Search NASA⌕ Search

SEARCH · Search NASA

Results for “Accelerator Design, Technology, Operations”

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 1,135 records · Page 63

PANTHER: A Programmable Architecture for Neural Network Training Harnessing Energy-Efficient ReRAM

The wide adoption of deep neural networks has been accompanied by ever-increasing energy and performance demands due to the expensive nature of training them. Additionally, numerous special-purpose architectures have been proposed to accelerate training: both digital and hybrid digital-analog using resistive RAM (ReRAM) crossbars. ReRAM-based accelerators have demonstrated the effectiveness of ReRAM crossbars at performing matrix-vector multiplication operations that are prevalent in training. However, they still suffer from inefficiency due to the use of serial reads and writes for performing the weight gradient and update step. A few works have demonstrated the possibility of performing outer products in crossbars, which can be used to realize the weight gradient and update step without the use of serial reads and writes. However, these works have been limited to low precision operations which are not sufficient for typical training workloads. Moreover, they have been confined to a limited set of training algorithms for fully-connected layers only. To address these limitations, we propose a bit-slicing technique for enhancing the precision of ReRAM-based outer products, which is substantially different from bit-slicing for matrix-vector multiplication only. We incorporate this technique into a crossbar architecture with three variants catered to different training algorithms. To evaluate our design on different types of layers in neural networks (fully-connected, convolutional, etc.) and training algorithms, we develop PANTHER, an ISA-programmable training accelerator with compiler support. Our design can also be integrated into other accelerators in the literature to enhance their efficiency. Our evaluation shows that PANTHER achieves up to 8.02×, 54.21×, and 103× energy reductions as well as 7.16×, 4.02×, and 16× execution time reductions compared to digital accelerators, ReRAM-based accelerators, and GPUs, respectively.

42 ENGINEERING↗

Advanced light source-Upgrade accumulator-ring gas bremsstrahlung production

Lawrence Berkeley National Laboratory is designing a new electron synchrotron for scientific research using synchrotron radiation. An upgrade of the Advanced Light Source (ALS) would position the nation to retain its international competitiveness in the soft X-ray sciences. The ALS is a 1.9 GeV storage ring operating at 500 mA of beam current. It is optimized to produce intense beams of soft X-rays, which offer spectroscopic contrast, nanometer-scale resolution, and broad temporal sensitivity. Upgrading the ALS storage ring from a triple-bend-achromat to a multibend-achromat (MBA) lattice design will provide a soft X-ray source that is up to 100–1,000 times brighter than today’s ALS and will provide a significantly higher fraction of coherent light in the soft X-ray region than is currently available at the ALS. Further, the upgraded ALS will offer the highest coherent flux of any existing or planned storage-ring facility, worldwide, up to a photon energy of 3.5 keV. The addition of an accumulator ring that enables on-axis, swap-out, recovery, and exchange of bunch trains is required to achieve this coherence and brightness. Analysis of the radiation generated and shielding required by the addition of this new accelerator is presented in this paper.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

Understanding neutrinos with accelerator beams and liquid argon time-projection chambers: ICARUS and DUNE

A global program of experiments has worked towards characterizing neutrino oscillation over the past few decades. However, important parameters remain to be measured, and mysteries remain to be elucidated. Current and upcoming experiments are targeting the open questions and probing the consistency of the neutrino oscillation paradigm. Likewise, the liquid argon (LAr) time-projection chamber (TPC) has emerged as a sensitive particle detection technology for neutrino experiments. A current generation of LAr TPC detectors are being used to study neutrinos while also gaining important experience in operating and analyzing with this technology. SBND, MicroBooNE, and ICARUS have collected or are collecting data from beams at Fermilab (near Chicago) to explore the possibility of a sterile neutrino and/or other beyond Standard Model (BSM) physics. SBND and ICARUS will be used to conduct a two-detector analysis as part of the Short Baseline Neutrino (SBN) Program. Additionally, these detectors are enabling important neutrino interaction studies necessary to prepare for the next generation of oscillation experiments. One such oscillation experiment that will come online over the next years is the Deep Underground Neutrino Experiment (DUNE), which will install multiple 10 kiloton LAr TPCs underground in South Dakota (south of Saskatchewan) to conduct oscillation measurements with neutrinos originating in a beamline at Fermilab. A detector complex will be installed at Fermilab as well, to study the beam before the expected flavour oscillations. This “near detector” will also employ a LAr TPC, with a segmented and pixel-based design, as well as other technologies to constrain uncertainties in the oscillation measurement by characterizing the beam and neutrino interactions. This talk will discuss the ICARUS and DUNE experiments, the LAr TPC detector technology, and the efforts to realize and leverage these experiments to better understand the properties of neutrinos.

