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.

SourceResearch connectionFrom yearThrough year

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 703 records · Page 39

All sources NASA NTRS DOE OSTI

DOE OSTI2020

Scalable All-pairs Shortest Paths for Huge Graphs on Multi-GPU Clusters

We present an optimized Floyd-Warshall (Floyd-Warshall) algorithm that computes the All-pairs shortest path (APSP) for GPU accelerated clusters. The Floyd-Warshall algorithm due to its structural similarities to matrix-multiplication is well suited for highly parallel GPU architectures. To achieve high parallel efficiency, we address two key algorithmic challenges: reducing high communication overhead and addressing limited GPU memory. To reduce high communication costs, we redesign the parallel (a) to expose more parallelism, (b) aggressively overlap communication and computation with pipelined and asynchronous scheduling of operations, and (c) tailored MPI-collective. To cope with limited GPU memory, we employ an offload model, where the data resides on the host and is transferred to GPU on-demand. The proposed optimizations are supported with detailed performance models for tuning. Our optimized parallel Floyd-Warshall implementation is up to 5x faster than a strong baseline and achieves 8.1 PetaFLOPS/sec on 256~nodes of the Summit supercomputer at Oak Ridge National Laboratory. This performance represents 70% of the theoretical peak and 80% parallel efficiency. The offload algorithm can handle 2.5x larger graphs with a 20% increase in overall running time.

Sao, Piyush↗

DOE OSTI2025

Neuromorphic heat transport effects in a molecular junction

Understanding energy transport at the nanoscale is an open and fundamental challenge in the molecular sciences with direct implications for the design of new electronics, computing devices, and materials. While nanoscale energy transport under steady-state conditions has been studied extensively, there is much less known about energy transport under time-dependent driving forces, particularly in the far-from-equilibrium regime. In this work, we use nonequilibrium molecular dynamics simulations and stochastic thermodynamics to investigate energy transport in a well-studied nanoscale system—a molecular junction—subjected to a time-periodic temperature gradient. The primary observation is that molecular junctions can exhibit heat transport hysteresis, a phenomenon in which the heat flux through a system depends not only on the instantaneous value of a time-dependent temperature bias but also on the temporal history of that bias. The presented findings illustrate that molecular junctions can exhibit the specific memory effect—heat transport hysteresis—that is essential for the design of thermal neuromorphic computers. This work elucidates a potential pathway toward the realization of such devices.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

DOE OSTI2021

NeuralCubes: Deep Representations for Visual Data Exploration

Visual exploration of large multi-dimensional datasets has seen tremendous progress in recent years, allowing users to express rich data queries that produce informative visual summaries, all in real time. Techniques based on data cubes are some of the most promising approaches. However, these techniques usually require a large memory footprint for large datasets. To tackle this problem, we present NeuralCubes: neural networks that predict results for aggregate queries, similar to data cubes. NeuralCubes learns a function that takes as input a given query, for instance, a geographic region and temporal interval, and outputs the result of the query. The learned function serves as a real-time, low-memory approximator for aggregation queries. Our models are small enough to be sent to the client side (e.g. the web browser for a web-based application) for evaluation, enabling data exploration of large datasets without database/network connection. Here, we demonstrate the effectiveness of NeuralCubes through extensive experiments on a variety of datasets and discuss how NeuralCubes opens up opportunities for new types of visualization and interaction.

97 MATHEMATICS AND COMPUTING↗

DOE OSTI2024

Advancing attenuation estimation through integration of the Hessian in multiparameter viscoacoustic full-waveform inversion

Accurate seismic attenuation models of subsurface structures not only enhance subsequent migration processes by improving fidelity, resolution, and facilitating amplitude-compliant angle gather generation but also provide valuable constraints on subsurface physical properties. Leveraging full-wavefield information, multiparameter viscoacoustic full-waveform inversion ( Q-FWI) simultaneously estimates seismic velocity and attenuation ( Q) models. However, a major challenge in Q-FWI is the contamination of crosstalk artifacts, where inaccuracies in the velocity model are mistakenly mapped to the inverted attenuation model. While incorporating the Hessian is expected to mitigate these artifacts, the explicit implementation is prohibitively expensive due to its formidable computational cost. In this study, we formulate and develop a Q-FWI algorithm via the Newton-conjugate gradient (CG) framework, where the search direction at each iteration is determined through an internal CG loop. In particular, the Hessian is integrated into each CG step in a matrix-free fashion using the second-order adjoint-state method. We find through synthetic experiments that our Newton-CG Q-FWI significantly mitigates crosstalk artifacts compared with the limited-memory Broyden-Fletcher-Goldfarb-Shanno method and the CG method, albeit with a notable computational cost. In the discussion of several key implementation details, we also determine the significance of the approximate Gauss-Newton Hessian, the second-order adjoint-state method, and the two-stage inversion strategy.

