Intel Burn vs Prime95 For Stress Testing Overclock

Bint Box

Bint Box

Bint Box

Associate
Joined
22 Dec 2009
Posts
1,390
Location
Upper Skurt
Hi,

I have been stress testing a 4.2GHz overclock with an i7-2600k CPU in a sandybridge board using Prime95 and Intel Burn and wondered at what point do I decide I have a stable system?

With Prime95 (small FFT) I can run the overclock for hours without crashing using a vCore of 1.275v. However, if I then test the idential overclock with Intel Burn using stress Level of "Maximum" it crashes after one or 2 GFlops.

Testing with Intel Burn at a lower stress level of "Standard" runs OK without crashes although I have only run it up to 40 GFlops.

Returning to a "stable system", the PC will be used for video encoding and rendering at an overclock of 4.2GHz, do I really need to get it stable with Intel Burn at stress level maximum for a minimum of 40 GFlops or is my Prime95 (small FFT) stability test OK for the real world use of the PC?


Rgds
Bintos
 
IBT on Maximum uses all the RAM it can gets it's grubby mits on, Prime95 small fft doesn't and also doesn't stress the CPU as much.

I would look at checking out your RAM and IMC voltage first.
 
Use them both and then run a game or as above rendering to check stability. I had a supposedly stable overclock until i ran some games.
 
Concorde Rules said:
It isn't stable then, simple as that.

If you notice it in practice is another matter.

Personally i'd bring the OC back or up the V.
Click to expand...



I bitched up the vCore to 1.285v and it seems to be stressing out OK with IntelBurm set on stress level of Maximum. I will add one more notch to the vCore just to keep the juice flowing nicely then poke it with a bit of encoding and rendering.
 
You must log in or register to reply here.
Back
Top Bottom