Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Accelerator Applications”

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 163 records · Page 9
All sourcesNASA NTRSDOE OSTI
DOE OSTI2024

SmartFuse: Reconfigurable Smart Switches to Accelerate Fused Collectives in HPC Applications

Communication switches have sometimes been augmented to process collectives (e.g., the IBM BlueGene project and the Mellanox SHArP switch). In this work, we find that there is a great acceleration opportunity through the further augmentation of switches to accelerate more complex functions that combine communication with computation. We consider three types of such functions. The first is fully-fused collectives built by fusing multiple existing collectives like Allreduce with Alltoall. The second is semi-fused collectives built by combining a collective with another computation. The third we refer to as higher-order collectives built by combining multiple computations and communications, such as to perform matrix-matrix multiply (PGEMM). In this work, we propose a framework called SmartFuse to accelerate fused collective functions. The core of SmartFuse is a reconfigurable smart switch to support these operations. The semi/fully fused collectives are implemented with a CGRAlike architecture, while higher-order collectives are implemented with a more specialized computational unit that can also schedule communication. Supporting our framework is software to evaluate and translate relevant parts of the input program, compile them into a control data flow graph, and then map this graph to the switch hardware. The proposed framework, once deployed, has the strong potential to accelerate existing HPC applications transparently by encapsulation within an MPI implementation. Experimental results show that this approach improves the performance of the PGEMM kernel, MINIFE, and AMG by, on average, 94%, 15%, and 13%, respectively.

Haghi, Pouya↗
DOE OSTI2020

Tensile behavior of dual-phase titanium alloys under high-intensity proton beam exposure: radiation-induced omega phase transformation in Ti-6Al-4V

A high-intensity proton beam exposure with 181 MeV energy has been conducted at Brookhaven Linac Isotope Producer facility on various material specimens for accelerator targetry applications, including titanium alloys as a beam window material. The radiation damage level of the analyzed capsule was 0.25 dpa at the beam center region with an irradiation temperature around 120 oC. Tensile tests showed increased hardness and a large decrease in ductility for the dual a+ß-phase Ti-6Al-4V Grade-5 and Grade-23 extra low interstitial alloys, with the near a-phase Ti-3Al-2.5V Grade-9 alloy still exhibiting uniform elongation of a few % after irradiation. Transmission Electron Microscope analyses on Ti-6Al-4V indicated clear evidence of a high-density of defect clusters with size less than 2 nm in each a-phase grain. The ß-phase grains did not contain any visible defects such as loops or black dots, while the diffraction patterns clearly indicated ?-phase precipitation in an advanced formation stage. The radiation-induced ?-phase transformation in the ß-phase could lead to greater loss of ductility in Ti-6Al-4V alloys in comparison with Ti-3Al-2.5V alloy with less ß-phase.google scho

Tensile Test, titanium, acclerator, proton beam↗
DOE OSTI2026

Power Electronic Development Challenges for High Power Density Integrated Electric Drive Applications

The ever-increasing demand for compact, efficient, and high-performance traction drive systems for transportation applications has accelerated the development of integrated electric drives. In these systems, the electric machine and inverter are integrated within a single housing, offering significant advantages in electrical performance, volume, weight, cost, and overall system efficiency. Despite these benefits, such high levels of integration introduce a new set of challenges, particularly in power electronic design and component selection. This article presents an in-depth investigation of a highly integrated electric drive architecture with an internal stator-mounted inverter, highlighting key design considerations and trade-offs aimed exclusively at maximizing system power density.Based on a comprehensive review of the literature, the power density of a voltage-source-inverter-driven electric drive is primarily governed by the volumetric contributions of the power modules, heat sinks, and DC-link capacitors. Accordingly, this work focuses on the optimization of these three critical components to enhance the inverter's overall power density. The proposed design achieves a power density of 100 kW/L, demonstrating the effectiveness of the presented approach for next-generation integrated electric drive systems.

