ip35 pro Q6600

[WingZero30]

Thanks for that WingZero, hopwfully someone can shed some light.

Hopefully I wont destroy my rig whilst we are trying to resolve this. ha

. I am glad I spotted the Cpu-z 'Core VID' early on which could have mislead you.

 

[NoIdea]

So any thoughts from anyone else?

 

[Diggsy]

The lower voltage readings on the older versions of CPU-Z when idle are because you have C1E and EIST enabled, which lowers the core speed and voltage when not needed, hence saving power.

If you have it set to 1.43v in the bios and you are seeing 1.38v or so when at load, then that is probably about right as these boards do suffer quite a bit of vdroop. For instance, mine drops 0.1v when loaded, yours seems to be about half that if I am reading it correctly.

Best way to check is load up HWMonitor and then start Prime95 and watch the Vcore reading.

 

[NoIdea]

Top one is idel and second one is load.

slight difference between cou-z and hwmonitor.

 

[Hairybudda]

They're just refreshing at different times, afaik they both read from the same place

 

[NoIdea]

So are we happy that all looks ok?

Looks ok from my perspective, I guess proof is in the pudding. Touch wood I had no problems with it yet.

 

[redshadows]

Hmm OP it is very strange indeed. Your voltages are 100% correct but they should be the other way round. My Q6600 has exactly the same VID of 1.2625v. I think we need the input of fellow overclockers to shed light on this issue especially

Marine-RX179
redshadows
Rroff

These guys helped me in the past with my E6600 and Q6600 queries and their input proved very valuable. If you guys are on this thread please shed your light on the issue of lower idle core voltage and higher load core voltage OP is having.

Thanks Wing

I have seen see some very peculiar things but this isnt one:

In my experiences HWmonitor is more accurate.
Different vcore for load vs idle is normal
vcore droop is the answer

your running a q6600 @3.5 GHz on a Abit IP35 Pro, p35 needs that extra little bit on vcore for stable especially on that board.

Some chips also need the push, 1.39 (HWreadings) is cutting it close for a 1.1GHz overclock on that board mate.

Another indicter for needing more vcore is "after a couple of hours, a couple of cores fail prime"
Your cooler is keeping things well within spec, push the bios vcore until HW is reading 1.44 idle.

 

[NoIdea]

Thanks Redshadows, when yo say

"Some chips also need the push, 1.39 (HWreadings) is cutting it close for a 1.1GHz overclock on that board mate."

I'm assuming you mean push the vcore higher, and cutting it close refers to underpowering it?

In relation to my earlier comment about a couple of cores failing, ive not had that since i altered my settings to:

FSB 390 x 9 =3.5
DDR2 1:1 780
PCI E 100mhz
Vcore 1.4325
DDR2 2.00
VTT 1.23
MCH 1.29
ICH 1.05
ICHIO 1.50

Its all been stable since that and no failures with Prime or Everest.

 

[WingZero30]

In my case for [email protected], HWmonitor and Cpu-z core voltages correspond with each other during idle and load with slight variation presumably due to rounding off . Lol although I don't understand why HWmonitor is giving 'Trial' values. Is it some sort of bug?

Bios Vcore 1.2625v (stock voltage)
Idle Core voltage 1.216v
Load Core voltage 1.184v

Idle

Load

Edit: I downloaded older version of HWmonitor and there are no 'trial' values anymore lol.

 

[Diggsy]

^ You are running HWMonitor Pro, which is a trial unless you buy it.