Q6600@4ghz on air potential

[elal30]

Basically i left bios settings at those for my 3.2ghz(8x400) overclock(most settings on auto). I only changed voltage from 1.2.000v(1.2250v in bios) to 1.5750v(1.6v in bios), ram voltage from auto(1.8) to 2.2v, and Nb voltage from auto to 1.70v(max).
Cpu and system fans at full speed.
Windows(xp) was good for average use, of course rebooted 10 sec after prime was started.
Room temp: 16C
system: 30C
Idle core temps around 59C!
Average load: 79C+
Max load: ???!

Highest stable overclock for this cpu at 3.5ghz(9x389) 1.328v

Question: Will it run 100% stable with water cooling?

System specs:
Q6600 Go @3.200(8x400) 1.2v temps 39c-66c
Xigmatek xp-s964 heat-pipe cooler
Asus p5k-e wi/fi
2gb 800 crucial ram 1.8v
Asus 8800gs 550/1320/800@700/1750/860
HD seagate 160gb + hitachi 80gb usb
Asus dvdrw lightscribe
5 system fans
Cordless Logitech keyboard+mouse
PSU Zalman noiseless 460w
22'' LG monitor/tv

 

[Lupiron]

I'd like to know the VID of your quad! Run come temp 0.96.1 with speed step disabled, and then the vcore and fsb and all that on auto, so you get stock settings and list the VID thats listed in core temp!
If its a 1.2000 in the bios, your VID should be like 1.1xxx And I wanna know!

If thats the case, you prolly have a 4.0 or so processor! Then again, your listed temps are still high already.

Good water would help, I am sure, bad wouldnt!

Your board can soar, because I have the same one with a g0 at 3.8! I get 72 c at that speed.

--Lupi

 

[elal30]

If I'm not wrong VID is 1.2375v
On auto in bios at stock speed (2.4ghz) my idle voltage is 1.1200v and goes up to 1.1600v on load.

Prime95 failed after 2,5 hours at 3.600ghz with max temps at 79C. Voltage auto in bios(which is about 1.39v), 1.36v idle in winxp, and 1.344v under load.

 

[Lupiron]

Wow, yeah. I dont get that hot at 3.8, lol, with a vcore of 1.52500 so its 1.5 in windows! At 3.6 I get 62-63 c on a hot day, because its always hot in my room! To test it I just opened my door, lol, went from 44c to 32c lows!

So yours is hot! I dont know why it would be that hot... Hmm. Is the cooler seated right?

--Lupi

 

[elal30]

When I first placed the cooler it was loose on one corner and i got all temps 20c degrees higher! But I fixed that. I think I just need a better cooler, or water.

 

[Evilonigiri]

If you're, taking your CPU above 3.8GHz, water cooling becomes a viable solution.

4GHz on air is rather difficult to do on air safely. Currently, the best air-cooling is the Thermalright Ultra 120 Extreme (TRUE), and Anandtech got to 3.9GHz on the TRUE. The only difference was that they were using the QX6850 instead of the Q6600. According to them, the temps were around 60C on load, but perhaps their lab is cold.

 

[Perp]

I have a chip with the same VID and I get 3.6 (400x9) @ 1.36v.

My core temps are lower (air) but case/cpu cooling ect come into play.

 

[Lupiron]

Either of you two wanna sell your chip? ::Grins evil-like!:: What?? Greed is good! And pay pal works wonders!

I should get my paws on 2 or 3 of them here in the next couple weeks. I cant wait. I didnt OC when I was in a position to get more. So I dont have a clue what the VIDs or stepping of the few I sold were! Waaaaahh! Fate left me a decent G0, and a flamer B3! Waaaahhh!! Dunno a thing about the Phenom 9500.

--Lupi!

 

[elal30]

I can have it stable at 3.6 and 1.36v but i don't appreciate the temps. Besides, I'd like to keep this chip as long as possible... Performance wise(windows-gaming etc) it won't reach its full potential for some time yet.

 

[Lupiron]

I wanna know about your first post, you set the VCore to 1.6 etc... for what? What did you try and do? What speed? Was that the 4.0 on air try?

You just need a better cooler for that processor. 79c peaks at 3.6 is hot, hehe.

