Poor Q6600 temps

aznofazns

Limp Gawd
Joined
Sep 14, 2008
Messages
273
I don't why but my G0 Q6600 gets pretty high temps even with a Xigmatek HDT-S1283 (non-lapped with stock fan) with Arctic Silver 5 thermal compound spread thinly and evenly over the heatsink's base.

q6600temps.jpg


I don't remember the VID of my cpu but I'm pretty sure it's below 1.30v... I know that 1.45v is pretty high but most people with similar conditions aren't reaching 73C like I am. I've reseated and reapplied AS5 three times now but there's never a difference. I'm also pretty sure that's it's due to poor case airflow because I've tested with the side window open before without too much change in temps.
 
Are those load temps? if they are then they aren't too much out of control.
 
where do you live you might have a high ambient temp. your gpu looks abit high too. thats a pretty high overlock for q6600 on air though. post your non oc temps
 
That's nuthin!! My B3 pegs 74°C and doesn't crash. Unless you experience instability I would worry about it.
 
Those temps are great if 1.45v is accurate. What does core temp or real temp say?

I'm getting 67-69C at 1.39v 8x445.
 
considering the voltage + that you're prime is running 8k's, that looks about normal to me.
 
i run my q6600 at 1.45 ish when im at 3.4 Ghz and temps dont go past 63. gotta upgrade to the TRUE for ultimate cooling
 
My voltage is set to 1.45v in the BIOS (vdroop control turned on), and those temps are at max load with Blend (NOT small FFTs) so the temps could potentially be even higher. I turned my A/C on to help the ambient temp, which is probably roughly 70 degrees F. Do you guys think I can push for a higher overclock? My goal is ultimately to reach 3.6ghz with stability but I don't want to push my voltage up to 1.5v.
 
Start doing Small FFT's and see what you get. Also consider dropping the multiplier to 8 and running a high FSB. It might make a difference. Keep the voltage where it is for now and push the cpu until you get errors, make sure your memory is set to the lowest speed for now.
 
i run my q6600 at 1.45 ish when im at 3.4 Ghz and temps dont go past 63. gotta upgrade to the TRUE for ultimate cooling

Q6600 @1.45 3.6Ghz for me hit 46 after 9 hours of prime, I love water.

anyway those temps seem normal for air. you could lap everything, or swap out the cooler.
 
Start doing Small FFT's and see what you get. Also consider dropping the multiplier to 8 and running a high FSB. It might make a difference. Keep the voltage where it is for now and push the cpu until you get errors, make sure your memory is set to the lowest speed for now.

What's the benefit of 8x multi, exactly?
 
This is interesting... I set the multi to 8x and raised the FSB to match the 3.44ghz I had set previously with the 9x multi... and after 15 minutes of prime95 Blend the max temps are 9C lower... and I have no idea why.

EDIT: Nvm, 5 minutes later the temps jumped from 63C to 69C... but that's still lower than the 73C I was seeing with 9x multi.
 
Is this safe for extended use? The most load I'll ever place on my CPU in everyday activities will probably be Crysis or other demanding games.

q6600temps.jpg
 
Those temps are perfectly fine. No need to worry about that. There really is no need to worry about temps unless they are 85c+ or are causing instability. The chip is rated for 100c max.

Run it, run it hard. Q6600's are beasts that can be beaten on like no other.
 
Yeah those temps are pretty good imo, for that vcore and speed, do you still have the stock fans in the Rocketfish case?

What vid was your q6600?
 
I agree with everybody else, those temps are well within lines of every Q6600 I've seen between 3.4 and 3.6ghz so I wouldn't worry too much.
You mentioned case airflow in your first post? What is your complete cooling setup? number and type of fans, any little tweaks you've done to try and lower temps, etcetc

I am also curious what you are running the ram at now with your cpu at 8x450, and what gskill kit it is that you are running?
 
Thanks for your replies. I feel a lot more comfortable keeping my CPU at its current clocks with your confirmation. Keep in mind that I've only been running Blend tests so far, so i'm not pushing to its max yet.

I'm running a stock Rocketfish case with stock Rocketfish fans (front lower intake fan, rear upper exhaust fan, side PCI fan removed because it's nearly useless) and stock S1283 heatsink with fan. I'm running my 2x2gb G.Skill DDR2 800 kit at 2x450mhz (900mhz) overvolted from 1.8v to 1.88v with loosened timings of 6-6-6-18 (stock 5-5-5-15). I plan on tightening these timings once I get my CPU 100% stable. So far my Prime95 Blend tests have been failing after at least an hour but I'm not sure which voltage needs to pushed to achieve stability. My NB is already at 1.3v (from 1.1v), my CPU core is set to just under 1.49v, FSB termination left at auto (should I change this?), and CPU PLL overvolted to 1.58v from 1.50v. Any suggestions?
 
