Search NASA⌕ Search

SEARCH · Search NASA

Results for “Kernel Designs”

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 55 records · Page 3
All sourcesNASA NTRSDOE OSTI
DOE OSTI2021

Fast and Scalable Sparse Triangular Solver for Multi-GPU Based HPC Architectures

Designing efficient and scalable sparse linear algebra kernels on modern multi-GPU based HPC systems is a daunting task due to significant irregular memory references and workload imbalance across the GPUs. This is particularly the case for \textit{Sparse Triangular Solver (SpTRSV)} which introduces additional two-dimensional computation dependencies among subsequent computation steps. Dependency information is exchanged and shared among GPUs, thus warrant for efficient memory allocation, data partitioning, and workload distribution as well as fine-grained communication and synchronization support. In this work, we demonstrate that directly adopting unified memory can adversely affect the performance of SpTRSV on multi-GPU architectures, despite linking via fast interconnect like NVLinks and NVSwitches. Alternatively, we employ the latest NVSHMEM technology based on Partitioned Global Address Space programming model to enable efficient fine-grained communication and drastic synchronization overhead reduction. Furthermore, to handle workload imbalance, we propose a malleable task-pool execution model which can further enhance the utilization of GPUs. By applying these techniques, our experiments on the NVIDIA multi-GPU supernode V100-DGX-1 and DGX-2 systems demonstrate that our design can achieve on average 3.53x (up to 9.86x) speedup on a DGX-1 system and 3.66x (up to 9.64x) speedup on a DGX-2 system with 4-GPUs over the Unified-Memory design. The comprehensive sensitivity and scalability studies also show that the proposed zero-copy SpTRSV is able to fully utilize the computing and communication resources of the multi-GPU system.

Xie, Chenhao↗
DOE OSTI2021

MFDn Kernels v1.0

MFDn kernels is a collection of small subroutines designed to capture essential characteristics of the Many-Fermion Dynamics---nuclear (MFDn) application (DOI: 10.1016/j.procs.2010.04.012). It features multiple implementations with different programming models/languages of simplified routines analogous to those found in the full application. The main use is to explore the performance and utility of different programming models, languages and compilers.

Cook, Brandon↗
DOE OSTI2023

Multi-scale fission product release model with comparison to AGR data

TRistructural ISOtropic (TRISO) particle fuel is central to several advanced, high-temperature reactor designs. Each particle consists of a fuel kernel encapsulated by three layers of carbon and ceramics that prevent the release of fission products and ensure physical integrity. Despite outstanding retention properties, fission product release has been observed from intact particles. To better understand and quantify fission product release from TRISO particles, a multiscale, mechanistic model of fission product transport is being developed by the Nuclear Energy Advanced Modeling and Simulation (NEAMS) program. Previous work focused on silver (Ag) transport and improved Ag release predictions. The work described in this report builds on this experience to better understand cesium (Cs) transport in silicon carbide (SiC), the main barrier to the release of fission products. Atomistic simulations provide bulk and grain boundary (GB) Cs diffusivities in SiC, which are used by phase field simulations in the mesoscale code Marmot to determine the temperature, microstructure, and irradiation-dependent Cs diffusivity at the mesoscale in SiC. This approach attributes the different temperature regimes experimentally observed for Cs diffusivities in SiC to a transition from bulk-dominated diffusivity at high temperatures to a GB-dominated regime at low temperatures, providing new insight. The multiscale, mechanistic effective diffusivity is then implemented in the fuel performance code BISON and further validated by comparing Cs release predictions from Advanced Gas Reactor (AGR)-1 and AGR-2 post-irradiation measurements. The new model improves BISON’s predictability. This document also reports improvements made on Ag transport modeling by accounting for different GB types having different diffusivities. Moreover, this report details preliminary efforts to model palladium (Pd) attack of the SiC at the mesoscale using a phase field approach. Pd attack and its impact on accelerated Ag transport remains a misunderstood phenomenon, and we use the model to demonstrate that the formation of lamellae that has been observed in experiments can be explained by the reaction of Pd with SiC to form alternating layers of graphite and Pd 2 Si. This effort aims to improve our understanding of the reaction and eventually provide a model for BISON to account for Pd penetration and its effects on fission product release.