Trade you! 3.6 @ 1.38VCore, max temp 63c! Tried and proven and only 3 months old!

--Lupi

 

[zenmaster]

3.4 to 3.6 is a Nice Reasonable OC for the Q6600, especially if you want to keep it a long time.

Usually pushing it to 3.8 or 4.0 is going to require voltages that are considered borderline for CPU health.
I would just set the CPU to 3.4 and drop the volt a tad from your 3.5 stable level and you should be set.

I seriously doubt you are being limited much @ 3.4ghz on your Q6600 except maybe in benchmarks.

 

[elal30]

Yes, that was for 4ghz. I did some research before applying the voltage, some people agreed on 1.575v, or around about that. Besides, after 3.2ghz the volts required, rise steeply.
I think my cpu is great. My venture into the 4ghz zone is an indication that it can be done safely as long as better cooling is applied.

 

[Perp]

Well good luck to you then.

I'm hoping to push mine to a stable 500x8 for 4ghz, but not untill I can lap the processor and do a few more case mods to keep the heat down.

 

[Lupiron]

Waaaaahhhh!

 

[elal30]

500x8? Now that, I haven't tried! But won't you just be putting more stress on your mobo?

 

[elal30]

Anything higher than stock speed is always good value for money, but a 66% overclock @4ghz is even better! Especially when the technology at hand gives you the opportunity to try it.
For the time being though, 3.2ghz with just 1.200v is more than enough power for me...

 

[Lupiron]

Not to brag or anything, but check my other post, my VID is BIGGER than yours! ::Sticks out his tongue.:: Err.. I mean littler-er than yours!

Hehe, I wanna see results!!

--Lupi!

 

[elal30]

That's a great vid. Let's see if you can do 4ghz comfortably.
I must say however that at default speeds(2.4ghz) my voltage in windows is 1.1200 idle and goes up to 1.1600 on load which is indeed very low.
At 3.6ghz your lower vid cpu of course is pretty cool.

 

[Lupiron]

Hmmm, I can boot at 4.0 and run stuff at 4.0, I cant seem to get prime stable, and I had my 1.2625 at 3.8 @ 1.51 volts in windows, and the 1.2000 has a bunch more voltage notches than the higher VID processor, my only conclusion so far is its the boar, so i will try a low multiplier OC here soon. AND I have a workroom that is 60c and isnt gonna change.

I'll let you know, the maximus formula also has a fully operational Load Line Calibration! It defeats VDroop totally, and may even add +.0015 volts while under load, thats excellent for OCing!

--Lupi

 

[elal30]

What voltage did you try for @4ghz?

 

[Lupiron]

I have revised my conclusion. It appears that to go from 3.6 to 3.8 you need a HUGE amount of VCore to get prime stable.

You will be windows stable pretty easily, but for the fake situation you place the cores under while testing, you're gonna need a huge jump in VCore beyond what you think ya need. like 6 notches more.

Hmm... thats neat!

--Lupi

 

[Lupiron]

Oh, sorry. I didnt really try and test it for actual prime stability yet. Bet I could now, it needs at least 1.440 volts to pass primes first 8 k test. So thats 1 or 2 more to be "stable" enough to test higher.

But since I went from 3.6 to 3.8 with like 10 voltage notches to be stable! Gonna need more than 10 more! (But I really dont know if its just that initial vcore jump, or will it stick to the same pattern as 3.6 to 3.8 did? Thats gonna hurt!

--Lupi!

 

[Lupiron]

Oh, sorry. I didnt really try and test it for actual prime stability yet. Bet I could now, it needs at least 1.440 volts to pass primes first 8 k test. So thats 1 or 2 more to be "stable" enough to test higher.

But since I went from 3.6 to 3.8 with like 10 voltage notches to be stable! Gonna need more than 10 more! (But I really dont know if its just that initial vcore jump, or will it stick to the same pattern as 3.6 to 3.8 did? Thats gonna hurt!)

--Lupi!

 

[Evilonigiri]

You can say that again!

 

[elal30]

Just out of curiosity I tried at 3.6ghz again. Changed cpu voltage to auto and left everything else pretty much the same. In windows idle was 1.368v and load 1352v.
The second core failed prime 5 hours later!
Highest temp 81C.