• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

Linux performance issues

nitrobass24

[H]ard|DCer of the Month - December 2009
2FA
Joined
Apr 7, 2006
Messages
10,477
So ive got SMP going in ubuntu but it does not seem to be working right.

its maxing out all 16 cores so i know its running, but im only getting about 6k PPD :(

Ive got 2x E5530's @2.4ghz so i should get at a minimum of 10k ppd
 
Is it one of the kernels that is known to have slow performance with folding? I'm no Linux guru, but I've read on here from someone that there's specific Linux kernels that are considerably better for folding than others. I'll try to post back after I find it.

Edit: Here is what I'm talking about. Look at Apollo's post about a 1/3 down the page and following for more info. Hope this helps.
 
Last edited:
yea im running the 2.6.31-16 kernel so thats not it.

i dunno this linux stuff is wearing me out.

I cant get anything to work right it seems :(
 
I rember trying to get the original linux smp working on one of my rigs and it took like a week...

I said forget it...but hang in there, you'll get it :D
 
i just dont get it. i got performance running on 8 cpus in a VM then i am right now natively.
 
Regarding the Linux kernel version, I believe only the -bigadv WUs are sensitive to that although I could be wrong about this. The kernel version can affect scaling above 4 threads. It don't believe it becomes a concern until you approach 8 threads. This is for -bigadv but now a lot of people are running -smp 7 or 8 with standard SMP WUs as well. It may or may not make a difference, I really don't know.

Someone ought to do thorough kernel benchmarking. To be effective, this would require the same system and the same distro except for kernel versions. Anything to get those TPDs down.
 
Back
Top