[Basy40]

Hello and greetings,
My Computer won't POST after two weeks of stability. I loaded the default settings last night, then loaded my O.C. profile through my ASUS P6T O.C. profile in the Bios, rebooted about 6-7 times to make sure it would POST (which it did), did not see any black artifacts whatsoever, passed 12.5 hours of stress testing last night with Prime95 and now this morning it DID NOT POST. Below are my System Components and Bios Settings. Please help me find out what is going on with this random instability.

Coolermaster Storm Sniper
I7 930 O.C. to 3.57
Asus P6T Mobo
1000w Rosewill power supply (which is two years old and is only drawing 611w to power everything in my case)
Radeon HD 5970
Soundblaster X-Fi sound card
x4 200mm fans (Intake), x1 140mm fan (Intake), x1 120mm rear exhaust fan
Corsair H100 (with x2 120 mm fans in pull config.)
WD 7200 RPM 1 TB HD
12 Gigs Kingston HyperX Genesis 1600Mhz Ram

BIOS SETTINGS
x21 Multiplier
170 Bclock
1023 DRAM Freq (6-6-6-18 timings)
2046 Unclock Freq
1.325 Vcore
1.35 QPI/VTT
1.1 IOH
1.5 DRAM BUS Freq
All C-state/power saving settings disabled
Load Line Calibration enabled
Everything else is on auto settings

I followed Miahallen's 3 Step Overclocking guide to a T and I still have this random not being able to POST issue after 2 weeks of stability

[redduc900]

Raise the QPI/DRAM Core voltage to 1.40V, and take the DRAM multiplier off of x6. Bump it up to x8 = an effective DRAM freq. of DDR3-1360, w/ the UCLK freq. set to 2720MHz and QPI Link Data Rate at 6120MT/s

[Basy40]

Ok, i did what you said and it failed to POST during the 4th hard boot from shutdown. Going to increase Vcore to 1.35 and decrease QPI to 1.35 since that didn't help and will see how many successful POSTS I can get. Will leave all other settings the same that you recommended.

P.S.
Does having Load Line Calibration Enabled make a difference as far as stability?

[Basy40]

Ok so increasing the Vcore from 1.325 to 1.35 with QPI at 1.35 made no difference either. It failed to POST on the 5th hard boot from shutdown.

[muddocktor]

Not to sound like I am criticizing or anything, but how long are you letting it sit with the no post situation? The reason I ask is that I have a P6T also and I've noticed before that sometimes it seemed like it took a while for it to get to the post screen. I was using this board for my heatsink test system board until I also started getting some random vcore instability, where the vcore would slightly increase itself for no reason and with nothing changed in bios too. It didn't affect running too much but made for too much error for heatsink testing in my situation.

[Basy40]

I let it sit for 15-20 seconds. If it's going to POST then I'll see the Num Lock, Caps Lock, and Scroll Lock lights all flash. That's an indicator that it's definitely going to POST. If it doesn't and it just hangs at a black screen then I know the O.C. failed. Then I have to restart it manually with the power button . Once the post screen comes up it tells me that the overclock failed and to press F2 to load default values.

[Basy40]

BTW, my It 930 is a D0 version and from what I've heard that's a better one than the C0.

[Rotary]

remove you're sound baster, i had the same problem and it was the sound card

[Basy40]

How do you know that it was your sound card? I'm just trying to understand the logic behind that. I hope that's the case

[Rotary]

i tried everything and the last thing i did was remove the sound card then bam, it worked fine

[Basy40]

ok well i tried that at those settings and it still failed to POST. Do you know if overclocking can cause degradation of a hard drive? The only reason I ask is because I read that failure to POST is also a symptom of hard drive failure. If the hard drive is starting to fail due to overclocking, then a POST failure makes sense since everything is stored on the hard drive

[Basy40]

So does anyone else have any suggestions as to why my system would fail to POST after two weeks of stability as listed above? I also downloaded and ran the S.M.A.R.T long ctl test to make sure that all of my overclocking hadn't messed up my hard drive and every time I've run the long test it has passed with no errors. So I think my hard drive isn't causing the instability. I've also successfully passed 3 iterations of MEMTEST86 so my RAM should be good to go. Any help from another experienced overclocker would be greatly appreciated.

[Techjesse]

Update your BIOS, make sure you reset CMOS before updating. Your i7 930 will run at 4.0GHz with no problems. I'm using the same i7 930 that hit 5.0GHz http://valid.canardpc.com/show_oc.php?id=2048724

[Basy40]

I updated the BIOS for the P6T before I started O.C.'ing. I don't have to reset CMOS because if it doesn't POST, I do a hard restart with the power button on the case and the next POST screen will tell me to "press F2 to load default values" which works every time. So this works in place of manually using the CMOS. Every processor is different so you may have gotten a well made 930 that requires minimal voltage. Also you have a different, and probably better, MOBO which probably overclocks easier than my bargain P6T.

I currently have it O.C.'d to 3.36 with the following settings and it's rock solid stable.

Multi: x21
BCLOCK: 160
RAM: 1604 Mhz(9-9-9-27 timings, effective RAM multi. of 10. It also runs solid with the ram underclocked at 960. I had the ram on lowest setting to remove that from the overclocking equation so it wouldn't be a possible variable with instability)
Vcore: 1.25
QPI: 1.4
IOH: 1.1
DRAM voltage: 1.64 (recommended 1.65 by JEDEC but the P6T increase by .02 and i
don't want to go over 1.65 for safety sake)
All C-States disabled
Load Line Calibration enabled
All spread spectrums disabled.

Are there any other voltages that I could adjust which would noticeably increase stability like PLL? All other voltages are on AUTO

[Basy40]

Does anyone else have a clue? Are there any other voltages that I should adjust that could fix this random instability issue?

[redduc900]

Again bump the VTT (QPI/DRAM Core voltage) from 1.40 to 1.45V.

[Basy40]

Isn't that beyond Intel's specifications for QPI? I don't know about you but that seems pretty high to me. I don't feel like destroying my CPU

[Basy40]

I just found the max QPI settings from Intel's website:

http://download.intel.com/design/pro...hts/320834.pdf

Table 2-6 says that the Max QPI Voltage is 1.35, although if you choose the XMP profile in order to run my RAM at 1600 MHz, it automatically increases QPI to 1.4. Personally I wouldn't want to push it beyond 1.4.

If you look at the notes section under table 2-6, it says that excessive overshoot or undershoot on any signal will likely result in permanent damage to the processor.

I did notice that my stock Vcore voltage was around .92 in CPUZ. It could be that I'm giving it too much Vcore voltage at 1.325 or 1.35 and that a QPI of 1.35 is actually fine. I guess I'll have to try decreasing Vcore incrementally and leaving QPI at 1.35.

Could too much Vcore cause instability?

[Basy40]

Does anyone else have a clue?

[evilrad]

Try setting the QPI/DRAM voltage to stock (I believe it's 1.1) and work your way up. Too much volts is just as bad as too little.
I have a thread that I started a while back on stabilizing my 920. Take a look at that thread and you'll see the "fun" I had. Everyone was saying, "More voltage" when, in fact, it was less that was needed.
I'd follow the OC guide in the stickies, except the part where it's says to start out with the QPI/DRAM voltage at 1.3, start with stock.

Edit: One day I need to update my sig.