Search NASA⌕ Search

SEARCH · Search NASA

Results for “Accelerators,OTHER”

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 271 records · Page 15
All sourcesNASA NTRSDOE OSTI
DOE OSTI2017

Impact of Linearity and Write Noise of Analog Resistive Memory Devices in a Neural Algorithm Accelerator

Resistive memory (ReRAM) shows promise for use as an analog synapse element in energy-efficient neural network algorithm accelerators. A particularly important application is the training of neural networks, as this is the most computationally-intensive procedure in using a neural algorithm. However, training a network with analog ReRAM synapses can significantly reduce the accuracy at the algorithm level. In order to assess this degradation, analog properties of ReRAM devices were measured and hand-written digit recognition accuracy was modeled for the training using backpropagation. Bipolar filamentary devices utilizing three material systems were measured and compared: one oxygen vacancy system, Ta-TaO x , and two conducting metallization systems, Cu-SiO 2 , and Ag/chalcogenide. Analog properties and conductance ranges of the devices are optimized by measuring the response to varying voltage pulse characteristics. Key analog device properties which degrade the accuracy are update linearity and write noise. Write noise may improve as a function of device manufacturing maturity, but write nonlinearity appears relatively consistent among the different device material systems and is found to be the most significant factor affecting accuracy. As a result, this suggests that new materials and/or fundamentally different resistive switching mechanisms may be required to improve device linearity and achieve higher algorithm training accuracy.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2016

Towards Achieving Performance Portability Using Directives for Accelerators

In this paper we explore the performance portability of directives provided by OpenMP 4 and OpenACC to program various types of node architectures with attached accelerators, both self-hosted multicore and offload multicore/GPU. Our goal is to examine how successful OpenACC and the newer offload features of OpenMP 4.5 are for moving codes between architectures, how much tuning might be required and what lessons we can learn from this experience. To do this, we use examples of algorithms with varying computational intensities for our evaluation, as both compute and data access efficiency are important considerations for overall application performance. We implement these kernels using various methods provided by newer OpenACC and OpenMP implementations, and we evaluate their performance on various platforms including both X86_64 with attached NVIDIA GPUs, self-hosted Intel Xeon Phi KNL, as well as an X86_64 host system with Intel Xeon Phi coprocessors. In this paper, we explain what factors affected the performance portability such as how to pick the right programming model, its programming style, its availability on different platforms, and how well compilers can optimize and target to multiple platforms.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2016

Cerium migration during PEM fuel cell accelerated stress testing

Cerium is a radical scavenger which improves polymer electrolyte membrane (PEM) fuel cell durability. During operation, however, cerium rapidly migrates in the PEM and into the catalyst layers (CLs). In this work, membrane electrode assemblies (MEAs) were subjected to accelerated stress tests (ASTs) under different humidity conditions. Cerium migration was characterized in the MEAs after ASTs using X-ray fluorescence. During fully humidified operation, water flux from cell inlet to outlet generated in-plane cerium gradients. Conversely, cerium profiles were flat during low humidity operation, where in-plane water flux was negligible, however, migration from the PEM into the CLs was enhanced. Humidity cycling resulted in both in-plane cerium gradients due to water flux during the hydration component of the cycle, and significant migration into the CLs. Fluoride and cerium emissions into effluent cell waters were measured during ASTs and correlated, which signifies that ionomer degradation products serve as possible counter-ions for cerium emissions. Fluoride emission rates were also correlated to final PEM cerium contents, which indicates that PEM degradation and cerium migration are coupled. Lastly, it is proposed that cerium migrates from the PEM due to humidification conditions and degradation, and is subsequently stabilized in the CLs by carbon catalyst supports.

30 DIRECT ENERGY CONVERSION↗
DOE OSTI2013

Coherent seeding of self-modulated plasma wakefield accelerators

The growth of the beam self-modulation and hosing instabilities initiated by a seed wakefield is examined. Although the growth rates for the self-modulation and hosing instabilities are comparable, it is shown that an externally excited wakefield can be effective in selectively seeding the beam radial self-modulation, enabling the beam to fully modulate before strong beam hosing develops. Methods for coherent seeding are discussed.

43 PARTICLE ACCELERATORS↗
DOE OSTI2014

Studies of basic limitations on production, transport and acceleration of the high intensity polarized H{sup −} beam in the RHIC polarized ion source

Basic limitations on the high-intensity polarized H{sup −} ion beam production and transport were experimentally studied in charge-exchange collisions of the neutral atomic hydrogen beam in the Na-vaporjet ionizer cell. The energy dependence of space-charge effects on the beam instabilities and losses were studied and described in the model of synthetic H{sup +} — H{sup −} beam transport. A xenon gas admixture to the H{sup −} ion production cell and beam transport line greatly improves the space-charge compensation, which is also successfully described in the simulations. These studies are the part of the polarized source upgrade project for RHIC.

43 PARTICLE ACCELERATORS↗
DOE OSTI2018

Direct numerical simulations of reacting flows with detailed chemistry using many-core/GPU acceleration

A new direct numerical simulation (DNS) code for multi-component gaseous reacting flows has been developed at KAUST, with the state-of-the-art programming model for next generation high performance computing platforms. The code, named KAUST Adaptive Reacting Flows Solver (KARFS), employs the MPI+X programming, and relies on Kokkos for “X” for performance portability to multi-core, many-core and GPUs, providing innovative software development while maintaining backward compatibility with established parallel models and legacy code. The capability and potential of KARFS to perform DNS of reacting flows with large, detailed reaction mechanisms is demonstrated with various model problems involving ignition and turbulent flame propagations with varying degrees of chemical complexities.

Computer Science↗