In the numerical treatment of population dynamic models a great number of large linear systems must be solved at each time step. Thus the application of parallel algorithms may result to be convenient. In this paper we analyze how it can be done and the problems that arise in doing it. We propose two different ways of parallel implementation for a numerical method suitable to solve the model considered. As numerical tests we have considered the speedup obtained for a distributed memory parallel computer and the speedup extimated for shared memory parallel computers. These numerical results show high speedup for the second kind of computers
PARALLEL METHODS IN THE NUMERICAL TREATMENT OF POPULATION-DYNAMIC MODELS
LOPEZ, Luciano;
1992-01-01
Abstract
In the numerical treatment of population dynamic models a great number of large linear systems must be solved at each time step. Thus the application of parallel algorithms may result to be convenient. In this paper we analyze how it can be done and the problems that arise in doing it. We propose two different ways of parallel implementation for a numerical method suitable to solve the model considered. As numerical tests we have considered the speedup obtained for a distributed memory parallel computer and the speedup extimated for shared memory parallel computers. These numerical results show high speedup for the second kind of computersI documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