Howard, Bruce L. [Fermilab]↗

Design and development of permanent magnet based focusing lens for J-Band Klystron

Applying permanent magnet technology to beam focusing in klystrons can reduce their power consumption and increase their reliability of operation. Electromagnetic design of the beam focusing elements, for high frequency travelling wave tubes, is very critical. The magnitude and profile of the magnetic field need to match the optics requirement from beam dynamics studies. The rise of the field from cathode gun region to the uniform field region (RF section) is important as the desired transition from zero to peak axial field must occur over a short axial distance. Confined flow regime is an optimum choice to minimize beam scalloping but demands an axial magnetic field greater than 2 - 3 times the Brillouin flow field. This necessitates optimization in the magnet design achieve high magnetic field within given spatial constraints. Electromagnetic design and simulations were done using 3D Finite element method (FEM) analysis software. A permanent magnet based focusing lens for a miniature J-Band klystron has been designed and developed at Control Instrumentation Division, BARC. This paper presents the design, simulation studies, beam transmission and RF tests results for J Band klystron with permanent magnet focusing lens. (author)

43 PARTICLE ACCELERATORS↗

A threshold gas Cerenkov detector for the spin asymmetries of the nucleon experiment

In this study, we report on the design, construction, commissioning, and performance of a threshold gas Cerenkov counter in an open configuration, which operates in a high luminosity environment and produces a high photo-electron yield. Part of a unique open geometry detector package known as the Big Electron Telescope Array, this Cerenkov counter served to identify scattered electrons and reject produced pions in an inclusive scattering experiment known as the Spin Asymmetries of the Nucleon Experiment E07-003 at the Thomas Jefferson National Accelerator Facility (TJNAF) also known as Jefferson Lab. The experiment consisted of a measurement of double spin asymmetries A || and A ⊥ of a polarized electron beam impinging on a polarized ammonia target. The Cerenkov counter's performance is characterised by a yield of about 20 photoelectrons per electron or positron track. Thanks to this large number of photoelectrons per track, the Cerenkov counter had enough resolution to identify electron-positron pairs from the conversion of photons resulting mainly from π 0 decays.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

Status and operation of the Linac4 ion source prototypes

CERN's Linac4 45 kV H{sup −} ion sources prototypes are installed at a dedicated ion source test stand and in the Linac4 tunnel. The operation of the pulsed hydrogen injection, RF sustained plasma, and pulsed high voltages are described. The first experimental results of two prototypes relying on 2 MHz RF-plasma heating are presented. The plasma is ignited via capacitive coupling, and sustained by inductive coupling. The light emitted from the plasma is collected by viewports pointing to the plasma chamber wall in the middle of the RF solenoid and to the plasma chamber axis. Preliminary measurements of optical emission spectroscopy and photometry of the plasma have been performed. The design of a cesiated ion source is presented. The volume source has produced a 45 keV H{sup −} beam of 16–22 mA which has successfully been used for the commissioning of the Low Energy Beam Transport (LEBT), Radio Frequency Quadrupole (RFQ) accelerator, and chopper of Linac4.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Controlled Pyroshock Test Transients Can Be Used To Better Match Operational Shock Transients and Reduce Artificial Pyroshock Test Failures [Slides]

Shock analysis is typically done via SRS (shock response spectrum). Oftentimes shock analysis is not performed because most loads specifications only supply SRS requirements with no representative shock transients that can be used for analysis. SRS analysis is too conservative. There is no structural dynamics involved as all modes are assumed to have peak responses simultaneously and in-phase with each other. It is possible that all other loads and combination of loads show the design to be good, whereas SRS analysis (1000 g’s are involved) results may require redesign. Multiple shock transients can satisfy a SRS, yet most loads documents don’t provide a representative shock transient that could be used for shock analysis. Shock testing is less common than harmonic and random vibration testing. Engineers typically go into shock tests with very little knowledge of what to expect. A pass/fail type of approach with very little engineering done to prepare for the test or to control the type of shock transient is used. Most engineers will not model the shock test fixture due to lack of knowledge about the details of the test setup. Shock analysis and testing is often an open loop type situation with no test setup modeling, no test correlation, and no control over the test shock transient. In this work, if we can create applied test force transient that comes close to matching calculated force transient then a more reasonable acceleration transient will be created at the test reference accelerometer that better matches operational shocks.

