Prime suddenly stops working around 4 mins into the test

BruceOTB

Distinguished
Mar 5, 2009
318
0
18,780
comp.jpg


E5300 (RO) at 3.65ghz, stock voltage or auto voltage.
GA-G31M-ES2C

when i run prime, 4 minutes into the test, it suddenly stops. just like in the pic above. why?

also, how long should i run prime to be able to know if my OC is already stable?
 

orangegator

Distinguished
Mar 30, 2007
1,163
0
19,310


Just because you haven't had it happen, does not mean it can't. A bsod is really just a major error that windows can't recover from. This looks like an error that windows can recover from. Vista shuts down the program. It's more than just a rounding error in prime95. My guess is that if he was using xp, and not vista, then windows would crash from the error, as it doesn't try to recover like vista does.
 

Dangerous Beans

Distinguished
Jun 12, 2009
25
0
18,540


I agree with the penguin with a rocket launcher. Drop it back to stock clocks and see if prime fails.

I've never heard of prime crashing. It isn't a 32bit version on a 64bit os?
 
hmm, i dont think so. its just around 48-50 when it stops.

What I meant was that the processor was too hot for the clock speed that you are running it at. Your maximum overclock is temperature and voltage dependent, so when the processor reaches a certain temperature it will fail. Upping the voltage a bit may help, but it will make the processor get hotter so won’t improve matters much. Better cooling is the way to go or reduce the clock speed.
I of course assumed that you had it working correctly at stock speeds.
 

You get all kinds of answers: 2, 4, 6, 8, 12, or 24 hours. If the system is truly stable, it will run forever. I used to test for 12 hours. I let one system I was building - well, overclocking - run overnight. Next morning, I discovered that it failed at 14 hours. So now I test for 24 hours.