Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “compute in memory”

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 937 records · Page 52
All sourcesNASA NTRSDOE OSTI
DOE OSTI2020

QMCPACK: Advances in the development, efficiency, and application of auxiliary field and real-space variational and diffusion quantum Monte Carlo

We review recent advances in the capabilities of the open source ab initio Quantum Monte Carlo (QMC) package QMCPACK and the workflow tool Nexus used for greater efficiency and reproducibility. The auxiliary field QMC (AFQMC) implementation has been greatly expanded to include k-point symmetries, tensor-hypercontraction, and accelerated graphical processing unit (GPU) support. These scaling and memory reductions greatly increase the number of orbitals that can practically be included in AFQMC calculations, increasing the accuracy. Advances in real space methods include techniques for accurate computation of bandgaps and for systematically improving the nodal surface of ground state wavefunctions. Results of these calculations can be used to validate application of more approximate electronic structure methods, including GW and density functional based techniques. To provide an improved foundation for these calculations, we utilize a new set of correlation-consistent effective core potentials (pseudopotentials) that are more accurate than previous sets; these can also be applied in quantum-chemical and other many-body applications, not only QMC. These advances increase the efficiency, accuracy, and range of properties that can be studied in both molecules and materials with QMC and QMCPACK.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
DOE OSTI2024

Single-shot quantum error correction in intertwined toric codes

We construct a subsystem code in three dimensions that exhibits single-shot error correction in a user-friendly and transparent way. As this code is a subsystem version of coupled toric codes, we call it the intertwined toric code (ITC). Although previous codes share the property of single-shot error correction, the ITC is distinguished by its physically motivated origin, geometrically straightforward logical operators and errors, and a simple phase diagram. The code arises from three-dimensional (3D) stabilizer toric codes in a way that emphasizes the physical origin of the single-shot property. In particular, starting with two copies of the 3D toric code, we add check operators that provide for the confinement of pointlike excitations without condensing the loop excitations. Geometrically, the bare and dressed logical operators in the ITC derive from logical operators in the underlying toric codes, creating a clear relationship between errors and measurement outcomes. The syndromes of the ITC resemble the syndromes of the single-shot code by Kubica and Vasmer, allowing us to use their decoding schemes. We also extract the phase diagram corresponding to ITC and show that it contains the phases found in the Kubica-Vasmer code. Lastly, we suggest various connections to Walker-Wang models and measurement-based quantum computation.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗
DOE OSTI2021

Semantic segmentation with a sparse convolutional neural network for event reconstruction in MicroBooNE

We present the performance of a semantic segmentation network, SparseSSNet, that provides pixel-level classification of MicroBooNE data. The MicroBooNE experiment employs a liquid argon time projection chamber for the study of neutrino properties and interactions. SparseSSNet is a submanifold sparse convolutional neural network, which provides the initial machine learning based algorithm utilized in one of MicroBooNE's ν e -appearance oscillation analyses. The network is trained to categorize pixels into five classes, which are re-classified into two classes more relevant to the current analysis. The output of SparseSSNet is a key input in further analysis steps. This technique, used for the first time in liquid argon time projection chambers data and is an improvement compared to a previously used convolutional neural network, both in accuracy and computing resource utilization. Here, the accuracy achieved on the test sample is ≥ 99%. For full neutrino interaction simulations, the time for processing one image is ≈ 0.5 sec, the memory usage is at 1 GB level, which allows utilization of most typical CPU worker machine.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗
DOE OSTI2020

PANTHER: A Programmable Architecture for Neural Network Training Harnessing Energy-Efficient ReRAM

