Search NASA⌕ Search

SEARCH · Search NASA

Results for “hardware accelerators”

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 505 records · Page 28

Transportation Vibration Data for Klystron Experimental Hardware

This memo discusses the over-the-road vibration and shock data that was captured while transporting a klystron unit from the Stanford Linear Accelerator (SLAC) facility in Menlo Park, CA to the Lawrence Livermore National Laboratory (LLNL) main site (B381). The klystron was shipped over the road in an air ride truck on May 1, 2012 in a specially designed wooden crate with an aluminum support frame for reinforcement. The klystron unit, itself, was suspended from the frame using wire cable spring isolators to help mitigate unwanted shock and vibration.

42 ENGINEERING↗

Massively Multi-core Acceleration of a Document-Similarity Classifier to Detect Web Attacks

This paper describes our approach to adapting a text document similarity classifier based on the Term Frequency Inverse Document Frequency (TFIDF) metric to two massively multi-core hardware platforms. The TFIDF classifier is used to detect web attacks in HTTP data. In our parallel hardware approaches, we design streaming, real time classifiers by simplifying the sequential algorithm and manipulating the classifier's model to allow decision information to be represented compactly. Parallel implementations on the Tilera 64-core System on Chip and the Xilinx Virtex 5-LX FPGA are presented. For the Tilera, we employ a reduced state machine to recognize dictionary terms without requiring explicit tokenization, and achieve throughput of 37MB/s at slightly reduced accuracy. For the FPGA, we have developed a set of software tools to help automate the process of converting training data to synthesizable hardware and to provide a means of trading off between accuracy and resource utilization. The Xilinx Virtex 5-LX implementation requires 0.2% of the memory used by the original algorithm. At 166MB/s (80X the software) the hardware implementation is able to achieve Gigabit network throughput at the same accuracy as the original algorithm.

97 MATHEMATICS AND COMPUTING↗

Impacts of electrode coating irregularities on polymer electrolyte membrane fuel cell lifetime using quasi in-situ infrared thermography and accelerated stress testing

In-line quality control diagnostics for roll-to-roll (R2R) manufacturing techniques will play a key role in the future commercialization of the polymer electrolyte membrane fuel cell (PEMFC) used in automotive applications. These diagnostics monitor the fabrication of the membrane electrode assembly (MEA), which detect and flag any non-uniformity that may potentially harm PEMFC performance and/or lifetime. This will require quantitative thresholds and a clear distinction between harmful defects and harmless coating irregularities. Thus, novel fuel cell hardware with quasi in-situ infrared (IR) thermography capabilities is utilized to understand how bare spots in the cathode electrode impact MEA lifetime. An accelerated stress test (AST) simulates chemical and mechanical degradation modes seen in vehicular operation. The actual open circuit voltage and rate of change of this voltage are used as in-situ indicators for MEA failure, enabling capture of the progression of failure point development. Bare spot coating irregularities located at the center of the electrode were found to have no impact on MEA lifetime when compared to a pristine MEA. However, MEA lifetime was found to be considerably shortened when these same irregularities are located at the cathode inlet and, especially, the anode inlet regions of the fuel cell.

36 MATERIALS SCIENCE↗

From Cell to System: Accelerated hpc Simulations of BESS Aging under Frequency Regulation and Arbitrage use cases

Lithium-ion battery energy storage systems (BESS) packs have emerged as a leading solution for grid-scale energy storage, enhancing resiliency and balancing load fluctuations. Yet, experimental characterization of large-format LIB packs-particularly to assess performance and degradation over hundreds of cycles - demands substantial hardware investment and multi-year testing campaigns. In this work, we couple a hierarchical, physics-based modeling framework agnostic to electrode chemistries with high-performance computing to accelerate systems level evaluation by upto two orders of magnitude. Building on the open-source liionpack platform, we implement cell, module, and pack-scale electrochemical models enriched with mechanistic aging mechanisms and deploy them on an HPC cluster to simulate 150−200kWh systems over 500 - 1,000 cycles with in days. We subject these virtual B ESS to both constant-current cycling and realistic grid service profiles spanning frequency regulation, ramp-rate support, and energy arbitrage-and quantify the resulting degradation patterns. Our results reveal that localized cell aging can induce substantial nonuniformity at module and pack levels, with service-specific cycling protocols driving distinct aging modes. This rapid, multiscale modeling approach provides a powerful design-space exploration tool for optimizing electrical architecture, control strategies, and operational schedules to prolong pack lifetime and lower total cost of ownership.