11 NUCLEAR FUEL CYCLE AND FUEL MATERIALS↗
DOE OSTI2024

Evaluation of Radiography for TRISO Buffer Layer Density Measurement

Tristructural isotropic (TRISO) fuel particles consist of a central uranium-bearing kernel and a series of coating layers designed to retain fission products and to ensure fuel performance. Several parameters such as thickness and density must be measured for these coating layers to show that they conform with fuel specifications. Current methods for measuring the density of pyrolytic carbon and silicon carbide layers (liquid gradient density column) and the buffer layer (mercury porosimetry) generate Resource Conservation and Recovery Act (RCRA) radiological-mixed waste. In addition, measurement of buffer and inner pyrolytic carbon layer densities require hot sampling or interrupted coating runs and the mercury porosimetry method used for buffer density measurement only measures the mean buffer density, not the interparticle distribution. A new approach has been evaluated to measure the density of coating layers in TRISO particles based on the dependence of x-ray attenuation in radiographs on material density. This method does not generate RCRA mixed waste, measures density on a particle-by-particle basis, and in principle is capable of measuring the density of all coating layers in a single process. Initial results using thinned TRISO particle sections to evaluate radiography measurement of density as a quality control characterization method are reported herein. In this work, the primary focus is on measurement of the density of the buffer layer; however, with appropriate calibration the method should be applicable to other coating layers. Improvements to the initial method and a full demonstration of the method on the remaining coating layers may be pursued as a future effort.

11 NUCLEAR FUEL CYCLE AND FUEL MATERIALS↗
DOE OSTI2024

Post-irradiation Heating Tests of As-Irradiated AGR-3/4 TRISO Fuel Compacts

Four post-irradiation heating tests of fuel compacts from the U.S. Advanced Gas Reactor (AGR)-3/4 irradiation experiment were completed. In addition to tristructural isotropic (TRISO)-coated driver fuel, each compact contained designed-to-fail (DTF) particles with fuel kernels coated only in pyrocarbon so as to simulate exposed kernels. Tests at 1600/1700°C, 1400°C, and 1200°C were performed to measure fission product release as a function of time and temperature. Silver release was highest in the 1200°C test, supporting the observation that silver release rates are highest in the 1100–1300°C range. Compared to tests of AGR-1 compacts with no exposed kernels, the Cs-134 and Kr-85 releases were noticeably higher in AGR-3/4. The exposed kernels’ contributions to Eu and Sr release are inconclusive, due to the difficulty in distinguishing among the combined effects of higher irradiation temperatures in these particular AGR-3/4 compacts, the presence of the DTF particles, and the Fuel Accident Condition Simulator (FACS) test temperatures. These data can be used to make inferences about fission product retention in exposed kernels as a function of time and temperature.