The wide adoption of deep neural networks has been accompanied by ever-increasing energy and performance demands due to the expensive nature of training them. Additionally, numerous special-purpose architectures have been proposed to accelerate training: both digital and hybrid digital-analog using resistive RAM (ReRAM) crossbars. ReRAM-based accelerators have demonstrated the effectiveness of ReRAM crossbars at performing matrix-vector multiplication operations that are prevalent in training. However, they still suffer from inefficiency due to the use of serial reads and writes for performing the weight gradient and update step. A few works have demonstrated the possibility of performing outer products in crossbars, which can be used to realize the weight gradient and update step without the use of serial reads and writes. However, these works have been limited to low precision operations which are not sufficient for typical training workloads. Moreover, they have been confined to a limited set of training algorithms for fully-connected layers only. To address these limitations, we propose a bit-slicing technique for enhancing the precision of ReRAM-based outer products, which is substantially different from bit-slicing for matrix-vector multiplication only. We incorporate this technique into a crossbar architecture with three variants catered to different training algorithms. To evaluate our design on different types of layers in neural networks (fully-connected, convolutional, etc.) and training algorithms, we develop PANTHER, an ISA-programmable training accelerator with compiler support. Our design can also be integrated into other accelerators in the literature to enhance their efficiency. Our evaluation shows that PANTHER achieves up to 8.02×, 54.21×, and 103× energy reductions as well as 7.16×, 4.02×, and 16× execution time reductions compared to digital accelerators, ReRAM-based accelerators, and GPUs, respectively.

42 ENGINEERING↗
DOE OSTI2022

Reduced Dynamics of Full Counting Statistics

We present a theory of modified reduced dynamics in the presence of counting fields. Reduced dynamics techniques are useful for describing open quantum systems at long emergent timescales when the memory timescales are short. However, they can be difficult to formulate for observables spanning the system and its environment, such as those characterizing transport properties. A large variety of mixed system--environment observables, as well as their statistical properties, can be evaluated by considering counting fields. Given a numerical method able to simulate the field-modified dynamics over the memory timescale, we show that the long-lived full counting statistics can be efficiently obtained from the reduced dynamics. We demonstrate the utility of the technique by computing the long-time current in the nonequilibrium Anderson impurity model from short-time Monte Carlo simulations.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
DOE OSTI2022

Access Patterns and Performance Behaviors of Multi-layer Supercomputer I/O Subsystems under Production Load

Scientific computing workloads at HPC facilities have been shifting from traditional numerical simulations to AI/ML applications for training and inference while processing and producing ever-increasing amounts of scientific data. To address the growing need for increased storage capacity, lower access latency, and higher bandwidth, emerging technologies such as non-volatile memory are integrated into supercomputer I/O subsystems. With these emerging trends, we need a better understanding of the multilayer supercomputer I/O systems and ways to use these subsystems efficiently. In this work, we study the I/O access patterns and performance characteristics of two representative supercomputer I/O subsystems. Through an extensive analysis of year-long I/O logs on each system, we report new observations in I/O reads and writes, unbalanced use of storage system layers, and new trends in user behaviors at the HPC I/O middleware stack.

Bez, JL↗
DOE OSTI2023

High Fidelity Computational Model for Fluidized Bed Experiments (Final Report)

MFiX was integrated with preconditioners and linear solver packages in Trilinos via MFIX, Fortran, C, and C++ wrappers. The MFIX wrapper interpreted the structure of the matrix and vector from MFiX while the Fortran wrapper transfers this information to a low-level language, C. C-wrapper transforms the memory semantic between Fortran and C++ language. C++ wrapper passes the matrix and vector to a Trilinos package and gets the solution. This solution was transferred to MFiX via C++, C, Fortran, MFIX wrapper. The framework is used to integrate first, second and third generation linear solvers in Trilinos with MFiX.

