*Official Retail G0 Q6600 Overclocking Thread*

[Kyo`]

yay finally got my quad stable at 3.6ghz, 1.5475v set in bios, 1.44v real

temps running OCCT cpu test 75-75-72-72

mult 8x
fsb 450
mch voltage +0,2v
fsb voltage +0,2v

That's some high temps, around 10C too hot for my taste... And it's on AIR, it's gonna get even worse when summer comes, I would so something about it if you don't want to fry the chip.

[Exal]

its 32deg celcius in my room year round, so I don't worry about the temps at all, and frankly I don't care as long as it's stable, I highly doubt it's gonna fry or anything...

[Kyo`]

That's some high ambient temps... Well yeah, it propably won't fry, but I don't think it's gonna last a long time running 24/7 like this.

[Hans Gruber]

I found that uses 1:1 dividers provides no better overclocking than using memory dividers. Every "GO" should be able to hit 3.6Ghz, after that it takes big voltage or what I like to call brute force to gain a stable overclock over 3.6Ghz.

My highest Core hit 68C for a brief moment of online gaming.

Has anybody hit anything above 3.6Ghz stable on air? 3.6 seems to be the wall for air cooling. I'm not saying you can't push it higher. I'm talking about 24x7 stable overclocks.

Alot of people post overclocks that they've achieved but they forget to mention that they are not 24x7 overclocks.

It looks like the 45nm will allow 4Ghz+ overclocks on air.

[WizardofOz]

Has anybody hit anything above 3.6Ghz stable on air? 3.6 seems to be the wall for air cooling. I'm not saying you can't push it higher. I'm talking about 24x7 stable overclocks.

Alot of people post overclocks that they've achieved but they forget to mention that they are not 24x7 overclocks.

I run mine 24/7 at 3.7Ghz and I am selling it.

[lowtcell]

I ran mine @ 3.8 for a day or 2 with 1.35v (CPUZ), I think it was around 1.38 in Bios. Idle temps were around 39c in RealTemp, 44c CoreTemp. Prime killed it within 15 minutes, but normal use was fine. Now running 3.6 @ 1.3 Bios, Prime stable.

[[XC] 4X4N]

I found that uses 1:1 dividers provides no better overclocking than using memory dividers. Every "GO" should be able to hit 3.6Ghz, after that it takes big voltage or what I like to call brute force to gain a stable overclock over 3.6Ghz.

My highest Core hit 68C for a brief moment of online gaming.

Has anybody hit anything above 3.6Ghz stable on air? 3.6 seems to be the wall for air cooling. I'm not saying you can't push it higher. I'm talking about 24x7 stable overclocks.

Alot of people post overclocks that they've achieved but they forget to mention that they are not 24x7 overclocks.

It looks like the 45nm will allow 4Ghz+ overclocks on air.

Here is 3.8 on air running WCG 24/7. P5K Premium, 2x1gb Team Xtreem, TRUE with 2 Yate Loon medium speed fans in push/pull, Antec P182. Just took this shot a few minutes ago. About 65f ambient. Notice it has been up for 18 days. I got this processor from another member who was able to get to 4.0 with water.



Don't waste a lot of time with prime, but here is 2 hours from a previous shot.

[JackeShan]

Has anybody hit anything above 3.6Ghz stable on air? 3.6 seems to be the wall for air cooling. I'm not saying you can't push it higher. I'm talking about 24x7 stable overclocks.

Alot of people post overclocks that they've achieved but they forget to mention that they are not 24x7 overclocks.

I have my Q6600 at 3.8GHz with 1.49Vcore 24/7.. I had it at 3.9GHz earlier but it was too hot. I don't care at all if it is 24/7 PRIME stable... what matters to me is that it is at least stable for games and so on.. I've had my computer on for 3 days now without reboot and I have been playing games for several hours. Enough for me.. I don't think I'd ever stress the CPU 100% for 24 hours

[KaOSoFt]

I thought there would be a flurry of results based on all the price drops recently. Picked one up at Frys for $190. My first Quad since price/performance were never aligned above the $200 mark.

Decided to take a break from the conga line of air coolers crossing my bench recently and just go for a simple water kit the Swiftech H20-220 Compact. I thought about breaking out my old more robust water gear but I remember getting tired of wrestling with hoses when I last had a water loop on my bench. This kit's reservoir is in the rad and the pump is in the block. Super simple mounting as well. But I digress ...

L744B144 pack date 2/26/08. Initially testing looked good with low voltage, but core 0 kept crapping out while other cores Primed happily. Is that common? It didn't help that my Blood Iron is droopier than ever.

Here's a shot of OCCT at the tail end of the default 1-hour run. Good starting point?

Wow, 1.2000v VID... GIVE IT TO ME OR DIE!

[Hans Gruber]

I ran mine @ 3.8 for a day or 2 with 1.35v (CPUZ), I think it was around 1.38 in Bios. Idle temps were around 39c in RealTemp, 44c CoreTemp. Prime killed it within 15 minutes, but normal use was fine. Now running 3.6 @ 1.3 Bios, Prime stable.

