Overclocking an E2220, stability problems

[Bigga Lexx]

Hello everyone:
I just changed my E1200 yesterday radbox with a E2220 with Cooler Master Hyper Z600 and the Noctua S12. "Small" problem: my E2220 can not spend 30 minutes of OCCT if not at its reference frequency of 2.4GHz (or below of course).

Well, I was a little accustomed errors OCCT my E1200@3.2GHz, 1.35V, register an error after 28 minutes / 30 to OCCT, but that's after 10 minutes even Unless I make mistakes. + Sometimes BSOD, refusal to boot Windows and other types of crashes

The rest of my setup:

EP35-DS3R
2x1Go Corsair TWIN2X2048-6400 CAS5 ~ ~> 450MHz it takes in cas5 ... if I met earlier, many errors in memtest 86 +
Twintech XT OC 9600GT
DD 250GB
Antec NSK4480B
Noctua NF-R8

Well, I read a letter in the PCU No. 35, which spoke of an oc of E2220 which could not exceed 320MHz FSB with a voltage of 1.4V, and talked of a problem if power the PC could not boot up windows then BSOD when you have a big CPU solicitation.

Incidentally, I just do a 3dmark 2006 with my E2220@3.2GHz / 1.30V, both CPU tests are passed, but what is it slow ... (1 FPS average, I got the same result with my E1200@3.2GHz). The total score is 10433, instead of ~ 9500 with my E1200.

So my questions:

• Are the symptoms listed (OCCT errors in some BSOD, Windows refused to boot) may indicate weaknesses in my diet, a Earthwatts 380W that came with my camera, bought it about 5 months ago?
• Knowing that E2220 is clocked at 2.4GHz origin, 1.25V and a E1200 is clocked at 1.6GHz, 1.2375V, is it normal that I should increase both the voltage (+0.05 V) CPU for that it may take only 10 minutes of OCCT? Compared to my E1200 which takes 28 minutes (which gave a 100% frequency level, all for an increase of +0.1225 V), it seems huge, because my E2220 has "only" a 33% for a increase of +0.05 V
• Is that ultimately, the stability under OCCT is it so important? XD

[Hockson]

If you do not want 30min to OCCT it means that your CPU is unstable, you may have to crash.
To be sure of the stability of your processor, launching OCCT for at least 8h. Try also other software (Prime95, bunks, CPU Stress, CPU Burn ...)

It is possible that it's your diet that limits, try to obtain another worry for testing.

What is your temperature ?

[Viceman]

OCCT bench is a very demanding, which can sometimes go up to detect errors on a configuration that has not overclocked . I counsel you to use another software type orthos to see if you have the problems.
otherwise, are you sure it is the cpu that creates a problem ? did you put the ram ratio to a minimum?

[Bigga Lexx]

I put my ram to 3:1 ratio, since my CPU is x12 coeff for 3.2GHz just need a FSB of 266MHz. I checked with my E1200, my ram is 450MHz (900MHz therefore in my bios). For tension, I tried with three different voltages, below 1.30 OCCT bugue when it is more idle, with 1.30 the test is 10 minutes without error. I'll see what orthos can do with my problem.