11 NUCLEAR FUEL CYCLE AND FUEL MATERIALS↗
DOE OSTI2023

Physical Properties of Moist, Fermented Corn Kernels

A novel approach to producing corn stover biomass feedstock has been investigated. In this approach, corn grain and stover are co-harvested at moisture contents much less than typical corn silage. The grain and stover are conserved together by anaerobic storage and fermentation and then separated before end use. When separated from the stover, the moist, fermented grain had physical characteristics that differ from typical low-moisture, unfermented grain. A comprehensive study was conducted to quantify the physical properties of this moist, fermented grain. Six corn kernel treatments, either fermented or unfermented, having different moisture contents, were used. Moist, fermented kernels (26 and 36% w.b. moisture content) increased in size during storage. The fermented kernels’ widths and thicknesses were 10% and 15% greater, respectively, and their volume was 28% greater than the dry kernels (15% w.b.). Dry basis particle density was 9% less for moist, fermented kernels. Additionally, the dry basis bulk density was 29% less, and the dry basis hopper-discharged mass flow rate was 36% less. Moist, fermented grain had significantly greater kernel-to-kernel coefficients of friction and angles of repose compared to relatively dry grain. The friction coefficient on four different surfaces was also significantly greater for fermented kernels. Fermented corn kernels had lower individual kernel rupture strengths than unfermented kernels. These physical differences must be considered when designing material handling and processing systems for moist, fermented corn grain.

09 BIOMASS FUELS↗
DOE OSTI2021

Achieving performance portability in Gaussian basis set density functional theory on accelerator based architectures in NWChemEx

The numerical integration of the exchange–correlation (XC) potential is one of the primary computational bottlenecks in Gaussian basis set Kohn–Sham density functional theory (KS-DFT). To achieve optimal performance and accuracy, care must be taken in this numerical integration to preserve local sparsity as to allow for near linear weak scaling with system size. This leads to an integration scheme with several performance critical kernels which must be hand optimized for each architecture of interest. As the set of available accelerator hardware goes more diverse, a key challenge for developers of KS-DFT software is to maintain performance portability across a wide range of computational architectures. In this article, we examine a modular software design pattern which decouples the implementation details of performance critical kernels from the expression of high-level algorithmic workflows in a device-agnostic language such as C++; thus allowing for developers to target existing and emerging accelerator hardware within a single code base. We consider the efficacy of such a design pattern in the numerical integration of the XC potential by demonstrating its ability to achieve performance portability across a set of accelerator architectures which are representative of those on current and future U.S. Department of Energy Leadership Computing Facilities.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2022

Representation Learning via Quantum Neural Tangent Kernels

Variational quantum circuits are used in quantum machine learning and variational quantum simulation tasks. Designing good variational circuits or predicting how well they perform for given learning or optimization tasks is still unclear. Here we discuss these problems, analyzing variational quantum circuits using the theory of neural tangent kernels. We define quantum neural tangent kernels, and derive dynamical equations for their associated loss function in optimization and learning tasks. We analytically solve the dynamics in the frozen limit, or lazy training regime, where variational angles change slowly and a linear perturbation is good enough. We extend the analysis to a dynamical setting, including quadratic corrections in the variational angles. We then consider a hybrid quantum classical architecture and define a large-width limit for hybrid kernels, showing that a hybrid quantum classical neural network can be approximately Gaussian. The results presented here show limits for which analytical understandings of the training dynamics for variational quantum circuits, used for quantum machine learning and optimization problems, are possible. These analytical results are supported by numerical simulations of quantum machine-learning experiments.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
DOE OSTI2022

Non-Blind Deblurring for Fluorescence: A Deformable Latent Space Approach with Kernel Parameterization

We report N\non-blind deblurring (NBD) is a modeling method of the image deblurring problem in computer vision, where the blurring kernel is known or can be externally estimated. In this paper, we attempt to solve a parametric NBD problem, inspired by the simultaneous acquisition of ptychography and fluorescent imaging (FI). Ptychography is an imaging method that favors larger probes, i.e. convolutional kernels, while FI relies on a small probe for high resolution. Also, the kernel can be solved during ptychographic reconstruction. With Ptycho-FI using the same larger kernel, we can perform NBD on the blurred fluorescent images to achieve high-resolution FI, and thus speed up the experiments. To this end, we design a deep latent space deformation network that is directly parameterized by the kernel. The network consists of three components: encoder, deformer, and decoder, where the deformer is specifically meant to rectify the latent space representations of blurred images to a standard latent space, regardless of the kernel. The deformation network is trained with a two-stage training scheme. We conduct extensive experiments to confirm that our parametric model can adapt to drastically different blurring kernels and perform robust deblurring.

