Lets roll back the clocks - Q6600 Overclocking help needed

Akagi · 22 Sep 2013 at 05:27

Hey guys

I'm having a spot of bother overclocking my Q6600, the way the system reacts to the settings I dial in doesnt seem to make sense. For example, I can get into the BIOS at 4.0GHz with 1.55v, but I can't get into the BIOS at 3.8GHz with 1.55v.

I can boot into the OS and start stress testing at 3.7GHz, but I couldnt even boot into the OS at 3.6GHz (or even 3.4GHz) with the same voltage used for 3.7GHz. All in all it seems rather strange.

Whats more the system locked up after a few hours of Prime95 Blend at 3.15GHz with 1.45v, which I thought should be more than enough given the fact that an awful lot of Q6600's didn't even need that much voltage for 3.6GHz.

Can anyone make any sense of this? Are there settings i'm missing or something?

I have the multiplier set to the default and highest value (9x), load line calibration is disabled (enabling it didnt seem to make much difference), enhanced speed stepping is disabled (enabling it helps me boot into the OS at higher clocks because it isn't running at the overclocked speed initially until I start stressing the system) RAM speed and FSB speed are unlinked.

Is there anything I can try?

The system specs are:

Core 2 Quad Q6600 (G0) w' Asus Royal Knight
Asus ROG Striker II Formula
4GB Kingston HyperX Blue 800MHz (in single channel mode because 2x RAM slots do not work)
ATI Radeon HD4870X2 2GB
Asus ROG SupremeFX II
160GB Barracuda 7200.7
600W FSP 80+ (it is actually a half decent PSU, 504W on the 12v, nice stable voltages, runs cool and quiet, plenty of connectors, nice and weighty with decent quality wires. It is also not very old)
Antec 300

coupe69 · 22 Sep 2013 at 08:33

1.55V

krisboats · 22 Sep 2013 at 13:18

Have you dropped the memory and cpu multipliers down and had a go at stress testing the FSB at different speeds to find out what is stable for your motherboard? Some boards had an FSB deadzone where the system was unstable, but pushing past that and the motherboard was stable for a good deal higher.
EG:

300-380 fsb stable
380-410 fsb unstable
410-460 fsb stable

If that were the case you simply push the fsb higher but drop the cpu multi if the cpu can't handle that high of an overclock.

Speed · 22 Sep 2013 at 14:39

coupe69 said:
1.55V

1.55v is considered pretty safe for a Q6600. You have to remember that its based on a much larger 65nm process compared to today's processors and is therefore more robust. It handles higher voltages, voltages that would kill modern processors.

Deus Ex · 22 Sep 2013 at 17:36

Just thought I'd chime in and agree with you, in my biased opinion FSP psu's are decent performers at a good price.

Akagi · 23 Sep 2013 at 22:15

I have bigger problems now

It has decided to become a dual core. Only two cores are working.

*edit* ooh thats that problem fixed, and i think the issue was Vdroop. 1.5125v dialed into the BIOS translates to 1.4v. *sigh*

Nocky24 · 24 Sep 2013 at 08:19

acme15 said:
I have bigger problems now It has decided to become a dual core. Only two cores are working.

*edit* ooh thats that problem fixed, and i think the issue was Vdroop. 1.5125v dialed into the BIOS translates to 1.4v. *sigh*

This is where a bit of llc usually helps. Did you fix your cores?

Akagi · 24 Sep 2013 at 18:43

Nocky24 said:
This is where a bit of llc usually helps. Did you fix your cores?

Fixed the cores, and I have all but given up with this now.

I have always wanted a Q6600, but now i'm cursing their existance. It cant even sit at 3.2GHz with 1.42v (1.5125v dialed in in BIOS) and LLC without locking up on blend. But it is fine at stock.

JonSmith · 24 Sep 2013 at 20:47

How's the cooling on your chipset? Might want to try repasting the chipset heatsink or adding some extra airflow.

