I test it at stock for 24 hours last week and at 460 x 6 last week and it was fine... I did all that trying to make sure that it runs. Prime is stable @ 3.6 (400x9) everything on auto... I don't get it.
I test it at stock for 24 hours last week and at 460 x 6 last week and it was fine... I did all that trying to make sure that it runs. Prime is stable @ 3.6 (400x9) everything on auto... I don't get it.
My $0.02, I think you need some more NB. Probably somewhere around 1.3v +/-
I think your vcore & vtt look about right. No idea on the GTL though...
I think your vcore & vtt look about right. No idea on the GTL though...
Dude, shoulda got a Gigabyte board instead of the Asus, 450fsb on the UD3(r/p) is a walk in the park. You don't have to take my word for it, or my overclock as evidence. Gary would seem to agree.
"We have absolutely no concerns about recommending this board for 24/7 500FSB+ use. This board overclocks quad-core chips just as easily as it does dual-core offerings, which is a feat that most P45 boards cannot accomplish."
1.3625 Intel Specs
I read both were good... I would just try that board but then I'm on an OEM version of Vista and I've already swapped mobo's once and had to call them. The only thing that is original from the initial installation of Vista is the PSU & Case.
Anyways even on my ASUS, I'm stable at 460 FSB so I don't think the mobo is the problem.
BonzaiDuck
Lifer
- Jun 30, 2004
- 16,632
- 2,027
- 126
Like I've said before as a matter of strong suspicion from looking at "customer reviews," threads and posts here, etc.: the way the UD3R or UD3P boards ship with current BIOS, it is very likely that they work like a charm with Wolfie and Yorkie cores. So . . . . "Q9650?" I'm inclined to think that could be an awesome combination. I just don't know firsthand.
Alright! I think I'm getting somewhere. I started using IBT for stress testing because it seems faster.
I IBT stable after 5 passes with these settings:
1.3750 vCore ~ 1.58 vPLL ~ 1.4 vVTT ~ 2.1 vDRAM ~ .69x GTL (0/2) ~ .69x GTL ~ 1.30 vNB ~ .64x GTL ref
IntelBurnTest Stable 5 Passes @ 4GHz (445x9)
***EDIT*** Going to run 20 passes now and see what happens.
I IBT stable after 5 passes with these settings:
1.3750 vCore ~ 1.58 vPLL ~ 1.4 vVTT ~ 2.1 vDRAM ~ .69x GTL (0/2) ~ .69x GTL ~ 1.30 vNB ~ .64x GTL ref
IntelBurnTest Stable 5 Passes @ 4GHz (445x9)
***EDIT*** Going to run 20 passes now and see what happens.
n7
Elite Member
- Jan 4, 2004
- 21,281
- 4
- 81
NB voltage is pretty low...you likely need more.
If you got to 20 passes stably though, you're obviously very very close to stable.
I've gotten lucky for my 4 GHz operation & am only using 1.24v, but when i had my Q9550 or have run my Q9650 over 450+, i need a lot more NB voltage.
BTW, have you run HCI Memtest for a few hundred % or higher to ensure RAM is fully stable?
It's a good test of the RAM/NB that's far more sensitive than Memtest86+ is.
If you got to 20 passes stably though, you're obviously very very close to stable.
I've gotten lucky for my 4 GHz operation & am only using 1.24v, but when i had my Q9550 or have run my Q9650 over 450+, i need a lot more NB voltage.
BTW, have you run HCI Memtest for a few hundred % or higher to ensure RAM is fully stable?
It's a good test of the RAM/NB that's far more sensitive than Memtest86+ is.
Just realized I didn't update the NB, it's now at 1.3 I edited post above to show it accurately. I cut and pasted from an earlier post.
n7
Elite Member
- Jan 4, 2004
- 21,281
- 4
- 81
Oh btw, use LinX: http://www.xtremesystems.org/f...howthread.php?t=201670
IBT doesn't show error till after the run is done, which is retarded, since you can run 100 times & never realize it has errored after the first one.
LinX will stop & let you know when that happens...
IBT doesn't show error till after the run is done, which is retarded, since you can run 100 times & never realize it has errored after the first one.
LinX will stop & let you know when that happens...
- Sep 13, 2008
- 8,226
- 3,131
- 146
Ya, you may need more voltage for the NB, especially with your 4 sticks of ram. I don't know what your safe range is on the p45, but from my days of ocing the 780i FTW, upping north bridge V is important, especially with more than 2 sticks of ram. anyways, mine is with a q6600, so again, different. since the fsb is going to be very high due to a lower multi on the yorkfields in general, VTT or FSB voltage will also be important.
How high do you think I should go up on the NB? I'm pretty much maxed on VTT and vCore @ 1.4 and 1.3750 respectively. I've gone as high as 1.46 on NB. If it's a GTL issue, then I'm completely lost. I've tried everything from .635 of all three settings up. Obviously I may have missed a few but for the most part I've hit pretty much everything. I've been at this for nearly 3 weeks and I haven't been able to get stable at anything higher then 3.6. I think I just got a bad chip...