77 NANOSCIENCE AND NANOTECHNOLOGY↗
DOE OSTI2022

Motion tomography via occupation kernels

The goal of motion tomography is to recover a description of a vector flow field using measurements along the trajectory of a sensing unit. In this paper, we develop a predictor corrector algorithm designed to recover vector flow fields from trajectory data with the use of occupation kernels developed by Rosenfeld et al. [9,10]. Specifically, we use the occupation kernels as an adaptive basis; that is, the trajectories defining our occupation kernels are iteratively updated to improve the estimation in the next stage. Initial estimates are established, then under mild assumptions, such as relatively straight trajectories, convergence is proven using the Contraction Mapping Theorem. We then compare the developed method with the established method by Chang et al. [5] by defining a set of error metrics. Here, we found that for simulated data, where a ground truth is available, our method offers a marked improvement over [5]. For a real-world example, where ground truth is not available, our results are similar results to the established method.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2026

Intelligent Experiments through Real-Time AI: Fast Data Processing and Autonomous Detector Control for High-Energy Nuclear Experiments

The aim of this project is to develop software and hardware for fast real-time data processing and autonomous detector control and calibration for the sPHENIX and the future EIC experiments. Below summarizes Georgia Tech team efforts in the past year: 1. We developed a real-time clustering algorithm and FPGA-based pipeline architecture for processing fired pixel data from ALPIDE sensors in sPHENIX experiments. Our Columnar Clustering Co-Design introduces a hardware-aware, stream-friendly approach that segments pixel data by column pairs using a Column Pair Clustering (CPC) strategy, followed by Cluster Stitching to merge adjacent subclusters. Implemented in Vitis HLS, the pipeline comprises five stages—read-in, subclustering, stitching, analysis, and write-out—connected by tagged HLS streams with custom end-of-event signaling for robust synchronization. We designed a pipelined dataflow model optimized for throughput, low latency, and minimal buffering, enabling scalable clustering across events of arbitrary size. Our system maintains spatial precision via center-of-mass and shape key extraction and efficiently handles edge cases such as fragmented or nested clusters. Compared against DBSCAN in both software and hardware, our approach demonstrates competitive performance under FPGA constraints. 2. We also conducted a comprehensive algorithm-to-hardware co-design of connected component analysis tailored for sPHENIX experiments, focusing on real-time, low-latency processing using FPGAs and High-Level Synthesis (HLS). Starting from a Python-based particle tracking pipeline, the team translated the core logic—graph traversal via DFS and Union-Find—into an HLS-compatible C++ model, replacing dynamic memory and recursion with static arrays and pipelined control flow. The final design includes a fully streamed and dataflow-compatible Union-Find kernel optimized across five iterations, incorporating loop pipelining, array partitioning, AXI/FIFO interface tuning, and function flattening. Experimental results show up to 14.8× speedup over the CPU baseline, reducing per-graph latency to 1.58 μs and demonstrating strong resource efficiency with only ~7k LUTs and zero BRAM usage. The design maintains functional correctness against the Python reference using a Python-based C-simulation framework and Mean Squared Error metrics. This work validates the potential of HLS-driven FPGA designs for edge-level HEP data acquisition, laying a scalable foundation for future integration with real-time detector pipelines and multi-graph processing systems.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗
DOE OSTI2021

