Search NASA⌕ Search

SEARCH · Search NASA

Results for “ACCELERATION - MEASUREMENTS”

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 91 records · Page 5
All sourcesNASA NTRSDOE OSTI
DOE OSTI2020

FPDeep: Scalable Acceleration of CNN Training on Deeply-Pipelined FPGA Clusters

In this paper, we propose a framework called FPDeep, which uses a hybrid of model and layer paral- lelism to configure distributed reconfigurable clusters to train DNNs. This approach has numerous benefits. First, the design does not suffer from batch size growth. Second, novel workload and weight partitioning leads to balanced loads of both among nodes. And third, the entire system is fine-grained pipeline. This leads to high parallelism and utilization and also minimizes the time features need to be cached while waiting for back-propagation.

Wang, Tianqi↗
DOE OSTI2023

Scalable Incremental Checkpointing using GPU-Accelerated De-Duplication

Writing large amounts of data concurrently to stable storage is a typical I/O pattern of many HPC workflows. This pattern introduces high I/O overheads and results in increased storage space utilization especially for workflows that need to capture the evolution of data structures with high frequency as checkpoints. In this context, many applications, such as graph pattern matching, perform sparse updates to large data structures between checkpoints. For these applications, incremental checkpointing techniques that save only the differences from one checkpoint to another can dramatically reduce the checkpoint sizes, I/O bottlenecks, and storage space utilization. However, such techniques are not without challenges: it is non-trivial to transparently determine what data has changed since a previous checkpoint and assemble the differences in a compact fashion that does not result in excessive metadata. State-of-art data reduction techniques (e.g., compression and de-duplication) have significant limitations when applied to modern HPC applications that leverage GPUs: slow at detecting the differences, generate a large amount of metadata to keep track of the differences, and ignore crucial spatiotemporal checkpoint data redundancy. This paper addresses these challenges by proposing a Merkle tree-based incremental checkpointing method to exploit GPUs' high memory bandwidth and massive parallelism. Experimental results at scale show a significant reduction of the I/O overhead and space utilization of checkpointing compared with state-of-the-art incremental checkpointing and compression techniques.

Tan, Nigel↗
DOE OSTI2025

Novel Concepts for High Gradient Acceleration (Final Technical Report)

We have conducted an intensive, pioneering program to demonstrate novel concepts for achieving high gradient acceleration at frequencies from the conventional microwave bands up to the millimeter wave /THz bands. High gradient accelerators hold the promise of smaller and less costly accelerators for applications ranging from the largest scale accelerators used for discovery science down to the smallest accelerators used for industrial, homeland security and medical applications. The research consisted of two major research thrusts: 1.) Structure-based wakefield accelerator (SWFA) research in collaboration with the Argonne Wakefield Accelerator research group and 2.) Millimeter Wave / THz high gradient acceleration in collaboration with SLAC. The specific goals of the research program were: Design novel metallic metamaterial structures that increase the beam-wave coupling for the accelerator mode and reduce the effect of high order modes; test novel metamaterial structures to achieve higher output power, > 1 GW at X-Band (11.7 GHz), in test at the Argonne Wakefield Accelerator (AWA); determine the experimental breakdown threshold for nanosecond-scale pulses at X-Band in testing at the AWA; test a 110 GHz accelerator structure with a field emission electron gun, built at SLAC, using pulses from a 1 MW, 110 GHz gyrotron; design, build and test a 110 GHz quasi-optical, resonant-ring pulse compressor to compress microsecond pulses from the 1 MW gyrotron into > 20 MW, 5 ns output pulses for accelerator structure testing. The proposed research program built on our successes in our research program including: Generation of 510 MW, 2.1 ns (FWHM) pulses at 11.7 GHz from a metallic metamaterial structure in test at the Argonne Wakefield Accelerator using a train of eight 65 MeV electron bunches spaced at 1.3 GHz with a total charge of 280 nC. The metamaterial structure consisted of 100 copper unit cells each consisting of a “wagon-wheel” plate and a spacer plate with a total structure length of 0.2 m. The 510 MW pulse generated an on-axis wakefield of 130 MV/m that could be used to accelerate a trailing witness bunch. Demonstration of coupling of an unprecedented rf power level of 575kW into a 110 GHz accelerator structure using a quasi-optical setup. The standing structure consisted of a central copper cavity located between two matching cavities fed by a TM01 mode. The 6 ns input pulses were sliced from 3 microsecond pulses from the gyrotron using a laser-driven silicon switch. We obtained an unprecedented high gradient up to 230MV/m corresponding to a peak surface electric field of more than 520 MV/m.

