skip to main content
OSTI.GOV title logo U.S. Department of Energy
Office of Scientific and Technical Information

Title: Improving Power and Performance in HPC Networks.

Abstract

Abstract not provided.

Authors:
Publication Date:
Research Org.:
Sandia National Lab. (SNL-NM), Albuquerque, NM (United States)
Sponsoring Org.:
USDOE National Nuclear Security Administration (NNSA)
OSTI Identifier:
1371618
Report Number(s):
SAND2016-6556PE
644879
DOE Contract Number:
AC04-94AL85000
Resource Type:
Conference
Resource Relation:
Conference: Proposed for presentation at the Tech Talk @ AMD held July 8, 2016 in Austin, Texas, United States.
Country of Publication:
United States
Language:
English

Citation Formats

Groves, Taylor. Improving Power and Performance in HPC Networks.. United States: N. p., 2016. Web.
Groves, Taylor. Improving Power and Performance in HPC Networks.. United States.
Groves, Taylor. 2016. "Improving Power and Performance in HPC Networks.". United States. doi:. https://www.osti.gov/servlets/purl/1371618.
@article{osti_1371618,
title = {Improving Power and Performance in HPC Networks.},
author = {Groves, Taylor},
abstractNote = {Abstract not provided.},
doi = {},
journal = {},
number = ,
volume = ,
place = {United States},
year = 2016,
month = 7
}

Conference:
Other availability
Please see Document Availability for additional information on obtaining the full-text document. Library patrons may search WorldCat to identify libraries that hold this conference proceeding.

Save / Share:
  • The backpropagation learning algorithm has proven to be a robust method for training feedforward multilayer neural networks to map the relationships between input/output patterns. However, as with many gradient descent optimization methods, the rate of convergence of the backpropagation algorithm decreases the closer it gets to the solution, and it requires judicious selection of the learning and momentum constants to achieve reasonable convergence and avoid oscillations about the optimum solution. In this paper, the discussion focuses on how the method of conjugate gradients can be combined with the backpropagation algorithm to improve and accelerate learning in neural networks and eliminatemore » the process of selecting parameters. The proposed method was used to train a neural network to classify nuclear power plant transients, and it significantly expedited the learning process. 5 refs., 1 fig.« less
  • Since the vast majority of power plants that have operated for several years are not operating even close to their optimum performance level, some other approach or technique is needed to improve the power plant's performance. This paper reports that the disciplined approach needed for most power plants to reach that optimum performance is reliability centered maintenance. Obviously the composite effects of design, operation, maintenance, and management yield the power plant's operating performance. The process of reliability centered maintenance uses a number of improvement techniques to develop the maintenance tasks. Conceptually the inputs to the disciplined approach are economic goalsmore » and safety/environmental goals. The outputs of the approach are a well-defined preventive maintenance program and capital costs for highly selective design changes against minimal power plant operating costs.« less
  • Significant attention is turning to the need for energy security and diversity in light of the Gulf War and increasing regulatory requirements, (including the Clean Air Act Amendments of 1990), and a national desire for fuel conservation. All of the above have contributed to directing utilities attention to improving the overall performance of their fossil power plants. This paper discusses various methods for successfully developing and implementing an overall performance program in a typical fossil power plant. Furthermore, this paper will discuss specific technical concepts, state-of-the-art techniques, as well as proven and demonstrated methods related to both environmental and efficiencymore » programs, and implementation of integrated concept.« less
  • The basic responsibility of the electric utility industry is to supply reliable, efficient, and economical electric service to the customer. However, meeting these responsibilities has become increasingly complex because of the new technical, economical, environmental, political, and social considerations. The effect of these considerations/constraints has been to increase the uncertainty in the planning process and raise operating and capital costs for the electric utility industry. Therefore, utilities have sought means to control the rate of peak load to improve the efficiency/utilization of present electric energy systems and to reduce requirements for additional generation, transmission, and distribution investments. These methods formmore » the set of functions known as load management. In this paper, application of the residential load management program of Florida Power and Light (FPL) is discussed. The load management program in South Florida is evaluated from both, the utility and customer`s perspective. This includes system configuration, establishment of residential load management program, customer acceptance, utility costs and benefits, and demographics.« less