42 ENGINEERING↗

Safety-Related Instrumentation and Control Pilot Upgrade (Initial Scoping Phase Implementation Report)

This research report (1) describes the process followed and products developed during the SR I&C Pilot Project Initial Scoping Phase, and (2) captures lessons learned. Exelon Generation and LWRS collaborated to develop a Digital Transformation Strategy as part of a larger Advanced Concept of Operations. The proposed SR I&C Pilot Upgrade provides a foundation stone for this Digital Transformation that will improve plant safety, reliability, and operational performance while lowering plant Total Cost of Ownership (TCO). Initial Scoping Phase activities for this Pilot Project have been performed in accordance with industry processes that have been adapted to better support digital upgrades. These processes include IP-ENG-001, Standard Design Process (SDP), NISP-EN-04, Standard Digital Engineering Process (SDEP), and Electric Power Research Institute (EPRI) Report 3002011816, Digital Engineering Guide (DEG). Completing Initial Scoping Phase Engineering and Operations, Licensing, and Project Management Activities was necessary to sufficiently bound the scope, schedule, and estimated cost of the Project to enable utility management to authorize moving into the Conceptual Design Phase. A significant finding of the Business Case Analysis (BCA) methodology developed and applied as part of this effort was that the growth rate of material costs for sustaining the operation of obsolete SR I&C equipment is accelerating. This directly contributed to the Project Economic Analysis created to justify continuing the Project. Project Initial Scoping Phase lessons learned have also been captured to assist the larger industry in understanding the Digital Transformation Strategy and SR I&C Pilot Project Initial Scoping Phase efforts. This is in keeping with the public/private partnership that has been established between the Department of Energy (DOE) and Exelon for this effort with engagement from the NRC. By addressing first-of-a-kind (FOAK) risks and capturing lessons learned, the SR I&C Pilot Upgrade Project addresses technical, regulatory, and business risks to enable subsequent implementers of similar upgrade.

22 GENERAL STUDIES OF NUCLEAR REACTORS↗

Initial Characterization of the NREL Large-Amplitude Motion Platform

The Large Amplitude Motion Platform (LAMP) at NREL represents a significant advancement in the controlled testing of Wave Energy Converters (WECs) under laboratory conditions. Originally designed by E2M as a six-degree-of-freedom (DOF) Stewart platform for flight simulation, LAMP has been adapted by NREL to facilitate the mounting and evaluation of WECs. This adaptation enables dry testing of WECs using motion profiles similar to the ocean, facilitating the iterative design, testing, and validation of WEC performance prior to ocean deployments. This report presents the initial work completed to characterize LAMP, with particular emphasis on its stability and operational capabilities across various single and multi-degree-of-freedom (DOF) motion profiles. The report includes planned comparisons at three distinct mass payloads, aimed at assessing the platform's positional accuracy, frequency response, and endurance over extended runtime periods. These experimental tests are critical for establishing the platform's limitations and ensuring that the data generated during WEC validation is both accurate and reproducible. The outcomes of this study not only contribute to a deeper understanding of LAMP's capabilities but also lay the groundwork for future advancements in WEC testing methodologies. By providing robust and reliable performance data within a controlled laboratory setting, the findings are expected to significantly enhance the development and commercialization of marine energy technologies. This report presents the initial findings of the LAMP Characterization work and proposed steps to further understand and characterize LAMP. Data collected during this work can be found on MHKDR at: https://mhkdr.openei.org/submissions/602 Note that this report shares the measured/found instantaneous maximum operating range of LAMP. For most applications the maximum operating range cannot be used for system health and longevity. The operating range and capabilities of LAMP will be evaluated on a case-by-case basis, single-DOF position, velocity, and acceleration values presented in Table 4a-c, and Table 5 should be taken as instantaneous absolute maximum values. Future use of LAMP will likely be limited to smaller values.

