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 721 records · Page 40
All sourcesNASA NTRSDOE OSTI
DOE OSTI2021

Skyrme pseudopotentials at next-to-next-to-leading order: Construction of local densities and first symmetry-breaking calculations

There is an ongoing quest to improve on the spectroscopic quality of nuclear energy density functionals (EDFs) of the Skyrme type through extensions of its traditional form. One direction for such activities is the inclusion of terms of higher order in gradients in the EDF. We report on exploratory symmetry-breaking calculations performed for an extension of the Skyrme EDF that includes central terms with four gradients at next-to-next-to-leading order (N2LO) and for which the high-quality parametrization SN2LO1 has been constructed recently. Up to now, the investigation of such functionals with higher-order terms was limited to infinite matter and spherically symmetric configurations of singly and doubly magic nuclei. We address here nuclei and phenomena that require us to consider axial and nonaxial deformation, both for reflection-symmetric and also reflection-asymmetric shapes, as well as the breaking of time-reversal invariance. Achieving these calculations demanded a number of formal developments. These all resulted from the formulation of the N2LO EDF requiring the introduction of new local densities with additional gradients that are not present in the EDF at NLO. Their choice is not unique, but can differ in the way the gradients are coupled. While designing a numerical implementation of N2LO EDFs in Cartesian three-dimensional coordinate-space representation, we have developed a novel definition and a new unifying notation for normal and pair densities that contain gradients at arbitrary order. Besides having mnemonic advantages, the new notation allows for the easy identification of redundancies and reducibilities in a given set of local densities, and the new definition makes it straightforward to construct densities that automatically adopt the symmetries of the many-body state they are constructed from. The resulting scheme resolves several issues with some of the choices that have been made for local densities in the past, in particular when breaking time-reversal symmetry. Guided by general practical considerations, we propose an alternative form of the N2LO contribution to the Skyrme EDF that is built from a different set of densities. It has exactly the same physics content, but is much more efficient to handle in formal discussions and, compared to the original formulation, leads to a substantial reduction of computational cost and memory requirements in deformed codes. As representative examples for the performance of SN2LO1, we have chosen the ground states of even-even Kr and Nd isotopes, the fission barrier of 240 Pu as well as the superdeformed rotational band of 194 Hg. Overall, for the nuclei and phenomena studied here, the SN2LO1 parametrization does not yet present a systematic improvement over standard NLO parametrizations. This finding calls for improved fit protocols that better discriminate between NLO and N2LO terms and better exploit the unique features of the additional degrees of freedom offered by the latter.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗
DOE OSTI2022

Throughput-Oriented and Accuracy-Aware DNN Training with BFloat16 on GPU

Deep Neural Networks (DNNs) have transformed the field of artificial intelligence and achieved extraordinary success in many areas. The training of DNNs is commonly compute and memory-intensive, which has resulted in several optimizations in the training phase. Among them, reduced precision is a typical and widely used technique to accelerate DNN training and reduce memory requirements. However, applying a widely adopted reduced precision format such as Float16 to all involved operations in DNN training is not optimal as the use of Float16 in some operations can hurt model accuracy. Meanwhile, additional optimizations including loss scaling and autocast techniques can mitigate the accuracy loss but lead to inherent overhead and inadequate use of reduced precision. In this work, we leverage another reduced precision format, BFloat16, and introduce a throughput-oriented and accuracy-aware approach to maximize the performance potential of DNN training. Since the high throughput provided by BFloat16 format is accompanied by low precision of the floating-point representation, this approach achieves high throughput by using BFloat16 on all DNN operations and avoids the accuracy loss through a customized accuracy-aware normalization. Results show that our approach outperforms the state-of-the-art mixed-precision training by 1.21x on an NVIDIA A100 GPU.

Xie, Zhen↗
DOE OSTI2021

Really Embedding Domain-Specific Languages into C++

The following topics are dealt with: program compilers; optimising compilers; parallel processing; software engineering; learning (artificial intelligence); multiprocessing systems; shared memory systems; optimisation; computational complexity; specification languages.