20 FOSSIL-FUELED POWER PLANTS↗
DOE OSTI2024

Modeling the Impacts of Material Properties on Oscillatory Neuron Behavior

In this study, neuromorphic computing, which mimics the functions of biological brains, offers improvements in both latency and energy efficiency over typical von Neumann computing architectures. Spiking neural networks can be especially power-efficient because they encode information temporally and can use more sparse electrical inputs. Here, we study the design of volatile memristors (variable resistors with memory) for neuronal devices, with particular consideration toward the feasibility of all-on-chip oscillation using built-in capacitance. We use circuit simulations to model the behavior of oscillator neurons with a range of realistic material properties. We find that energy inputs increase with insulating-phase resistivity, thermal conductivity, and device aspect ratio. However, we also find that the minimum capacitance needed for oscillation decreases with increasing insulating-phase resistivity, which opposes the constraints for power efficiency. Based on published data on NbO 2 , VO 2 , and EuNiO 3 , we find that existing materials can be engineered for all-on-chip spiking using their parasitic capacitance.

36 MATERIALS SCIENCE↗
DOE OSTI2020

Noisy Intermediate-Scale Quantum Applications on a Pathfinder System

Work performed under this one-year LDRD was concerned with estimating resource requirements for small quantum test beds that are expected to be available in the near future. This work represents a preliminary demonstration of our ability to leverage quantum hardware for solving small quantum simulation problems in areas of interest to the DOE. The algorithms enabling such studies are hybrid quantum-classical variational algorithms, in particular the widely-used variational quantum eigensolver (VQE). Employing this hybrid algorithm, in which the quantum computer complements the classical one, we implemented an end-to-end application-level toolchain that allows the user to specify a molecule of interest and compute the ground state energy using the VQE approach. We found significant limitations attributable to the classical portion of the hybrid system, including a greater than greater-than-quartic power scaling of the classical memory requirements with the system size. Current VQE approaches would require an exascale machine for solving any molecule with size greater than 150 nuclei. Our findings include several improvements that we implemented into the VQE toolchain, including a new classical optimizer that is decades old but hadn't been considered before in the VQE ecosystem. Our findings suggest limitations to variational hybrid approaches to simulation that further motivate the need for a gate-based fault-tolerant quantum processor that can implement larger problems using the fully digital quantum phase estimation algorithm.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
DOE OSTI2021

Dimensionality reduction of the many-body problem using coupled-cluster subsystem flow equations: classical and quantum computing perspective

We discuss reduced-scaling strategies employing recently introduced sub-system embedding sub-algebras coupled-cluster formalism (SES-CC) to describe many-body systems. These strategies utilize properties of the SES-CC formulations where the equations describing certain classes of sub- systems can be integrated into a computational flows composed coupled eigenvalue problems of reduced dimensionality. Additionally, these flows can be defined at the level of the CC Ansatz defined by selected classes of cluster amplitudes, which define the wave function ”memory” of possible partitionings of the many-body system into constituent sub-systems. One of the possible ways of solving these coupled problems is through implementing procedures, where the information is passed between the sub-systems in a self-consistent manner. As a special case, we consider local flow formulations where the so-called local character of correlation effects can be closely related to properties of sub-system embedding sub-algebras employing localized molecular basis. We also generalize flow equations to the time domain and to downfolding methods utilizing double exponential unitary CC Ansatz (DUCC), where reduced dimensionality of constituent sub-problems offer a possibility of efficient utilization of limited quantum resources in modeling realistic systems.

Electron correlation, quantum chemistry, quantum c↗
DOE OSTI2020

Organismic materials for beyond von Neumann machines

The elementary basis of intelligence in organisms with a central nervous system includes neurons and synapses and their complex interconnections forming neural circuits. In non-neural organisms such as slime mold with gel-like media, viscosity modulation enables adaptation to changing environments. At a larger scale, collective intelligence emerges via social interactions and feedback in animal colonies. Learning and memory are therefore multi-scale features that evolve as a result of constant interactions with the environment. There is growing interest in emulating such features of intelligence in computing machines and autonomous systems. Materials that can respond to their environment in a manner similar to organisms (referred to as "organismic materials") therefore may be of interest as hardware components in artificial intelligence machines. In this brief review, we present a class of semiconductors called correlated oxides as candidates for learning machines. The term "correlated" refers to the fact that electrons in such lattices strongly interact and the ground state is not what is predicted by classical band theory. Such materials can undergo insulator-metal transitions at near ambient conditions under external stimuli such as thermal or electrical fields, strain, and chemical doping. Depending on the mechanism driving the transition, intermediate states can be metastable with different volatilities, and the time scales of phase change can be controlled over many orders of magnitude. The change in electronic properties can be sharp or gradual, leading to digital or analog behavior. These properties enable the realization of artificial neurons and synapses and emulate the associative and non-associative learning characteristics found in various organisms. We examine microscopic properties concerning electronic and structural transitions leading to collective behavior and theoretical treatments of the ground state and dynamical response, showcasing VO 2 as a model system. Next, we briefly review algorithms designed from the plasticity demonstrated by phase changing systems. Finally, we conclude the brief review with suggestions for future research toward realizing non-von Neumann machines.

