|
|
|
|
Subject: Re: ciradoz
From: francois_bousquet (francois_bousquet@hotmail.com)
Date: jeu jun 13 2002 - 11:52:25 CEST
Well, it depends on the complexity of your agents. Recently I runned a simple cellular automata 1000*1000= 1 million entities on a pentium4 2GhZ, 1Ghz memory. It was not so slow.
Some years ago we had the Djemiong model with more than 5000 agents, moving, reproducing.
So, I guess 5000 agents is feasible, but it depends on the individual amount of computation
----- Original Message -----
From: Pascal Perez
To: Cormas
Sent: Thursday, June 13, 2002 1:12 PM
Subject: ciradoz
Hi mates,
Has anybody tried to run a CORMAS application with 5 000 agents (households in our case) ?
This an important preliminary issue for us before deciding whther we scale down our model to 1/10 or not.
thank you for your help.
cheers
Pascal and Anne
Dr Pascal PEREZ
CIRAD Australia
Research School of Pacific & Asian Studies
RMAP Program - room 5027 - Coombs Building
Canberra ACT 0200
tel: (61) (0)2 61 25 87 05
pascal@coombs.anu.edu.au