Not that anything is wrong with it, just that it can't do very well on the OC side of things.
Not that anything is wrong with it, just that it can't do very well on the OC side of things.
You are risking a lot by setting volts that high. I don't think it's worth it if you think you've got a "less gifted" chip yourself. Would you rather have another 400MHZ and a soon-dead CPU or one that is 0.5% slower in real life performance and lasts approx. 4 years longer?
I think my chip is burnt up... it works at 3.6Ghz just fine but it just BSOD on voltage configurations that were passing 3, 5 13 passes on Linx... I think putting it over 1.4 on NB was a bad idea. Even though I'm seeing it done in other places.
Every chip is different. And the power supply is the second most important part when overclocking, equal to the RAM. Sorry if it's been answered before but which PSU are you using anyway?
I'm using Thermaltake 700w SLI certified. I think 700 should be enough... at on point I had it running at 4ghz and it completed 2 runs for 3dMark Vantage w/the video card OC'd aswell... maybe it is a problem though I don't know.
Yeah it's toughpower. Even at 4ghz though the performance like you said isn't much better so I think I'm going to leave it at 3.6 for the time being.
I'm somewhat convinced that my mobo or ram or cpu is done at this point... Linpak is not stable past 3.0Ghz though 3.6 and 4.0 seem to work on everyday apps... I think I f'd something up... I'm leaning towards the NB but I really have no basis for that maybe nothing is wrong and the Bios is just all confused so my next step is to flash the bios and see if that helps.
N7, what Bios are you using? I flashed from the oldest bios 601 to the newest 2001 at the beginning of this adventure...
What do you think I was more likely to burn up? How can you tell?
N7, what Bios are you using? I flashed from the oldest bios 601 to the newest 2001 at the beginning of this adventure...
What do you think I was more likely to burn up? How can you tell?
n7
Elite Member
- Jan 4, 2004
- 21,281
- 4
- 81
I've used a number of different BIOSes since i got my P5Q-D...
You have the P5Q-E though...the versions may not be comparable between our boards; i don't remember.
I used 1406 for a while; been on 1805 for the last couple months.
Reset your CMOS man...if everything seems funny, it's likely a setting you've changed along the way & forgotten that you did.
TBH, as i tried to explain, you cannot get the P5Qs w/ 8 GB stable at high FSBs simply by pumping in the voltage...it just does not work.
Believe me, i've tried that method...it's not the path to success.
You must get VTT/vNB + GTL perfect, or it will not happen.
Like i said earlier, TBH, if you cannot get GTLs figured out, you're going to want to settle with a lower OC.
It's not an easy or fun process at all, which is why many people prefer boards like the Gigabyte UD3R/P, which seem to run just fine if you just pump up the voltage w/o messing with GTLs much.
That said, the big thing to remember is that by far, the majority of people out there OCing only run two DIMMs, & it's so much easier OCing with only 2x2 GB or 2x1 GB than 4x1 GB/4x2 GB, it's not even funny.
So what comes easily to most overclockers...well those of us with more RAM sometimes have to really fight for.
You have the P5Q-E though...the versions may not be comparable between our boards; i don't remember.
I used 1406 for a while; been on 1805 for the last couple months.
Reset your CMOS man...if everything seems funny, it's likely a setting you've changed along the way & forgotten that you did.
TBH, as i tried to explain, you cannot get the P5Qs w/ 8 GB stable at high FSBs simply by pumping in the voltage...it just does not work.
Believe me, i've tried that method...it's not the path to success.
You must get VTT/vNB + GTL perfect, or it will not happen.
Like i said earlier, TBH, if you cannot get GTLs figured out, you're going to want to settle with a lower OC.
It's not an easy or fun process at all, which is why many people prefer boards like the Gigabyte UD3R/P, which seem to run just fine if you just pump up the voltage w/o messing with GTLs much.
That said, the big thing to remember is that by far, the majority of people out there OCing only run two DIMMs, & it's so much easier OCing with only 2x2 GB or 2x1 GB than 4x1 GB/4x2 GB, it's not even funny.
So what comes easily to most overclockers...well those of us with more RAM sometimes have to really fight for.
Yeah, I couldn't even get it stable with only 2 dimms and then when I put all 4 back they didn't work so I cleared CMOS and it worked but I was still getting BSOD's on settings that were close to stable...
TRENDING THREADS
-
Discussion Zen 5 Speculation (EPYC Turin and Strix Point/Granite Ridge - Ryzen 9000)- Started by DisEnchantment
- Replies: 25K
-
Discussion Intel Meteor, Arrow, Lunar & Panther Lakes + WCL Discussion Threads
- Started by Tigerick
- Replies: 22K
-
Discussion Intel current and future Lakes & Rapids thread- Started by TheF34RChannel
- Replies: 23K
-
-
News NVIDIA and Intel to Develop AI Infrastructure and Personal Computing Products- Started by poke01
- Replies: 411
Facebook
Twitter