Ayalasomayajula, Surya [ORNL] (ORCID:0009000860788↗

Neutralized Drift Compression Experiment (NDCX) - II Quarterly Report

LBNL has received American Recovery and Reinvestment Act (ARRA) funding to construct a new accelerator at Lawrence Berkeley National Laboratory (LBNL) to significantly increase the energy on target, which will allow both the Heavy Ion Fusion (HIF) and Warm Dense Matter (WDM) research communities to explore scientific conditions that have not been available in any other device. For NDCX-II, a new induction linear accelerator (linac) will be constructed at Lawrence Berkeley National Laboratory (LBNL). NDCX-II will produce nano-second long ion beam bunches to hit thin foil targets. The final kinetic energy of the ions arriving at the target varies according to the ion mass. For atomic mass unit of 6 or 7 (Lithium ions), useful kinetic energies range from 1.5 to 5 or more MeV. The expected beam charge in the 1 ns (or shorter) pulse is about 20 nanoCoulombs. The pulse repetition rate will be about once or twice per minute (of course, target considerations will often reduce this rate). Our approach to building the NDCX-II ion accelerator is to make use of the available induction modules and 200 kV pulsers from the retired ATA electron linac at LLNL. Reusing this hardware will maximize the ion energy on target at a minimum cost. Some modification of the cells (e.g., reduce the bore diameter and replace with higher field pulsed solenoids) are needed in order to meet the requirements of this project. The NDCX-II project will include the following tasks: (1) Physics design to determine the required ion current density at the ion source, the injector beam optics, the layout of accelerator cells along the beam line, the voltage waveforms for beam acceleration and compression, the solenoid focusing, the neutralized drift compression and the final focus on target; (2) Engineering design and fabrication of the accelerator components, pulsed power system, diagnostic system, and control and data acquisition system; (3) Conventional facilities; and (4) Installation and integration. The project will be considered completed when the accelerator and pulsed power systems are in place and tested, and we begin beam production and acceleration. The period of performance for this project is July 7, 2009 to Mar 31, 2012, and the total funding, including contingency, is $11.0 M.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

An X-ray Intensity Operations Monitor (AXIOM) (Final LDRD Project Report)

The Saturn accelerator has historically lacked the capability to measure time-resolved spectra for its 3-ring bremsstrahlung x-ray source. This project aimed to create a spectrometer called AXIOM to provide this capability. The project had three major development pillars: hardware, simulation, and unfold code. The hardware consists of a ring of 24 detectors around an existing x-ray pinhole camera. The diagnostic was fielded on two shots at Saturn and over 100 shots at the TriMeV accelerator at Idaho Accelerator Center. A new Saturn x-ray environment simulation was created using measured data to validate. This simulation allows for timeresolved spectra computation to compare the experimental results. The AXIOM-Unfold code is a new parametric unfold code using modern global optimizers and uncertainty quantification. The code was written in Python, uses Gitlab version control and issue tracking, and has been developed with long term code support and maintenance in mind.

