Q6600 @ 67C

E Blade

n00b
Joined
Apr 8, 2007
Messages
51
Running Prime @ 3GHz with the room temp being 75F. Idle temp is around 40C. Is it time for a new heatsink? :(
 

6foot4geek

[H]ard|Gawd
Joined
Apr 27, 2008
Messages
1,287
that is kinda toasty...... what are your load temps?? are they 67c???

if there that high at 3.0GHz yes!!! you need a new HS, are you using the stock cooler???
 

UtopiA

2[H]4U
Joined
Oct 13, 2007
Messages
3,509
So you run Prime95 for fun?
Or do your other games/apps ever actually hit 67C? Or just Prime95?

Why do you care about your peak Prime95 temp? Nothing you ever do *for real* with that processor will ever load you that much. 40C idle is fine. If you're satisfied with that OC (You don't want to OC higher) then your current cooler is fine.

If you want to OC more then you should probably consider a new one.
 

E Blade

n00b
Joined
Apr 8, 2007
Messages
51
I run Prime just to make sure it's stable. My original idea was to push it as high as it can go on air. :D
 

silent-circuit

[H]F Junkie
Joined
Sep 18, 2005
Messages
16,136
Well then you need a better HSF as 3Ghz is nowhere near its max on air. Probably closer to 3.6Ghz.
 

GotNoRice

[H]F Junkie
Joined
Jul 11, 2001
Messages
11,054
that is kinda toasty...... what are your load temps?? are they 67c???

if there that high at 3.0GHz yes!!! you need a new HS, are you using the stock cooler???

Where do people come up with this stuff? How is 40C idle toasty? What are you basing that assumption on?

My first attempt to overclock my B3 Q6600 was done using the stock heatsink, overclocked to 3Ghz with 50C+ idle temps and load temps as high as 85C using Prime95. It made me nervous but it none the less was stable for over a week in that configuration. I got a Thermaltake Mini-typhoon as a temporary replacement; I was able to raise my overclock to my current 3.3Ghz while bringing my temps down to ~77C under Prime95. It ran like that just fine and perfectly stable for several weeks. I now use a Tuniq Tower 120 and Prime95 load temperatures are usually 65-70C depending on ambient. I’ll still see 45C+ idle temps on a normal day. And 50C+ idle temps during a hot day.

It’s really difficult to get an accurate feel for what your chip “should” be running especially from just looking at what others are running. For one, don’t underestimate the effect ambient temperature has. My idle temps are 45-50C but when I went to a small LANparty up in the santa cruz mountains where it’s much cooler my idle temp never went above 40C the whole time.

Also, temperatures from the thermal sensor on the core are calculated by subtracting the sensor value from the TjMax value. On the older Core2 chips the TjMax was 85C while on most newer chips the TjMax is 100C. Some older versions of programs such as CoreTemp, speedfan, etc would tend to assume that all chips had a 85C TjMax, which would cause your temperatures to be reported as 15C too low if you did in fact have a 100C TjMax chip. Some people have continued to use the wrong version of the program with their chip either due to ignorance or because they simply want to believe that their chip is just “that good” that it runs 15C lower than everyone else’s chip, but unfortunately it doesn’t do anyone any favors when they toss out their temp values for comparison, which are obviously bogus. Additionally, others who employ watercooling enjoy tossing their temps out in obvious air-cooling threads merely for the purposes of boosting their ego when it helps no one.

Bottom line is that your temps are fine. Prime95 does not represent a real-world load senario. You *really* only need to worry about keeping your chip about ~20C under your TjMax under worst case scenario.
 

RamonGTP

Supreme [H]ardness
Joined
Nov 9, 2005
Messages
8,150
Where do people come up with this stuff? How is 40C idle toasty? What are you basing that assumption on?

Perhaps he is refferring to the title of the thread which stated 67C in which case I'd tend to agree that that's a bit too high. Overclocked processors don't have the luxury of being able to run as hot as a stock one and still maintain stability. So while a stock clocked Q6600 will be fun at running ~20C under TJ Max, that is not the case for an OCed processor.
 

