Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “training”

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 1,171 records · Page 65
All sourcesNASA NTRSDOE OSTI
DOE OSTI2022

Efficient quantum state tomography with mode-assisted training

Neural networks (NNs) representing quantum states are typically trained using Markov chain Monte Carlo based methods. However, unless specifically designed, such samplers only consist of local moves, making the slow-mixing problem prominent even for extremely simple quantum states. Here, we propose to use mode-assisted training that provides global information via the modes of the NN distribution. Further, applied to quantum state tomography using restricted Boltzmann machines, this method improves the quality of reconstructed quantum states by orders of magnitude. The method is applicable to other types of NNs and may efficiently tackle problems previously unmanageable.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
DOE OSTI2023

Generation of multiple attosecond sub-bursts in femtosecond pulse trains by long-wavelength driving lasers

High-order harmonics generated by a long intense femtosecond laser are known experimentally to create attosecond pulse trains (APTs). In the time domain, an APT consists of a sequence of sharp attosecond bursts that are equally separated by each half optical cycle. Here, in this work, we show that such well-known features can be modified when a longer wavelength driving laser is used. From our simulations, we show that multiple shorter attosecond sub-bursts exist in the femtosecond pulse train within each half optical cycle and the duration of each sub-burst scales approximately as $λ^{–2}_0$ with the driving laser wavelength λ 0 . We show that such sub-bursts can be found using quantitative rescattering model for harmonics generated from a single atom, and their origin is due to the interference of the quantum orbits from first two returns of the recombining electron. We further show that such sub-bursts can be phase matched under proper laser focusing condition and the position of the gas cell, thus, such new features should be observable experimentally.

47 OTHER INSTRUMENTATION↗
DOE OSTI2011

Doppler cooling with coherent trains of laser pulses and a tunable velocity comb

We explore the possibility of decelerating and Doppler cooling an ensemble of two-level atoms by a coherent train of short, nonoverlapping laser pulses. We derive analytical expressions for mechanical force exerted by the train. In frequency space the force pattern reflects the underlying frequency comb structure. The pattern depends strongly on the ratio of the atomic lifetime to the repetition time between the pulses and pulse area. For example, in the limit of short lifetimes, the frequency-space peaks of the optical force wash out. We propose to tune the carrier-envelope offset frequency to follow the Doppler-shifted detuning as atoms decelerate; this leads to compression of atomic velocity distribution about comb teeth and results in a ''velocity comb''--a series of narrow equidistant peaks in the velocity space.

74 ATOMIC AND MOLECULAR PHYSICS↗
DOE OSTI2011

Observation of cold Rb{sub 2} molecules trapped in an optical dipole trap using a laser-pulse-train technique

In this work, we have developed and characterized a laser-pulse-train technique to observe cold Rb{sub 2} molecules trapped in an optical dipole trap. The molecules are produced in a magneto-optical trap, and then loaded into a crossed optical dipole trap. The time evolution of the molecular population is obtained by applying a laser pulse train, which photoionizes the ground-state molecules through intermediate molecular bands. Our results show that this technique allows us to obtain a faster data acquisition rate of the time evolution of the molecule population than other techniques.

74 ATOMIC AND MOLECULAR PHYSICS↗
DOE OSTI2023

Tensor train continuous time solver for quantum impurity models

The simulation of strongly correlated quantum impurity models is a significant challenge in modern condensed matter physics that has multiple important applications. Thus far, the most successful methods for approaching this challenge involve Monte Carlo techniques that accurately and reliably sample perturbative expansions to any order. However, the cost of obtaining high precision through these methods is high. Recently, tensor train decomposition techniques have been developed as an alternative to Monte Carlo integration. In this study, we apply these techniques to the single-impurity Anderson model at equilibrium by calculating the systematic expansion in power of the hybridization of the impurity with the bath. Furthermore, we demonstrate the performance of the method in a paradigmatic application, examining the first-order phase transition on the infinite-dimensional Bethe lattice, which can be mapped to an impurity model through dynamical mean field theory. Our results indicate that using tensor train decomposition schemes allows the calculation of finite-temperature Green's functions and thermodynamic observables with unprecedented accuracy. The methodology holds promise for future applications to frustrated multiorbital systems, using a combination of partially summed series with other techniques pioneered in diagrammatic and continuous time quantum Monte Carlo.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗
DOE OSTI2024