Those temps seem pretty ok, (quite similair to mine @3.6ghz, prior to lapping) perhaps as suggested look at replacing your case fans for something a bit more powerful, NB voltage seems ok as does CPU-PLL, id suggest taking the FSB-TERM voltage of auto and setting it to 1.34v (if stable try lowering it) Test with small fft to ensure cpu stability, then give blend a try.
 
Ah ok, that's the same set of GSkill DDR2 I am using right now.
I've have them running right this minute at 900mhz at 5-5-5-15 2T 2.0v
I wonder how close I can get to the GSkill PQ 1000mhz kit. :D
 
I've found a setup that's endured 2.5 hours of small FFTs testing and here are the results:

q6600hot.jpg


77C seems to be really pushing it, but according to you guys it's still within the acceptable range for the Q6600, right?
 
1.488v under load? how much voltage are you setting it in the bios and how much shows while you are idling in windows? just curious as my cpu drops to 1.29v under IBT load from 1.4v in bios(1.35 in windows)

btw at 8x450mhz 1.4v I load IBT at 68/65/68/65c
 
In the BIOS it's set to a tad over 1.49v but I turned on the option that reduces vdroop. I'm not sure why my vcore doesn't drop during load anymore... it used to when my setup was 9x378 (3.4ghz) at 1.45v...
 
Have you experimented with lowering your vcore at all? 1.49v/1.45v seems pretty high just for 3.4ghz. You should try to get at least under 1.4v at that OC, at least in my experience. That will work wonders for your temps.

My Q6600 is stable with 400x9 at 1.35v, just for reference. Maybe you should try 1.35v for your oc and stress test it again.
 
My Q6600 G0 runs at 9x400, 1.425V in bios, 1.416V reported by CPU-Z under load.

Running 8Ks in Prime temps are at 2x 68C, 2x 61C with Ultima90 and 120x38mm fan.

IntelBurnTest clears 20 cycles, temps climb to 2x 77C, 2x 69C.

IBT will tell you if you have stability probs in 10 mins instead of 8 hrs for prime95.
 
All Q6600's are different in regards to the amount of voltage needed for a stable OC. Mine requires 1.41 to be stable at 9x400 under load. Unfortunately my board has so much vdroop it's set to 1.55v in BIOS. At idle, with vdrop, it sits at 1.48. Pretty sad, but it works and stays cool.
 
You should try 8x425 for 3.4ghz. I got this with my old setup with 1.45V in bios. Also for that system temp I could be wrong but I beleive your board is a P45? I dont think the P45 gives NB temps. I ran into this issue with my gigabyte thinking I had good NB temps. As for paste application you can try the line method suggestd by the makers of Arctic Silver to see if your temps drop any. Make sure your fan is throttling up when the system gets hot. On my setup my fan at idle is around 900rpm but jumps to 1,900rpm under load. Yours should do something similar. Make sure the case has proper cooling as well. Also if you want to see how hot it can get try the Intel Burn in Test. IBT took my temps over what prime did in a matter of minutes.
 
The problem is that Prime95 always fails if I try a lower voltage. It required 1.49v to hit 8*450(3.6ghz), anything lower and Prime would fail after 1-2 hours.

I can hear my fan speed up when priming so I don't think anything is wrong with the fan.

I applied AS5 onto the CPU by putting a glob in the center and then using a plastic bag to create a thin even layer covering the entire surface... which is a tried-and-true method.

Regarding IBT, if Prime95 small FFTs test already peaks my CPU at 77C then I really don't want to know what IBT will do...
 
I applied AS5 onto the CPU by putting a glob in the center and then using a plastic bag to create a thin even layer covering the entire surface... which is a tried-and-true method.

Regarding IBT, if Prime95 small FFTs test already peaks my CPU at 77C then I really don't want to know what IBT will do...

Okay, the even spread method of heat paste was for direct-die cpus back in the Athlon XP days... Ever since the return of heat spreaders on cpus, the correct method is the size and shape of a rice grain. Place a thin line of paste about 1/2" long in the center of the cpu (ideally run parallel to the text for quad-cores) and then just clamp the heatsink down on top. The paste will distribute itself evenly.

IBT will tell you in 5 mins what prime95 takes 2 hours to do... makes overclocking testing much less time consuming.
 
Back
Top