Prime95 has stopped working

[PowerK]

Yeah, since I'm not aiming for extreme overclock, I always go for LLC off to play within Intel's Vdroop specifications.

 

[JAG87]

Bah... WHEA errors @ 4.4 with +0.010V offset. (during Prime95 custom blend).
I don't want to raise offset. I'll try 4.3 with +0.005V offset.

JAG87,

No WHEA errors on your end 4.3 with +0.005V offset ?

Nope. But I'm not using stupidly high clocked ram that adds unnecessary stress to the IMC. That could be another source of your problem.

 

[PowerK]

I thought 2133 was pretty much standard these days.

Almost every off-the-shelf Samsung 1033 DDR3 memory manufactured since last year runs at 2133/2400 without any problems.

 

[cmdrdredd]

It depends, not everything will run at that speed. I know the 30nm Samsung memory does, I run mine currently at DDR3-2133 10-10-10-25 with only 1.4v

 

[PowerK]

Ok.

At last, my 3770K is stable @ 4.5 GHz with 4x 4GB G.SKILL 2133 RAM at 9-11-10-28.

Voltages and temps are a lot better than before, too (due to lower offset).

Internal PLL Overvoltage = Enabled
And one notch bump in VTT voltage did the trick. (from 1.076V to 1.085V)

Those two above allowed me to lower my offset from +0.050V to +0.005V resulting in lower temp. (This is the lowest positive offset value in the BIOS)
With the lower offset at 4.5, CPU-Z shows 1.168V (under Prime95 Small FFTs and LinX load)

With the above settings,
- Passed 50 passes of LinX
- Passed 8 hours of Prime95 v27.7 small FFTs test
- Passed 10 hours of Prime95 v27.7 in-place large FFTs test
- Passed 6 hours of Prime95 v27.7 custom blend with min/max FFT of 1344/1344 with 90% of RAM.
- Passed 6 hours of Prime95 v27.7 custom blend with min/max FFT of 1792/1792 with 90% of RAM.
- Passed 15 hours of Prime95 v27.7 custom blend with 90% of RAM

And several hours of gaming including (Max Payne 3, Witcher 2, Crysis, Diablo3 etc) no problem at all. No BSOD, no app crash and most importantly no WHEA errors.

Thank you all for advice.

 

[tigersty1e]

Is anyone not getting any whea errors ever?

I seem to get a random whea error every 1-2 days.

 

[2is]

None here, but I'm also at a mild 4.2 GHz and 1600 memory that isnt being OC'd. Anyone getting these errors should go back to stock and see if it's indeed being caused by an unstable OC.

 

[Ferzerp]

I get none. 4.4Ghz If you get them, you need more vcore.

 

[PowerK]

I get none. 4.4Ghz If you get them, you need more vcore.

Not necessarily. As you can see in this thread, no matter what offset value I had it at (went as high as +0.060V), I got random WHEA errors.

It seems one notch bump in VTT voltage did the trick for me (as mentioned earlier). This actually allowed me to lower offset all the way to +0.005V. This is most likely because of 2133MHz CL9 G.SKILL RAM.

Is anyone not getting any whea errors ever?

I seem to get a random whea error every 1-2 days.

From what I've seen, core temperature also plays big role when it comes to WHEA errors. Check the time of WHEA error being issued and check the core temperature around the same time. See what kind of temperature you were getting at the time of WHEA errors being issued.

 

[Ferzerp]

From what I've seen, core temperature also plays big role when it comes to WHEA errors. Check the time of WHEA error being issued and check the core temperature around the same time. See what kind of temperature you were getting at the time of WHEA errors being issued.

I think it's more that the higher load increases the chances (more operations, more chances, etc, just like stress testing), and the higher load makes higher temps.

Not necessarily. As you can see in this thread, no matter what offset value I had it at (went as high as +0.060V), I got random WHEA errors. It seems one notch bump in VTT voltage did the trick for me (as mentioned earlier). This actually allowed me to lower offset all the way to +0.005V.

Interesting. They're the reason I have vcore about .05V higher than I had it at one point. I'll have to try that.