Overclocking E8400 on P5Q-Pro, stuck at 4.32GHz- Suggestions?

jctazzy

Limp Gawd
Joined
Nov 22, 2008
Messages
150
I've been overclocking my E0 E8400 incrementally since I built my new system last fall. I seem to be stuck at a plateau and I'm looking for any suggestions on the P5Q-Pro that can help me past it.


This is my "everyday" configuration and has been stable through many days of Prime 95, IntelBurnTest and months of gaming:

FSB 450 (4.05GHz)
PCIE 100
DRAM Freq 901MHz
CPU v 1.275
vdroop 1.26
DRAM v Auto
NB v Auto

Max temps (8 hours Prime 95)
C1 temp 44ºC
C2 temp 44ºC
Any BIOS settings not specified are AUTO


After some voltage tweaks and stress testing I came to this. Note the .05v jump required as well as 10ºC temp gains just for 270MHz.

*Stable 8 hours Prime 95*

FSB 475 (4.275GHz)
PCIE 100
DRAM Freq 951MHz
CPU v 1.3250
vdroop 1.312
DRAM v 2.05
NB v 1.26

Max temps (8+ hours Prime 95)
C1 temp 55ºC
C2 temp 53ºC
Any BIOS settings not specified are AUTO


Lots of incremental tweaking to get to the next stable overclock. Again a .05v jump and now a 7ºC temp jump just to get 45MHZ.


*Stable 8 hours Prime 95*

FSB 480 (4.32GHz)
PCIE 100
DRAM Freq 961
CPU v 1.3750
vdroop 1.360
DRAM v 2.10
NB v 1.28

Max temps (8+ hours Prime 95)
C1 temp 62ºC
C2 temp 60ºC
Any BIOS settings not specified are AUTO


I definitely seemed to have hit the point of diminishing returns. More vcores are returning far less an increase in CPU.

If I go above 480 FSB (say, 485 or 490) I can boot but after 16 minutes in Prime 95 I always error out. Specifically, I can always pass all iterations of this:

4000 Lucas-Lehmer iterations of Mxxxxxxxx using FFT length 1024k​

But the very first iteration of this always fails on both cores:

800000 Lucas-Lehmer iterations of Mxxxxxxxx using FFT length 8k​


I've raised NB v to 1.3 during the attempts at 485-490 FSB with no luck and have had CPU v up to 1.38125 BIOS/1.368 vdroop. Seeing as my last two increments both required .05v jumps I'm not quite ready to put 1.425v to my beloved E8400.

Any other mobo tweaks I can try to liberate a bit more? I was hoping to hit 4.5GHz but just can't seem to make it past this.

Thanks!
 
Last edited:

SnowBeast

[H]ard|Gawd
Joined
Aug 3, 2003
Messages
1,312
I can't see you getting any higher than that if temps are climbing that fast with voltage bump. You would see a better gains from 8x533 (faster bus, memory at its speed) than trying to throw more volts at it. Try it. Run the voltage at 1.35v and ram at 2.0v. Why is the ram voltage up so high on 1066(533mhz) memory anyways? I only see you needing to throw higher voltage at the memory if you were going above its 1066 rating.(533mhz) From what I can tell, you are only at 480mhz on memory rated at 533mhz.(DDR2-1066 in your sig)

I don't even touch the NB voltage and can hit only using DDR2-800 memory up to 450mhz, and this memory has to have higher voltages to reach that.

You are confusing me as to why you have Voltages turned up on the NB and the memory.
 
Last edited:

jctazzy

Limp Gawd
Joined
Nov 22, 2008
Messages
150
The RAM I have is 2.0-2.1v so I set it at that to rule it out. Typically I run it at AUTO and only set it manually when I hit this wall.

NB voltage is the same story... typically at AUTO and I only specified it to nail down the variable. I was previously running DDR2-800 at 450mhz very stable and swapped out with a friend for the 1066 RAM I have now to see if that was limiting me.

I didn't go 8x533 (4.26GHz) since the 9*480 (4.32GHz) was actually a bit faster but it's possible the lower multi would allow me more headroom to push the RAM a bit beyond its rating. Maybe 8x550 or so, but that's just speculation until I put it in BIOS and give it a whirl.
 

jctazzy

Limp Gawd
Joined
Nov 22, 2008
Messages
150
Failed at this, couldn't load OS:

CPU multi 8
FSB 533 (4.26GHz)
PCIE 100
DRAM Freq 1068MHz (auto-set by BIOS based upon FSB)
CPU v 1.35
DRAM v 2.0
NB v Auto

Will tweak a bit more but I re-booted three times and each time bombed out on the Windows splash screen.
 
Top