state-of-the-art, parameterized 3D models based on real-world images and sample from their parameters to increase the variance in synthetic image training data; Use machine learning explainability techniques to highlight and correct through targeted training the biases that result from training using completely synthetic images; and Validate our results by comparing the performance of the models trained on synthetic data to one another, and to a control model created by fine-tuning an existing ImageNet-trained model with a limited number (hundreds) of real-world images.
The Arctic is warming and feedbacks in the coupled Earth system may be driving the Arctic to tipping events that could have critical downstream impacts for the rest of the globe. In this project we have focused on analyzing sea ice variability and loss in the coupled Earth system Summer sea ice loss is happening rapidly and although the loss may be smooth and reversible, it has significant consequences for other Arctic systems as well as geopolitical and economic implications. Accurate seasonal predictions of sea ice minimum extent and long-term estimates of timing for a seasonally ice-free Arctic depend on a better understanding of the factors influencing sea ice dynamics and variation in this strongly coupled system. Under this project we have investigated the most influential factors in accurate predictions of September Arctic sea ice extent using machine learning models trained separately on observational data and on simulation data from five E3SM historical ensembles. Monthly averaged data from June, July, and August for a selection of ice, ocean, and atmosphere variables were used to train a random forest regression model. Gini importance measures were computed for each input feature with the testing data. We found that sea ice volume is most important earlier in the season (June) and sea ice extent became a more important predictor closer to September. Results from this study provide insight into how feature importance changes with forecast length and illustrates differences between observational data and simulated Earth system data. We have additionally performed a global sensitivity analysis (GSA) using a fully coupled ultra- low resolution configuration E3SM. To our knowledge, this is the first global sensitivity analysis involving the fully-coupled E3SM Earth system model. We have found that parameter variations show significant impact on the Arctic climate state and atmospheric parameters related to cloud parameterizations are the most significant. We also find significant interactions between parameters from different components of E3SM. The results of this study provide invaluable insight into the relative importance of various parameters from the sea ice, atmosphere and ocean components of the E3SM (including cross-component parameter interactions) on various Arctic-focused quantities of interest (QOIs).
Digital Systems Engineering strategies typically call for digital Systems Engineering models to be retained in repositories and certified as an authoritative source of truth (enabling model reuse, qualification, and collaboration). In order for digital Systems Engineering models to be certified as authoritative (credible), they need to be assessed - verified and validated - and with the amount of uncertainty in the model quantified (consider reusing someone else's model without knowing the author). Due to this increasing model complexity, the authors assert that traditional human-based methods for validating, verifying, and uncertainty quantification - such as human-based peer-review sessions - cannot sufficiently establish that a digital Systems Engineering model of a complex system is credible. Digital Systems Engineering models of complex systems can contain millions of nodes and edges. The authors assert that this level of detail is beyond the ability of any group of humans - even working for weeks at a time - to discern and catch every minor model infraction. In contrast, computers are highly effective at discerning infractions with massive amounts of information. The authors suggest that a better approach might be to focus the humans at what model patterns should be assessed and enable the computer to assess the massive details in accordance with those patterns - by running through perhaps 100,000 test loops. In anticipation of future projects to implement and automate the assessment of models at Sandia National Laboratories, a study was initiated to elicit input from a group of 25 Systems Engineering experts. The authors positioning query began with - 'What questions would a Systems Engineer ask to assess Systems Engineering models for credibility?" This report documents the results of that survey.
We propose to develop a computational sensitivity analysis capability for Monte Carlo sampling-based particle simulation relevant to Aleph, Cheetah-MC, Empire, Emphasis, ITS, SPARTA, and LAMMPS codes. These software tools model plasmas, radiation transport, low-density fluids, and molecular motion. Our report demonstrates how adjoint optimization methods can be combined with Monte Carlo sampling-based adjoint particle simulation. Our goal is to develop a sensitivity analysis to drive robust design-based optimization for Monte Carlo sampling-based particle simulation - a currently unavailable capability.
