[wudaddy]

So currently I have my Q6700 at 340x10 (3.4ghz) and wanted to run some stability tests to see if it's stable or not.

5 runs of IntelBurnTest passed.
1 hour of Orthos passed.
wPrime 1024M calculated with no errors.
3dmark06 passed.

However, sometimes when I run HyperPI I get an error and crashes. It's really weird...

 

[richardbb85]

3.4 at what volt

 

[wudaddy]

Quote:

Originally Posted by richardbb85 3.4 at what volt

1.375. With all other tests the temps don't go above 60C (my room is hot as hell so the ambient temp is pretty high). Only with IntelBurnTest does it reach low 70s. Mostly in games and such it's in the upper 50s.

 

[richardbb85]

sounds about right

my q6600 was stable at 3.6ghz with 1.39v in bios

why stop at 3.4? i think u got more room to OC

 

[I_dalder_I]

Its obviously not stable. None of the stability tests you listed ran long enough to prove if it were stable. and if super pi crashes with a error then that should tell you right there. try tweaking your volts up a bit

 

[richardbb85]

run Prime with small ftt for at least 6-10 hrs

 

[wudaddy]

Quote:

Originally Posted by I_dalder_I Its obviously not stable. None of the stability tests you listed ran long enough to prove if it were stable. and if super pi crashes with a error then that should tell you right there. try tweaking your volts up a bit

Yeah but the thing with HyperPI is that it wouldn't even start running yet and I'd get the error - it only calculates the digits of pi... I would think wPrime calculates in a similar fashion and that ran without flawlessly. I'm pretty sure IntelBurnTest stresses the CPUs more than HyperPI and yet it ran a good while before it passed the 5 runs. I was told somewhere that 5 runs was equal to hours of Prime95 testing and would be considered a stable overclock. What I was thinking was...perhaps there is a bug in HyperPI? What if the program itself was not stable? Everything else I throw at the CPU is stable, which is why I find this weird.

EDIT: Ok well I just set the CPU clocks and voltages back to stock and I'm still getting the crash from HyperPI, which shows that the actual software has a bug or something.

 

[FSF-Foxhound]

Could be your RAM, i doub't its bugged.

 

[PERSPOLIS]

There is a compatibility problem between vista 64 & super pi/hyper pi.There was a thread about this sometime ago here in OCN.Some workarounds were also suggested which may or may not work.
I say forget about hyperpi as a stability test in vista64.Just run IBT/ortho properly & your oc is fine.

 

[wudaddy]

Quote:

Originally Posted by PERSPOLIS There is a compatibility problem between vista 64 & super pi/hyper pi.There was a thread about this sometime ago here in OCN.Some workarounds were also suggested which may or may not work. I say forget about hyperpi as a stability test in vista64.Just run IBT/ortho properly & your oc is fine.

Thanks man, that explains everything.

 

[GBPS]

Surprisingly, I think your O/C is stable enough for real-world tests. If you get some time, run Prime95 for at least 5 hours (overnight) to see for sure if your OC is stable. Otherwise, I say you have around .2GHz more you can push that thing before stability comes into question.