43 PARTICLE ACCELERATORS↗
DOE OSTI2022

Challenges of Future Accelerators for Particle Physics Research

For over half a century, high-energy particle accelerators have been a major enabling technology for particle and nuclear physics research as well as sources of X-rays for photon science research in material science, chemistry and biology. Particle accelerators for energy and intensity Frontier research in particle and nuclear physics continuously push the accelerator community to invent ways to increase the energy and improve the performance of accelerators, reduce their cost, and make them more power efficient. The accelerator community has demonstrated imagination and creativity in developing a plethora of future accelerator ideas and proposals. The technical maturity of the proposed facilities ranges from shovel-ready to those that are still largely conceptual. At this time, over 100 contributed papers have been submitted to the Accelerator Frontier of the US particle physics decadal community planning exercise known as Snowmass’2021. These papers cover a broad spectrum of topics: beam physics and accelerator education, accelerators for neutrinos, colliders for Electroweak/Higgs studies and multi-TeV energies, accelerators for Physics Beyond Colliders and rare processes, advanced accelerator concepts, and accelerator technology for Radio Frequency cavities (RF), magnets, targets and sources. This paper provides an overview of the present state of accelerators for particle physics and gives a brief description of some of the major facilities that have been proposed, their perceived advantages and some of the remaining challenges.

43 PARTICLE ACCELERATORS↗
DOE OSTI2024

AXI4MLIR: User-Driven Automatic Host Code Generation for Custom AXI-Based Accelerators

Tensor algebra operations represent an important class of algorithms used across many applications, including machine learning, scientific computing, and data analytics. As a result, the efficient generation of custom accelerators for tensor operations has received increased attention. Previous efforts have produced automated tools enabling users to prototype and explore optimized accelerators. However, little effort has been focused on the host-accelerator interaction in these tools. Efficient use of hardware accelerators requires knowledge about the accelerator's capabilities (operations, data formats, and opcode support), the host CPU microarchitecture (e.g., memory hierarchy), the host-accelerator interface, and the application's features (which code regions should be mapped onto an accelerator). Manually rewriting the original applications to facilitate improved custom accelerator mapping is an error-prone and time-consuming endeavor. To cope with this, we propose AXI4MLIR, a new framework to automatically generate and optimize the communication between the host CPU and arbitrary accelerators that implement linear algebra algorithms. AXI4MLIR extends the MLIR compiler framework to automatically generate efficient host-accelerator driver code for accelerators with AXI-based interfaces. Our compiler extensions enable automatic driver code generation while carefully considering the host's memory hierarchy and target accelerator features. To demonstrate the flexibility and utility of AXI4MLIR, we test it with diverse use cases that include different types of accelerators, tiling scenarios, and dataflow schemes. We compare our experimental results to manual implementations of host-accelerator driver code and find that our approach can reduce CPU cache references by 56% and deliver up to a 1.65x speedup.

Bohm Agostini, Nicolas↗
DOE OSTI2026

Progress in the development of the community Particle Accelerator Lattice Standard (PALS)

The Particle Accelerator Lattice Standard (PALS) is a community effort to create an open standard to promote lattice information exchange for particle accelerators. PALS development is a community-wide international effort involving accelerator physicists from multiple institutions. While it started as a lattice standard for beam dynamics simulations, it is now being extended to support other particle accelerator activities, in particular accelerator operation. With new accelerators that are becoming more complex, larger collaborations and the increasing imprint of artificial intelligence in all accelerator activities (from design to operation to workforce development), the imperative for a common, standardized accelerator ontology has been transitioning from “nice-to-have” to “must-have”. We will present the status of the project, its relations to other projects, including to two of the particle accelerator projects of the newly announced US DOE Genesis Mission: the Multi-Office Accelerator Team (MOAT) project and the Nuclear physics AI-Ready Accelerator Data (NARAD) project.

Brynes, A. [Science and Technology Facilities Coun↗
DOE OSTI2023

Linear colliders based on laser-plasma accelerators

Laser-plasma accelerators are capable of sustaining accelerating fields of 10–100 GeV/m, 100–1000 times that of conventional technology and the highest fields produced by any of the widely researched advanced accelerator concepts. Laser-plasma accelerators also intrinsically accelerate short particle bunches, several orders of magnitude shorter than that of conventional technology, which leads to reductions in beamstrahlung and, hence, savings in the overall power consumption to reach a desired luminosity. These properties make laser-plasma accelerators a promising accelerator technology for a more compact, less expensive high-energy linear collider providing multi-TeV polarized leptons. In this submission to the Snowmass 2021 Accelerator Frontier, we discuss the motivation for a laser-plasma-accelerator-based linear collider, the status of the field, and potential linear collider concepts up to 15 TeV. Herein, we outline the research and development path toward a collider based on laser-plasma accelerator technology, and highlight near-term and mid-term applications of this technology on the collider development path. The required experimental facilities to carry out this research are described. We conclude with community recommendations developed during Snowmass.