We explore the use of multiple deep learning models for detecting flaws in software programs. Current, standard approaches for flaw detection rely on a single representation of a software program (e.g., source code or a program binary). We illustrate that, by using techniques from multimodal deep learning, we can simultaneously leverage multiple representations of software programs to improve flaw detection over single representation analyses. Specifically, we adapt three deep learning models from the multimodal learning literature for use in flaw detection and demonstrate how these models outperform traditional deep learning models. We present results on detecting software flaws using the Juliet Test Suite and Linux Kernel.
In this project, we investigate the use of neural networks for the prediction of molecular properties, namely the interatomic potential. We use the machine learning package Tensorflow to build a variety of neural networks and compare performance with a popular Fortran package - Atomic Energy Networks (aenet). There are two primary goals for this work: 1) use the wide availability of different optimization techniques in Tensorflow to outperform aenet and 2) use new descriptors that can outperform Behler descriptors.
Software flaw detection using multimodal deep learning models has been demonstrated as a very competitive approach on benchmark problems. In this work, we demonstrate that even better performance can be achieved using neural architecture search (NAS) combined with multimodal learning models. We adapt a NAS framework aimed at investigating image classification to the problem of software flaw detection and demonstrate improved results on the Juliet Test Suite, a popular benchmarking data set for measuring performance of machine learning models in this problem domain.
This report summarizes work done under the Laboratory Directed Research and Development (LDRD) project titled "Incorporating physical constraints into Gaussian process surrogate models?' In this project, we explored a variety of strategies for constraint implementations. We considered bound constraints, monotonicity and related convexity constraints, Gaussian processes which are constrained to satisfy linear operator constraints which represent physical laws expressed as partial differential equations, and intrinsic boundary condition constraints. We wrote three papers and are currently finishing two others. We developed initial software implementations for some approaches. This report summarizes the work done under this LDRD.
In a previous work, embedded ensemble propagation was proposed to improve the efficiency of sampling-based uncertainty quantification methods of computational models on emerging computational architectures. It consists of simultaneously evaluating the model for a subset of samples together, instead of evaluating them individually. A first approach introduced to solve parametric linear systems with ensemble propagation is ensemble reduction. In Krylov methods for example, this reduction consists in coupling the samples together using an inner product that sums the sample contributions. Ensemble reduction has the advantages of being able to use optimized implementations of BLAS functions and having a stopping criterion which involves only one scalar. However, the reduction potentially decreases the rate of convergence due to the gathering of the spectra of the samples. In this paper, we investigate a second approach: ensemble propagation without ensemble reduction in the case of GMRES. This second approach solves each sample simultaneously but independently to improve the convergence compared to ensemble reduction. This raises two new issues which are solved in this paper: the fact that optimized implementations of BLAS functions cannot be used anymore and that ensemble divergence, whereby individual samples within an ensemble must follow different code execution paths, can occur. We tackle those issues by implementing a high-performing ensemble GEMV and by using masks. The proposed ensemble GEMV leads to a similar cost per GMRES iteration for both approaches, i.e. with and without reduction. For illustration, we study the performances of the new linear solver in the context of a mesh tying problem. This example demonstrates improved ensemble propagation speed-up without reduction.
Computing uses energy. At the bare minimum, erasing information in a computer increases the entropy. Landauer has calculated %7E kBT log(2) Joules is dissipated per bit of energy erased. While the success of Moores law has allowed increasing computing power and efficiency for many years, these improvements are coming to an end. This project asks if there is a way to continue those gains by circumventing Landauer through reversible computing. We explore a new reversible computing paradigm, asynchronous ballistic reversible computing or ABRC. The ballistic nature of data in ABRC matches well with superconductivity which provides a low-loss environment and a quantized bit encoding the fluxon. We discuss both these and our development of a superconducting fabrication process at Sandia. We describe a fully reversible 1-bit memory cell based on fluxon dynamics. Building on this model, we propose several other gates which may also offer reversible operation.