16 TIDAL AND WAVE POWER↗

Research Needs and Impacts in Predictive Simulation for Internal Combustion Engines (PreSICE)

This report is based on a SC/EERE Workshop to Identify Research Needs and Impacts in Predictive Simulation for Internal Combustion Engines (PreSICE), held March 3, 2011, to determine strategic focus areas that will accelerate innovation in engine design to meet national goals in transportation efficiency. The U.S. has reached a pivotal moment when pressures of energy security, climate change, and economic competitiveness converge. Oil prices remain volatile and have exceeded $100 per barrel twice in five years. At these prices, the U.S. spends $1 billion per day on imported oil to meet our energy demands. Because the transportation sector accounts for two-thirds of our petroleum use, energy security is deeply entangled with our transportation needs. At the same time, transportation produces one-quarter of the nation’s carbon dioxide output. Increasing the efficiency of internal combustion engines is a technologically proven and cost-effective approach to dramatically improving the fuel economy of the nation’s fleet of vehicles in the near- to mid-term, with the corresponding benefits of reducing our dependence on foreign oil and reducing carbon emissions. Because of their relatively low cost, high performance, and ability to utilize renewable fuels, internal combustion engines—including those in hybrid vehicles—will continue to be critical to our transportation infrastructure for decades. Achievable advances in engine technology can improve the fuel economy of automobiles by over 50% and trucks by over 30%. Achieving these goals will require the transportation sector to compress its product development cycle for cleaner, more efficient engine technologies by 50% while simultaneously exploring innovative design space. Concurrently, fuels will also be evolving, adding another layer of complexity and further highlighting the need for efficient product development cycles. Current design processes, using “build and test” prototype engineering, will not suffice. Current market penetration of new engine technologies is simply too slow—it must be dramatically accelerated. These challenges present a unique opportunity to marshal U.S. leadership in science-based simulation to develop predictive computational design tools for use by the transportation industry. The use of predictive simulation tools for enhancing combustion engine performance will shrink engine development timescales, accelerate time to market, and reduce development costs, while ensuring the timely achievement of energy security and emissions targets and enhancing U.S. industrial competitiveness. In 2007 Cummins achieved a milestone in engine design by bringing a diesel engine to market solely with computer modeling and analysis tools. The only testing was after the fact to confirm performance. Cummins achieved a reduction in development time and cost. As important, they realized a more robust design, improved fuel economy, and met all environmental and customer constraints. This important first step demonstrates the potential for computational engine design. But, the daunting complexity of engine combustion and the revolutionary increases in efficiency needed require the development of simulation codes and computation platforms far more advanced than those available today. Based on these needs, a Workshop to Identify Research Needs and Impacts in Predictive Simulation for Internal Combustion Engines (PreSICE) convened over 60 U.S. leaders in the engine combustion field from industry, academia, and national laboratories to focus on two critical areas of advanced simulation, as identified by the U.S. automotive and engine industries. First, modern engines require precise control of the injection of a broad variety of fuels that is far more subtle than achievable to date and that can be obtained only through predictive modeling and simulation. Second, the simulation, understanding, and control of these stochastic in-cylinder combustion processes lie on the critical path to realizing more efficient engines with greater power density. Fuel sprays set the initial conditions for combustion in essentially all future transportation engines; yet today designers primarily use empirical methods that limit the efficiency achievable. Three primary spray topics were identified as focus areas in the workshop: The fuel delivery system, which includes fuel manifolds and internal injector flow, The multi-phase fuel–air mixing in the combustion chamber of the engine, and The heat transfer and fluid interactions with cylinder walls. Current understanding and modeling capability of stochastic processes in engines remains limited and prevents designers from achieving significantly higher fuel economy. To improve this situation, the workshop participants identified three focus areas for stochastic processes: Improve fundamental understanding that will help to establish and characterize the physical causes of stochastic events, Develop physics-based simulation models that are accurate and sensitive enough to capture performance-limiting variability, and Quantify and manage uncertainty in model parameters and boundary conditions. Improved models and understanding in these areas will allow designers to develop engines with reduced design margins and that operate reliably in more efficient regimes. All of these areas require improved basic understanding, high-fidelity model development, and rigorous model validation. These advances will greatly reduce the uncertainties in current models and improve understanding of sprays and fuel–air mixture preparation that limit the investigation and development of advanced combustion technologies. The two strategic focus areas have distinctive characteristics but are inherently coupled. Coordinated activities in basic experiments, fundamental simulations, and engineering-level model development and validation can be used to successfully address all of the topics identified in the PreSICE workshop. The outcome will be: New and deeper understanding of the relevant fundamental physical and chemical processes in advanced combustion technologies, Implementation of this understanding into models and simulation tools appropriate for both exploration and design, and Sufficient validation with uncertainty quantification to provide confidence in the simulation results. These outcomes will provide the design tools for industry to reduce development time by up to 30% and improve engine efficiencies by 30% to 50%. The improved efficiencies applied to the national mix of transportation applications have the potential to save over 5 million barrels of oil per day, a current cost savings of $500 million per day.