33 ADVANCED PROPULSION SYSTEMS↗
DOE OSTI2025

Experimental characterization of discharge plasma dynamics in a square capillary for prospective applications in laser wakefield acceleration

Research on plasma generated by electrical discharge in gas-filled capillaries plays an important role in advancing laser wakefield accelerators. We present a comprehensive study of the temporal and spatial distribution of plasma density within a short, 1 cm long, square-shaped capillary filled with hydrogen gas. Two transverse capillary sizes, 500 μm and 300 μm, were investigated achieving peak plasma densities of 0.9 × 10 18 cm −3 and 2.5 × 10 18 cm −3 , respectively, at the capillary center. In addition, we explore how these distributions depend on discharge parameters, specifically discharge current and gas flow through the capillary. Plasma density was determined by analyzing the Stark broadening of the hydrogen Balmer-alpha line. The results obtained were compared with the theoretical models and simulations. The comparison between the model predictions and the experimental data at the transient ionization stage of the discharge reveals a discrepancy of a factor of ∼1.3–2.1, depending on the capillary size, which is thoroughly discussed.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
DOE OSTI2021

In Situ Generation of High-Energy Spin-Polarized Electrons in a Beam-Driven Plasma Wakefield Accelerator

We report that in situ generation of a high-energy, high-current, spin-polarized electron beam is an outstanding scientific challenge to the development of plasma-based accelerators for high-energy colliders. In this Letter, we show how such a spin-polarized relativistic beam can be produced by ionization injection of electrons of certain atoms with a circularly polarized laser field into a beam-driven plasma wakefield accelerator, providing a much desired one-step solution to this challenge. Using time-dependent Schrödinger equation (TDSE) simulations, we show the propensity rule of spin-dependent ionization of xenon atoms can be reversed in the strong-field multiphoton regime compared with the non-adiabatic tunneling regime, leading to high total spin polarization. Furthermore, three-dimensional particle-in-cell simulations are incorporated with TDSE simulations, providing start-to-end simulations of spin-dependent strong-field ionization of xenon atoms and subsequent trapping, acceleration, and preservation of electron spin polarization in lithium plasma. We show the generation of a high-current (0.8 kA), ultralow-normalized-emittance (~ 37 nm ), and high-energy (2.7 GeV) electron beam within just 11 cm distance, with up to ~31% net spin polarization. Higher current, energy, and net spin-polarization beams are possible by optimizing this concept, thus solving a long-standing problem facing the development of plasma accelerators.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
DOE OSTI2023

Artificial Intelligence in Nuclear Physics

Artificial Intelligence (AI) and Machine Learning (ML) are rapidly developing fields providing data-driven algorithms to predict, classify, and make decisions based on data. Nuclear Physics Research is data-driven and AI/ML techniques have been implemented for experiment and accelerator control, in theoretical applications, and in data processing and analysis. These algorithms open possibilities for automation, thereby augmenting human capabilities. Additionally, Open Science is enabled by simultaneous analyses of multiple data sources, leading to scientific knowledge. This talk will summarize current applications of AI/ML in nuclear physics, as well as accelerator applications, and will cover upcoming initiatives and research in AI/ML.

Jeske, Torri↗
DOE OSTI2021

Union: A Unified HW-SW Co-Design Ecosystem in MLIR for Evaluating Tensor Operationson Spatial Accelerators

To meet the extreme compute demands for deep learning across commercial and scientific applications, dataflow accelerators are becoming increasingly popular. While these“domain-specific” accelerators are not fully programmable like CPUs and GPUs, they retain varying levels of flexibility with respect to data orchestration, i.e., dataflow and tiling optimizations to enhance efficiency. There are several challenges when designing new algorithms and mapping approaches to execute the algorithms for a target problem on new hardware. Previous works have addressed these challenges individually. To address this challenge as a whole, in this work, we present an HW-SW co-design ecosystem for spatial accelerators called Union within the popular MLIR compiler infrastructure. Our framework allows exploring different algorithms and their mappings on several accelerator cost models. Union also includes a plug-and-play library of accelerator cost models and mappers which can easily be extended. The algorithms and accelerator cost models are connected via a novel mapping abstraction that captures the map space of spatial accelerators which can be systematically pruned based on constraints from the hardware, workload, and mapper. We demonstrate the value of Union for the community with several case studies which examine offloading different tensor operations (CONV/GEMM/Tensor Contraction) on diverse accelerator architectures using different mapping schemes.

