Search NASA⌕ Search

SEARCH · Search NASA

Results for “COMPUTER ARCHITECTURE”

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 325 records · Page 18
All sourcesNASA NTRSDOE OSTI
DOE OSTI2025

Dynamic Low-Rank Training with Spectral Regularization: Achieving Robustness in Compressed Representations

Deployment of neural networks on resource-constrained devices demands models that are both compact and robust to adversarial inputs. However, compression and adversarial robustness often conflict. In this work, we introduce a dynamical low-rank training scheme enhanced with a novel spectral regularizer that controls the condition number of the low-rank core in each layer. This approach mitigates the sensitivity of compressed models to adversarial perturbations without sacrificing clean accuracy. The method is model- and data-agnostic, computationally efficient, and supports rank adaptivity to automatically compress the network at hand. Extensive experiments across standard architectures, datasets, and adversarial attacks show the regularized networks can achieve over 94\% compression while recovering or improving adversarial accuracy relative to uncompressed baselines.

Schotthoefer, Steffen [ORNL] (ORCID:00000002156965↗
DOE OSTI2025

Dynamical Low-Rank Compression of Neural Networks with Robustness under Adversarial Attacks

Deployment of neural networks on resource-constrained devices demands models that are both compact and robust to adversarial inputs. However, compression and adversarial robustness often conflict. In this work, we introduce a dynamical low-rank training scheme enhanced with a novel spectral regularizer that controls the condition number of the low-rank core in each layer. This approach mitigates the sensitivity of compressed models to adversarial perturbations without sacrificing clean accuracy. The method is model- and data-agnostic, computationally efficient, and supports rank adaptivity to automatically compress the network at hand. Extensive experiments across standard architectures, datasets, and adversarial attacks show the regularized networks can achieve over 94 compression while recovering or improving adversarial accuracy relative to uncompressed baselines.

Schotthoefer, Steffen [ORNL] (ORCID:00000002156965↗
DOE OSTI2025

Land-Based Wind Reference Architecture

This report describes a summary and the final deliverables for the Wind Reference Architecture project funded by the Department of Energy's (DOE) Wind Energy Technology Office (WETO). The project objective was to further refine the reference architecture of the existing wind power plant reference architecture developed by Idaho National Laboratory (INL) and Sandia National Laboratories (SNL) by expanding the wind turbine generator (WTG) into a separate individual wind turbine reference architecture. Additional details regarding the wind turbine system and how it integrates with a wind power plant were researched and reflected in the reference architecture. This addition is important because it allows researchers to understand the components and devices in a wind turbine, how they function and how a wind turbine integrates into a wind power plant to be able to perform cybersecurity evaluations on the system. The communication and control systems were the focus while developing this reference architecture to understand the cybersecurity posture of a wind turbine and power plant. The information technology (IT) and operational technology (OT) systems perspective are integral in helping improve the cybersecurity posture by creating a starting point to study how wind energy can be safely designed and deployed in our power grid from an integrated systems standpoint. A holistic wind power plant and turbine reference architecture and simulation was developed and made open-sourced for further research efforts.

17 WIND ENERGY↗
DOE OSTI2025

Summer Internship Report: ARA2 Benchmarking

Over the past decade, the RISC-V Instruction Set Architecture (ISA) has emerged as a significant player in both academic and industrial processor design due to its open-source nature, modular extension system, and versatility across domains ranging from microcontrollers to high-performance computing (HPC). One of its most important recent advancements is the RISC-V Vector Extension (RVV), which enables explicit data-level parallelism through vector registers and vectorized instructions. Unlike traditional SIMD (Single Instruction, Multiple Data) architectures that fix vector lengths at design time, RVV uses the concept of VLEN (vector register length) as a hardware-independent parameter and allows software to adapt dynamically to the available vector width. This flexible approach ensures portability across implementations while enabling scalable performance. The ARA2 core is a parameterizable RISC-V vector processor developed at the Integrated Systems Lab at ETH Zürich and the University of Bologna. Designed as a tightly-coupled accelerator to a scalar RISC-V core, ARA2 implements the RVV 1.0 specification and offers tunable architectural parameters such as the number of vector lanes, VLEN, and cache sizes.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2024