33 ADVANCED PROPULSION SYSTEMS↗

Utilization of the CW magnetrons as coherent RF sources for superconducting RF accelerators

CW magnetrons designed and optimized for industrial heaters, driven by an injection-locking signal, were suggested to power Superconducting RF (SRF) cavities. However, the CW magnetrons, which are regenerative generators, are intended to operate in self-excitation mode, so, for reliable starting they use a large regenerative gain, causing a large regenerative instability in operation. Thus, power supply ripple, even shot noise, etc., lead to multiple sidebands forming a quasi-continuous noise spectrum at the carrier frequency. Therefore, in the traditionally used mode of operation and control, such magnetrons are inapplicable for powering high Q-factor SRF cavities. A new approach to operation and control of CW magnetrons considering non-stationary processes during start-up and operation makes it possible to find a mode of almost coherent RF generation of tubes with a significant reduction in regenerative instability and noise and increased efficiency. In our work is presented substantiation for the newly developed mode with an analysis of the new approach based on experimental results. •In this article, for the first time, non-stationary, regenerative processes are considered during the start-up and operation of CW magnetrons for powering high-Q superconducting RF cavities.•An analytical study, confirmed by experiments, showed that the traditionally used modes of starting and operating CW magnetrons are not suitable for powering high-Q superconducting RF cavities.•In fact, to make CW magnetrons suitable for such cavities, the physics of starting and operating CW magnetrons must be changed. This has been demonstrated in our work.•In developing the Stimulated Coherent Generation Mode [NIM A 980 (2020) 164366], we have shown the suitability of this mode for starting and operating CW magnetrons for high-Q superconducting RF cavities.•Analytical analysis, verified experimentally, substantiates in our article the suitability of the “Mode of stimulated coherent generation” of CW magnetrons for powering high-Q superconducting cavities in pulsed and continuous modes.

43 PARTICLE ACCELERATORS↗

Enhanced life ion source for germanium and carbon ion implantation

Germanium and carbon ions represent a significant portion of total ion implantation steps in the process flow. Very often ion source materials that used to produce ions are chemically aggressive, especially at higher temperatures, and result in fast ion source performance degradation and a very limited lifetime [B.S. Freer, et. al., 2002 14th Intl. Conf. on Ion Implantation Technology Proc, IEEE Conf. Proc., p. 420 (2003)]. GeF{sub 4} and CO{sub 2} are commonly used to generate germanium and carbon beams. In the case of GeF{sub 4} controlling the tungsten deposition due to the de-composition of WF{sub 6} (halogen cycle) is critical to ion source life. With CO{sub 2}, the materials oxidation and carbon deposition must be controlled as both will affect cathode thermionic emission and anti-cathode (repeller) efficiencies due to the formation of volatile metal oxides. The improved ion source design Extended Life Source 3 (Eterna ELS3) together with its proprietary co-gas material implementation has demonstrated >300 hours of stable continuous operation when using carbon and germanium ion beams. Optimizing cogas chemistries retard the cathode erosion rate for germanium and carbon minimizes the adverse effects of oxygen when reducing gas is introduced for carbon. The proprietary combination of hardware and co-gas has improved source stability and the results of the hardware and co-gas development are discussed.

