Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “AUTOS”

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 1,243 records · Page 69

WE-H-202-04: Advanced Medical Image Registration Techniques

Deformable image registration has now been commercially available for several years, with solid performance in a number of sites and for several applications including contour and dose mapping. However, more complex applications have arisen, such as assessing response to radiation therapy over time, registering images pre- and post-surgery, and auto-segmentation from atlases. These applications require innovative registration algorithms to achieve accurate alignment. The goal of this session is to highlight emerging registration technology and these new applications. The state of the art in image registration will be presented from an engineering perspective. Translational clinical applications will also be discussed to tie these new registration approaches together with imaging and radiation therapy applications in specific diseases such as cervical and lung cancers. Learning Objectives: To understand developing techniques and algorithms in deformable image registration that are likely to translate into clinical tools in the near future. To understand emerging imaging and radiation therapy clinical applications that require such new registration algorithms. Research supported in part by the National Institutes of Health under award numbers P01CA059827, R01CA166119, and R01CA166703. Disclosures: Phillips Medical systems (Hugo), Roger Koch (Christensen) support, Varian Medical Systems (Brock), licensing agreements from Raysearch (Brock) and Varian (Hugo).; K. Brock, Licensing Agreement - RaySearch Laboratories. Research Funding - Varian Medical Systems; G. Hugo, Research grant from National Institutes of Health, award number R01CA166119.; G. Christensen, Research support from NIH grants CA166119 and CA166703 and a gift from Roger Koch. There are no conflicts of interest.

60 APPLIED LIFE SCIENCES↗

TH-CD-201-03: A Real-Time Method to Simultaneously Measure Linear Energy Transfer and Dose for Proton Therapy Using Organic Scintillators

Purpose: The light generated in organic scintillators depends on both the radiation dose and the linear energy transfer (LET). The LET dependence leads to an under-response of the detector in the Bragg peak of proton beams. This phenomenon, called ionization quenching, must be corrected to obtain accurate dose measurements of proton beams. This work exploits the ionization quenching phenomenon to provide a method of measuring LET and auto correcting quenching. Methods: We exposed simultaneously four different organic scintillators (BCF-12, PMMA, PVT, and LSD; 1mm in diameter) and a plane parallel ionization chamber in passively scattered proton beams to doses between 32 and 43 cGy and fluence averaged LET values from 0.47 to 1.26 keV/µm. The LET values for each irradiation condition were determined using a validated Monte Carlo model of the beam line. We determined the quenching parameter in the Birk’s equation for scintillation in BCF-12 for dose measurements. One set of irradiation conditions was used to correlate the scintillation response ratio to the LET values and plot a scintillation response ratio versus LET calibration curve. Irradiation conditions independent from the calibration ones were used to validate this method. Comparisons to the expected values were made on both the basis of dose and LET. Results: Among all the scintillators investigated, the ratio of PMMA to BCF-12 provided the best correlation to LET values and was used as the LET calibration curve. The expected LET values in the validation set were within 2%±6%, which resulted in dose accuracy of 1.5%±5.8% for the range of LET values investigated in this work. Conclusion: We have demonstrated the feasibility of using the ratio between the light output of two organic scintillators to simultaneously measure LET and dose of therapeutic proton beams. Further studies are needed to verify the response in higher LET values.

60 APPLIED LIFE SCIENCES↗

Laser excitation of long-lived oscillatory states in a dusty-plasma trap

Vertical oscillations of a single particle in a monolayer dusty-plasma sheath have been excited by a laser radiation pulse. The particle oscillates in a self-sustained regime at a characteristic frequency of about {nu} = 25 Hz. It is established that phenomena related to the particle recharge and the delay of its charge variation relative to the equilibrium value cannot account for the self-sustained regime of oscillations. It is suggested that the experimentally observed vertical auto-oscillations of a single particle take place due to resonant pumping of the kinetic energy of the chaotic motion of particles in the dusty-plasma sheath to the energy of vertical oscillations.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Optical Properties of InGaAs/InAlAs Metamorphic Nanoheterostructures for Photovoltaic Converters of Laser and Solar Radiation

Reflectance spectroscopy has been used to determine the refractive indices of nanoscale In{sub x}Al{sub y}Ga{sub 1–x–y}As and In{sub x}Al{sub 1–x}As layers with indium and aluminum concentrations x = 0.21–0.24 and y = 0, 0.1, and 0.2 on specially created Bragg-reflector heterostructures at wavelengths in the range 700–2000 nm. It was demonstrated that the method based on an analysis of the auto- and cross-correlation coefficients of the wavelength derivatives of the dependence of the reflectance of structures of this kind in order to determine the dispersion curves of the materials forming a reflector. It was found that raising the concentration of indium in InGaAs and AlInAs leads to a substantial rise in the refractive index, with a preserved spectral run of the refractive indices, which is characteristic of gallium arsenide and aluminum arsenide.

