Stubborn E6600 OC problems (BSOD at startup)...

[Dew itt right]

Well, the systems in the sig. I recently replaced my e6300 and 4x1Gb Corsair XMS2 DDR2-800 cas4's (FOR SALE HERE) with the E6600 and DDR2-1066 Dominators in the sig.

System's OC history...
**Ran old E6300 daily at 3.0GHz (429x7) and RAM OC'd to 860MHz (4-4-4-12)
**Topped out the E6300 at 3.15GHz (450x7) with RAM at 900MHz (4-4-4-12)
-------At max, E6300 reaches 58C

**Installed new E6600 and 2Gb 1066MHz Dominators
**Runs great at 3.0Hz (427x7) with RAM at 1068MHz
**Runs great at 3.2GHz (356x9) with RAM at 1068MHz
**BSOD's (shown below) at bootup at 3.4GHz (427x8) with RAM at 1068MHz
**Backed it down to 3.3MHz and Prime fails in seconds.

The part I don't get is that temps after 7 hours of Prime at the max OC so far (3.2GHz shown below) only reached a cool 44C and the voltage (set to 1.4V in BIOS) only hits 1.296V. But it seems I have to be maxing out the CPU since the FSB has definitly been higher before and the RAM isn't changing speeds at all. Am I missing something here or did I just get a lemon of an E6600? Any input is appriciated...



Screenshot of stable 3.2GHz OC with new E6600...

BSOD at startup when booting higher than 3.3GHz...

 

[unclewebb]

If a bios setting of 1.4 volts is only getting you 1.296 volts while running Orthos then you are likely going to need some more CPU voltage for stability at 3400 MHz. Your temps are great so you don't need to worry about that. I need approximately 1.40 volts at 3400 MHz while running Orthos so try bumping your core voltage. I find SpeedFan does the best job of reporting core voltage at the higher levels. CPUz is good up to a point but then it starts reporting garbage.

 

[Dew itt right]

Well, at 3.4GHz (427x8) w/ RAM still at 1068MHz I bumped the BIOS Vcore up to 1.425V and I actually made it to the desktop before BSOD'ing. I bumped it up to 1.45V and Vista fully loaded, ran Prime and failed in seconds. Bumped it up again to 1.475V and I was having problems booting. I probably could have gotten it to boot but at that point its pretty much told me all I need to know about it's stability.

I bumped Vcore back down to 1.45V and tried for 3.28GHz (410x8). Prime ran for about 8-10 minutes befor failing. Temps again only in the low 50's.

 

[RossY]

So based on your ram clock you are running some sort of divider to max your ram. Try and overclock using a 1:1 divider and see if that gives you some more stability at the higher FSB levels. There have been people that have issues using a divider other than 1:1 when trying to do anything other than a mild overclock and of course we don't want you to do just a mild overclock. I don't know if you've run a different divider before the e6600 so it may be the memory controller on the board acting funny rather than the cpu.

 

[SR45]

I see your E6600 Multiplier is set at x8, but should be set at x9 for the E6600. This may be an issue as well

 

[oneil]

I see your E6600 Multiplier is set at x8, but should be set at x9 for the E6600. This may be an issue as well

Beat me to it.. I'm getting the E6600 and I dont see why a 3.6Ghz OC isint possible with stock ram @ 400FSB

 

[unclewebb]

Why not run MemTest86+ and experiment with your OC settings and voltages before heading into Windows?

If you are looking to maximize your cpu overclock then you might have to sacrifice some memory speed by running it at 1:1 instead of 4:5. How much memory voltage are you using as well as northbridge voltage?

 

[Dew itt right]

So based on your ram clock you are running some sort of divider to max your ram. Try and overclock using a 1:1 divider and see if that gives you some more stability at the higher FSB levels. There have been people that have issues using a divider other than 1:1 when trying to do anything other than a mild overclock and of course we don't want you to do just a mild overclock. I don't know if you've run a different divider before the e6600 so it may be the memory controller on the board acting funny rather than the cpu.

I did try a 1:1 at 3.4GHz (427x8) with the RAM crawling along at 854MHz and it still failed. I automatically assumed it couldn't be the RAM's fault becase the PC is 100% stable at 3.0GHz (427*7) but completely UNstable at 3.4GHz (427*8) with the same FSB and the same memory divider and speed.

I see your E6600 Multiplier is set at x8, but should be set at x9 for the E6600. This may be an issue as well

I'll try OC'ing it a bit more with the 9x next. Obviously I was looking for some higher FSB clock but we'll see if I can't atleats get a bit higher with a higher multi and lower FSB's.

Why not run MemTest86+ and experiment with your OC settings and voltages before heading into Windows?

If you are looking to maximize your cpu overclock then you might have to sacrifice some memory speed by running it at 1:1 instead of 4:5. How much memory voltage are you using as well as northbridge voltage?

At the 3.4Ghz (427x8), I tried running memtest with the RAM at 1068MHz (4:5 divider) and it ran fine for a couple hours. The RAM appears to be doing fine. I'm also going to try maxing out the individual components (FSB, RAM, & CPU) with everything else significantly underclocked and see what kind of numbers I get. I believe I've been using 2.3Vdimm and 1.4Vnb as well.

 

[unclewebb]

Memory voltage looks good. You might want to try boosting the northbridge up a notch.