What's the longest time you have ran prime for before getting an error?

TechnoCorruption

TechnoCorruption

TechnoCorruption

Soldato
Joined
9 May 2005
Posts
2,511
I have just been running prime blend on my rig for most of the day and after 6 hours 25 minutes an error has stopped it. I have a feeling it is ram as that is the newest addition to the system, specs-

E2160 @ 3.42GHz 1.52v (Passed 8hours of Small FFT's)
Asus P5K-E / Wifi-AP @ 380x9 (Northbridge - 1.25v)
2GB Crucial Ballistix @ 1140MHz 2v
Coolermaster Extreme 460w

Now I will obviously try the ram at 2.05v and then 2.10v, if this doesn't solve it then I will up the northbridge to 1.4v and try again.

My question is though what is the longest you have had prime running before the dreaded error has popped up? I tend to run for 8 hour sessions, I start with a 8 hour session of small fft's and then a 8 hour session of blend, throw a couple of passes of memtest into the mix and I then consider it stable.

My other question is... if for example you left a prime session running indefinitely would it fail at some point? Could random factors like change in room temperature or small drops of power on the home electrics give errors?
 
The prime we know and love was a distributed computing project but is now just a stress-test tool (they got bored of getting dodgy work-units back from us overclockers).
 
Have you run memtest? if prime failed due to ram, chances are it will be crapping out constantly in memtest.

Dont use prime small fft or blend to test ram stability, at least use a custom large fft only prime or run memtest for an hour or two.
 
6 months+ by mistake on a machine i used mainly as a file server and occasional media pc, Opteron 144@ 2.8Ghz i think, shows how often i watch films :D
 
It was the NB which was the culprit in the end, I had already passed a session of small fft to rule out the cpu and had done passes on memtest for the memory.

I had tried taking the NB and SB off auto volts and that was where the problem was, I tried setting it to NB-1.25v and SB-1.05v which caused the fail, I set NB-1.4v and SB-1.05v and it passed no problems :)
 
Justintime said:
6 months+ by mistake on a machine i used mainly as a file server and occasional media pc, Opteron 144@ 2.8Ghz i think, shows how often i watch films :D
Click to expand...

That's a real good selling point if you ever get rid of it...

Potential Buyer - I am interested in your Opteron, was it stable at those clocks, got any screenshots.

You - How does a 6month pass on Orthos sound to you??? :D

LOL, good job though mate, that is a really impressive result you have there :)
 
Agr3sive said:
It was the NB which was the culprit in the end, I had already passed a session of small fft to rule out the cpu and had done passes on memtest for the memory.

I had tried taking the NB and SB off auto volts and that was where the problem was, I tried setting it to NB-1.25v and SB-1.05v which caused the fail, I set NB-1.4v and SB-1.05v and it passed no problems :)
Click to expand...

Congrats :)
 
You must log in or register to reply here.
Back
Top Bottom