Geochemistry & Geophysics↗

DOE OSTI2024

AI-Enhanced Co-Design for Next-Generation Microelectronics: Innovating Innovation (Workshop Report)

The Artificial Intelligence Enhanced Co-Design for Next Generation Microelectronics virtual workshop was held April 4-5, 2023, and attended by subject matter experts from universities, industry, and national laboratories. This was the third in a series of workshops to motivate the research community to identify and address major challenges facing microelectronics research and production. The 2023 workshop focused on a set of topics from materials to computing algorithms, and included discussions on relevant federal legislation and such as the Creating Helpful Incentives to Produce Semiconductors and Science Act (CHIPS Act) which was signed into law in the summer of 2022. Talks at the workshop included edge computing in radiation environments, new materials for neuromorphic computing, advanced packaging for microelectronics, and new AI techniques. We also received project updates from several of the Department of Energy (DOE) microelectronics co-design projects funded in the fall of 2021, and from three of the Energy Frontier Research Centers (EFRCs) that had been funded in the fall of 2022. The workshop also conducted a set of breakout discussions around the five principal research directions (PRDs) from the 2018 Department of Energy workshop report: 1) define innovative material, device, and architecture requirements driven by applications, algorithms, and software; 2) revolutionize memory and data storage; 3) re-imagine information flow unconstrained by interconnects; 4) redefine computing by leveraging unexploited physical phenomena; 5) reinvent the electricity grid through new materials, devices, and architectures. We tasked each breakout group to consider one primary PRD (and other PRDs as relevant topics arose during discussions) and to address questions such as whether the research community has embraced co-design as a methodology and whether new developments at any level of innovation from materials to programming models requires the research community to reevaluate the PRDs developed back in 2018.

97 MATHEMATICS AND COMPUTING↗

DOE OSTI2022

Reducing Memory Consumption in Calico with Shared Memory

This document details the work to reduce memory consumption in Calico. Calico is SimTools’ Constructive Solid Geometry (CSG) and geometry painting library. It is primarily used to paint material volume fractions in the Eulerian meshes of the physics codes. Calico provides point-in-body checks for the geometry supplied by an Oso model, which are then aggregated by the host codes. In addition, Calico can be used to build Oso models and is used by Ingen for that purpose. Oso models, and thus Calico, provide support for various CSG primitives such as spheres, cylinders, surfaces generated by rotating tabular curve data, and STL files as well as binary combinations of those primitives. Prior to refactoring Calico will run out of memory on CTS-1 machines when 36 MPI ranks are used per node when reading STL models on the order of 1.5 GB. This limitation is a bottleneck in designer workflow. This problem has been alleviated through the use of data structures to both reduce memory consumption and to leverage MPI-3 shared memory. This report details the data structures targeted for refactoring in Calico, the methods and implementation details for reducing memory consumption and leveraging shared memory, and results for one test problem. Results show a memory reduction when loading a 1 GB STL file by a factor of 27.5, from 93.4 to 3.4 GB.

97 MATHEMATICS AND COMPUTING↗

DOE OSTI2026

GPU-Accelerated Analytic Simulation of Sparse Ionization Signal Formation in Pixelated Projection Detector

This paper presents a GPU-accelerated simulation package, TRED, for next-generation neutrino detectors with pixelated charge readout, leveraging community-driven software ecosystems to ensure adaptability and extensibility. We introduce two generic contributions: (i) an effective-charge representation based on Gaussian quadrature rules, in which the linear- interpolation factors for the field response inside each voxel are absorbed into the effective charge, and (ii) a sparse, block- binned tensor representation that enables efficient FFT-based computation of induced signals on readout electrodes for sparsely activated detector volumes. The former captures structure inside a voxel without dense sampling, while the latter achieves low memory usage and scalable runtime, as demonstrated in bench- mark studies. The underlying data representation is applicable to large-scale detectors and to other computational problems involving sparse activity.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

DOE OSTI2025

Enabling Multireference Calculations on Multimetallic Systems with Graphic Processing Units

Modeling multimetallic systems efficiently enables faster prediction of desirable chemical properties and the design of new materials. This work describes an initial implementation for performing multireference wave function method localized active-space self-consistent field (LASSCF) calculations through the use of multiple graphics processing units (GPUs) to accelerate time-to-solution. Density fitting is leveraged to reduce memory requirements, and we demonstrate the ability to fully utilize multi-GPU compute nodes. Performance improvements of 5–10x in total application runtime were observed in LASSCF calculations for multimetallic catalyst systems up to 1200 AOs and an active space of (22e,40o) using up to four NVIDIA A100 GPUs. Furthermore, written with performance portability in mind, a comparable performance is also observed in early runs on the Aurora exascale system using Intel Max Series GPUs.

