Detail výsledku
Runge-Kutta Based Parallel Computations
Vopěnka Václav, Ing., FIT (FIT), UITS (FIT)
Kraus Michal, Ing., Ph.D., FIT (FIT)
An important part of contemporary characteristic problems in science and technology is based on simulating and analysing complex systems such as economical models, weather
forecast models or technology process control.
Solving most of those problems leads to complex set of linear and non-linear differential and partial differential equations with time changing parameters and large sets of algebraic and transcendental equations. For solving such problems with great emphasis on high accuracy or
high speed we need high-performance computer platforms.
The processor's clock frequency is slowly reaching it's physical limits. The increase of performance is now possible only by including multiple parallel computing units. This approach
is supported by form of some problems. It appears that contemporary sequential methods of solving some problems is not natural.
Dynamic modelling, Numerical solution of dierential equations, Runge-Kutta methods, Arithmetic logic unit
@misc{BUT192790,
author="Jiří {Kunovský} and Václav {Vopěnka} and Michal {Kraus}",
title="Runge-Kutta Based Parallel Computations",
booktitle="MATHMOD VIENNA 2012 - 7th Vienna Conference on Mathematical Modelling",
year="2012",
series="Report no. S38",
pages="390--390",
publisher="ARGE Simulation News",
address="Vienna",
note="Abstract"
}