Nice to See people are being honest now. I can run my at 3.8Ghz and a low vcore, but I would be shut down by prime 95 within 10 minutes.

Priming a computer is overkill. If your chip can complete 32M of Pi, you're ok. The guy with 1.49v on air has got balls to run that hot.

[Leeghoofd]

If your chip can complete 32M of Pi, you're ok. The guy with 1.49v on air has got balls to run that hot.

Well you are in for some nasty suprises then if you think a 32mb PI is a reference for a rig to being stable

Best thing to do is indeed a core adapted 32mb pi run, some OCCT 1 hour tests and run some 3dmark06 loops... running solely Superpi doesn't stress ya CPU/hardware enough, best to run a Superpi for each core, that's something completely different...

Why should 1.5 volts be hot ? each CPU is different, some pump out more heat at 1.35 volts than some at 1.5 volts... if he stresses it and it runs below eg 65°C (that's my safe limit some go further) no worries then... your CPU core hits 68°C during gaming that's 70 plus under Prime or co... it's all about ambient temps, airflow, cooler, IHS, CPU luck ,... there's more to heat then just voltages or VIDs...

But to each his own...

[jhonyl]

Hi,

I wonder how stable are these OCs of the Q6600, and if they are good enough for a work environment?

Is it stable enough to be OCed say by an admin, and used by a normal user, who doesn't understand OCing, and who does heavy production work (100% CPU most of the time), and needs his machine to be reliable?


I am also wondering (though in the same context), what type of coolers work for what speed? Can I do 3.2Ghz on air, and with which heat sinks? and from what speed would I need a liquid system? And is a liquid cooled PC safe to be given to a normal user?

[Exal]

almost all q6600 and xeon x3220's do 3.2ghz on stock voltage with stock heatsink or any aftermarket heatsink.. it's after that it get's harder and requires more juice... Most can do 3.8-4.0ghz on a good water system, but unless the person you're selling it to has some knowledge of watercooling, in which case he probaly wouldnt be buying a prebuild system, don't sell someone a watercooling system, they can be a real hassle..

stability is pretty subjective, if it was for a customer or something I probaly wouldnt send it off before it had done OCCT cpu for like 48 hours, mine passed 12 hours OCCT so I'm pretty confident it's stable

[JackeShan]

Well you are in for some nasty suprises then if you think a 32mb PI is a reference for a rig to being stable

Best thing to do is indeed a core adapted 32mb pi run, some OCCT 1 hour tests and run some 3dmark06 loops... running solely Superpi doesn't stress ya CPU/hardware enough, best to run a Superpi for each core, that's something completely different...

Why should 1.5 volts be hot ? each CPU is different, some pump out more heat at 1.35 volts than some at 1.5 volts... if he stresses it and it runs below eg 65°C (that's my safe limit some go further) no worries then... your CPU core hits 68°C during gaming that's 70 plus under Prime or co... it's all about ambient temps, airflow, cooler, IHS, CPU luck ,... there's more to heat then just voltages or VIDs...

But to each his own...

Indeed. My cores doesn't go above 65c when I'm playing games for a long time. And that is totally acceptable as long as it is stable. I don't care if I would kill my CPU in a year because I'm upgrading then anyway. Each to his own..

[cyrixMII300]

Did you mount your TRUE properly? Maybe try lapping it? From the difference of temps on cores I think it has a VERY concave surface... Around 60C is fairly normal on the settings you're using, but 77C is wicked, you definitely need to fix this.

thanks mate.. just lapped all of them. was very much tempted to lap the G80 core on the 8800GTS but decided to skip it since this was my very first lapping job. otherwise everything else was lapped including the processor and boy am i glad that i did it. didnt get a truly flat surface on the 120 extreme but so far it seems acceptable.

[JackeShan]

thanks mate.. just lapped all of them. was very much tempted to lap the G80 core on the 8800GTS but decided to skip it since this was my very first lapping job. otherwise everything else was lapped including the processor and boy am i glad that i did it. didnt get a truly flat surface on the 120 extreme but so far it seems acceptable.

How is your temperature compared to earlier? (Load/Idle)? I'm thinking of lapping my TRUE and Q6600 too.

[Celcius]

I finally got my upgrades done and got my G0 to 3.4ghz stable:

idle temps:


load temps:

It gets a bit hot, but as long as it's stable. I don't think it'll be as hot playing games anyway.

'06 daily score:

I gained 2.3k on the cpu score and 2k on the overall score by going from a dual-core to a quad-core at the same clockspeed.

It did 3.2ghz on stock vcore (occt 30min stable) but needed 1.368 for 3.4ghz to be completely stable (VID of 1.2750). Doing the vdroop mod on the mobo really helped.

[fordf250]

I found that uses 1:1 dividers provides no better overclocking than using memory dividers. Every "GO" should be able to hit 3.6Ghz, after that it takes big voltage or what I like to call brute force to gain a stable overclock over 3.6Ghz.

My highest Core hit 68C for a brief moment of online gaming.

Has anybody hit anything above 3.6Ghz stable on air? 3.6 seems to be the wall for air cooling. I'm not saying you can't push it higher. I'm talking about 24x7 stable overclocks.