Training toward significance with the decorrelated event classifier transformer neural network

Experimental particle physics uses machine learning for many tasks, where one application is to classify signal and background events. This classification can be used to bin an analysis region to enhance the expected significance for a mass resonance search. In natural language processing, one of the leading neural network architectures is the transformer. In this work, an event classifier transformer is proposed to bin an analysis region, in which the network is trained with special techniques. The techniques developed here can enhance the significance and reduce the correlation between the network’s output and the reconstructed mass. It is found that this trained network can perform better than boosted decision trees and feed-forward networks. Published by the American Physical Society 2024

Astronomy & Astrophysics↗
DOE OSTI2011

Suppression of Beam-Ion Instability in Electron Rings with Multi-Bunch Train Beam Fillings

The ion-caused beam instability in the future light sources and electron damping rings can be serious due to the high beam current and ultra-small emittance of picometer level. One simple and effective mitigation of the instability is a multi-bunch train beam filling pattern which can significantly reduce the ion density near the beam, and therefore reduce the instability growth rate up to two orders of magnitude. The suppression is more effective for high intensity beams with low emittance. The distribution and the field of trapped ions are benchmarked to validate the model used in the paper. The wake field of ion-cloud and the beam-ion instability is investigated both analytically and numerically. We derived a simple formula for the build-up of ion-cloud and instability growth rate with the multi-bunch-train filling pattern. The ion instabilities in ILC damping ring, SuperKEKB and SPEAR3 are used to compare with our analyses. The analyses in this paper agree well with simulations.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗
DOE OSTI2024

MassiveGNN: Efficient Training via Prefetching for Massively Connected Distributed Graphs

Graph Neural Networks (GNN) are indispensable in learning from graph-structured data, yet their rising computational costs, especially on massively connected graphs, pose significant challenges in terms of execution performance. To tackle this, distributed-memory solutions such as partitioning the graph to concurrently train multiple replicas of GNNs are in practice. However, approaches requiring a partitioned graph usually suffer from communication overhead and load imbalance, even under optimal partitioning and communication strategies due to irregularities in the neighborhood minibatch sampling. This paper proposes practical trade-offs for improving the sampling and communication overheads for representation learn- ing on distributed graphs (using popular GraphSAGE architecture) by developing a parameterized prefetch and eviction scheme on top of the state-of-the-art Amazon DistDGL distributed GNN framework, demonstrating about 15–40% improvement in end-to-end training performance on the NERSC Perlmutter supercomputer for various OGB datasets.

Machine Leanring, high performance comptuing, grap↗
DOE OSTI2025

Transfer Learning Trained LSTM Models for Household Load Profile Forecasting

Grid edge renewable energy resources, such as rooftop solar photovoltaics, closely interact with consumer load profiles. Therefore, forecasting future electricity demand, ideally at the individual household level, is indispensable. In this paper, we present a transfer learning enhanced household load profile forecasting method. First, we tune a long short-term memory forecasting model to perform day-ahead prediction of household electricity load profiles. Then we improve these individualized models using transfer learning, and we use k-means clustering to create optimal source data sets. We find average improvements of 4.38% (largest improvement of 10.71%) when the entire data set was used to train the source model and 2.45% (largest improvement of 11.57%) in the mean absolute error when households were first clustered and used to train separate source models for each cluster. We find that transfer learning with clustered data can effectively boost the forecasting performance of the LSTM models. We use realistic household power measurements for 148 real residential households in Austin, Texas.

deep learning↗
DOE OSTI2025

Integrated On-Board Charger for Dual Motor Based Electric Vehicle Power Train

This paper presents an integrated on-board charger for dual motor based electric vehicle (EV) power train with silicon carbide semiconductor switch-based traction inverters. The proposed charger architecture reconfigures the traction inverters to be used in the battery charging mode, thereby eliminating several high frequency switches and their accessories. The motor windings are reused as line inductors for the power factor correction (PFC) rectifier stage in the charging mode thereby reducing the requirements of magnetic components. The architectures reduce the cost, weight, and volume of the EV power train. Simulation results are presented to demonstrate the working principle of the proposed architecture.