14 SOLAR ENERGY↗

Autodifferentiable Ensemble Kalman Filters

Data assimilation is concerned with sequentially estimating a temporally evolving state. This task, which arises in a wide range of scientific and engineering applications, is particularly challenging when the state is high-dimensional and the state-space dynamics are unknown. This paper introduces a machine learning framework for learning dynamical systems in data assimilation. Here, our auto-differentiable ensemble Kalman filters (AD-EnKFs) blend ensemble Kalman filters for state recovery with machine learning tools for learning the dynamics. In doing so, AD-EnKFs leverage the ability of ensemble Kalman filters to scale to high-dimensional states and the power of automatic differentiation to train high-dimensional surrogate models for the dynamics. Numerical results using the Lorenz-96 model show that AD-EnKFs outperform existing methods that use expectation-maximization or particle filters to merge data assimilation and machine learning. In addition, AD-EnKFs are easy to implement and require minimal tuning.

autodifferentiation↗

Multiplicity dependence of (multi-)strange hadron production in proton-proton collisions at $\sqrt{s}$s = 13 TeV

The production rates and the transverse momentum distribution of strange hadrons at mid-rapidity ( | y | < 0.5 ) are measured in proton-proton collisions at $\sqrt{s}$ = 13 TeV as a function of the charged particle multiplicity, using the ALICE detector at the LHC. The production rates of K$^0_S$, Λ, &#x039E;<!-- Ξ --> , and Ω ncrease with the multiplicity faster than what is reported for inclusive charged particles. The increase is found to be more pronounced for hadrons with a larger strangeness content. Possible auto-correlations between the charged particles and the strange hadrons are evaluated by measuring the event-activity with charged particle multiplicity estimators covering different pseudorapidity regions. When comparing to lower energy results, the yields of strange hadrons are found to depend only on the mid-rapidity charged particle multiplicity. Several features of the data are reproduced qualitatively by general purpose QCD Monte Carlo models that take into account the effect of densely-packed QCD strings in high multiplicity collisions. However, none of the tested models reproduce the data quantitatively. This work corroborates and extends the ALICE findings on strangeness production in proton-proton collisions at 7 TeV.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

A design methodology for domain-optimized power-efficient supercomputing

As power has become the pre-eminent design constraint for future HPC systems, computational efficiency is being emphasized over simply peak performance. Recently, static benchmark codes have been used to find a power efficient architecture. Unfortunately, because compilers generate sub-optimal code, benchmark performance can be a poor indicator of the performance potential of architecture design points. Therefore, we present hardware/software cotuning as a novel approach for system design, in which traditional architecture space exploration is tightly coupled with software auto-tuning for delivering substantial improvements in area and power efficiency. We demonstrate the proposed methodology by exploring the parameter space of a Tensilica-based multi-processor running three of the most heavily used kernels in scientific computing, each with widely varying micro-architectural requirements: sparse matrix vector multiplication, stencil-based computations, and general matrix-matrix multiplication. Results demonstrate that co-tuning significantly improves hardware area and energy efficiency - a key driver for next generation of HPC system design.

97 MATHEMATICS AND COMPUTING↗

GPU code optimization using abstract kernel emulation and sensitivity analysis

In this paper, we develop a novel approach to GPU kernel optimization by focusing on identification of bottleneck resources and determining optimization parameters that can alleviate the bottleneck. Performance modeling for GPUs is done by abstract kernel emulation along with latency/gap modeling of resources. Sensitivity analysis with respect to resource latency/gap parameters is used to predict the bottleneck resource for a given kernel's execution. Here, the utility of the bottleneck analysis is demonstrated in two contexts: 1) Enhancing the OpenTuner auto-tuner with the new bottleneck-driven optimization strategy. Experimental results on all kernels from the Rodinia suite and GPU tensor contraction kernels from the NWChem computational chemistry suite demonstrate effectiveness. 2) Manual code optimization: two case studies illustrate the use of the bottleneck analysis to iteratively improve performance of code from state-of-the-art DSL code generators.