Alot of people post overclocks that they've achieved but they forget to mention that they are not 24x7 overclocks.

It looks like the 45nm will allow 4Ghz+ overclocks on air.

I have change from air to water and back with cpu,s from 2 diff mb,s and never seen a cpu need more volts when going from water to air so if your cpu is stable at 3.8G with 1.4V on water it will be stable at 1.4V at 3.8G on air it your heatsink can get that heat out. I ran the first q6600 I had at 3.8G and 1.725V for 68C 1hour OCCT on air. When I put it under water it did not clock any better and still needed the same volts just running 20C cooler.

The q6600 I have now in the same rig runs 4G at 1.55V and 51C all the time and runs 3.8G easy at 1.4V even at 475fsbx8 on 780i.

[blazin-asian]

my latest results:

http://xs.to/xs.php?h=xs125&d=08141&...v_vista914.jpg

im old school, still running dual orthos with 2 cores assigned to each

im currently running it again with a lower NB voltage to see if its still stable. i need to reseat my Big Typhoon for some better temps. when i try for 3.6, it goes up to 73C!

[Elio]

I have a question about degradetion of the Q6600, Did anyone encounter it after long term OCing?
does the only way to know degradation is if i need more voltage to operate on same frequencies and be stable?

[blazin-asian]

I have a question about degradetion of the Q6600, Did anyone encounter it after long term OCing?
does the only way to know degradation is if i need more voltage to operate on same frequencies and be stable?

unless you are pumping insane volts into your chip, i think you will upgrade by the time you notice degradation. as long as temps are good, the chip should live a happy life

[Hans Gruber]

I have change from air to water and back with cpu,s from 2 diff mb,s and never seen a cpu need more volts when going from water to air so if your cpu is stable at 3.8G with 1.4V on water it will be stable at 1.4V at 3.8G on air it your heatsink can get that heat out. I ran the first q6600 I had at 3.8G and 1.725V for 68C 1hour OCCT on air. When I put it under water it did not clock any better and still needed the same volts just running 20C cooler.

The q6600 I have now in the same rig runs 4G at 1.55V and 51C all the time and runs 3.8G easy at 1.4V even at 475fsbx8 on 780i.

You and I are essentially saying the same thing. I was not implying that you needed water cooling for better overclocks. You need watercooling for 24x7 overclocks to get up to 4Ghz. When I see vcore between 1.55v-1.7v with water cooling, I know it's stable and will work. I admire watercooling but I'm too scared to try it. It takes a lot of expertise to do watercooling better than the best air cooling solution.

I just think 3.6Ghz is the sweetspot or the max overclock stable 24x7 with air. Mine will run 3.2Ghz at stock voltage as well.

We all have different motherboard but seem to be hitting the same wall regardless of motherboard. The P35's seem to handle the Q6600 the best.

Running at 1.5v or higher is a waste of energy and provides diminished returns. But this is extreme forums so the sky is the limit.

I'm looking forward to 45nm Quad Cores which should provide 4.0Ghz with air.

[fordf250]

You and I are essentially saying the same thing. I was not implying that you needed water cooling for better overclocks. You need watercooling for 24x7 overclocks to get up to 4Ghz. When I see vcore between 1.55v-1.7v with water cooling, I know it's stable and will work. I admire watercooling but I'm too scared to try it. It takes a lot of expertise to do watercooling better than the best air cooling solution.

I just think 3.6Ghz is the sweetspot or the max overclock stable 24x7 with air. Mine will run 3.2Ghz at stock voltage as well.

We all have different motherboard but seem to be hitting the same wall regardless of motherboard. The P35's seem to handle the Q6600 the best.

Running at 1.5v or higher is a waste of energy and provides diminished returns. But this is extreme forums so the sky is the limit.

I'm looking forward to 45nm Quad Cores which should provide 4.0Ghz with air.

Motherboards are controling how far they go. With Q6600 and the Max formula my max overclock was 3900mhz@1.5V at 60+C. Now same ram, gup, cooling and cpu with warmer ambient (spring) and coretemp vers. with the striker II Formula 4G@1.55V is easy at 50C and I am trying to get 4.1G stable. 100mhz+ more and 10C cooler with the same parts. I move cpus around all the time so this is not a seating problem but constantly 10c or more diff.

[Ic3man]

hey, what is your water temp? That is so incredicle cool running cpu that I wonder!!

My Apogee with E6600 and I hit 70s with that Vcore

No idea of water temps. The Coretemp fluctuates between 60-70c depending on what fft length its performing at the time. I've not tried going any higher than 4.1ghz atm, not seen any non extreme cooled Q6600 get higher so I'm happy

[cyrixMII300]

How is your temperature compared to earlier? (Load/Idle)? I'm thinking of lapping my TRUE and Q6600 too.

take a look at the following. just got them this morning after running a good 23hours.

i guess the difference is substantial. a good 10°C in average. but mind you that i also lapped the processor IHS.

here's the link for my previous unlapped conditions.

http://www.xtremesystems.org/forums/...152603&page=97