36 MATERIALS SCIENCE↗

Microscopic Measurement: Need for Nuclear Material Sample Specimen Characterization Capability

The Surveillance, Fabrication, and Off-Site Operations team at Lawrence Livermore National Laboratory is beginning an update to the nuclear material machining lab in Superblock. Plutonium and other nuclear materials tend to accelerate the aging process for equipment and machines that would otherwise be maintainable for long lives. They have their lifespans shortened from deteriorated seals, embrittled polymer components, accelerated corrosion, and damage to electronics from exposure to radioactive materials. The cost of installation of equipment into a glove box (GB) tends to outstrip the cost of the equipment itself. Due to the magnification of cost, the historical approach to equipment installs has been to purchase and install the highest-quality and most robust equipment available to maximize time between failures and updates. To make room for the installation of a new lathe, the inspection GB is being moved to a different location within the machining lab. As part of the overall effort, the equipment within the inspection GB is being revisited. Current measurement capabilities are antiquated and starting to become problematic due to age. Advances in metrology are sought for more comprehensive characterization of scientific samples. This document aims to identify several options for the replacement of current measurement equipment and identify a system for purchase and eventual installation. The intention is not to select the cheapest system, but rather to identify a system that best meets the needs of both the operators and principal investigators (PI). Following the convention of the SYSM-5620 Design Thinking and System Engineering, it was determined that a laser confocal measurement system from Keyence is the most suitable option for our programmatic need. Form factor, capability, and maintainability were considered. Along with stakeholder input, to ensure the best decision for the Nuclear Material Technology Program.

36 MATERIALS SCIENCE↗

Assessment of the Viability of Scaled Annular Pellet Fabrication Technologies

The DOE-NE AFC program broadly supports the creation of novel fuel forms including intermetallics, composites, and dopants for high uranium density compounds. While data can be obtained for key performance indicators such as thermal expansion and oxidation resistance in steam, fuel qualification remains a time intensive process. As such, accelerated fuel qualification methods are a recent focus of the program. By first looking at systems with known historical irradiation testing information, like UO 2 , the method of accelerated burnup testing can be validated, followed by use on novel materials with less information available, like silicides and nitrides. The biggest time delay in doing irradiation testing is reaching a “high burnup” fuel, to effectively model how the fuel will behave when it has been fully used. Reaching this point conventionally could take upwards of 30 years. Accelerated burnup would be approaching the threshold of “high burnup”, something on the order or 50GWd/MTU, more quickly, which although achievable in test reactors requires a decade to achieve using current irradiation testing schemes. To achieve this faster rate, the fuel is likely going to be subject to thermal effects commiserate with the increased heat by more fission. In a normal fuel pellet, this increased temperature would operate less efficiently because the thermal diffusivity of the pellet only allows it to transfer heat so quickly, thus causing a high centerline temperature and noticeable stress along the diameter of the pellet. Solutions could include smaller pellets, which would have a greater surface area giving off heat relative to their volume, or “shaped” fuel designs modifying the thermal behavior. The concept that drives the annular pellet design is reduction of the thermal gradient caused from the high “centerline” temperature at the center of a normal pellet to its surface. By removing the center of the pellet, the new centerline will instead be closer to an edge (whether that be the inner or outer edge), allowing cooling to wick away more heat than the traditional design. In an ideal world, the center of the pellet could also have a heat wick running through it allowing heat to be transported away from throughout the pellet. This increase in heat removal would allow the pellet to be subjected to higher energy output, and thus accelerate the burn-up process. UO 2 has been the staple fuel form of nuclear energy for decades, and replacement of this functionally understood material can only be done with sufficient fuel qualification. Increasing the burnup allows us to characterize new fuels with limited to no operational experience behind them, including carbides, nitrides, and silicide’s. Production of the unique geometries detailed herein will allow us to cut qualification times from multi-decadal experiments into the sub-decadal range, vastly increasing the quantity and quality of information available.

11 NUCLEAR FUEL CYCLE AND FUEL MATERIALS↗

Even Higher-Level Synthesis: An Exploration of AI Hardware Accelerators using HLS4ML

