Conroe Overclocking on P45

Soldato
Joined
1 Jan 2008
Posts
11,421
I recently swapped my Gigabyte GA-P35C-DS3R for an Asus P5Q-Pro, but am having some trouble overclocking.

Spec:
E6750 (Conroe, E0, 6-8 multi, VID 1.350)
Asus P5Q-Pro (BIOS 2002)
4x1GB Geil PC2-6400 (4-4-4-12, 1.9-2.1v)
GTX280
Rest in sig.

The gigabyte board was able to take this setup to 3.2Ghz on all stock volts with RAM at 1.9v (+0.1v), stable. It could do 3.4 but would sometimes just fail to boot despite being stable once it had.

My aim was to get to 3.6 with this new board, and I have had it boot and game for a few mins at 3.6, but it was not prime (blend) stable. However, it was stable in prime with small FFT test. This made me think the RAM/NB was causing a problem. CPU temps never seemed to go over 50 deg at load.

The NB voltage requires a small bump to boot with 4 sticks (5-5-5-18) at CPU stock (1.20v thats +0.1v), and is prime blend/FFT stable. However, when increasing the FSB, seemingly by anything, the thing will fail a prime blend test within a few minutes, despite still being small FFT stable.

I've had the NB volts to 1.34v, which lets it run for a few more mins before failing the blend test again. Tried booting NB at 1.36v, and it bluescreened before I could get into windows. The DRAM volts I have tried at 1.9-2.02v with no success (at 5-5-5-18).

Can anyone help me here, i'm not sure what to try next besides taking out 2 sticks of RAM? It looks like the CPU is fine, but the NB/RAM is struggling here. I'm about to run a memtest too, though the RAM has been perfectly fine in the other mobo for a long time (didn't require any increase in NB volts either!).

Thanks in advance!
 
Funnily enough i was about to make a new thread about a similar problem.

I also have the P5Q Pro and a Q6600 G0. I can get it to 3.0Ghz prime stable at:

1.3950v CPU
2.2v vDimm
1.38v NB

but to get it any higher (even 3.2Ghz), it doesn't want 1.450v CPU, 1.45v NB, 2.2v vDimm

:eek:

Now i understand that there is always a sweet spot in terms of FSB and voltage, but this seems bizarre and something that i think is down to the BIOS.

@alexhull24 - Do you think a BIOS update will sort things out. Is your board new like mine?
 
Interesting.

Not sure what a BIOS update could do, i'm on the latest (2002), but there may be a better version for overclocking.

My board was bought from b-grade (twas a bargain!), i'm pretty sure it's an RMA return, so I assume it's been tested working etc.

ozzie, what RAM configuration do you have? 1.38v on the NB seems high to me, I assume you have 4 sticks?
 
Nope, just two sticks of memory. Last time i checked with prime, any lower and it fails. :o

I'm going to dig a little more, see what i can find. :)
 
I just whipped out 2 of my sticks and i'm still having the same problem, though I lowered the NB volts to 1.26v this time, still 0.16v higher than stock.

Let me know if you find anything relevant. I've searched a quite a few threads, but every board/chip/ram combo is different!
 
What cpu pll and fsb-term voltages are you using, as theese usually need a nudge or two as well.
 
I've tried upping them one notch each, but it made no difference.

There are some modded BIOS available here, which suggest they improve overclocking and memory compatibility, which made me think, perhaps my memory is not fully compatible with the board. I know the P5Q's can be very picky, but didn't think it'd be an issue on the latest BIOS.

I just had a random bluescreen while typing this too, darn. I was trying 3.2 (fsb 400), and it was definitely fft stable as that passed for ages, just randomly happened about 30mins after testing. The blend test still doesn't pass. I'm going to predict that memtest will flag up some errors too with any sort of OC, perhaps even at stock.

It was also suggested that prime wasn't a good program to test this board with as it might not be compatible? Never heard of that myself but was on the forum i've linked to.

EDIT: All 4 of my sticks are back in now as it didn't make an iota of difference!
 
Tried the custom BIOS 2002 at 3GHz and it's prime blend stable for 1hr. Not really good enough.

The tweaking continues....

EDIT: So it's prime blend test stable for 1hr at 3ghz with nb volts at 1.10v, then if I up it to 3.2ghx and the nb volts to 1.30v, it fails within seconds. Not quite sure where to go from here except trying different memory?
 
Last edited:
Back
Top Bottom