43 PARTICLE ACCELERATORS↗

Exploiting first-class arrays in Fortran for accelerator programming

Emerging architectures for high performance computing often are well suited to a data parallel programming model. This paper presents a simple programming methodology based on existing languages and compiler tools that allows programmers to take advantage of these systems. We will work with the array features of Fortran 90 to show how this infrequently exploited, standardized language feature is easily transformed to lower level accelerator code. Our transformations are based on a mapping from Fortran 90 to C++ code with OpenCL extensions. The sheer complexity of programming for clusters of many or multi-core processors with tens of millions threads of execution make the simplicity of the data parallel model attractive. Furthermore, the increasing complexity of todays applications (especially when convolved with the increasing complexity of the hardware) and the need for portability across hardware architectures make a higher-level and simpler programming model like data parallel attractive. The goal of this work has been to exploit source-to-source transformations that allow programmers to develop and maintain programs at a high-level of abstraction, without coding to a specific hardware architecture. Furthermore these transformations allow multiple hardware architectures to be targeted without changing the high-level source. It also removes the necessity for application programmers to understand details of the accelerator architecture or to know OpenCL.

43 PARTICLE ACCELERATORS↗

LCLS-II Gun/Buncher LLRF System Design

For a free electron laser, the stability of injector is critical to the final electron beam parameters, e.g., beam energy, beam arrival time, and eventually it determines the photon quality. The LCLS-II project's injector contains a VHF copper cavity as the gun and a two-cell L-band copper cavity as its buncher. The cavity designs are inherited from the APEX design, but requires more field stability than demonstrated in APEX operation. The gun LLRF system design uses a connectorized RF front end and low noise digitizer, together with the same general purpose FPGA carrier board used in the LCLS-II SRF LLRF system. The buncher LLRF system directly adopts the SRF LLRF chassis design, but programs the controller to run the normal conducting cavities. In this paper, we describe the gun/buncher LLRF system design, including the hardware design, the firmware design and bench test.

43 PARTICLE ACCELERATORS↗

Reimagining Codesign for Advanced Scientific Computing: Report for the ASCR Workshop on Reimagining Codesign

In March 2021, the U.S. Department of Energy’s Advanced Scientific Computing Research program convened the Workshop on Reimagining Codesign. The workshop, also known as ReCoDe, was organized around discussions on eight topic areas: (1) codesign for traditional high-performance computing workloads; (2) codesign of memory/storage systems; (3) codesign of machine learning, neuromorphic, quantum, and other non-von Neumann accelerators; (4) codesign for edge computing and processing at experimental instruments; (5) codesign for security and privacy; (6) hardware design tools and open-source hardware for high-productivity codesign; (7) tools, software stack, and programming languages for high-productivity codesign; and (8) quantitative tools and data collection for modeling and simulation for codesign. The panels identified four Priority Research Directions from these deliberations: (1) breakthrough computing capabilities with targeted heterogeneity and rapid design; (2) software and applications that embrace radical architecture diversity; (3) engineered security and integrity, from transistors to applications; and (4) design with data-rich processes.

97 MATHEMATICS AND COMPUTING↗

Beam and Spin Optics Simulation Tutorials, Using Zgoubi Exercises and Solutions 2 - RHIC Run 22 Spin Transparency Experiment

High electron and ion polarizations are paramount design requirements for the future Electron Ion Collider (EIC) at BNL. A “spin transparency” mode of operation, inherited from the “figure-8” ring design [1], allows preservation and control of electron and ion spin polarization in a collider or storage ring [2, 3]. It makes the ring lattice “invisible” to the spin and allows for polarization control by small quasi-static magnetic fields with practically no effect on beam orbital parameters. Beam polarization can then be fully controlled by small adjustments of snake axis orientation (a few degrees) and snake strengths (a few percent); polarization orientation at IPs can be controlled without spin rotators, spins can be adiabatically flipped without the need of ac magnets. The transparent spin mode has been demonstrated in simulations [2] for the purpose of being tested experimentally during RHIC Run 22. This experiment was aimed at providing data to validate the principles. Yellow RHIC ring (Fig. 1.1) was configured in the transparent spin mode by aligning the axes of its two helical snakes (located respectively in the 3 o’clock and 9 o’clock sextants) parallel to the longitudinal axis. Experiment goals and expected results, ad hoc procedures and hardware requirements, are addressed in Refs. [2, 3].

