I've been reading some threads about Fedora@Home and noticed some guys are running more than one client in a single machine. I remember reading somewhere that Stanford project doesn't recommend using more than one client per cpu.
So where is the truth? Can I optimize my folding work by running more than one client? If so, how many clients should I run simultaneously? Is there any significant difference between running one client and more than one?
As far as I know, my CPU usage is 100% using just one client, so I suppose I'm already giving all my processing capacity to the project. Am I right?
That's my configuration: FC3, P4 Prescott, 3 Ghz, 1Gb (ram). (Although I've got a Prescott, I never pulled off to run the smp kernel)
Thank you guys.