The component structure of a self-adapting numerical software system

Victor Eijkhout; Erika Fuentes; Thomas Eidson; Jack Dongarra

doi:10.1007/s10766-005-3577-3

The component structure of a self-adapting numerical software system

Victor Eijkhout, Erika Fuentes, Thomas Eidson, Jack Dongarra

Research output: Contribution to journal › Article › peer-review

Abstract

Self-Adapting Numerical Software (SANS) systems aim to automate some of the laborious human decision making involved in adapting numerical algorithms to problem data, network conditions, and computational platform. In this paper we describe the structure of a SANS system that tackles automatic algorithm choice, based on dynamic inspection of the problem data. We describe the various components of such a system, and their interfaces. © 2005 Springer Science+Business Media, Inc.

Original language	English
Pages (from-to)	137-143
Number of pages	6
Journal	International Journal of Parallel Programming
Volume	33
Issue number	2-3
DOIs	https://doi.org/10.1007/s10766-005-3577-3
Publication status	Published - Jun 2005

Keywords

Adaptive systems
Component frameworks
Linear system solving

Access to Document

10.1007/s10766-005-3577-3

Cite this

@article{0eeb6095ce4d482d8a315f64e16c6b50,

title = "The component structure of a self-adapting numerical software system",

abstract = "Self-Adapting Numerical Software (SANS) systems aim to automate some of the laborious human decision making involved in adapting numerical algorithms to problem data, network conditions, and computational platform. In this paper we describe the structure of a SANS system that tackles automatic algorithm choice, based on dynamic inspection of the problem data. We describe the various components of such a system, and their interfaces. {\textcopyright} 2005 Springer Science+Business Media, Inc.",

keywords = "Adaptive systems, Component frameworks, Linear system solving",

author = "Victor Eijkhout and Erika Fuentes and Thomas Eidson and Jack Dongarra",

year = "2005",

month = jun,

doi = "10.1007/s10766-005-3577-3",

language = "English",

volume = "33",

pages = "137--143",

journal = "International Journal of Parallel Programming",

issn = "0885-7458",

publisher = "Springer New York",

number = "2-3",

}

TY - JOUR

T1 - The component structure of a self-adapting numerical software system

AU - Eijkhout, Victor

AU - Fuentes, Erika

AU - Eidson, Thomas

AU - Dongarra, Jack

PY - 2005/6

Y1 - 2005/6

N2 - Self-Adapting Numerical Software (SANS) systems aim to automate some of the laborious human decision making involved in adapting numerical algorithms to problem data, network conditions, and computational platform. In this paper we describe the structure of a SANS system that tackles automatic algorithm choice, based on dynamic inspection of the problem data. We describe the various components of such a system, and their interfaces. © 2005 Springer Science+Business Media, Inc.

AB - Self-Adapting Numerical Software (SANS) systems aim to automate some of the laborious human decision making involved in adapting numerical algorithms to problem data, network conditions, and computational platform. In this paper we describe the structure of a SANS system that tackles automatic algorithm choice, based on dynamic inspection of the problem data. We describe the various components of such a system, and their interfaces. © 2005 Springer Science+Business Media, Inc.

KW - Adaptive systems

KW - Component frameworks

KW - Linear system solving

U2 - 10.1007/s10766-005-3577-3

DO - 10.1007/s10766-005-3577-3

M3 - Article

SN - 0885-7458

VL - 33

SP - 137

EP - 143

JO - International Journal of Parallel Programming

JF - International Journal of Parallel Programming

IS - 2-3

ER -

The component structure of a self-adapting numerical software system

Abstract

Keywords

Access to Document

Fingerprint

Cite this