Mukherjee, Subho [ORNL] (ORCID:0009000672297925)↗
DOE OSTI2025

Impulse Response Functions for Characterizing Pulse-to-Pulse Junction Temperature Behavior in Laser Diode Pulse Trains

A method is proposed for determination of the pulse-to-pulse variation in junction temperature and emission wavelength of a semiconductor laser diode during a train of pulses. Here, this approach, based on impulse response functions, enables predictions for pulse trains with arbitrary pulse-to-pulse variations in output power, pulse width, and pulse delay using a limited set of experimental characterization data. The use of this approach is illustrated by application to a particular device structure.

Deri, Robert J. [Lawrence Livermore National Labor↗
DOE OSTI2021

Distributed Training for High Resolution Images: A Domain and Spatial Decomposition Approach

In this work we developed two Pytorch libraries using the PyTorch RPC interface for distributed deep learning approaches on high resolution images. The spatial decomposition library allows for distributed training on very large images, which otherwise wouldn’t be possible on a single GPU. The domain parallelism library allows for distributed training across multiple domain unlabeled data, by leveraging the domain separation architecture. Both of those libraries where tested on the Summit supercomputer at Oak Ridge National Laboratory at a moderate scale.

Tsaris, Aristeidis (aris)↗
DOE OSTI2021

Visualization of shocked material instabilities using a fast-framing camera and XFEL four-pulse train

Many questions regarding dynamic materials could be answered by using time-resolved ultra-fast imaging techniques to characterize the physical and chemical behavior of materials in extreme conditions and their evolution on the nanosecond scale. In this work, we perform multi-frame phase-contrast imaging (PCI) of micro-voids in low density polymers under laser-driven shock compression. At the Matter in Extreme Conditions (MEC) Instrument at the Linac Coherent Light Source (LCLS), we used a train of four x-ray free electron laser (XFEL) pulses to probe the evolution of the samples. To visualize the void and shock wave interaction, here, we deployed the Icarus V2 detector to record up to four XFEL pulses, separated by 1-3 nanoseconds. In this work, we image elastic waves interacting with the micro-voids at a pressure of several GPa. Monitoring how the material’s heterogeneities, like micro-voids, dictate its response to a compressive wave is important for benchmarking the performances of inertial confinement fusion energy materials. For the first time in a single sample, we have combined an ultrafast x-ray framing camera and four XFEL pulse train to create an ultrafast movie of micro-void evolution under laser-driven shock compression. Eventually, we hope this technique will resolve the material density as it evolves dynamically under laser shock compression.

fusion↗
DOE OSTI2019

Exchange Bias Training Effect in GdBaCo{sub 2}O{sub 5.5} Cobaltite

The nature of the training effect in exchange bias compounds, which consists in a decrease in the exchange bias field H{sub EB} under magnetic field cycling, is clarified. It is shown that the training effect in GdBaCo{sub 2}O{sub 5 + δ} cobaltite with δ ≈ 0.5 is caused by a decrease in magnetization. Depending on time, magnetization M(t) decreases sharply after the first cycle and slowly in subsequent N cycles. A direct relationship (proportionality) is established between the decrease in the magnetization M(t) of ferromagnetic particles and the change of the field H{sub EB}(t). It is assumed that the cooling of a sample in a magnetic field gives rise to small (single-domain) and large ferromagnetic clusters in an antiferromagnetic environment. The sharp decrease in M(t) and H{sub EB}(t) and the deviation of the function H{sub EB}(N) from the well-known power-law relation ΔH{sub EB} ∝ N{sup –1/2} with N > 1 is explained by the transformation of large ferromagnetic clusters into a multidomain state under magnetic field cycling. The results are consistent with the phenomenological model of Meiklejohn and Bean, which predicts a linear dependence of the exchange bias on the ferromagnetic magnetization of a ferromagnet/antiferromagnet structure.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗
DOE OSTI2021

Low Power Hardware-In-The-Loop Neuromorphic Training Accelerator

The training process for spiking neural networks can be very computationally intensive. Approaches such as evolutionary algorithms may require evaluating thousands or millions of candidate solutions. In this work, we propose using neuromorphic cores implemented on a Xilinx Zynq system on chip to accelerate and improve the energy efficiency of the evaluation step of an evolutionary training approach. We demonstrate this can significantly reduce the required energy to evolve a network with some cases showing greater than 10 times improvement as compared to a CPU-only system.