Jeong, Geonhwa↗
DOE OSTI2026

Recent developments and operation of polarized photocathodes at Jefferson Lab

Spin-polarized electron sources are critical to a wide range of accelerator-based applications for nuclear and particle physics. At Thomas Jefferson National Accelerator Facility, they play a central role in delivering high-quality polarized beams for precision nuclear physics experiments and next-generation parity-violation measurements, where stringent control of systematic uncertainties is essential. These sources are also expected to be key components of other initiatives, including the Electron-Ion Collider and the potential future positron capabilities at Jefferson Lab. In this talk, I will present ongoing research and development efforts at Jefferson Lab focused on the design, fabrication and optimization of spin-polarized photocathodes. This includes growth using molecular beam epitaxy (MBE) or metal-organic chemical vapor deposition (MOCVD), along with detailed characterization of their performance metrics, such as quantum efficiency (QE), electron spin polarization and QE anisotropy, all of which are increasingly important metrics for polarized electron sources at Jefferson Lab and the Electron-Ion Collider. Strategies to mitigate QE anisotropy, which is critical for reducing helicity-correlated beam asymmetries in precision experiments such as MOLLER will be highlighted. Finally, I will present recent efforts aimed at improving the operational lifetime of spin-polarized photocathodes in injector environments, particularly under high-voltage conditions in DC electron guns. These developments are essential for enabling reliable, high-performance operation of polarized sources for current and future accelerator programs.

Kachwala, Alimohammed [Thomas Jefferson National A↗
DOE OSTI2025

Fabrication and Test of C3a: A Six-Layer Subscale Canted $\cos \theta$ Dipole Magnet Using High-Temperature Superconducting corc Wires

REBCO coated conductors have a strong potential for high-field magnet applications. The REBCO technology, however, is still in its infancy for accelerator magnet applications. As part of the U.S. Magnet Development Program (MDP), we developed a six-layer canted cos θ dipole magnet, C3a, using CORC® wires developed by Advanced Conductor Technologies LLC. All the layers were wound using a semi-automated winding machine. Three layers of the magnet used CORC® wires containing the SuperPower “AP” REBCO tapes and the remaining layers used the wires containing the “HM” tapes. At 77 K, both kinds of CORC® wires showed 5% to 10% degradation, after bending to a minimum bend radius of 30 or 35 mm, with respect to the self-field critical current measured before winding. At 4.2 K, the magnet reached 9.5 kA at a ramp rate of 9 A s -1 and generated a dipole field of 1.4 T. Further, the critical current of one layer degraded by 4% after a current transient up to 10.5 kA ramped in an averaged rate of 175 kA s -1 or 20 T s -1 . We confirmed the HM CORC® wire can carry a higher current than the AP CORC® wire at 4.2 K. The test results of the C3a magnet showed that the fabrication and assembly procedure can be used for the upcoming full-scale C3 magnet.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗
DOE OSTI2023

Principles and applications of x-ray light sources driven by laser wakefield acceleration

One of the most prominent applications of modern particle accelerators is the generation of radiation. In a synchrotron or an x-ray free electron laser (XFEL), high energy electrons oscillating in periodic magnetic structures emit bright x rays. In spite of their scientific appeal that will remain evident for many decades, one limitation of synchrotrons and XFELs is their typical mile-long size and their cost, which often limits access to the broader scientific community. Here, this tutorial reviews the principles and prospects of using plasmas produced by intense lasers as particle accelerators and x-ray light sources, as well as some of the applications they enable. A plasma is an ionized medium that can sustain electrical fields many orders of magnitude higher than that in conventional radio frequency accelerator structures and can be used to accelerate electrons. When short, intense laser pulses are focused into a gas, it produces electron plasma waves in which electrons can be trapped and accelerated to GeV energies. This process, laser-wakefield acceleration (LWFA), is analogous to a surfer being propelled by an ocean wave. Many radiation sources, from THz to gamma-rays, can be produced by these relativistic electrons. This tutorial reviews several LWFA-driven sources in the keV-MeV photon energy range: betatron radiation, inverse Compton scattering, bremsstrahlung radiation, and undulator/XFEL radiation. X rays from laser plasma accelerators have many emerging applications. They can be used in innovative and flexible x-ray imaging and x-ray absorption spectroscopy configurations, for use in biology, industry, and high-energy density science.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
DOE OSTI2021

Characterization of Dosimetric Differences in Strut-Adjusted Volume Implant Treatment Plans Calculated With TG-43 Formalism and a Model-Based Dose Calculation Algorithm

To comprehensively characterize dosimetric differences between calculations with a commercial model-based dose calculation algorithm (MBDCA) and the TG-43 formalism in application to accelerated partial breast irradiation (APBI) with the strut-adjusted volume implant (SAVI) applicator.