This report summarizes the work performed under a one-year LDRD project aiming to enable accurate and robust numerical simulation of partial differential equations for meshes that are of poor quality. Traditional finite element methods use the mesh to both discretize the geometric domain and to define the finite element shape functions. The latter creates a dependence between the quality of the mesh and the properties of the finite element basis that may adversely affect the accuracy of the discretized problem. In this project, we propose a new approach for defining finite element shape functions that breaks this dependence and separates mesh quality from the discretization quality. At the core of the approach is a meshless definition of the shape functions, which limits the purpose of the mesh to representing the geometric domain and integrating the basis functions without having any role in their approximation quality. The resulting non-conforming space can be utilized within a standard discontinuous Galerkin framework providing a rigorous foundation for solving partial differential equations on low-quality meshes. We present a collection of numerical experiments demonstrating our approach in a wide range of settings: strongly coercive elliptic problems, linear elasticity in the compressible regime, and the stationary Stokes problem. We demonstrate convergence for all problems and stability for element pairs for problems which usually require inf-sup compatibility for conforming methods, also referring to a minor modification possible through the symmetric interior penalty Galerkin framework for stabilizing element pairs that would otherwise be traditionally unstable. Mesh robustness is particularly critical for elasticity, and we provide an example that our approach provides a greater than 5x improvement in accuracy and allows for taking an 8x larger stable timestep for a highly deformed mesh, compared to the continuous Galerkin finite element method. The report concludes with a brief summary of ongoing projects and collaborations that utilize or extend the products of this work.
Computing stands to be radically improved by neuromorphic computing (NMC) approaches inspired by the brain's incredible efficiency and capabilities. Most NMC research, which aims to replicate the brain's computational structure and architecture in man-made hardware, has focused on artificial intelligence; however, less explored is whether this brain-inspired hardware can provide value beyond cognitive tasks. We demonstrate that high-degree parallelism and configurability of spiking neuromorphic architectures makes them well-suited to implement random walks via discrete time Markov chains. Such random walks are useful in Monte Carlo methods, which represent a fundamental computational tool for solving a wide range of numerical computing tasks. Additionally, we show how the mathematical basis for a probabilistic solution involving a class of stochastic differential equations can leverage those simulations to provide solutions for a range of broadly applicable computational tasks. Despite being in an early development stage, we find that NMC platforms, at a sufficient scale, can drastically reduce the energy demands of high-performance computing platforms.
Energy transfer through anharmonically-coupled vibrations influences the earliest chemical steps in shockwave-induced detonation in energetic materials. A mechanistic description of vibrational energy transfer is therefore necessary to develop predictive models of energetic material behavior. We performed transient broadband infrared spectroscopy on hundreds of femtoseconds to hundreds of picosecond timescales as well as density functional theory and molecular dynamics simulations to investigate the evolution of vibrational energy distribution in thin film samples of pentaerythritol tetranitrate (PETN) , 1,3,5 - trinitroperhydro - 1,3,5 - triazine (RDX) , and 2,4,6 - triamino 1,3,5 - trinitrobenzene (TATB). Experimental results show dynamics on multiple timescales, providing strong evidence for coupled vibrations in these systems, as well as material-dependent evolution on tens to hundreds of picosecond timescales. Theoretical results also reveal pathways and distinct timescales for energy transfer through coupled vibrations in the three investigated materials, providing further insight into the mechanistic underpinnings of energy transfer dynamics in energetic material sensitivity.
This report summarizes the goals and findings of eight research projects conducted under the Computing and Information Sciences (CIS) Research Foundation and related to the COVID- 19 pandemic. The projects were all formulated in response to Sandia's call for proposals for rapid-response research with the potential to have a positive impact on the global health emergency. Six of the projects in the CIS portfolio focused on modeling various facets of disease spread, resource requirements, testing programs, and economic impact. The two remaining projects examined the use of web-crawlers and text analytics to allow rapid identification of articles relevant to specific technical questions, and categorization of the reliability of content. The portfolio has collectively produced methods and findings that are being applied by a range of state, regional, and national entities to support enhanced understanding and prediction of the pandemic's spread and its impacts.
The overall goal of this work was to accelerate simulations supporting the nuclear deterrence (ND) mission through improved performance of key algorithms in the ASC IC Sierra multi-physics application suite. This work focused on porting and optimizing algorithms for the graphics processing units (GPU) on the second ASC advanced technology system (ATS-2), while maintaining or improving performance on commodity technology systems (CTS) and ATS-1. Furthermore, these algorithmic developments used the ASC developed Kokkos performance portability abstraction library to maintain high performance across platforms using identical code, and enable sustainable reduced-cost migration and performance optimization to emerging hardware.
