Did you test stability before overclocking? did you try running stability tests again? the same ones you passed to deem current OC stable. And what were those tests and for how long did you run them exactly please?

I had a ud4 as well.Make sure you have a beta bios that has pll overide and make sure it is enabled. As others have said, you may need more voltage. The one thing with SB, what ever the memory I have I would just enable 'quick' settings tab in the bios and manually put in the memory. Never gave me an issue on the ud4 or ud5, make sure you enable 'standard' as memory type not the 'ehanced' etc mode memory tab.

I was running 1.34 vcore (in bios) for about 1.5 weeks when I got my first BSOD. It would fluctuate SLIGHTLY to around 1.35 under full load. The cause of the BSOD certainly could be something simple such as too low of vcore.

Are you sure about this? The ud4 has more vdroop than that as it has only enabled or off as options for llc in bios. I would have to run my voltage very high in bios, 1.38+ (to get to 4.7 ghz) to get an idle voltage in windows of 1.365 to get a load voltage of 1.34 or so...that was with all beta and stock gig bios this was only possible to be bsod free using a bios with pll overide and having it enabled in the bios. I don't see how you could have that small of vdroop on a ud4? Voltage readings right?Edit..btw I have never touched the actual pll voltage setting, l have always left it on auto without any bsod issues on both boards. I just enabled pll overide and that is it.

Are you sure about this? The ud4 has more vdroop than that as it has only enabled or off as options for llc in bios. I would have to run my voltage very high in bios, 1.38+ (to get to 4.7 ghz) to get an idle voltage in windows of 1.365 to get a load voltage of 1.34 or so...that was with all beta and stock gig bios this was only possible to be bsod free using a bios with pll overide and having it enabled in the bios. I don't see how you could have that small of vdroop on a ud4? Voltage readings right?Edit..btw I have never touched the actual pll voltage setting, l have always left it on auto without any bsod issues on both boards. I just enabled pll overide and that is it.

LLC is MUCH improved on the new UD4-B3. Here's a table of voltages from the P67A-UD4 thread at xtremesystems as reported by EasyTune 6 and actually measured using a digital multimeter (DMM). The voltage column is the voltage set in the BIOS. The voltage range we are really interested in is 1.30 to 1.45LLC Enabled (P67A-UD4-B3, F3b bios):There are a few conclusions we can make from this table-Comparing ET6 voltage readings with DMM voltage readings, it looks as if the sensors are underreporting voltage (by approximately 0.02v)-According to the DMM readings at load, voltage is practically dead on with what is set in the BIOS (approximately 0.01v overvoltage or less is good)-Almost no vdroop according to the DMM readings (maybe 0.003v)Question is, how close are the DMM voltages to the actual voltage at the CPU?

Thanks Corey, did not realize he in on a b3 board. That is hell of alot better than my old b2 ud4 board llc. Maybe he needs more vdimm...some ppl with the low volt ram (venegance) have had issues with stability...I suspect he will be bsod free at stock...some chips need more vcore to get to 4.5, it just depends...seems odd that it was fine initally. I doubt degradation at that voltage though for sure.

They all occurred when I was running FSX. Though they didn't start up until a few days ago. This is what gets me... why now?ET6 numbers:3.3v - 3.363 idle5v - 5.021 idle12v - 12.319 idledram - 1.5 (all the time)cpu - 1.236 - idle

Ryan, is the et6 cpu idle value above at stock, not oc'd? It is showing here 1.236? I thought your idle voltages are 1.35?

Did you test stability before overclocking? did you try running stability tests again? the same ones you passed to deem current OC stable. And what were those tests and for how long did you run them exactly please?

Just completed the Memtest stuff and OCCT run with STOCK settings (except changing RAM to XMP profile 1 which runs them at the correct timings) while at work.Memtest ran for two passes both sticks (about 2 hrs long), with NO errors. OCCT ran for a little over 8 hrs, NO errors. Highest temp 50C (core 3 via RealTemp)Considering in the 1.5 weeks I went without errors at 4.5 GHZ, and rarely got to 100% load (then turned my PC off each day), I deem 8 hrs of OCCT to be stable at default settings. My theory on the BSOD's while running 4.5GHz is that in the time between turning off my PC for the night, and running idle (probably 75% of the time it was on), it never really got to "that" point where some component was unhappy. The breaking point probably occurred recently, where I've been flying a LOT of flight sim, with very demanding FSX settings/planes, forgot to turn off the PC a couple of nights in a row etc etc. I only ran OCCT for a few hours on that 4.5 GHz setting. I figured if the system would hiccup it would have been right away - wrong pffff!

Ryan, is the et6 cpu idle value above at stock, not oc'd? It is showing here 1.236? I thought your idle voltages are 1.35?

It WAS, but I posted earlier in the thread I'm back at stock settings for testing purposes.

Looks like you just need to bump up your Vcore a notch or two

Well I had another "idle" BSOD today - first in about a week. Come to think of it, every one of my "idle" BSODs has occured while watching youtube - so I guess I have the youtube BSODs.

Well I had another "idle" BSOD today - first in about a week. Come to think of it, every one of my "idle" BSODs has occured while watching youtube - so I guess I have the youtube BSODs.

Ugh, that's nuts. Could be anything, join the club! :PI'll run the system how it is for another week and make sure I don't pick up the you tube BSOD bug LOL!p.s. I can't even play Portal 2 without it CTD thanks to some stupid saved game bug GRRRR!p.s.s - Seasonic wrote back and said "for fun" I could try testing with their X850 - he made it sound like I could just swap the 850 for the 750 for free :Batting Eyelashes:

There's nothing wrong with having a little too much power. Do it!!!Did you see Gigabyte's EFI bios? I really didn't like it when I saw the still images of it - It looks a bit cheesy. Now that I have seen it in action, I have to admit that it looks pretty cool and definitely seems to improve on the functionality of a regular bios.

There's nothing wrong with having a little too much power. Do it!!!I agree. At least it will eliminate it as a suspect, if nothing else.Did you see Gigabyte's EFI bios? I really didn't like it when I saw the still images of it - It looks a bit cheesy. Now that I have seen it in action, I have to admit that it looks pretty cool and definitely seems to improve on the functionality of a regular bios.

Saw the video, looks easy. A number of my friends are really scared of their Gigabyte Bios settings. This should help.Kind regards,

Saw the video, looks easy. A number of my friends are really scared of their Gigabyte Bios settings. This should help.Kind regards,

The old school look is the only thing I'm used to lol!Can't wait to get this thing back to 4.5 GHz - ugh, I really notice the difference, Flight1 Mustang VC at FSDT KORD with 8.5 LOD and very dense autogen.... it's like 16 fps vs 26 fps

Yeah, I can't even run it at stock any more while used to 5. 3.6 is like almost half the performance only. LOL...We are so spoiled.

The old school look is the only thing I'm used to lol!Can't wait to get this thing back to 4.5 GHz - ugh, I really notice the difference, Flight1 Mustang VC at FSDT KORD with 8.5 LOD and very dense autogen.... it's like 16 fps vs 26 fps

Me too; I plan never to go into my bios again now things are working. I could care less what it looks like:)