62 RADIOLOGY AND NUCLEAR MEDICINE↗
DOE OSTI2023

Linearization of an Electron Beam’s Longitudinal Phase Space Using a Hollow-Channel Plasma

The removal of undesired beam nonlinear energy chirp (time-energy correlation) or linearization of the beam longitudinal phase space (LPS) is crucial for high-brightness linac-based scientific applications, such as x-ray free-electron lasers. In this paper, we propose that a low-density hollow channel plasma can be used as a near-ideal passive linearizer to significantly linearize the beam LPS and, at the same time, preserve the beam emittance. Physically, the passage of the beam through the hollow plasma channel excites a strong quasicosinoidal longitudinal wakefield that acts to mitigate the beam nonlinear energy chirp by superimposing a reverse chirp on the beam. The theoretical analyses and large-scale three-dimensional start-to-end simulations confirm that the beam longitudinal phase space can be almost completely linearized without noticeable beam emittance growth. In conclusion, application of such a near-ideal linearizer may significantly improve the performance of numerous accelerator-based applications.

43 PARTICLE ACCELERATORS↗
DOE OSTI2024

Toward a Holistic Performance Evaluation of Large Language Models Across Diverse AI Accelerators

Artificial intelligence (AI) methods have become critical in scientific applications to help accelerate scientific discovery. Large language models (LLMs) are being considered a promising approach to address some challenging problems because of their superior generalization capabilities across domains. The effectiveness of the models and the accuracy of the applications are contingent upon their efficient execution on the underlying hardware infrastructure. Specialized Al accelerator hardware systems have recently become available for accelerating Al applications. However, the comparative performance of these AI accelerators on large language models has not been previously studied. In this paper, we systematically study LLMs on multiple AI accelerators and GPUs and evaluate their performance characteristics for these models. We evaluate these systems with (i) a micro-benchmark using a core transformer block, (ii) a GPT-2 model, and (iii) an 1,I,M-driven science use case, GenSLM. We present our findings and analyses of the models' performance to better understand the intrinsic capabilities of AI accelerators. Furthermore, our analysis takes into account key factors such as sequence lengths, scaling behavior, and sensitivity to gradient accumulation steps.

Emani, Murali↗
DOE OSTI2022

Hvac: Removing I/O Bottleneck for Large-Scale Deep Learning Applications

Scientific communities are increasingly adopting deep learning (DL) models in their applications to accelerate scientific discovery processes. However, with rapid growth in the computing capabilities of HPC supercomputers, large-scale DL applications have to spend a significant portion of training time performing I/O to a parallel storage system. Previous research works have investigated optimization techniques such as prefetching and caching. Unfortunately, there exist non-trivial challenges to adopting the existing solutions on HPC supercomputers for large-scale DL training applications, which include non-performance and/or failures at extreme scale, lack of portability and generality in design, complex deployment methodology, and being limited to a specific application or dataset. To address these challenges, we propose High-Velocity AI Cache (HVAC), a distributed read-cache layer that targets and fully exploits the node-local storage or near node-local storage technology. HVAC seamlessly accelerates read I/O by aggregating node-local or near node-local storage, avoiding metadata lookups and file locking while preserving portability in the application code. We deploy and evaluate HVAC on 1,024 nodes (with over 6000 NVIDIA V100 GPUS) of the Summit supercomputer. In particular, we evaluate the scalability, efficiency, accuracy, and load distribution of HVAC compared to GPFS and XFS-on-NVMe. With four different DL applications, we observe an average 25 % performance improvement atop GPFS and 9% drop against XFS-on-NVMe, which scale linearly and are considered the performance upper bound. We envision HVAC as an important caching library for upcoming HPC supercomputers such as Frontier.

Khan, Awais↗