GA-EX58-UD3R,,,and odd BIOs behaviour after OCing

[Roffey123]

I'll warn you now...chances are I've commited some nubbins mistakes - so try and keep the face-palms down to a minimum...phweeeeeaasssseeee?

*imagine cute puppy picture here*

Now that's out the way, I'll explain the situation.

A couple of weeks ago I managed to get myself an i7 mobo [EX58-UD3R]/CPU [i7 920]/RAM [G.Skill 3GB] setup, along with a third party cooler [ThermalRight Ultra-120A with LGA1366 adapter]. Now the intention was to grab the new CPU and do a mild OC, temps willing. To begin with...that was easily achieved, a stable 3.2GHz OC with max temps of 70C (I had read up on maximum temperature ranges of the i7, looks like it can handle a toasty 80C) when running Prime95 - all by adjusting the BClk to 160 (again, research on the interwebs says that this was perfectly possible without the CPU voltages being adjusted).

Now the mystery quirks kick in.

First off, the computer started double booting everytime I turned the PC on. I suppose I should of taken the hint *slaps self* and looked into my BIOs settings further - but since it didn't affect the PC at the time...I was willing to live with the inconvience. All was fine until I decided to boot the PC up about a week after OCing...and the thing kept BIOs posting and then restarting. Odd...I had thought...as I prepared to open the innards and see if it was the RAM that was causing the trouble. Thankfully, since there's a Backup BIOs on the UD3R, it managed to use the backup to revive the main BIOs before I started butchering the insides.

"Hmm," I thought, "perhaps I should up the the voltage slightly to stop that double boot, its probably the reasonn why the BIOs got corrupted." This I did, not by much (less than 0.1V on both the CPU and QPI voltages); and lo, the double boot problem was cured. "Sorted," I thought, as I reinstated the overclock - negating to Stability test, reasoning that the OC was stable, since only the voltages were upped from the previous OC...I can feel the facepalms and gloats already...

Well, needless to say, another week and a half passes (and a couple of rare...for me....Vista BSoDs)...and then it happens again. Another BIOs corruption...

Now I'm sure that power isn't the issue (a 650W Antec TruPower Trio clears this issue up), and the RAM isn't running over-spec, the fact that the CPU can OC well means it isn't a dud - which leds to two conclusions...either the BIOs doesn't like the OC, and to OC without corrupting it means a BIOs flash to a more recent version, OR, my stability tests were obviously not thorough enough.

Any ideas and suggestions to help with curing the problem would be greatly appreciated.

 

[Cache]

My personal experience says that a lot of BSOD's are due to memory issues, and given that the I7 is persnickety about the voltages going to it's RAM... well, my first gut reaction is to ask what your RAM voltage looks like. But--and I'm assuming that it ran just fine doing anything you want when it was running. Could you post some of your current settings? Also, is it having issues when you bring everything back to stock?

I know that most motherboards are coming out with new BIOS updates every few weeks on the I7's, so keeping current on that is always highly recommended.

 

[Roffey123]

Well...in answer to your RAM question: it's running at 1.5V; and yes and it ran fine once the OS was up and running, hell, I even put it through its paces by rendering a scene in 3DS Max while OC'd. Worked a dream (love the fact that a scene that once took an hour on my e6600 now takes 19 mins). At stock it runs fine, no issues what so ever. I'll post my OC BIOs settings later on...I'm suffering from a cookie defeciency.

 

[Roffey123]

BUMP.

Can't believe I forgot about this for a week.

Anyway I've looked into my settings, and these are the ones I have changed:

Multiplier: 20x
QPI Link: x36
BCLK: 160
Memory Multiplier: 8
Memory Clock: 1280MHz (certified for 1333MHz, memory voltages left alone)
CPU vCore: 1.23v (increased to stop double boot)
QPI voltage: 1.215v (again, same reason as above).