97 MATHEMATICS AND COMPUTING↗

Benchmarking a Proof-of-Concept Performance Portable SYCL-based Fast Fourier Transformation Library

ABSTRACT In this paper, we present an early version of a SYCL-based FFT library, capable of running on all major vendor hardware, including CPUs and GPUs from AMD, ARM, Intel and NVIDIA. The current limitations of our library is it supports single-dimension FFTs up to 211 in length and base-2 input sequences. Although preliminary, the aim of this work is to seed further developments for a rich set of features for calculating FFTs. The library has the advantage over existing portable FFT libraries in that it is single-source, and there- fore removes the complexities that arise due to abundant use of pre-processor macros and auto-generated kernels to target different architectures. We exercise two SYCL-enabled compilers, Codeplay ComputeCpp and Intel's open-source LLVM project, to evaluate performance portability of our SYCL-based FFT on various hetero- geneous architectures.We provide studies comparing our portable library with highly optimized vendor-specific FFT libraries, and discuss potential sources hindering performance.

97 MATHEMATICS AND COMPUTING↗

Revisiting Temporal Blocking Stencil Optimizations

Iterative stencils are used widely across the spectrum of High Performance Computing (HPC) applications. Many efforts have been put into optimizing stencil GPU kernels, given the prevalence of GPU-accelerated supercomputers. To improve the data locality, temporal blocking is an optimization that combines a batch of time steps to process them together. Under the observation that GPUs are evolving to resemble CPUs in some aspects, we revisit temporal blocking optimizations for GPUs. We explore how temporal blocking schemes can be adapted to the new features in the recent Nvidia GPUs, including large scratchpad memory, hardware prefetching, and device-wide synchronization. We propose a novel temporal blocking method, EBISU, which champions low device occupancy to drive aggressive deep temporal blocking on large tiles that are executed tile-by-tile. We compare EBISU with state-of-the-art temporal blocking libraries: STENCILGEN and AN5D. We also compare with state-of-the-art stencil auto-tuning tools that are equipped with temporal blocking optimizations: ARTEMIS and DRSTENCIL. Over a wide range of stencil benchmarks, EBISU achieves speedups up to 2.53x and a geometric mean speedup of 1.49x over the best state-of-the-art performance in each stencil benchmark.

Zhang, Lingqi↗

Agentic AI vs ML-Based Autotuning: A Comparative Study for Loop Reordering Optimization

High Performance Computing (HPC) applications rely heavily on code optimizations to achieve good performance on modern CPU and GPU architectures. Traditional Machine Learning auto-tuning approaches have demonstrated success in exploring high-dimensional spaces, but they often require expensive compile-run evaluations and lack adaptability for large HPC applications. The recent advances in Large Language Models (LLMs) and Agentic AI systems raise intriguing questions about the potential of these approaches to address specific optimization methodologies. This work aims to answer an essential question for the HPC community: “How Agentic AI Systems Compare to Traditional ML Autotuning Techniques?” To address this question, we present a comparative analysis between a traditional ML-based optimization approach and an Agentic AI system, evaluating their respective capabilities and limitations for loop-level optimization. In addition, we introduced a new Agentic AI system named LoopGen-AI using three different Large Language Models: GPT-4.1, Claude 4.0, and Gemini 2.5. A key finding is that LoopGen-AI achieves competitive per-formance with only a few program runs, the reasoning logs from the agents revealed that their decisions rely heavily on the combination of semantic understanding of the target kernel with dynamic feedback from the environment, highlighting a promising new dimension in performance tuning. In contrast, ML-based autotuners focus on statistical exploration, and require orders of magnitude more runs to reach peak performance. Additionally, our analysis shows that prompt engineering, particularly using Persona + Context Manager patterns, significantly impacts the effectiveness of Agentic AI. Our results indicate that while Agentic AI systems are not yet a complete replacement for ML-based autotuners, it can effectively complement traditional methods.

Rosas, Miguel Romero↗

DMS Update Manager

The DMS Update Manager is used to keep automated processing manager software up-to-date. Although it is part of PRISM (the Proteomics Research Information and Management System), it is a standalone program and does not contact any databases. The program compares the files on a central distribution share to files on the local computer, updating any files that are out of date. It supports rolling back local files to an older version and auto-deleting extra files.