Neuromorphic overparameterisation and few-shot learning in multilayer physical neural networks

Abstract Physical neuromorphic computing, exploiting the complex dynamics of physical systems, has seen rapid advancements in sophistication and performance. Physical reservoir computing, a subset of neuromorphic computing, faces limitations due to its reliance on single systems. This constrains output dimensionality and dynamic range, limiting performance to a narrow range of tasks. Here, we engineer a suite of nanomagnetic array physical reservoirs and interconnect them in parallel and series to create a multilayer neural network architecture. The output of one reservoir is recorded, scaled and virtually fed as input to the next reservoir. This networked approach increases output dimensionality, internal dynamics and computational performance. We demonstrate that a physical neuromorphic system can achieve an overparameterised state, facilitating meta-learning on small training sets and yielding strong performance across a wide range of tasks. Our approach’s efficacy is further demonstrated through few-shot learning, where the system rapidly adapts to new tasks.

Science & Technology - Other Topics↗
DOE OSTI2025

Solving the Bernstein-Vazirani problem using Majorana-based topological quantum algorithms

Executing quantum algorithms using Majorana zero modes—a major milestone for the field of topological quantum computing—requires a platform that can be scaled to large quantum registers, can be controlled in real time and space, and a braiding protocol that uses the unique properties of these exotic particles. Here, we demonstrate the first successful simulation of a Majorana-based, fault-tolerant quantum algorithm to solve the Bernstein-Vazirani problem in two-dimensional magnet-superconductor hybrid structures from initialization to read-out of the final many-body state. Utilizing the Majorana zero modes’ topological properties, we introduce an optimized braiding protocol for the algorithm and a scalable architecture for its implementation with an arbitrary number of qubits. We visualize the algorithm protocol in real time and space by computing the non-equilibrium density of states, which is proportional to the time-dependent differential conductance, and the non-equilibrium charge density, which assigns a unique signature to each final state of the algorithm.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗
DOE OSTI2026

Knowledge-guided learning with curated prior genetic biomarkers for robust model interpretation

Abstract Motivation Knowledge-guided learning offers effective and robust model training strategies in data-scarce settings by incorporating established domain knowledge, thereby enhancing generalization, robustness, and interpretability. By contrast, conventional deep learning approaches rely purely on data-driven learning, which can limit robust model interpretability, particularly in high-dimensional settings with limited size samples. In computational biology, knowledge-guided learning has primarily leveraged network- and structural-based knowledge, leading to biologically interpretable representations and enhanced predictive performance compared to conventional approaches. However, curated biomarkers, one of the most accessible forms of biological knowledge, remain largely unexplored within knowledge-guided paradigms. Results In this study, we propose a model-agnostic training paradigm, Biomarker-driven Explainable Prior-guided Learning (BioExPL), that can be applied to any neural networks that incorporates curated prior knowledge. BioExPL enforces neural networks to reflect curated biomarker priors in their latent representations through a novel knowledge-alignment loss. BioExPL consistently demonstrated significantly improved predictive performance and enhanced model interpretability with minimized computational overhead in simulation studies and intensive experiments on multiple cancer datasets. BioExPL not only integrates prior curated knowledge into the model but also accurately identifies unknown associated signals additionally. BioExPL is model-agnostic and domain-independent, enabling its integration into diverse neural network architectures. Availability and implementation The open-source is publicly available at: https://github.com/datax-lab/BioExPL.