Finkel, Hal J.↗
DOE OSTI2021

O3BNN-R: An Out-Of-Order Architecture for High-Performance and Regularized BNN Inference

Binarized Neural Networks (BNN) have drawn tremendous attention due to significantly reduced computational complexity and memory demand. They have especially shown great potential in cost- and power-restricted domains, such as IoT and smart edge-devices, where reaching a certain accuracy bar is often sufficient, and real-time is highly desired.In this work, we demonstrate that the highly-condensed BNN model can be shrunk significantly further by dynamically pruning irregular redundant edges. Based on two new observations on BNN-specific properties, an out-of-order (OoO) architecture – O3BNN-R, can curtail edge evaluation in cases where the binary output of a neuron can be determined early. Similar to Instruction-Level-Parallelism(ILP), these fine-grained, irregular, runtime pruning opportunities are traditionally presumed to be difficult to exploit. In order to increase the pruning opportunities, we also optimize the training process by adding 2 regularization items in the loss function (1) for pooling pruning and (2) for threshold pruning. We evaluate our design on an FPGA platform using three well-known networks, including VggNet-16, AlexNet for ImageNet, and a VGG-like network for Cifar-10.

Geng, Tong↗
DOE OSTI2026

The Memory Scaling of Reverse-Mode Differentiation in Particle Accelerator Simulations with Space Charge

The recent development of differentiable simulation codes for particle accelerators has enabled gradient-based workflows that promise finer control and more realistic modeling of accelerator facilities. However, when using reverse-mode automatic differentiation, the memory usage continuously increases during the simulation, and can potentially exceed the available hardware memory - especially when costly space charge computation is included. To study the memory requirements for differentiable simulations, we have implemented space charge in Cheetah, a PyTorch-based beam tracking code that supports reverse-mode differentiation. We find that the memory usage for reverse-mode differentiation grows linearly with the number of macroparticles and cells, and that it is proportional to the number of space charge kicks involved in the simulation. This general scaling can be used to evaluate whether a given differentiable simulation is feasible given hardware memory constraints.

Dhamrait, Arjun↗
DOE OSTI2024

AMReX v2024

The software framework, AMReX, supports the development of block-structured adaptive mesh refinement (AMR) algorithms for solving systems of partial differential equations. AMR reduces the computational cost and memory footprint compared to a uniform mesh while preserving the essential local descriptions of different physical processes in complex multiphysics algorithms. AMR uses a hierarchical representation of the solution at multiple levels of resolution where the solution on each level is defined on the union of data containers at that resolution. These data containers, which represent the solution over a logically rectangular subregion of the domain, can contain field data defined on a mesh, Lagrangian particles or combinations of both. In addition to these basic data types, AMReX supports a multilevel embedded boundary representation of complex geometry; linear solvers for cell-centered and nodal data; asynchronous I/O in a native format readable by ParaView, VisIt and yt; and interfaces to hypre and PETSc solvers. AMReX enables applications to run on distributed memory architectures with multicore CPUs and with GPU accelerators. AMReX uses a lightweight abstraction layer that effectively hides the details of the architecture from the application. The framework currently supports CUDA, HIP and SYCL for GPU acceleration and OpenMP for multi-core CPU architectures.

Almgren, Ann↗
DOE OSTI2020

PARAFAC_T1

This is a method of performing trilinear analysis on large data sets using a modification of the PARAFAC-ALS algorithm. It iteratively decomposes the data matrix into a core matrix and three loading matrices based on the Tucker1 model. The algorithm is particularly useful for data sets that are too large to upload into a computer?s main memory. While the performance advantage in utilizing our algorithm is dependent on the number of data elements and dimensions of the data array, we have seen a significant performance improvement over operating PARAFAC-ALS on the full data set. SAND2020-12649 M Sandia National Laboratories is a multimission laboratory managed and operated by National Technology & Engineering Solutions of Sandia, LLC, a wholly owned subsidiary of Honeywell International Inc., for the U.S. Department of Energy's National Nuclear Security Administration under contract DE-NA0003525.

Van Benthem, Mark↗
DOE OSTI2021

AMReX: Block-structured adaptive mesh refinement for multiphysics applications

Block-structured adaptive mesh refinement (AMR) provides the basis for the temporal and spatial discretization strategy for a number of Exascale Computing Project applications in the areas of accelerator design, additive manufacturing, astrophysics, combustion, cosmology, multiphase flow, and wind plant modeling. AMReX is a software framework that provides a unified infrastructure with the functionality needed for these and other AMR applications to be able to effectively and efficiently utilize machines from laptops to exascale architectures. AMR reduces the computational cost and memory footprint compared to a uniform mesh while preserving accurate descriptions of different physical processes in complex multiphysics algorithms. AMReX supports algorithms that solve systems of partial differential equations in simple or complex geometries and those that use particles and/or particle–mesh operations to represent component physical processes. In this article, we will discuss the core elements of the AMReX framework such as data containers and iterators as well as several specialized operations to meet the needs of the application projects. In addition, we will highlight the strategy that the AMReX team is pursuing to achieve highly performant code across a range of accelerator-based architectures for a variety of different applications.

