*Official Retail G0 Q6600 Overclocking Thread*

[Dagalidis]

Nice MEM(Dominator 8500) perf here..

[Loser777]

I get a huge difference between 3.6GHz volts and 3.7GHz volts...
I need 1.392v for 9x400...
I need 1.496v for 9x411....
Should I lower multi or something?

[dnottis]

I get a huge difference between 3.6GHz volts and 3.7GHz volts...
I need 1.392v for 9x400...
I need 1.496v for 9x411....
Should I lower multi or something?

Thats what happens you hit a point where alot of voltage is needed to move up a little. Thats just how CPUs are, nothing will "fix" it.

You have to ask yourself, is it worth running that much voltage for so little gain?

[nfm]

Thats what happens you hit a point where alot of voltage is needed to move up a little. That's just how CPUs are, nothing will "fix" it.

You have to ask yourself, is it worth running that much voltage for so little gain?

When working with quads, from my experiences you need to adjust GTL Reference Volts or cool PWM, MOS-FET area _very well_ of your motherboard or don't go beyond a certain CPU temperature. I have trouble priming my quad +3.6Ghz if I don't cool my digital PWM or they go beyond 55C (I have copper ramsinks on CPL chips, applied MX-2 under the stock heatpipe, and transpiper attached). If I want +3.85Ghz I need to cool back of the motherboard because it is melting. I can successfully prime my quad at 3.95Ghz with 1.53v on DFI LP P35 motherboard, 80F (26.6C) ambient! During the winter my ambient was 17C and didn't need a single fan and was priming at 3.97Ghz. So when your CPU happens to be doing great at 3.6GHz, and for the next 100Mhz or so needs a lot more voltage then you should know the reason. I don't believe in what dnottis said anymore, even though it proved true too many time for me on many CPUs. I always worked my way out.

[OVERK|LL]

When working with quads, from my experiences you need to adjust GTL Reference Volts or cool PWM, MOS-FET area _very well_ of your motherboard or don't go beyond a certain CPU temperature. I have trouble priming my quad +3.6Ghz if I don't cool my digital PWM or they go beyond 55C (I have copper ramsinks on CPL chips, applied MX-2 under the stock heatpipe, and transpiper attached). If I want +3.85Ghz I need to cool back of the motherboard because it is melting. I can successfully prime my quad at 3.95Ghz with 1.53v on DFI LP P35 motherboard, 80F (26.6C) ambient! During the winter my ambient was 17C and didn't need a single fan and was priming at 3.97Ghz. So when your CPU happens to be doing great at 3.6GHz, and for the next 100Mhz or so needs a lot more voltage then you should know the reason. I don't believe in what dnottis said anymore, even though it proved true too many time for me on many CPUs. I always worked my way out.

Interesting, mine as well started requiring absolutely stupid voltage to get 3.6Ghz, yet hits 3.43 with relative ease. You think that adjusting GTL Reference is going to make that much of a difference?

[Loser777]

I don't have a GTL Reference... 750is BIOS options are a lot different.

What is also weird is that the motherboard won't post at 3.8GHz with an 8x multi... on my old chip it did that just fine...

Wait I found a GTLVRef... and it gives an option to add mv...
I tried adding a shiatload of mv to that... didn't work
I also tried changing it all to 0... didn't work.

[KiDMaioRKz]

is 1.7v danger for 445x9 ?

[Loser777]

Run that for a week and you'll find out.

[KiDMaioRKz]

I run that for 2 weeks already ... and nothing seems to be happen ...
but few days ago my mobo died without a reason ... it showed up that "Chassis Intruded" but I've already checked the chassis jumper ..the jumper was in the proper place (sure , i've never touch it a bit !).
Also I flashed the BIOS .. and it wouldn't help.

is this problem came from 1.7v ???

Thanks

[finnsopp]

Yes =) Got my q6600 L807A420 Up`n Runnin at 3.6ghz on air =)

I use a Value MB its Gigabyte GA-P31-DS3L @ 400mhz fsb

Im currently at 3600mhz on air, and my Full Load temps are 61 C soo i belive its room for moore =)

Anyways, been working on this all morning =) cheers

Prime 95 have only been runnin for 15 mins on that pic, but has tested it for 2 hours, will try a whole night run tonite too C if its 24/7 Stable.



Will try too push it further later today Hopefully i can get 3.7, 3.8ghz out of it on air

[fatguy1992]

I just lowered my volts even more

Was.



Now.





I'll run some better stress test later.

[fatguy1992]

BTW i now its not a giant decrease but still

[fatguy1992]

YAY even lower.

[Loser777]

If you want to test using pi, run FOUR instances.

[Loser777]

I don't get this, with my last (crappy) chip, I did some 3dm06 runs @ 3.85GHz and 8x multi, while on this one I can't even post at 3.8GHz 8x multi, regardless of voltages...
I think 3.6GHz 1.384v and 3.7GHz 1.496v is too big of a gap to be caused only by CPU.

[nfm]

Does your chip run hotter than the one before at similar settings?

[S1mon-]

I revisited my L737B and I have it running at 3,8GHz now at 1,384V

[nfm]

I revisited my L737B and I have it running at 3,8GHz now at 1,384V

Congrats
Max FSB? We want more

[Loser777]

Does your chip run hotter than the one before at similar settings?

Yeah, it does. Though it may be because of my bracket... I've remounted that already... so that's unlikely.

[dnottis]

Are any of these overclocks prime stable? I can run 3.85 @ 1.4v but it's not prime stable. 3.6 @ 1.35v IS prime stable for 8+ hours however.

[Loser777]

Mine are.

[S1mon-]

Congrats
Max FSB? We want more

Max FSB is kinda low, well normal for a Q6600. The max I did on air was 4,15GHz which is ~461 FSB... 465 FSB will crash while loading Windows...

[cincyrob]

im thinking of getting a Q6600. is the newer ones as good as the original ?
this is the one im looking at is it a good one that will OC to say 3.2ghz

BX80562Q6600
SL9UM
core stepping B3

[Loser777]

B3s aren't very good...

I'm hitting a serious FSB Wall here, 450MHz... maybe less.

[OC_Newbee]

Still testing max FSB, don't think my ram can get high than 520mhz (gskill 2 x 2gb pc8000). Here's a little ss @ 500fsb under load (still testing but for a short moment).