Baek, Beomsu [Department of Computer Science, Univ↗
DOE OSTI2025

Dependable classical-quantum computing systems engineering

Increasing evidence suggests quantum computing (QC) complements traditional High-Performance Computing (HPC) by leveraging its unique capabilities, leading to the emergence of a new, hybrid paradigm, QHPC. However, this integration introduces new challenges, with dependability–defined by reproducibility, resiliency, and security and privacy–emerging as a central concern for building trustworthy systems that provide an advantage to the users. This paper proposes a framework for dependable QHPC system design, organized around these three pillars. We identify integration challenges, anticipate roadblocks, and highlight productive synergies across QC, HPC, cloud platforms, and network security. Drawing from both classical computing principles and quantum-specific insights, we present a roadmap for co-design that supports robust hybrid architectures. Our approach offers concrete metrics for assessing dependability, provides design guidance for engineers working at the QC-HPC interface, and surfaces new engineering questions around complexity, scale, and fault tolerance. Ultimately, designing for dependability is key to realizing practical, scalable QHPC systems and accelerating the broader quantum ecosystem capable of translating quantum promises into actual application delivery.

HPC↗
DOE OSTI2024

Hybrid Storage Solution

With the rise of artificial intelligence and machine learning, data sets used to train models have become increasingly large. The availability, accessibility and integrity of large data sets has become important to the research conducted at Los Alamos National Laboratory. Ceph is a storage solution suitable for use with critical data because of its distributed nature and ability to keep multiple copies of a file in different locations. The amount of data means that bandwidth, latency, and cost are important factors and the reason most storage solutions are on-premises. However, there are distinct advantages to hosting services in the cloud, namely scalability and ease-of-use. In this paper, we explore the possibility of provisioning a hybrid Ceph cluster that leverages the benefits of both cloud architectures and on-premise performance.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2025

Performance Portability Evaluation of Fluid-Structure Interaction Simulations on Heterogeneous Platforms

The rapid proliferation of heterogeneous programming languages and multi-vendor hardware has underscored the critical need to evaluate the performance portability of scientific applications. In this work, we present the systematic porting and optimization of a massively parallel fluid-structure interaction code across multiple heterogeneous programming frameworks for deployment on leadership-class supercomputers from major vendors. Our analysis focuses on at-scale performance for simulations involving hundreds of millions of deformable cells, executed on a combination of CPUs and GPUs spanning thousands of nodes on exascale machines. We benchmark the performance of each implementation, highlighting the trade-offs inherent in adopting diverse programming models. Key insights regarding the portability of CUDA on multi-vendor platforms, the superior multi-core CPU performance from SYCL, and architectural considerations on performance optimization are distilled from our experience, offering guidance to other users of high performance computing based on our findings.

Martin, Aristotle [Duke University]↗
DOE OSTI2025

Spatially Aware Linear Transformer (SAL-T) for Particle Jet Tagging

Transformers are very effective in capturing both global and local correlations within high-energy particle collisions, but they present deployment challenges in high-data-throughput environments, such as the CERN LHC. The quadratic complexity of transformer models demands substantial resources and increases latency during inference. In order to address these issues, we introduce the Spatially Aware Linear Transformer (SAL-T), a physics-inspired enhancement of the linformer architecture that maintains linear attention. Our method incorporates spatially aware partitioning of particles based on kinematic features, thereby computing attention between regions of physical significance. Additionally, we employ convolutional layers to capture local correlations, informed by insights from jet physics. In addition to outperforming the standard linformer in jet classification tasks, SAL-T also achieves classification results comparable to full-attention transformers, while using considerably fewer resources with lower latency during inference. Experiments on a generic point cloud classification dataset (ModelNet10) further confirm this trend. Our code is available at https://github.com/aaronw5/SAL-T4HEP.

Wang, Aaron [Illinois U., Chicago] (ORCID:00000003↗
DOE OSTI2024

COSMIC DAWN: Distributed Analysis of Wireless at Nextscale

Distributed Analysis of Wireless at Nextscale (DAWN) is a novel simulation framework for large-scale design-space exploration (DSE) of unmodified software-defined radio (SDR) applications interacting in a scalable, high-fidelity, virtual physics environment. The software-defined nature of the coupled software-physics simulation leverages hardware emulation to permit in-depth examination and modification of not only the electromagnetic environment, including each signal in flight, but also the precise state of system software and components. DAWN supports modular, customizable physics environments allowing realistic propagation effects so that computationally efficient empirical models, reduced order/surrogate models, or large-scale, high-fidelity, site-specific simulations can be used as a propagation medium based on scenario requirements. This paper introduces DAWN’s design and initial implementation, detailing key architectural components, including the Physics Realization Engine (PhyRE), Runtime Infrastructure for Simulation Environments (RISE), and the design space exploration (DSE) suite. It concludes with demonstrations using unmodified 4G/LTE software available from srsRAN on computing resources ranging from a small cluster to ORNL’s Frontier Exascale system.