Algorithms↗

DOE OSTI2024

Ab initio generalized Langevin equation

We introduce a machine learning–based approach called ab initio generalized Langevin equation (AIGLE) to model the dynamics of slow collective variables (CVs) in materials and molecules. In this scheme, the parameters are learned from atomistic simulations based on ab initio quantum mechanical models. Force field, memory kernel, and noise generator are constructed in the context of the Mori–Zwanzig formalism, under the constraint of the fluctuation–dissipation theorem. Combined with deep potential molecular dynamics and electronic density functional theory, this approach opens the way to multiscale modeling in a variety of situations. Here, we demonstrate this capability with a study of two mesoscale processes in crystalline lead titanate, namely the field-driven dynamics of a planar ferroelectric domain wall, and the dynamics of an extensive lattice of coarse-grained electric dipoles. In the first case, AIGLE extends the reach of ab initio simulations to a regime of noise-driven motions not accessible to molecular dynamics. In the second case, AIGLE deals with an extensive set of CVs by adopting a local approximation for the memory kernel and retaining only short-range noise correlations. The scheme is computationally more efficient than molecular dynamics by several orders of magnitude and mimics the microscopic dynamics at low frequencies where it reproduces accurately the dominant far-infrared absorption frequency.

97 MATHEMATICS AND COMPUTING↗

DOE OSTI2021

eQuilibrator 3.0: a database solution for thermodynamic constant estimation

Abstract eQuilibrator (equilibrator.weizmann.ac.il) is a database of biochemical equilibrium constants and Gibbs free energies, originally designed as a web-based interface. While the website now counts around 1,000 distinct monthly users, its design could not accommodate larger compound databases and it lacked a scalable Application Programming Interface (API) for integration into other tools developed by the systems biology community. Here, we report on the recent updates to the database as well as the addition of a new Python-based interface to eQuilibrator that adds many new features such as a 100-fold larger compound database, the ability to add novel compounds, improvements in speed and memory use, and correction for Mg2+ ion concentrations. Moreover, the new interface can compute the covariance matrix of the uncertainty between estimates, for which we show the advantages and describe the application in metabolic modelling. We foresee that these improvements will make thermodynamic modelling more accessible and facilitate the integration of eQuilibrator into other software platforms.

59 BASIC BIOLOGICAL SCIENCES↗

DOE OSTI2021

Co-design of Advanced Architectures for Graph Analytics using Machine Learning

A graph is an excellent way of representing relationships among entities. We can use graph analytics to synthesize and analyze such relational data, and extract relevant features that are useful for various tasks such as machine learning. Considering the crucial role of graph analytics in various domains, it is important and timely to investigate the right hardware configurations that can achieve optimal performance for graph workloads on future high-performance computing systems. Design space exploration studies facilitate the selection of appropriate configurations (e.g. memory) to achieve a desired system performance. Recently, the approach of accelerating graph analytics using persistent non-volatile memory has gained a lot of attention. Traditional system simulators such as Gem5 and NVMain can be used to explore the design space of these advanced memory architectures for graph workloads. However, these simulators are slow in execution thus limiting the efficiency of design space exploration studies. To overcome this challenge, we proposed a machine learning based approach to co-design advanced memory architectures for graph workloads. We tested our approach with DRAM, non-volatile memory, and hybrid memory (DRAM+NVM) using a breadth first search benchmark algorithm. Our results showed the applicability of the proposed machine learning based approach to the co-design of the advanced memory architectures. In this paper, we provide recommendations on selecting advanced memory architectures to achieve desired performance for graph workloads. We also discuss the performances of different machine learning models that were considered in this study.

Kurte, Kuldeep↗

DOE OSTI2021

Accelerating Binarized Neural Networks via Bit-Tensor-Cores in Turing GPUs

Despite foreseeing tremendous speedups over conventional deep neural networks, the performance advantage of binarized neural networks (BNNs) has merely been showcased on general-purpose processors such as CPUs and GPUs. In fact, due to being unable to leverage bit-level-parallelism with a word-based architecture, GPUs have been criticized for extremely low utilization (1%) when executing BNNs. Consequently, the latest tensorcores in NVIDIA Turing GPUs start to experimentally support bit computation. In this work, we look into this brand new bit computation capability and characterize its unique features. We show that the stride of memory access can significantly affect performance delivery and a data-format co-design is highly desired to support the tensorcores for achieving superior performance than existing software solutions without tensorcores. We realize the tensorcore-accelerated BNN design, particularly the major functions for fully-connect and convolution layers — bit matrix multiplication and bit convolution. Evaluations on two NVIDIA Turing GPUs show that, with ResNet-18, our BTC-BNN design can process ImageNet at a rate of 5.6K images per second, 77% faster than state-of-the-art. Our BNN approach is released on https://github.com/pnnl/TCBNN.

