I'm working on a 1 million node model by using fluent6 on a linux system pc. It is so slow that I can't get a result. So I want to try the parallel process function in fluent6. I will get a multi-processor machine in near future. But right now, I want to try using network to using all the computer sources in our group to fulfill this function. Any one has ideas about how to fulfill it?

If you have a parallel license then all you need to do is to start fluent with something like:

fluent 3d -t# -pnmpi -cnf=hosts

Where # is the number of CPUs that you want to run on and hosts is a file in which you have written the host-names of the machines that you want to run on (one name on each line). Run this command when you are logged in to the first machine in you hosts file. A 1-million cell model should run well on up to something like 20 CPUs if you have a switched fast-ethernet network.

1). using RSHD setup RSHD service fluent 3d -t# -pnet -cnf=hosts 2). using pvm

I think u shall read the the Fluent6 parallel User's Guide first.

(È˳±È˺£ÖУ¬ÓÖ¼ûµ½Ä㣬......¹þ¹þ£¬»Ø¹úÄÚ°ÉÐÖµÜ)

Jonas is right about the command you have to type. But before that you have to ensure that following things are OK: 1. Your home directory on all computers is on the same path, for example: /home/jcai 2. You have the same password for all the computers 3. All computers are accessible by rlogin 4. You cannot use both Linux and WinNT machines (although I think Fluent says it is possible, but it is not) Good luck

Hi : My name is Sergio and im looking for FLUENT 6.0 for linux. where can i get for it ? and how much it is cost ? I really need run a code with more than 1 million nodes.

Thank you very much for the information.

Sergio.