Arcygenical

Fully [H]
Joined
Jun 10, 2005
Messages
25,068
Perhaps he is refferring to the title of the thread which stated 67C in which case I'd tend to agree that that's a bit too high. Overclocked processors don't have the luxury of being able to run as hot as a stock one and still maintain stability. So while a stock clocked Q6600 will be fun at running ~20C under TJ Max, that is not the case for an OCed processor.

No, it is the case for an overclocked processor. These chips are rated to 105c maximum TJ with a thermal HALT of 120c.

Temperature has relatively zero effect on stability with the C2D series until you get into their throttle zones.
 

RamonGTP

Supreme [H]ardness
Joined
Nov 9, 2005
Messages
8,150
No, it is the case for an overclocked processor. These chips are rated to 105c maximum TJ with a thermal HALT of 120c.

Temperature has relatively zero effect on stability with the C2D series until you get into their throttle zones.

No, it isn't. My TJ Max is 100C, and I KNOW i can't get anywerhe near 80C and still maintain stability at overclocked speeds. I coudln't even hit 3.2GHz stable with the stock cooler, all the while still well below 80C. My Zalmaon 9500 does 3.2GHz stable. Again, that rating is for default speeds, not overclocked and overvolted specs.
 

unclewebb

[H]ard|Gawd
Joined
Jun 21, 2006
Messages
1,807
Temperature has relatively zero effect on stability with the C2D series until you get into their throttle zones.

I've never heard something farther from the truth. The core temperature you run your processor at has a direct effect on stability when you are overclocking. As the MHz goes up, there is no way you will be able to run reliably unless you lower your core temps.

During extensive testing with my E6400 - B2 stepping I found this for Prime stability:

3200 MHz - 80C
3400 MHz - 70C
3600 MHz - 60C

When air cooled and overclocked to 3600 MHz, there was no way the core temp could go over 60C for any length of time without quickly losing stability. At lower MHz, I could run Prime stable right up to the throttling point and even beyond without any stability issues. You need to do more testing before posting stuff like that.
 

Arcygenical

Fully [H]
Joined
Jun 10, 2005
Messages
25,068
You need to do more testing before posting stuff like that.

I've done loads, and loads of testing with upwards of 10 Q6600's, 15 E6300s and many X2 AMD64 chips...

The only ones that failed while overclocking were the X2's which would fail p95 (2.9ghz) if their temperature passed the 55-60c mark. I've run my current Q6600 @ 3600mhz all the way up to 94c with no stability issues. Additionally, when searching for my Q6600 from the box of them I got (finally settled on one with a VID of 1.15v!!!) I pushed every chip to the absolute limit so I could "bin" them for customers.

As I said, I noticed no loss in stability at 80c 3600mhz on any of the G0 Quads (although, some that would post at 4ghz, wouldn't prime over 3.9ghz when above 75c)... Additionally, most watercoolers will agree that watercooling will not increase your maximum overclock, even if it decreases your temperatures.
 

Buckster

n00b
Joined
Oct 6, 2006
Messages
26
I agree with those who say Temp makes no difference to stability (up to a point)

my Q6600 gets up to 74-75C - still stable under Prime95.

the ONLY thing that effects stability is Vcore ... not enough and unstable. Temp makes no difference - unless you get to thermal throttling

it could DECREASE the chips life though - but no-where near enough to worry an overclocker !!

1.15V VID Q6600 - now that is rather special !!! mine is a joke at 1.325V :mad: :mad: :mad: :mad:
 

silent-circuit

[H]F Junkie
Joined
Sep 18, 2005
Messages
16,136
1.15v VID Q6600?
Gimme!

Below a certain threshold (~1.2) they've been shown to do very well up to a point, usually 3.4-3.5 Ghz, then just top out. So if you're going for max clock, lowest VID is not necessarily your goal. If you want reasonably high clock and low voltage/heat, yes, that's ideal.

My X3220's VID is 1.18v. It seems to top out around 3.6, 3.7Ghz even under water, and ideal is closer to 3.4. The voltage and heat jump from 3.4 to 3.6 is ridiculous.
 

Arcygenical

Fully [H]
Joined
Jun 10, 2005
Messages
25,068
1.15V VID Q6600 - now that is rather special !!! mine is a joke at 1.325V :mad: :mad: :mad: :mad:

I'll post my final undervolt results when I optimize everything... I'm looking for a clock around 3ghz, so, I'll keep you posted if you're interested.