Wise, Mike [ORNL] (ORCID:0000000266120641)↗
DOE OSTI2026

Rasterization with Data-Parallel Primitives

Parallel rasterization can suffer from race conditions during fragment generation, which is traditionally addressed by using specialized hardware accessible via vendor graphics APIs. Unfortunately, graphics APIs are increasingly problematic on high-performance computers, either because they are not provided or because of concerns about dependencies with in situ visualization. In response, we present a hardware-agnostic rasterization algorithm that handles race conditions using only data-parallel primitives (DPPs), enabling efficient rendering on HPC systems without graphics API dependencies and aligning with recent efforts to deliver visualization software with DPPs. Our evaluation consists of three phases: (1) evaluating portability across different CPU and GPU architectures, (2) evaluating competitiveness with a community standard, and (3) evaluating performance across varying workloads and available parallelism. The supporting experiments run on both AMD and NVIDIA GPUs, considering data sets as large as 460 million triangles and 160 million pixels. While performance generally falls short of graphics API baselines, it achieves interactive frame rates on most workloads. As a result, we conclude our approach is a viable solution for rasterization on high-performance computers since our approach is portably performant across different architectures without the need for specialized vendor support.

Buckley, Makani [University of Oregon] (ORCID:0009↗
DOE OSTI2024

A fast and accurate domain decomposition nonlinear manifold reduced order model

Here, this paper integrates nonlinear-manifold reduced order models (NM-ROMs) with domain decomposition (DD). NM ROMs approximate the full order model (FOM) state in a nonlinear-manifold by training a shallow, sparse autoencoder using FOM snapshot data. These NM-ROMs can be advantageous over linear-subspace ROMs (LS-ROMs) for problems with slowly decaying Kolmogorov n-width. However, the number of NM-ROM parameters that need to be trained scales with the size of the FOM. Moreover, for “extreme-scale” problems, the storage of high-dimensional FOM snapshots alone can make ROM training expensive. To alleviate the training cost, this paper applies DD to the FOM, computes NM-ROMs on each subdomain, and couples them to obtain a global NM-ROM. This approach has several advantages: Subdomain NM-ROMs can be trained in parallel, involve fewer parameters to be trained than global NM-ROMs, require smaller subdomain FOM dimensional training data, and can be tailored to subdomain specific features of the FOM. The shallow, sparse architecture of the autoencoder used in each subdomain NM-ROM allows application of hyper-reduction (HR), reducing the complexity caused by nonlinearity and yielding computational speedup of the NM-ROM. This paper provides the first application of NM-ROM (with HR) to a DD problem. In particular, this paper details an algebraic DD reformulation of the FOM, training a NM-ROM with HR for each sub domain, and a sequential quadratic programming (SQP) solver to evaluate the coupled global NM-ROM. Theoretical convergence results for the SQP method and a priori and a posteriori error estimates for the DD NM-ROM with HR are provided. The proposed DD NM-ROM with HR approach is numerically compared to a DD LS-ROM with HR on the 2D steady-state Burgers’ equation, showing an order of magnitude improvement in accuracy of the proposed DD NM-ROM over the DD LS-ROM.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2026

An Integrated Framework for Memory-Centric Analysis: From Trace Collection to Co-Design