Zhang, Weiqun↗
DOE OSTI2023

End-to-end GPU acceleration of low-order-refined preconditioning for high-order finite element discretizations

In this article, we present algorithms and implementations for the end-to-end GPU acceleration of matrix-free low-order-refined preconditioning of high-order finite element problems. The methods described here allow for the construction of effective preconditioners for high-order problems with optimal memory usage and computational complexity. The preconditioners are based on the construction of a spectrally equivalent low-order discretization on a refined mesh, which is then amenable to, for example, algebraic multigrid preconditioning. The constants of equivalence are independent of mesh size and polynomial degree. For vector finite element problems in H(curl) and H(div) (e.g., for electromagnetic or radiation diffusion problems), a specially constructed interpolation–histopolation basis is used to ensure fast convergence. Detailed performance studies are carried out to analyze the efficiency of the GPU algorithms. The kernel throughput of each of the main algorithmic components is measured, and the strong and weak parallel scalability of the methods is demonstrated. The different relative weighting and significance of the algorithmic components on GPUs and CPUs is discussed. Results on problems involving adaptively refined nonconforming meshes are shown, and the use of the preconditioners on a large-scale magnetic diffusion problem using all spaces of the finite element de Rham complex is illustrated.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2025

Efficient Hamiltonian encoding algorithms for extracting quantum control mechanism as interfering pathway amplitudes in the Dyson series

Hamiltonian encoding is a methodology for revealing the mechanism behind the dynamics governing controlled quantum systems. In this paper, following Mitra and Rabitz \cite{abhra_1}, we define mechanism via pathways of eigenstates that describe the evolution of the system, where each pathway is associated with a complex-valued amplitude corresponding to a term in the Dyson series. The evolution of the system is determined by the constructive and destructive interference of these pathway amplitudes. Pathways with similar attributes can be grouped together into pathway classes. The amplitudes of pathway classes are computed by modulating the Hamiltonian matrix elements and decoding the subsequent evolution of the system rather than by direct computation of the individual terms in the Dyson series. The original implementation of Hamiltonian encoding was computationally intensive and became prohibitively expensive in large quantum systems. This paper presents two new encoding algorithms that calculate the amplitudes of pathway classes by using techniques from graph theory and algebraic topology to exploit patterns in the set of allowed transitions, greatly reducing the number of matrix elements that need to be modulated. These new algorithms provide an exponential decrease in both computation time and memory utilization with respect to the Hilbert space dimension of the system. To demonstrate the use of these techniques, they are applied to two illustrative state-to-state transition problems.

Abrams, Erez [Princeton University, Massachusetts ↗
DOE OSTI2024

Enhancing Lattice Kinetic Schemes for Fluid Dynamics with Lattice-Equivariant Neural Networks

A new class of equivariant neural networks is presented, hereby dubbed lattice-equivariant neural networks (LENNs), designed to satisfy local symmetries of a lattice structure. The approach develops within a recently introduced framework aimed at learning neural network-based surrogate models’ lattice Boltzmann collision operators. Whenever neural networks are employed to model physical systems, respecting symmetries and equivariance properties has been shown to be key for accuracy, numerical stability, and performance. Here, hinging on ideas from group representation theory, trainable layers are defined whose algebraic structure is equivariant with respect to the symmetries of the lattice cell. In this work, the presented method naturally allows for efficient implementations, in terms of both memory usage and computational costs, supporting scalable training/testing for lattices in two spatial dimensions and higher (in which the size of symmetry group grows). The approach is validated and tested considering 2D and 3D flowing dynamics, both in laminar and turbulent regimes. It is compared with group-averaged-based symmetric networks and with plain, nonsymmetric, networks, showing how the presented approach unlocks the (a posteriori) accuracy and training stability of the former models and the train/inference speed of the latter networks. (LENNs are about one order of magnitude faster than group-averaged networks in 3D.) The work in this paper opens toward practical use of machine learning-augmented lattice Boltzmann CFD in real-world simulations.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2025

Space-Time Finite Element Tensor Network Approach for the Time-Dependent Convection–Diffusion–Reaction Equation with Variable Coefficients