With the rise of artificial intelligence, the popularization of deep learning, and a constantly evolving industry, the demand for flexible and efficient tools has never been greater. As algorithms grow more complex, their runtime and energy consumption increase exponentially. Customized hardware accelerators, long used for specific mathematical operations, remain essential for managing modern applications' computational and power demands. Hardware accelerators can speed up complex computations by orders of magnitude, but their manual design and verification processes are often challenging and time-consuming. High-Level Synthesis (HLS) provides a solution by transforming high-level algorithm descriptions, typically written in C++ or SystemC, into synthesizable RTL suitable for hardware implementation. This approach reduces development time for RTL engineers while offering flexibility beyond what traditional handwritten RTL can provide. We extended this capability to the machine-learning domain with the open-source framework hls4ml, which allows neural networks trained in Python frameworks like Tensorflow or PyTorch to be synthesized into efficient hardware representations for the traditional FPGA and ASIC flows. This breakthrough addresses the growing need for reduced design turnaround and easy verification of ML hardware accelerators with low latency and power efficiency constraints. During this tutorial, we will demonstrate how Python complements HLS by simplifying the ML design process, bridging the gap between software and hardware development. Attendees will explore how we translate neural networks modeled in Python into fixed-point C++ models suitable for HLS workflows. We will dive into strategies like Value-Range Analysis and Quantization-Aware Training, which optimize these designs for deployment and evaluate their accuracy, power consumption, and energy efficiency. To exemplify these concepts, experts from Fermilab will share their experiences applying this technology to high-energy physics experiments, where real-time, low-latency processing is critical. Over the years, Fermilab engineers have demonstrated how deep neural networks, optimized for hardware using hls4ml, can meet the stringent requirements of trigger systems at the CERN Large Hadron Collider. These systems rely on rapid decision-making to process immense data volumes while retaining only the most relevant events for further analysis. The application of hls4ml has also been extended to innovative technologies like smart pixel arrays. These smart pixels integrate ML inference capabilities directly into sensor devices, enabling localized data processing at the pixel level. This approach drastically reduces the need to transmit raw data to external processing units, significantly decreasing power consumption and latency. By embedding neural networks within the pixel architecture, the smart pixels can identify and prioritize relevant data in real time, providing a highly efficient solution for edge computing in scenarios such as particle detectors and imaging systems. Fermilab's work highlights the potential of hardware-accelerated ML in scenarios where both speed and power efficiency are mission-critical. Through this tutorial, attendees will gain valuable insights into the challenges and solutions of deploying ML in hardware. Understanding how HLS and hls4ml streamline the development of neural network-based hardware accelerators is fundamental for the industry's future. Participants will learn how these technologies are shaping the future of AI and scientific computing.

Di Guglielmo, Giuseppe [Fermilab]↗

Assessment of Laser-Driven Pulsed Neutron Sources for Poolside Neutron-based Advanced NDE – A Pathway to LANSCE-like Characterization at INL

A variety of opportunities for characterization of fresh nuclear fuels using thermal (~25meV) and epithermal (~10eV) neutrons have been documented at Los Alamos National Laboratory. They include spatially resolved non-destructive characterization of features, isotopic enrichment, chemical heterogeneity and stoichiometry. The LANSCE spallation neutron source is well suited in neutron fluence and temporal characteristics for studies of fuels. However, recent advances in high power short pulse lasers suggest that compact neutron sources might, over the next decade, become viable at a price point that would permit their consideration for poolside characterization on site at irradiation facilities. In a laser-driven neutron source the laser is used to accelerate deuterium ions into a beryllium target where neutrons are produced. At this time, the technology is new and their total neutron production is approximately four orders of magnitude less than a facility like LANSCE. However, recent measurements on a sub-optimized system demonstrated >10 10 neutrons in sub-nanosecond pulses in predominantly forward direction. The compactness of the target system compared to a spallation target may allow exchanging the target during a measurement to e.g. characterize a highly radioactive sample with thermal, epithermal, and fast neutrons as well as hard X-rays, thus avoiding sample handling. At this time several groups are working on laser-driven neutron production and are advancing concepts for lasers, laser targets, and optimized neutron target/moderator systems. Advances in performance sufficient to enable poolside fuels characterization with LANSCE-like fluence on sample within a decade may be possible. This report describes the underlying physics and state-of-the-art of the laser-driven neutron production process from the perspective of the DOE/NE mission. It also discusses the development and understanding that will be necessary to provide customized capability for characterization of irradiated fuels. Potential operational advantages compared to a spallation neutron source include reduced shielding complexity, reduced energy requirements, and a production target free of fission products. Contributors to this report include experts in laser-driven neutron production (Roth, Fernandez), laser design (Haefner, Siders, Leemans), laser target design (Glenzer), spallation target/moderator design (Mocko), neutron instrumentation and characterization applications (Vogel, Bourke).