Akagi · 25 Sep 2013 at 01:39

JonSmith said:
How's the cooling on your chipset? Might want to try repasting the chipset heatsink or adding some extra airflow.

I already did both of those things only a few days ago to cool down the northbridge.

To put things into perspective so you can appreciate why i'm confused, it locks up on prime95 at 3.2GHz with 1.425v (which is excessive for that clock anyway) but it is prime stable at 3.0GHz with 1.28v...

P4Clock · 26 Sep 2013 at 11:22

so you have a chip that dont want to clock any higher, you buy a 2.6ghz chip you should only ever expect it to go 2.6ghz. anything over that is a bonus. Not all Q6600 will clock to 4ghz.

krisboats · 28 Sep 2013 at 08:20

P4Clock said:
so you have a chip that dont want to clock any higher, you buy a 2.6ghz chip you should only ever expect it to go 2.6ghz. anything over that is a bonus. Not all Q6600 will clock to 4ghz.

Indeed, mines struggling to get past 3.6 on a decent cpu water loop. What is cpu-z showing the vcore to be while its stress testing? My ds3r has a terrible vdroop, 1.55 in bios translates to a max of 1.48 in cpu-z while under load.

Akagi · 1 Oct 2013 at 12:20

P4Clock said:
so you have a chip that dont want to clock any higher, you buy a 2.6ghz chip you should only ever expect it to go 2.6ghz. anything over that is a bonus. Not all Q6600 will clock to 4ghz.

It is a 2.4GHz chip, but true, not all will do 4GHz stable, but a-lot will. I want *a decent overclock* of some description, which I can't attain. Ethermaster rekon's it is because of a problem with the motherboard, so I might have to stick with 3GHz.

virdee2 · 2 Oct 2013 at 22:06

i had similar problems with my Q6600 and like you i waited and waited for it and got a really good bargain at the time.

I eventually got it stable at 3.2 and have been using it 24/7 for at least a year now.

I could get it posting and loading at higher but something didn't add up and i simply could increase anything high enough to satisfy a higher clock.

Sin_Chase · 3 Oct 2013 at 20:23

I managed 3.4 with 1.38Volts easily so it does not appear to be a lack of Vcore.

Mainboard or just a really trashy chip. That would be surprising for a G0 though.

From what I recall some mainboards had FSB walls. I could get 9 x 378 but that was about it. I could not get 9 x 400 no matter what I did or how much juice I gave the chip, a known FSB wall on the Abit I was using.

decto · 3 Oct 2013 at 22:51

Nvidia chipset ...... never had that much luck OC'ing on Nvidia

Best clocks always seem to be the premium intel boards

AD

AnthonyA85 · 5 Oct 2013 at 19:46

4.0 GHz on a Q6600?! What sorcery is this?

Last time i tried messing around with the FSB/Clock spees in BIOS to OC my Q6600, my system crashed. The max it goes now is 2.4 GHz, but idles about 1.6 (I've got CPU-Z open right now) averaging between 1.100 to 1.280 volts.

I do want to try overclocking mine, since i'd rather over-clock and try to get 3.0 GHz at least out of this thing, instead of spending money on a new stock 3.6 GHz CPU and new 1150/2011 socket board, but i'm afraid to really touch anything in the BIOS, since the P5K's bios lists settings that most overclocking guides either use different names for, or don't list, and i don't know what to alter.

Akagi · 4 Nov 2013 at 23:09

I have seen one at 4.3GHz before!

Mine would sit in the BIOS at 4.2GHz, boot at 4.0GHz, and would be prime stable for 3 hours at 3.8GHz in the end, but I stuck an easy overclock of 3.0GHz on it and sold the gear in the end.

I would recommend ditching it though. It is going to bottleneck the hell out of your 7950. I'm pretty sure my friends stock Q6600 bottlenecks his 6850 OC!