Mitchell, Parker↗
DOE OSTI2021

Training Spiking Neural Networks with Synaptic Plasticity under Integer Representation

Neuromorphic computing is emerging as a promising Beyond Moore computing paradigm that employs event-triggered computation and non-von Neumann hardware. Spike Timing Dependent Plasticity (STDP) is a well-known bio-inspired learning rule that relies on activities of locally connected neurons to adjust the weights of their respective synapses. In this work, we analyze a basic STDP rule and its sensitivity on the different hyperparameters for training spiking neural networks (SNNs) with supervision, customized for a neuromorphic hardware implementation with integer weights. We compare the classification performance on four UCI datasets (iris, wine, breast cancer and digits) that depict varying levels of complexity. We perform a search for optimal set of hyperparameters using both grid search and Bayesian optimization. Through the use of Bayesian optimization, we show the general trends in hyperparameter sensitivity in SNN classification problem. With the best sets of hyperparameters, we achieve accuracies comparable to some of the best performing SNNs on these four datasets. With a highly optimized supervised STDP rule we show that these accuracies can be achieved with just 20 epochs of training.

Kulkarni, Shruti↗
DOE OSTI2023

Software-Hardware Co-design of Heterogeneous SmartNIC System for Recommendation Models Inference and Training

Deep Learning Recommendation Models (DLRMs) are critical applications in various domains and have evolved as one of the single largest machine learning applications. Trillions of DLRM parameters exceed the on-chip memory capacity of GPUs. Large-scale multi-node systems are required for distributed DLRM inference and training, which suffer from the all-to-all communication bottleneck, mainly limiting the scalability of ever-growing DLRMs. In recent years, SmartNICs have evolved with coupled computation and communication capabilities providing opportunities for a powerful heterogeneous device in the system. However, there isn't such a distributed system that fully leverages the abundant smartNIC resources that resolve the scalability issue of DLRMs. In this work, we proposed a software-hardware co-design of a heterogeneous smartNIC system that resolves the communication bottleneck of distributed DLRMs, mitigates the memory bandwidth pressure, and improves computation efficiency. We provide a set of smartNIC designs of cache systems (including local cache and remote cache) and smartNIC computation kernels which reduce data movement, relieve memory lookup intensity, and improve the GPU's computation efficiency. In addition, we propose a graph algorithm that improves the data locality of queries within batches which optimizes the overall system performance with higher data reuse. Our evaluation shows that our system achieves 2.1x latency speedup for inference and 1.6x throughput speedup for training.

Guo, Anqi↗
DOE OSTI2023

FORGE: Pre-Training Open Foundation Models for Science

Large language models (LLMs) are poised to revolutionize the way we conduct scientific research. However, both model complexity and pre-training cost are impeding effective adoption for the wider science community. Identifying suitable scientific use cases, finding the optimal balance between model and data sizes, and scaling up model training are among the most pressing issues that need to be addressed. In this study, we provide practical solutions for building and using LLM-based foundation models targeting scientific research use cases. We present an end-to-end examination of the effectiveness of LLMs in scientific research, including their scaling behavior and computational requirements on Frontier, the first Exascale supercomputer. We have also developed for release to the scientific community a suite of open foundation models called FORGE with up to 26B parameters using 257B tokens from over 200M scientific articles, with performance either on par or superior to other state-of-the-art comparable models. We have demonstrated the use and effectiveness of FORGE on scientific downstream tasks. Our research establishes best practices that can be applied across various fields to take advantage of LLMs for scientific discovery.

Yin, Junqi↗