11 NUCLEAR FUEL CYCLE AND FUEL MATERIALS↗

The International Large Detector: Letter of Intent

The International Large Detector (ILD) is a concept for a detector at the International Linear Collider, ILC. The ILC will collide electrons and positrons at energies of initially 500 GeV, upgradeable to 1 TeV. The ILC has an ambitious physics program, which will extend and complement that of the Large Hadron Collider (LHC). The ILC physics case has been well documented, most recently in the ILC Reference Design Report, RDR. A hallmark of physics at the ILC is precision. The clean initial state and the comparatively benign environment of a lepton collider are ideally suited to high precision measurements. To take full advantage of the physics potential of ILC places great demands on the detector performance. The design of ILD, which is based on the GLD and the LDC detector concepts, is driven by these requirements. Excellent calorimetry and tracking are combined to obtain the best possible overall event reconstruction, including the capability to reconstruct individual particles within jets for particle flow calorimetry. This requires excellent spatial resolution for all detector systems. A highly granular calorimeter system is combined with a central tracker which stresses redundancy and efficiency. In addition, efficient reconstruction of secondary vertices and excellent momentum resolution for charged particles are essential for an ILC detector. The interaction region of the ILC is designed to host two detectors, which can be moved into the beam position with a 'push-pull' scheme. The mechanical design of ILD and the overall integration of subdetectors takes these operational conditions into account. The main features of ILD are outlined below. The central component of the ILD tracker is a Time Projection Chamber (TPC) which provides up to 224 precise measurements along the track of a charged particle. This is supplemented by a system of Silicon (Si) based tracking detectors, which provide additional measurement points inside and outside of the TPC, and extend the angular coverage down to very small angles. A Si-pixel based vertex detector (VTX) enables long lived particles such as b- and c-hadrons to be reconstructed. This combination of tracking devices, which has a large degree of redundancy, results in high track reconstruction efficiencies, and unprecedented momentum resolution and vertex reconstruction capabilities. One of the most direct measures of detector performance at the ILC is the jet-energy resolution. Precise di-jet mass reconstruction and separation of hadronically decaying W and Z bosons are essential for many physics channels. The ultimate jet energy resolution is achieved when every particle in the event, charged and neutral, is measured with the best possible precision. Within the paradigm of particle flow calorimetry, this goal is achieved by reconstructing charged particles in the tracker, photons in the electromagnetic calorimeter (ECAL), and neutral hadrons in the ECAL and hadronic calorimeter (HCAL). The ultimate performance is reached for perfect separation of charged-particle clusters from neutral particle clusters in the calorimeters. Thus, a highly granular calorimeter outside the tracker is the second key component of ILD. Sampling calorimeters with dense absorber material and fine grained readout are used. A tungsten absorber based electromagnetic calorimeter (ECAL) covers the first interaction length, followed by a somewhat coarser steel based sampling hadronic calorimeter (HCAL). Several ECAL and HCAL readout technologies are being pursued.

43 PARTICLE ACCELERATORS↗

Ferroelectric capacitors and field-effect transistors as in-memory computing elements for machine learning workloads

Abstract This study discusses the feasibility of Ferroelectric Capacitors (FeCaps) and Ferroelectric Field-Effect Transistors (FeFETs) as In-Memory Computing (IMC) elements to accelerate machine learning (ML) workloads. We conducted an exploration of device fabrication and proposed system-algorithm co-design to boost performance. A novel FeCap device, incorporating an interfacial layer (IL) and$$\text {Hf}_{0.5}\text {Zr}_{0.5}\text {O}_2$$ Hf 0.5 Zr 0.5 O 2 (HZO), ensures a reduction in operating voltage and enhances HZO scaling while being compatible with CMOS circuits. The IL also enriches ferroelectricity and retention properties. When integrated into crossbar arrays, FeCaps and FeFETs demonstrate their effectiveness as IMC components, eliminating sneak paths and enabling selector-less operation, leading to notable improvements in energy efficiency and area utilization. However, it is worth noting that limited capacitance ratios in FeCaps introduced errors in multiply-and-accumulate (MAC) computations. The proposed co-design approach helps in mitigating these errors and achieves high accuracy in classifying the CIFAR-10 dataset, elevating it from a baseline of 10% to 81.7%. FeFETs in crossbars, with a higher on-off ratio, outperform FeCaps, and our proposed charge-based sensing scheme achieved at least an order of magnitude reduction in power consumption, compared to prevalent current-based methods.

Science & Technology - Other Topics↗