The memory wall phenomenon—where advances in processor performance significantly outpace those in memory subsystems—poses a fundamental challenge for contemporary computing systems. In memory-bound applications, memory subsystem behavior dominates performance, yet existing analysis approaches present significant limitations: detailed microarchitectural simulators require days to weeks to simulate modest workloads; hardware performance counters provide only aggregate statistics that obscure temporal and spatial access patterns; and scaled simulation approaches face challenges in capturing certain behaviors that emerge at larger scales. These limitations reflect a processor-centric design philosophy increasingly misaligned with memory-bound workloads where detailed understanding of memory access patterns, cache hierarchy interactions, and contention is critical for effective optimization. This paper presents an integrated framework for memory-centric analysis that enables effective hardware-software co-design. We describe practical trace collection techniques, including hardware-assisted processor tracing with minimal overhead and portable software-based instrumentation with statistical sampling. We present multi-perspective analysis methods that examine memory behavior from temporal, sequential, spatial, and relational viewpoints, revealing distinct optimization opportunities invisible in aggregate metrics. We detail an architectural modeling framework that uses sampled traces with temporal interpolation and confidence-based filtering to evaluate cache and memory configurations. Evaluation on representative benchmarks demonstrates that this framework achieves practical accuracy (L2 cache errors of 2.64\%, confidence-filtered L3 errors of 9.92\%, bandwidth errors of 7.33\%) while providing substantial speedup (26.8×) over cycle-accurate simulation, enabling rapid design space exploration. We demonstrate how this integrated framework enables systematic identification of both hardware optimizations (memory controller tuning, bank partitioning, NUMA configuration) and software optimizations (data layout restructuring, prefetching strategies, memory-aware scheduling). Through this comprehensive treatment of the memory-centric analysis pipeline—from trace collection through architectural modeling to co-design application—we provide researchers and practitioners with practical techniques for addressing memory bottlenecks in contemporary computing systems.

Gajaria, Dhruv Mayur↗
DOE OSTI2024

A Pseudoreversible Normalizing Flow for Stochastic Dynamical Systems with Various Initial Distributions

Here, we present a pseudoreversible normalizing flow method for efficiently generating samples of the state of a stochastic differential equation (SDE) with various initial distributions. The primary objective is to construct an accurate and efficient sampler that can be used as a surrogate model for computationally expensive numerical integration of SDEs, such as those employed in particle simulation. After training, the normalizing flow model can directly generate samples of the SDE’s final state without simulating trajectories. The existing normalizing flow model for SDEs depends on the initial distribution, meaning the model needs to be retrained when the initial distribution changes. The main novelty of our normalizing flow model is that it can learn the conditional distribution of the state, i.e., the distribution of the final state conditional on any initial state, such that the model only needs to be trained once and the trained model can be used to handle various initial distributions. This feature can provide a significant computational saving in studies of how the final state varies with the initial distribution. Additionally, we propose to use a pseudoreversible network architecture to define the normalizing flow model, which has sufficient expressive power and training efficiency for a variety of SDEs in science and engineering, e.g., in particle physics. We provide a rigorous convergence analysis of the pseudoreversible normalizing flow model to the target probability density function in the Kullback–Leibler divergence metric. Numerical experiments are provided to demonstrate the effectiveness of the proposed normalizing flow model.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2025

The design space of E(3)-equivariant atom-centred interatomic potentials

Abstract Molecular dynamics simulation is an important tool in computational materials science and chemistry, and in the past decade it has been revolutionized by machine learning. This rapid progress in machine learning interatomic potentials has produced a number of new architectures in just the past few years. Particularly notable among these are the atomic cluster expansion, which unified many of the earlier ideas around atom-density-based descriptors, and Neural Equivariant Interatomic Potentials (NequIP), a message-passing neural network with equivariant features that exhibited state-of-the-art accuracy at the time. Here we construct a mathematical framework that unifies these models: atomic cluster expansion is extended and recast as one layer of a multi-layer architecture, while the linearized version of NequIP is understood as a particular sparsification of a much larger polynomial model. Our framework also provides a practical tool for systematically probing different choices in this unified design space. An ablation study of NequIP, via a set of experiments looking at in- and out-of-domain accuracy and smooth extrapolation very far from the training data, sheds some light on which design choices are critical to achieving high accuracy. A much-simplified version of NequIP, which we call BOTnet (for body-ordered tensor network), has an interpretable architecture and maintains its accuracy on benchmark datasets.

Computer Science↗