Best overclocking stabilty software ?

Soldato
Joined
19 Feb 2007
Posts
15,176
Location
Area 18, ArcCorp
So far ive been using OCCT so test my overclocks stability but was wondering if i should use others to test stability as ive heard OCCT isnt that reliable compared to others ?
Any info is appreciated :)
 
Last edited:
Prime95 imo

its best to use a mix of software not just one

i use linpack (Linx or IBT) initially for around 5-10 passes
then after that do a 60min OCCT run and finally ill do a 12 or 24 hour prime run
 
Please close the tread this will get nasty i can sense it lol
there all good.

You can pass hours on one and fail on another.

Best answer use them all
 
I use 20 passes of Linx to check a overclock then when i decide on a max speed i will run Prime for 24 hours. I have had several cpu's pass Linx but fail prime.
 
for me i try intel burn test on top setting (usually around 6.5gb of ram used) for 5 runs if it passes fine then onto prime 95 for a 24hour test, sometimes you will find itll pass intel burn test fine on max but after 8 hours on prime BSOD

only once you have got a 24hour prime test then you know its a gooden
 
for me i try intel burn test on top setting (usually around 6.5gb of ram used) for 5 runs if it passes fine then onto prime 95 for a 24hour test, sometimes you will find itll pass intel burn test fine on max but after 8 hours on prime BSOD

only once you have got a 24hour prime test then you know its a gooden

Have you found this personally? An OC that passes 5 runs at 80%+ GFLOPS and then fails prime? I'm looking for examples as I suspect it's impossible.
 
yes its happened a few times prime is still the ultimate stress tool i find, ive had them pass all kinds of programs then on prime bam! BSOD
 
Really struggle to believe that a CPU that lasts 2+ hours in prime won't last 24 hours, it doesn't really make sense unless it's REALLY on the edge, and then it won't pass IBT?

I've always run Linpack then OCCT, if I crash in either of them then things need to be changed.
 
Have you found this personally? An OC that passes 5 runs at 80%+ GFLOPS and then fails prime? I'm looking for examples as I suspect it's impossible.

I've found the same thing as well. The clock below always passes Linx, but always fails Prime Blend between 18 and 19 hrs in.

 
Really struggle to believe that a CPU that lasts 2+ hours in prime won't last 24 hours, it doesn't really make sense unless it's REALLY on the edge, and then it won't pass IBT?

I've always run Linpack then OCCT, if I crash in either of them then things need to be changed.

it really does happen tried and tested 2 hours prime isnt long enough for sure
 
Really struggle to believe that a CPU that lasts 2+ hours in prime won't last 24 hours, it doesn't really make sense unless it's REALLY on the edge, and then it won't pass IBT?

I've always run Linpack then OCCT, if I crash in either of them then things need to be changed.

The one i have now passes 20 rounds of Linx at 4.3Ghz but fails prime after 14 hours. I had the same with my Q9550 at 4.2Ghz before this. Yes 14 hours is a long time before failing but i would rather be sure it's rock solid than potentially have a bad time with a rig on the edge of stability.
 
I've found the same thing as well. The clock below always passes Linx, but always fails Prime Blend between 18 and 19 hrs in.


Interesting. 91 GFLOPS is below the 80% of max that I usually quote as being the target, but not by much (max would be 120 GFLOPS). I'd really love to know if giving it more memory and getting the GFLOPS over 96 would make it fail. I don't suppose you know the settings that were used for that screenshot and fancy doing a test? :)

Phase?
 
Last edited:
50 passes of intel burn test set the correct cores and normally your free memory amount if you have 4gb it's normally set to 2000 ish.
 
Back
Top Bottom