42 ENGINEERING↗
DOE OSTI2024

Lamellar: A Rust-based Asynchronous Tasking and PGAS Runtime for High Performance Computing

Cybersecurity is one of the largest concerns in modern computing, impacting and dictating how governments, private corporations, and individuals interact with and live in an increasingly digital world. The NSA has recently released a memo [ 1] on “Software Memory Safety” where they highlight that both Microsoft and Google have stated around 70% of software vulnerabilities were due to memory safety issues. Although languages such as C and C++ provide freedom and flexibility with memory management, guaran- teeing safety falls mostly on the developer. The NSA recommends using “memory safe” languages whenever possible. In this paper we introduce Lamellar, an asynchronous tasking and PGAS HPC runtime written in Rust, one such "memory safe" language. We describe the entire Lamellar stack, from network interfaces to high- level abstractions such as distributed LamellarArrays and Active Messages. We conclude by showing comparable performance to legacy PGAS runtimes (e.g. OpenSHMEM) on a subset of the BALE kernel suite while maintaining strong memory safety principles.

HPC Software Systems, Rust Programming Language, P↗
DOE OSTI2021

Real time trajectory optimization for hybrid energy management utilizing connected information technologies

The disclosure is directed to solving a full trajectory optimization problem in real-time for a hybrid electric vehicle (HEV) such that future driving conditions and energy usage may be fully considered in determining optimal engine energy usage and battery energy usage in real-time during a trip. An electronic control unit of the HEV may be configured to: receive route information for a route to be driven by the HEV; and after receiving the route information, iterating the operations of: measuring a current state of charge (SOC) of the battery; using at least the measured SOC and an initial co-state value stored in a memory, performing a process to iteratively update the co-state value to obtain an updated co-state value; using at least the updated co-state value, computing an updated control value; and applying the updated control value to control a usage of the battery and the internal combustion engine.

Huang, Mike X.↗
DOE OSTI2021

Hardware Evaluation Analytical Modeling and Node Simulation: Benefits of Tighter GPU Integration

In this report, we examine several emerging technologies of interest to the Department of Energy and its computational centers. These include: 1) quantifying the benefit of tighter CPU-GPU integration, 2) quantifying the appropriate CPU core:GPU ratio, 3) quantifying the penalty for CPU-GPU disaggregation, 4) quantifying the benefits of tighter GPU-GPU integration, 5) quantifying the benefits of unified memory, and 6) quantifying the benefits of tighter FPGA-GPU integration.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2025

Accelerating charge estimation in molecular dynamics simulations using physics-informed neural networks: corrosion applications