43 PARTICLE ACCELERATORS↗

Studies of the trigger performance of the ICARUS T600 detector at Fermilab

In recent years, experimental neutrino anomalies were reported: if confirmed, they could hint at the existence of additional sterile neutrino states participating in the mixing phenomenon. The future Short-Baseline Neutrino (SBN) project at Fermilab will pursue sterile neutrino searches, by exploiting three Liquid Argon (LAr) Time Projection Chamber (TPC) detectors located on-axis along the Booster Neutrino Beam (BNB). Aside from SBND and MicroBooNE, ICARUS will be the far detector and will also be interested by the NuMI beam (off-axis). The trigger system is a key component of the detector: it is based on the coincidence of prompt signals from scintillation light in the LAr-TPC (recorded by a system of PhotoMultiplier Tubes, PMTs) with the proton spill extraction of the beam. The present system exploits a majority-based logic and may be complemented by a trigger system based on adder boards, that add the analog signals of the PMTs in groups of 15. Triggering on the sum may help in identifying events closer to the TPC walls, in which there is plenty of light collected by few PMTs and for which the majority condition may not be satisfied. Many tests were carried out to characterize the adder boards, both on the hardware and software sides.

43 PARTICLE ACCELERATORS↗

LUNA: LUT-Based Neural Architecture for Fast and Low-Cost Qubit Readout

Qubit readout is a critical operation in quantum computing systems, which maps the analog response of qubits into discrete classical states. Deep neural networks (DNNs) have recently emerged as a promising solution to improve readout accuracy . Prior hardware implementations of DNN-based readout are resource-intensive and suffer from high inference latency, limiting their practical use in low-latency decoding and quantum error correction (QEC) loops. This paper proposes LUNA, a fast and efficient superconducting qubit readout accelerator that combines low-cost integrator-based preprocessing with Look-Up Table (LUT) based neural networks for classification. The architecture uses simple integrators for dimensionality reduction with minimal hardware overhead, and employs LogicNets (DNNs synthesized into LUT logic) to drastically reduce resource usage while enabling ultra-low-latency inference. We integrate this with a differential evolution based exploration and optimization framework to identify high-quality design points. Our results show up to a 10.95x reduction in area and 30% lower latency with little to no loss in fidelity compared to the state-of-the-art. LUNA enables scalable, low-footprint, and high-speed qubit readout, supporting the development of larger and more reliable quantum computing systems.

Farooq, M. A. [Arizona State U., Tempe]↗

P38 heterogeneous multi-tiled system with support for message queues (MoSAIC) v0.1

The proposed system is written in the hardware description language (HDL) verilog targeting an FPGA board. It is intended as a testbed to explore architecture tradeoffs in multi-tiled heterogeneous architectures. Although we target FPGAs, the system can be implemented as a monolithic SoC or a package comprised of many chiplets that are interconnected in the same package using a NoC. The proposed NoC is lightweight and follows an axi-lite interface. The endpoints of the NoC are a heterogeneous mix of "tiles" as endpoints that are general purpose processors, fixed function accelerators, and programmable accelerators. We assume that the network interfaces for the NoC endpoints are all addressable in a global name-space in that they represent an address range (for memory addresses) or a range of unique identifiers that are associated with each individual tile. This makes the functionality abstract from the standpoint of the NoC design details. Message queues offer a direct inter-processor interface between peer general purpose cores and diverse accelerators that comprise an SoC. Although they share the same NoC infrastructure for inter-tile communication within an SoC or SiP, the hardware message queues bypass the memory hierarchy and thus do not pollute the memory state or invoke the cache coherence mechanism.

Gonzalez, LouisaPatricia↗

Algorithms and data structures to accelerate network analysis

As the sheer amount of computer generated data continues to grow exponentially, new bottlenecks are unveiled that require rethinking our traditional software and hardware architectures. Here in this paper we present five algorithms and data structures (long queue emulation, lockless bimodal queues, tail early dropping, LFN tables, and multiresolution priority queues) designed to optimize the process of analyzing network traffic. We integrated these optimizations on R-Scope, a high performance network appliance that runs the Bro network analyzer, and present benchmarks showcasing performance speed ups of 5X at traffic rates of 10 Gbps.

