• 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.

For those with math experience.

C

CoW]8(0)

[H]ard|Gawd
Joined
Jul 25, 2005
Messages
1,427
I'm beginning to wonder if stability of a CPU in integer calculations would imply stability in floating point calculations.

Does anyone know what orthos small FFT's stresses? FFTs are Fast Fourier Transforms which are used to compute the Discrete Fourier Transform. Without knowing anything about these two terms, the word 'discrete' in my mind implies integer.

Does Orthos Small FFT's perform integer or floating point calculations?
 
if it is the same as prime95 then it is floating point

as far as I know the stability of your cpu has nothing to do with integer or floating point calculations but heat and voltage.

I believe orthos uses the Lucas-Lehmer test when you select small ffts. If you really want to learn the math of it, here is a link.

Now go get on your inner math geek

http://www.mersenne.org/math.htm
 
But stressing the CPU with only integer or floating point operations may not be a complete stability test.
 
i would say it is, if you load your cpu for a long time at 100% and it is fine, then yes. I mean define stable.... If you max out the heat and power consumption and you are ok, then what are you worried about?
 
Well, right now I seem to be getting stability issues with F@H but not Orthos.
 
Run TAT with your stability test(s) of choice. TAT seems to be the only program capable of fully loading Intel CPUs. http://www.techpowerup.com/downloads/392/Intel_Thermal_Analysis_Tool.html

I don't think it runs under Vista x64 if you happen to be running that OS, but it works on other versions and XP.

But since you already know F@H isn't stable, why not just lower the core speed? I use memtest86+ and it's pretty good at find obvious memory or overclocking errors (if the caches are unstable).
 
Wow pxc, you're right. I do get a couple C higher in temps with TAT than with Orthos small FFTs. I'm not sure if F@H showing my OC instability or just the instability of the client itself. The SMP client already has a ton of bugs...
 
You must log in or register to reply here.
Back
Top