Search NASA⌕ Search

SEARCH · Search NASA

Results for “architecture”

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 55 records · Page 3
All sourcesNASA NTRSDOE OSTI
DOE OSTI2021

The Architecture of Area 12 Camp - Nevada's Atomic Ghost Town. An Architectural Survey of Area 12 Camp, Nevada National Security Site, Nye County, Nevada

The U.S. Department of Energy, National Nuclear Security Administration Nevada Field Office (NNSA/NFO) plans to demolish 10 buildings and one structure in Area 12 Camp at the Nevada National Security Site (NNSS) in Nye County, Nevada, to meet current and future National Weapons Science, Global and Homeland Security Program, and Environmental Management mission requirements and to reduce dangers to site workers from some of these resources. These demolitions constitute an undertaking subject to review under Title 54 U.S.C. § 306108, commonly known as Section 106 of the National Historic Preservation Act, Title 54 U.S.C. § 300101, et seq., and its implementing regulations, 36 C.F.R. Part 800. These resources were recorded and evaluated as individually ineligible for the National Register of Historic Places (National Register or NRHP) by Menocal and Shaw (2019). Upon review by the Nevada State Historic Preservation Office (SHPO) and internal review by Desert Research Institute (DRI) and NNSA/NFO staff, it was determined that these resources needed to be rerecorded and evaluated in relation to what appeared to be a National Register-eligible Historic District made up of the entirety of Area 12 Camp. Based on preliminary information, it appeared that all 11 resources would likely be contributing elements of that historic district. With that assumption in mind, recording of the entire proposed Area 12 Camp Historic District was done for three purposes. First, it would result in a definitive significance evaluation of the 11 resources in question. Second, recording of the entire camp would be offered as an appropriate mitigation measure for the demolition of these resources. Third, it would present an essential background for future management of resources within the district boundaries. Prior to the present survey, Area 12 Camp had not yet been systematically recorded; therefore, DRI surveyed an area of approximately 229 acres for architectural resources. This effort resulted in the identification, recording, and evaluation of the Area 12 Camp Historic District (SHPO Resource No. D372), including the identification of its contributing components. This district is recommended eligible for the National Register of Historic Places under Criteria A and C. It is unevaluated under Criteria B and D. The Area 12 Historic District contains 71 Principal Resources, which include landscapes, buildings, and structures. Of these resources, 69 (including 10 of the 11 resources to be removed) are recommended as contributing elements of the historic district during its period of significance corresponding to nuclear testing from 1960 through 1992. During most of the Cold War, the government reservation now called the NNSS was named the Nevada Test Site (NTS).

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2022

Improved Architectures and Training Algorithms for Deep Operator Networks

Operator learning techniques have recently emerged as a powerful tool for learning maps between infinite-dimensional Banach spaces. Trained under appropriate constraints, they can also be effective in learning the solution operator of partial differential equations (PDEs) in an entirely self-supervised manner. In this work we analyze the training dynamics of deep operator networks (DeepONets) through the lens of Neural Tangent Kernel theory, and reveal a bias that favors the approximation of functions with larger magnitudes. To correct this bias we propose to adaptively re-weight the importance of each training example, and demonstrate how this procedure can effectively balance the magnitude of back-propagated gradients during training via gradient descent. We also propose a novel network architecture that is more resilient to vanishing gradient pathologies. Taken together, our developments provide new insights into the training of DeepONets and consistently improve their predictive accuracy by a factor of 10-50x, demonstrated in the challenging setting of learning PDE solution operators in the absence of paired input-output observations.

97 MATHEMATICS AND COMPUTING↗
NASA NTRS1976

Architecture and data processing alternatives for Tse computer. Volume 1: Tse logic design concepts and the development of image processing machine architectures

Digital computer components which perform two dimensional array logic operations (Tse logic) on binary data arrays are described. The properties of Golay transforms which make them useful in image processing are reviewed, and several architectures for Golay transform processors are presented with emphasis on the skeletonizing algorithm. Conventional logic control units developed for the Golay transform processors are described. One is a unique microprogrammable control unit that uses a microprocessor to control the Tse computer. The remaining control units are based on programmable logic arrays. Performance criteria are established and utilized to compare the various Golay transform machines developed. A critique of Tse logic is presented, and recommendations for additional research are included.

Rickard, D. A.↗
DOE OSTI2021

Accurate and scalable graph neural network force field and molecular dynamics with direct force architecture

Abstract Recently, machine learning (ML) has been used to address the computational cost that has been limiting ab initio molecular dynamics (AIMD). Here, we present GNNFF, a graph neural network framework to directly predict atomic forces from automatically extracted features of the local atomic environment that are translationally-invariant, but rotationally-covariant to the coordinate of the atoms. We demonstrate that GNNFF not only achieves high performance in terms of force prediction accuracy and computational speed on various materials systems, but also accurately predicts the forces of a large MD system after being trained on forces obtained from a smaller system. Finally, we use our framework to perform an MD simulation of Li 7 P 3 S 11 , a superionic conductor, and show that resulting Li diffusion coefficient is within 14% of that obtained directly from AIMD. The high performance exhibited by GNNFF can be easily generalized to study atomistic level dynamics of other material systems.

Chemistry↗