In this paper, we present a new space-time Galerkin-like method, where we treat the discretization of spatial and temporal domains simultaneously. This method utilizes a mixed formulation of the tensor-train (TT) and quantized tensor-train (QTT) (please see Section Tensor-Train Decomposition), designed for the finite element discretization (Q1-FEM) of the time-dependent convection–diffusion–reaction (CDR) equation. We reformulate the assembly process of the finite element discretized CDR to enhance its compatibility with tensor operations and introduce a low-rank tensor structure for the finite element operators. Recognizing the banded structure inherent in the finite element framework’s discrete operators, we further exploit the QTT format of the CDR to achieve greater speed and compression. Additionally, we present a comprehensive approach for integrating variable coefficients of CDR into the global discrete operators within the TT/QTT framework. The effectiveness of the proposed method, in terms of memory efficiency and computational complexity, is demonstrated through a series of numerical experiments, including a semi-linear example.

convection–diffusion–reaction equation↗
DOE OSTI2025

A Fortran–Python interface for integrating machine learning parameterization into earth system models

Abstract. Parameterizations in earth system models (ESMs) are subject to biases and uncertainties arising from subjective empirical assumptions and incomplete understanding of the underlying physical processes. Recently, the growing representational capability of machine learning (ML) in solving complex problems has spawned immense interests in climate science applications. Specifically, ML-based parameterizations have been developed to represent convection, radiation, and microphysics processes in ESMs by learning from observations or high-resolution simulations, which have the potential to improve the accuracies and alleviate the uncertainties. Previous works have developed some surrogate models for these processes using ML. These surrogate models need to be coupled with the dynamical core of ESMs to investigate the effectiveness and their performance in a coupled system. In this study, we present a novel Fortran–Python interface designed to seamlessly integrate ML parameterizations into ESMs. This interface showcases high versatility by supporting popular ML frameworks like PyTorch, TensorFlow, and scikit-learn. We demonstrate the interface's modularity and reusability through two cases: an ML trigger function for convection parameterization and an ML wildfire model. We conduct a comprehensive evaluation of memory usage and computational overhead resulting from the integration of Python codes into the Fortran ESMs. By leveraging this flexible interface, ML parameterizations can be effectively developed, tested, and integrated into ESMs.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2025

A Fortran-Python Interface for Integrating Machine Learning Parameterization into Earth System Models

Parameterizations in Earth System Models (ESMs) are subject to biases and uncertainties arising from subjective empirical assumptions and incomplete understanding of the underlying physical processes. Recently, the growing representational capability of machine learning (ML) in solving complex problems has spawned immense interests in climate science applications. Specifically, ML-based parameterizations have been developed to represent convection, radiation and microphysics processes in ESMs by learning from observations or high-resolution simulations, which have the potential to improve the accuracies and alleviate the uncertainties. Previous works have developed some surrogate models for these processes using ML. These surrogate models need to be coupled with the dynamical core of ESMs to investigate the effectiveness and their performance in a coupled system. In this study, we present a novel Fortran-Python interface designed to seamlessly integrate ML parameterizations into ESMs. This interface showcases high versatility by supporting popular ML frameworks like PyTorch, TensorFlow, and Scikit-learn. We demonstrate the interface's modularity and reusability through two cases: a ML trigger function for convection parameterization and a ML wildfire model. We conduct a comprehensive evaluation of memory usage and computational overhead resulting from the integration of Python codes into the Fortran ESMs. By leveraging this flexible interface, ML parameterizations can be effectively developed, tested, and integrated into ESMs.

54 ENVIRONMENTAL SCIENCES↗
NASA NTRS1966

Thin-film ferrites vapor deposited by one-step process in vacuum

Thin-film ferrites are formed by vapor deposition of a mixture of powdered ferrites and powdered boron oxide at controlled temperatures in a vacuum chamber. These films are used in memory devices for computers and as thin-film inductors in communications and telemetry systems.

Hacskaylo, M.↗
NASA NTRS1970

Multiport semiconductor devices

Device, made of a variety of semiconductors, incorporates three or more terminals. Between at least two terminals, switching action occurs. The other terminal pair performs either another switching function or a control function. This device is useful for computer-logic or memory applications.

Holmstrom, F. R.↗
NASA NTRS1979

A vectorization of the Hess McDonnell Douglas potential flow program NUED for the STAR-100 computer

The computer program NUED for analyzing potential flow about arbitrary three dimensional lifting bodies using the panel method was modified to use vector operations and run on the STAR-100 computer. A high speed of computation and ability to approximate the body surface with a large number of panels are characteristics of NUEDV. The new program shows that vector operations can be readily implemented in programs of this type to increase the computational speed on the STAR-100 computer. The virtual memory architecture of the STAR-100 facilitates the use of large numbers of panels to approximate the body surface.

Boney, L. R.↗