43 PARTICLE ACCELERATORS↗
DOE OSTI2023

High average power ultrafast laser technologies for driving future advanced accelerators

Large scale laser facilities are needed to advance the energy frontier in high energy physics and accelerator physics. Laser plasma accelerators are core to advanced accelerator concepts aimed at reaching TeV electron electron colliders. In these facilities, intense laser pulses drive plasmas and are used to accelerate electrons to high energies in remarkably short distances. A laser plasma accelerator could in principle reach high energies with an accelerating length that is 1000 times shorter than in conventional RF based accelerators. Notionally, laser driven particle beam energies could scale beyond state of the art conventional accelerators. LPAs have produced multi GeV electron beams in about 20 cm with relative energy spread of about 2 percent, supported by highly developed laser technology. This validates key elements of the US DOE strategy for such accelerators to enable future colliders but extending best results to date to a TeV collider will require lasers with higher average power. While the per pulse energies envisioned for laser driven colliders are achievable with current lasers, low laser repetition rates limit potential collider luminosity. Applications will require rates of kHz to tens of kHz at Joules of energy and high efficiency, and a collider would require about 100 such stages, a leap from current Hz class LPAs. This represents a challenging 1000 fold increase in laser repetition rates beyond current state of the art. Here, this whitepaper describes current research and outlook for candidate laser systems as well as the accompanying broadband and high damage threshold optics needed for driving future advanced accelerators.

47 OTHER INSTRUMENTATION↗
DOE OSTI2022

Exploring the Use of Novel Spatial Accelerators in Scientific Applications

Driven by the need to find alternative accelerators which can viably replace GPUs in next-generation Supercomputing systems, this paper proposes a methodology to enable agile application/hardware co-design. The application-first methodology provides the ability to come up with design of accelerators while working with real-world workloads, available accelerators, and system software. The iterative design process targets a set of kernels in a workload for performance estimates that can prune the design space for later phases of detailed architectural evaluations. To this effect, in this paper, a novel data-parallel device model is introduced that simulates the latency of performance-sensitive operations in an accelerator including data transfers and kernel computation using multi-core CPUs. The use of off-the-shelf simulators, such as pre-RTL simulator Aladdin or multiple tools available for exploring the design of deep neural network accelerators (e.g., Timeloop) is demonstrated for evaluation of various accelerator designs using applications with realistic inputs. Examples of multiple device configurations that are instantiable in a system are explored to evaluate the performance benefit of deploying novel accelerators. The proposed device is integrated with a programming model and system software to potentially explore the impacts of high-level programming languages/compilers and low-level effects such as task scheduling on multiple accelerators. We analyze our methodology for a set of applications that represent high-performance computing (HPC) and graph analytics. The applications include a computational chemistry kernel realized using tensor contractions, triangle counting, GraphSAGE and Breadth-first Search. These applications include kernels such as dense matrix-dense matrix multiplication, sparse matrix-spare matrix multiplication, and sparse matrix-dense vector multiplication. Our results indicate potential performance benefits and insights for system design by including accelerators that realize these kernels along-side general purpose accelerators.

AI, codesign, Accelerated Computing, Modeling and ↗
DOE OSTI2020

Compact, Achromatic Non-scaling FFAG Accelerator for HEP, Commercial and Medical Applications (CRADA FRA-2008-0005 Final Report)

Proton and light-ion accelerators have many research and medical applications, but the current state of technology in accelerators severely limits their widespread use: (1) synchrotrons have an intrinsic low duty cycle and become quite large for light ions; and (2) cyclotrons do not have energy variability and require complex superconducting magnets for high energy. A new concept in non-scaling Fixed-Field Alternating-Gradient (FFAGs) has been invented that incorporates both the energy variability of the synchrotron and the high duty cycle of the cyclotron. The concept uses normal-conducting, combined-function magnets that apply only constant (dipole) and linear-gradient (quadrupole) fields to stabilize the accelerator. This project will develop an optimized design for a new non-scaling and cost-effective innovation in FFAG accelerators. The accelerator design will have optics that can stably accelerate protons to 250 MeV (or higher) and carbon ions to =200 MeV, an energy considered ideal for radiation therapy. Phase I will optimize, fully simulate, and demonstrate the feasibility of this accelerator. Commercial Applications and other Benefits as described by the awardee: The fixed-field acceleration concept should eliminate some of the most pronounced technical difficulties, expense, maintenance, and required expertise faced in conventional proton and light-ion accelerators. In addition to the application for high energy physics, these accelerators should have application in cancer treatment, radiopharmaceuticals, and medical isotope production, and materials science.

43 PARTICLE ACCELERATORS↗
DOE OSTI2022

Accelerator and Beam Physics: Grand Challenges and Research Opportunities

Accelerator and beam physics (ABP) is the science of the motion, generation, acceleration, manipulation, prediction, observation and use of charged particle beams. It focuses on fundamental long-term accelerator and beam physics research and development. Accelerator and beam physics research has resulted in important advances in accelerator science, yet support for this research is declining. NSF has terminated its program in accelerator Science and funding by DOE through GARD and Accelerator Stewardship has been steady or declining. The declining support for accelerator research will slow advances and threaten student training and work-force development in accelerator science. We propose a robust and scientifically challenging program in accelerator and beam physics, which will position the field of US High Energy Physics to be productive and competitive for decades to come.

43 PARTICLE ACCELERATORS↗