Monroe, Matthew↗

Sequential Training ToolKit - STTK

Sequential Training ToolKit, or STTK, is a software toolkit plugin for the Unity development platform that enables software developers to easily develop structured Sequential Training applications for gaming and extended reality (XR) environments. STTK is architected to easily import the toolkit as a plugin for the Unity platform and then interact with the graphical user interface (GUI) for drag-and-drop functionality. STTK also includes easy add-on plugins for Prefabs and auto-builds a State Machine. Sandia National Laboratories is a multimission laboratory managed and operated by National Technology & Engineering Solutions of Sandia, LLC, a wholly owned subsidiary of Honeywell International Inc., for the U.S. Department of Energy’s National Nuclear Security Administration under contract DE-NA0003525. SAND2021-3059 O

Klein, BrandonThorin↗

Greedy Latent Space Dynamics Identification

This is a variant of LaSDI where it minimizes both reconstruction loss terms of auto encoder and dynamics regression loss terms simultaneously. Plus, an active learning framework is adopted so that a physics-informed greedy sampling point can be added to enhance the performance of gLaSDI. It is written in python and several examples are included in this software.

He, Xiaolong↗

Automatic Point Cloud Building Envelope Segmentation (AutoCuBES)

The Auto-CuBES algorithm is based on unsupervised machine learning that automatically labels 3D point cloud data and reduces the time spent in manual segmentation. The algorithm can process high-resolution point clouds and generate a wire-frame building envelope model with a small set of calibration parameters. The algorithm inputs a 3D point cloud generated by commonly available surveying equipment and outputs a wire-frame model of the building envelope. Unsupervised machine learning methods were used to identify facades, windows, and doors while minimizing the number of calibration parameters.

Puente, BryanMaldonado↗

Inventory of Composable Elements (ICE) v6.0.0

The Inventory of Composable Elements (ICE) is an open source registry software platform for managing information about biological parts. It is capable of recording information about plasmids, microbial host strains and seeds, as well as DNA parts. Includes features such as DNA sequence visualization, editing and annotation, auto-aligning sequencing trace files against reference templates, SBOL XML/RDF support, and web-of-registries functionality. The web of registries functionality provides strong support for distributed interconnected use and enables sharing and transfer of biological parts across various independent ICE instances. ICE adopts modern software development principles, leveraging component-base frameworks, offering a REST API for convenient third-party integration and emphasizing scalability, security, and service integrations for dynamic content availability. The source code is hosted at https://github.com/JBEI/ice. A public instance is available at public-registry.jbei.org, where users can try out features, upload parts or simply use it for their projects.

Plahar, Hector↗

OCHRE™ Gymnasium (ochre_gym) [SWR-23-47]

OCHRE™ Gymnasium is a Python framework for conducting reinforcement learning research on residential building energy management problems. It is a lightweight wrapper around OCHRE, a Python-based thermal-electric residential building simulator developed in-house at NREL. OCHRE Gymnasium adheres to the standard OpenAI Gym API for a reinforcement learning environment. Detailed Sphinx documentation will be auto-generated and accompany this codebase. This documentation will detail the API, provide examples of how to use the code, etc.

Emami, Patrick↗

Mneme

A simple tool allowing recording the execution of a GPU (CUDA) kernel and replaying that kernel as an independent executable. The tool operates in 3 phases. During compile time the user needs to apply a provided LLVM pass to instrument the code. The pass detects all device global variables and device functions and stores this information with the respective LLVM-IR in the global device memory. The compilation generates a record-able executable. The second phase involves running the application executable with a desired input and using LD_PRELOAD to enable recording. When recording before invoking a device kernel the pre-loaded library stores device memory in persistent storage and associates the memory with the device kernel and an LLVM IR file. At the end of the recorded execution the pre-load library generates a database in the form of a JSON file containing information regarding the LLVM-IR files and the snapshots of device memory. During the third and last phase the user can replay the execution of an kernel as a separate independent executable. Besides executing it the user can modify the LLVM IR file and auto-tune parameters such as kernel launch-bounds or kernel runtime execution parameters (e.g. Kernel Block and Grid Dimensions). Is

Parasyris, Konstantinos↗