Li, Ang↗

DOE OSTI2024

ZFP: A compressed array representation for numerical computations

HPC trends favor algorithms and implementations that reduce data motion relative to FLOPS. We investigate the use of lossy compressed data arrays in place of traditional IEEE floating point arrays to store the primary data of calculations. Simulation is fundamentally an exercise in controlled approximation, and error introduced by finite-precision arithmetic (or lossy compression) is just one of several sources of error that need to be managed to ensure sufficient accuracy in a computed result. We describe ZFP, a compressed numerical format designed for in-memory storage of multidimensional arrays, and summarize theoretical results that demonstrate that the error of repeated lossy compression can be bounded and controlled. Furthermore, we establish a relationship between grid resolution and compression-induced errors and show that, contrary to conventional floating point, ZFP reduces finite-difference errors with finer grids. We present example calculations that demonstrate data reduction by 4x or more with negligible impact on solution accuracy. Our results further demonstrate several orders-of-magnitude increase in accuracy using ZFP over IEEE floating point and Posits for the same storage budget.

Lindstrom, Peter↗

DOE OSTI2023

Real Space and Time Imaging of Collective Headgroup Dipole Motions in Zwitterionic Lipid Bilayers

Lipid bilayers are supramolecular structures responsible for a range of processes, such as transmembrane transport of ions and solutes, and sorting and replication of genetic materials, to name just a few. Some of these processes are transient and currently, cannot be visualized in real space and time. Here, we developed an approach using 1D, 2D, and 3D Van Hove correlation functions to image collective headgroup dipole motions in zwitterionic phospholipid bilayers. We show that both 2D and 3D spatiotemporal images of headgroup dipoles are consistent with commonly understood dynamic features of fluids. However, analysis of the 1D Van Hove function reveals lateral transient and re-emergent collective dynamics of the headgroup dipoles—occurring at picosecond time scales—that transmit and dissipate heat at longer times, due to relaxation processes. At the same time, the headgroup dipoles also generate membrane surface undulations due a collective tilting of the headgroup dipoles. A continuous intensity band of headgroup dipole spatiotemporal correlations—at nanometer length and nanosecond time scales—indicates that dipoles undergo stretching and squeezing elastic deformations. Importantly, the above mentioned intrinsic headgroup dipole motions can be externally stimulated at GHz-frequency scale, enhancing their flexoelectric and piezoelectric capabilities (i.e., increased conversion efficiency of mechanical energy into electric energy). In conclusion, we discuss how lipid membranes can provide molecular-level insights about biological learning and memory, and as platforms for the development of the next generation of neuromorphic computers.

59 BASIC BIOLOGICAL SCIENCES↗

DOE OSTI2020

Device for controlling additive manufacturing machinery

A computing device for controlling the operation of an additive manufacturing machine comprises a memory element and a processing element. The memory element is configured to store a three-dimensional model of a part to be manufactured, wherein the three-dimensional model defines a plurality of cross sections of the part. The processing element is in communication with the memory element. The processing element is configured to receive the three-dimensional model, determine a plurality of paths, each path including a plurality of parallel lines, determine a radiation beam power for each line, such that the radiation beam power varies non-linearly according to a length of the line, and determine a radiation beam scan speed for each line, such that the radiation beam scan speed is a function of a temperature of a material used to manufacture the part, the length of the line, and the radiation beam power for the line.

Barr, Christian↗

DOE OSTI2020

Device for controlling additive manufacturing machinery

A computing device for controlling the operation of an additive manufacturing machine comprises a memory element and a processing element. The memory element is configured to store a three-dimensional model of a part to be manufactured, wherein the three-dimensional model defines a plurality of cross sections of the part. The processing element is in communication with the memory element. The processing element is configured to receive the three-dimensional model, determine a path across a surface of each cross section, wherein the path includes a plurality of parallel lines, calculate a power for a radiation beam to scan each of the lines, such that the power varies from line to line non-linearly according to a length of the line, and calculate a scan speed for the radiation beam for each of the lines, such that the scan speed varies line to line non-linearly according to the power of the radiation beam.

Barr, Christian↗

DOE OSTI2020

Device for controlling additive manufacturing machinery

Barr, Christian↗

DOE OSTI2023

Systems and methods for enhanced power system model validation

A system for enhanced power system model validation is provided. The system includes a computing device including at least one processor in communication with at least one memory device. The at least one processor is programmed to store a plurality of models for a plurality of devices and a plurality of input files associated with the plurality of models, receive, from a user, a selection of model of the plurality of models to simulate, retrieve one or more input files of the plurality of input files, perform a model validity check on the selected model, if the selected model passed the model validity check, perform a model calibration on the selected model, and if the selected model passed the model calibration, perform a post evaluation on the selected model.

Wang, Honggang↗