97 MATHEMATICS AND COMPUTING↗

Neutrino Cross Sections II: Neutrino Boogaloo. UF Task P3, Accelerator-Based Experimental Neutrino Physics. Final Report

We report on progress in a series of different directions within high energy physics research. 1. Neutrino research in hardware and software on the Minerva and MiniBooNE experiments 2. Experimental particle physics at the hadron colliders, with emphasis on research and development and data analysis on the CMS experiment operating at the CERN LHC. This includes research on the discovery and properties on the Higgs Boson. 3. Educational outreach through the Quarknet program, taking physics research into High School classrooms. 4. Theoretical and Phenomenological High Energy research, covering a broad range of activities ranging from fundamental theoretical issues to areas of immediate phenomenological importance. 5. Experiment searches for the Axion, as part of the ADMX experiment.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

Design of a signal conditioner for the Fermilab Magnet Test Facility

This thesis describes the design of a remotely-programmable signal conditioner for the harmonic measurement of accelerator magnets. A 10-channel signal conditioning circuit featuring bucking capabilities was designed from scratch and implemented to the level of the printed circuit board layout. Other system components were chosen from those available on the market. Software design was started with the definition of routine procedures. This thesis is part of an upgrade project for replacing obsolescent automated test equipment belonging to the Fermilab Magnet Test Facility. The design started with a given set of requirements. Using a top-down approach, all the circuits were designed and their expected performances were theoretically predicted and simulated. A limited prototyping phase followed. The printed circuit boards were laid out and routed using a CAD software and focusing the design on maximum electromagnetic interference immunity. An embedded board was selected for controlling and interfacing the signal conditioning circuitry with the instrumentation network. Basic low level routines for hardware access were defined. This work covered the entire design process of the signal conditioner, resulting in a project ready for manufacturing. The expected performances are in line with the requirements and, in the cases where this was not possible, approval of trade-offs was sought and received from the end users. Part I deals with the global structure of the signal conditioner and the subdivision in functional macro-blocks. Part II treats the hardware design phase in detail, covering the analog and digital circuits, the printed circuit layouts, the embedded controller and the power supply selection. Part III deals with the basic hardware-related routines to be implemented in the final software.

43 PARTICLE ACCELERATORS↗

Ginkgo - A math library designed to accelerate Exascale Computing Project science applications

Large-scale simulations require efficient computation across the entire computing hierarchy. A challenge of the Exascale Computing Project (ECP) was to reconcile highly heterogeneous hardware with the myriad of applications that were required to run on these supercomputers. Mathematical software forms the backbone of almost all scientific applications, providing efficient abstractions and operations that are crucial to harness the performance of computing systems. Ginkgo is one such mathematical software library, nurtured by ECP, providing high-performance, user-friendly, and performance portable interfaces for applications in ECP and beyond. In this paper, we elaborate on Ginkgo’s philosophy of high-performance software that is sustainable, reproducible, and easy to use. We showcase the wide feature set of solvers and preconditioners available in Ginkgo and the central concepts involved in their design. We elaborate on four different ECP software integrations: MFEM, PeleLM + SUNDIALS, XGC, and ExaSGD that use Ginkgo to accelerate their science runs. Performance studies of different problems from these applications highlight the effectiveness of Ginkgo and the benefits incurred by these ECP applications.

Cojean, Terry↗

Enhancing a LArTPC with a magnetic field

This paper presents the results of the 2024 Summer Students Italian program at the Fermi National Accelerator Laboratory completed by participant Lorenzo Unich, which lasted between the 27th of July and the 29th of September. This project, which bears the same title of this paper, was supervised by Marco del Tutto, researcher at the Fermi National Accelerator Laboratory. In the first section, we introduce the instruments used throughout the project, detailing their origins and the goals of the traineeship. The second section presents the hardware work conducted, while the final section focuses on the simulations performed for the experimental apparatus. Unless otherwise cited, all pictures and graphs are original.

42 ENGINEERING↗