Physics-informed machine learning

Despite great progress in simulating multiphysics problems using the numerical discretization of partial differential equations (PDEs), one still cannot seamlessly incorporate noisy data into existing algorithms, mesh generation remains complex, and high-dimensional problems governed by parameterized PDEs cannot be tackled. Moreover, solving inverse problems with hidden physics is often prohibitively expensive and requires different formulations and elaborate computer codes. Machine learning has emerged as a promising alternative, but training deep neural networks requires big data, not always available for scientific problems. Instead, such networks can be trained from additional information obtained by enforcing the physical laws (for example, at random points in the continuous space-time domain). Such physics-informed learning integrates (noisy) data and mathematical models, and implements them through neural networks or other kernel-based regression networks. Moreover, it may be possible to design specialized network architectures that automatically satisfy some of the physical invariants for better accuracy, faster training and improved generalization. Furthermore, we review some of the prevailing trends in embedding physics into machine learning, present some of the current capabilities and limitations and discuss diverse applications of physics-informed learning both for forward and inverse problems, including discovering hidden physics and tackling high-dimensional problems.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2025

ExtremeMETA: High-speed Lightweight Image Segmentation Model by Remodeling Multi-channel Metamaterial Imagers

Deep neural networks (DNNs) have heavily relied on traditional computational units, such as CPUs and GPUs. However, this conventional approach brings significant computational burden, latency issues, and high power consumption, limiting their effectiveness. This has sparked the need for lightweight networks such as ExtremeC3Net. Meanwhile, there have been notable advancements in optical computational units, particularly with metamaterials, offering the exciting prospect of energy-efficient neural networks operating at the speed of light. Yet, the digital design of metamaterial neural networks (MNNs) faces precision, noise, and bandwidth challenges, limiting their application to intuitive tasks and low-resolution images. In this study, we proposed a large kernel lightweight segmentation model, ExtremeMETA. Based on ExtremeC3Net, our proposed model, ExtremeMETA maximized the ability of the first convolution layer by exploring a larger convolution kernel and multiple processing paths. With the large kernel convolution model, we extended the optic neural network application boundary to the segmentation task. To further lighten the computation burden of the digital processing part, a set of model compression methods was applied to improve model efficiency in the inference stage. The experimental results on three publicly available datasets demonstrated that the optimized efficient design improved segmentation performance from 92.45 to 95.97 on mIoU while reducing computational FLOPs from 461.07 MMacs to 166.03 MMacs. The large kernel lightweight model ExtremeMETA showcased the hybrid design’s ability on complex tasks.

large convolution kernel↗
DOE OSTI2022

Dynamic Response of an Electron Gas: Towards the Exact Exchange-Correlation Kernel

Precise calculations of dynamics in the homogeneous electron gas (jellium model) are of fundamental importance for design and characterization of new materials. In this work, we introduce a diagrammatic Monte Carlo technique based on algorithmic Matsubara integration that allows us to compute frequency and momentum resolved finite temperature response directly in the real frequency domain using series of connected Feynman diagrams. The data for charge response at moderate electron density are used to extract the frequency dependence of the exchange-correlation kernel at finite momenta and temperature. These results are as important for development of the time-dependent density functional theory for materials dynamics as ground state energies are for the density functional theory.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗
DOE OSTI2023

Digital Modeling on Large Kernel Metamaterial Neural Network

Deep neural networks (DNNs) utilized recently are physically deployed with computational units (e.g., CPUs and GPUs). Such a design might lead to a heavy computational burden, significant latency, and intensive power consumption, which are critical limitations in applications such as Internet of Things (IoT), edge computing, and usage of drones. Recent advances in optical computational units (e.g., metamaterial) have shed light on energy-free and light-speed neural networks. However, the digital design of the metamaterial neural network (MNN) is fundamentally limited by its physical limitations, such as precision, noise, and bandwidth during fabrication. Moreover, the unique advantages of MNN’s (e.g., light-speed computation) are not fully explored via standard 3×3 convolution kernels. In this paper, we propose a novel large kernel metamaterial neural network (LMNN) that maximizes the digital capacity of the state-of-the-art (SOTA) MNN with model re-parametrization and network compression, while also considering the optical limitation explicitly. The new digital learning scheme can maximize the learning capacity of MNN while modeling the physical restrictions of meta-optics. With the proposed LMNN, the computation cost of the convolutional front-end can be offloaded to fabricated optical hardware. The experimental results on two publicly available datasets demonstrate that the optimized hybrid design improved classification accuracy while reducing computational latency. In conclusion, the development of the proposed LMNN is a promising step towards the ultimate goal of energy-free and light-speed AI.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2021