Molecular Dynamics (MD) simulations are used to understand the effects of corrosion on metallic materials in salt brine. Reactive force fields in classical MD enable accurate modeling of bond formation and breakage in the aqueous medium and at the metal-electrolyte interface, while also facilitating dynamic partial charge equilibration. However, MD simulations are computationally intensive and unsuitable for modeling the long time scales characteristic of corrosive phenomena. To address this, we develop reduced-order machine learning models that provide accurate and efficient predictions of charge density in corrosive environments. Specifically, we use Long Short-Term Memory (LSTM) networks to forecast charge density evolution based on atomic environments represented by Smooth Overlap of Atomic Positions (SOAP) descriptors. A physics-informed loss function enforces charge neutrality and electronegativity equivalence. The atomic charges predicted by the deep learning model trained on this work were obtained two orders of magnitude faster than those from molecular dynamics (MD) simulations, with an error of less than 3% compared to the MD-obtained charges, even in extrapolative scenarios, while adhering to physical constraints. This demonstrates the excellent accuracy, computational efficiency, and validity of the developed model. Lastly, even though developed for corrosion, these protocols are formulated in a phenomenon-agnostic manner, allowing application to various variable-charge interatomic potentials and related fields.

Atomistic models↗
DOE OSTI2021

Real-Time and Adaptive Reservoir Computing With Application to Profile Prediction in Fusion Plasma

Nuclear fusion is a promising alternative to address the problem of sustainable energy production. The tokamak is an approach to fusion based on magnetic plasma confinement, constituting a complex physical system with many control challenges. Here we study the characteristics and optimization of reservoir computing (RC) for real-time and adaptive prediction of plasma profiles in the DIII-D tokamak. Our experiments demonstrate that RC achieves comparable results to state-of-the-art (deep) convolutional neural networks (CNNs) and long short-term memory (LSTM) models, with a significantly easier and faster training procedure. This efficient approach allows for fast and frequent adaptation of the model to new situations, such as changing plasma conditions or different fusion devices.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
DOE OSTI2024

On a Simplified Approach to Achieve Parallel Performance and Portability Across CPU and GPU Architectures

This paper presents software advances to easily exploit computer architectures consisting of a multi-core CPU and CPU+GPU to accelerate diverse types of high-performance computing (HPC) applications using a single code implementation. The paper describes and demonstrates the performance of the open-source C++ matrix and array (MATAR) library that uniquely offers: (1) a straightforward syntax for programming productivity, (2) usable data structures for data-oriented programming (DOP) for performance, and (3) a simple interface to the open-source C++ Kokkos library for portability and memory management across CPUs and GPUs. The portability across architectures with a single code implementation is achieved by automatically switching between diverse fine-grained parallelism backends (e.g., CUDA, HIP, OpenMP, pthreads, etc.) at compile time. The MATAR library solves many longstanding challenges associated with easily writing software that can run in parallel on any computer architecture. This work benefits projects seeking to write new C++ codes while also addressing the challenges of quickly making existing Fortran codes performant and portable over modern computer architectures with minimal syntactical changes from Fortran to C++. We demonstrate the feasibility of readily writing new C++ codes and modernizing existing codes with MATAR to be performant, parallel, and portable across diverse computer architectures.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2020

Data-driven recovery of hidden physics in reduced order modeling of fluid flows

In this article, we introduce a modular hybrid analysis and modeling (HAM) approach to account for hidden physics in reduced order modeling (ROM) of parameterized systems relevant to fluid dynamics. The hybrid ROM framework is based on using first principles to model the known physics in conjunction with utilizing the data-driven machine learning tools to model the remaining residual that is hidden in data. This framework employs proper orthogonal decomposition as a compression tool to construct orthonormal bases and Galerkin projection (GP) as a model to build the dynamical core of the system. Our proposed methodology hence compensates structural or epistemic uncertainties in models and utilizes the observed data snapshots to compute true modal coefficients spanned by these bases. The GP model is then corrected at every time step with a data-driven rectification using a long short-term memory (LSTM) neural network architecture to incorporate hidden physics. A Grassmannian manifold approach is also adopted for interpolating basis functions to unseen parametric conditions. The control parameter governing the system's behavior is thus implicitly considered through true modal coefficients as input features to the LSTM network. The effectiveness of the HAM approach is then discussed through illustrative examples that are generated synthetically to take hidden physics into account. Furthermore, our approach thus provides insights addressing a fundamental limitation of the physics-based models when the governing equations are incomplete to represent underlying physical processes.

42 ENGINEERING↗