CFD-Guided Evaluation of Spark-Assisted Gasoline Compression Ignition for Cold Idle Operation

A closed-cycle, three-dimensional (3D) computational fluid dynamics (CFD) analysis campaign was conducted to evaluate the performance of using spark plugs to assist gasoline compression ignition (GCI) combustion during cold idle operations. A conventional spark plug using single-sided J-strap design was put at a location on the cylinder head to facilitate spray-guided spark assistance. Ignition was modeled with an L-type energy distribution to depict the breakdown and the arc-to-glow phases during the energy discharge process. Several key design parameters were investigated, including injector clocking, number of nozzle holes, spray inclusion angle, number of fuel injections, fuel split ratio, and fuel injection timings. The study emphasized the region around the spark gap, focusing on flame kernel formation and development and local equivalence ratio distribution. Flame kernel development and the ignition process were found to correlate strongly with the fuel stratification and the flow velocity near the spark gap. The analysis results showed that the flame kernel development followed the direction of the local flow field. In addition, the local fuel stratification notably influenced early-stage flame kernel development due to varying injection spray patterns and the fuel injection strategies. Among these design parameters, the number of nozzle holes and fuel injection timing had the most significant effects on the engine combustion performance.

42 ENGINEERING↗
DOE OSTI2023

Multi-objective Bayesian alloy design using multi-task Gaussian processes

In design applications, correlations among material properties (such as the tendency for stronger materials to be less ductile) are often neglected. This approach is echoed in multi-objective optimization techniques which treat each performance characteristic as an independent objective, aiming to optimize scalar functions and find optimal Pareto fronts. However, this overlooks the statistical relationships between performance characteristics inherent in a material system. To address this, we propose the use of Bayesian optimization, a highly efficient black-box optimization algorithm known for constructing Gaussian processes (GPs) – uncorrelated surrogates - to model objective functions. Rather than evaluating multiple GPs for each objective function separately, we argue for a shift towards jointly modeling these objective functions, considering their statistical correlations. This integrated approach utilizes naturally occurring relationships among material properties, providing additional information to enhance the performance of the design framework. This requires the replacement of multiple independent GPs with a single multi-task GP, employing a correlation matrix to construct a multi-task kernel function, wherein each task corresponds to a single objective function. Here, we anticipate this refined methodology will better leverage material correlations, improving design optimization results.

36 MATERIALS SCIENCE↗
DOE OSTI2021

A High Performance Sparse Tensor Algebra Compiler in MLIR

Sparse tensor algebra is widely used in many applications, including scientific computing, machine learning, and data analytics. The performance of sparse tensor algebra kernels strongly depends on the intrinsic characteristics of the input tensors, hence many storage formats are designed for tensors to achieve optimal performance for particular applications/architectures, which makes it challenging to implement and optimize every tensor operation of interest on a given architecture. We propose a tensor algebra domain-specific language (DSL) and compiler framework to automatically generate kernels for mixed sparse-dense tensor algebra operations. The proposed DSL provides high-level programming abstractions that resemble the familiar Einstein notation to represent tensor algebra operations. The compiler introduces a new Sparse Tensor Algebra dialect built on top of LLVM's extensible MLIR compiler infrastructure for efficient code generation while covering a wide range of tensor storage formats. Our compiler also leverages input-dependent code optimization to enhance data locality for better performance. Our results show that the performance of automatically generated kernels outperforms the state-of-the-art sparse tensor algebra compiler, with up to 20.92x, 6.39x, and 13.9x performance improvement over state-of-the-art tensor algebra compilers, for parallel SpMV, SpMM, and TTM, respectively.

Tian, Ruiqin↗