Below a certain threshold (~1.2) they've been shown to do very well up to a point, usually 3.4-3.5 Ghz, then just top out. So if you're going for max clock, lowest VID is not necessarily your goal. If you want reasonably high clock and low voltage/heat, yes, that's ideal.

Yeah, I sampled 12 Q6600s, recorded their VID's and did a basic "will this boot up at 3.6ghz, 1.45v" test. 7 passed the 3.6ghz test, which were reduced to 3 that would post at 4ghz. All the ones with VID's lower than 1.2 (except mine, at 1.15) wouldn't post at 3.6ghz.

I then narrowed it down to the 2 chips that would prime at 4ghz, mine and another one... And decided on the 1.15v VID, even though it became unstable over 4ghz (the other primed all night at 4.2ghz, and was sold to a very lucky customer).

I'd like to get my load temps to around 40c (20c over ambient) and perhaps get the voltage down to 0.9825v, but I don't expect to get under 1v (and I'm not even sure if I have that option yet, aha).
 

RamonGTP

Supreme [H]ardness
Joined
Nov 9, 2005
Messages
8,150
I've done loads, and loads of testing with upwards of 10 Q6600's, 15 E6300s and many X2 AMD64 chips...

The only ones that failed while overclocking were the X2's which would fail p95 (2.9ghz) if their temperature passed the 55-60c mark. I've run my current Q6600 @ 3600mhz all the way up to 94c with no stability issues. Additionally, when searching for my Q6600 from the box of them I got (finally settled on one with a VID of 1.15v!!!) I pushed every chip to the absolute limit so I could "bin" them for customers.

As I said, I noticed no loss in stability at 80c 3600mhz on any of the G0 Quads (although, some that would post at 4ghz, wouldn't prime over 3.9ghz when above 75c)... Additionally, most watercoolers will agree that watercooling will not increase your maximum overclock, even if it decreases your temperatures.

I wish I had your good fortune. In the well over 10 years I've ben building PC's i've NEVER come across a CPU that was stable when approaching it's thermal threshold while over-clocked. I don't deny you've done your own tests as I too have seen OCed X2's fail at right around the 60C mark but I just have not come across a CPU, C2D or otherwise that was NOT affected by temps.
 

Javi

Weaksauce
Joined
Sep 2, 2005
Messages
97
I'm wanting to overclock my B3 Q6600 at home. I made a thread about this previously, and to some people, my CPU seems to be running warm at stock. I have an arctic freezer 7 pro, and my idle temps run at about 47C, 62C or so at load. Granted, I put in new thermal paste over the weekend, so we'll see what happens as the paste sets in.

Then I got a Dell G0 Q6600 at work, the temps here run at 39C idle. So much cooler.

I'm afraid to overclock because some people on here seem to think 62c load on stock speeds is too high...
 

qwakrz

Limp Gawd
Joined
Feb 23, 2004
Messages
138
One thing I have noticed.

With my B3 Q6600 on stock settings and with inline temp monitor in my water loop. At idle coretemp reports 14-16C over the true idle temp.

I believe core temp is using the wrong TjMAX for the B3 stepping as has been found with real temp. Using a TjMAX of 85C I get spot on idle and load temps when compairing my water temp to coretemp and a -15C offset (newest core temp allows setting a TjMAX offset in the ini file). The TjMAX for the G0 chips are close to what they shold be which is why they appear to have a lower idle temp when using CoreTemp

I have managed to run my B3 all the way upto 10C before TjMAX without problems before but it does draw alot of power from the PSU, worked out that at 3.4GHz and using 1.515v I was pulling somewhere near 200W on the CPU alone.
 

aethelwulf

[H]ard|Gawd
Joined
Nov 20, 2007
Messages
1,938
So you run Prime95 for fun?
Or do your other games/apps ever actually hit 67C? Or just Prime95?

Why do you care about your peak Prime95 temp? Nothing you ever do *for real* with that processor will ever load you that much. 40C idle is fine. If you're satisfied with that OC (You don't want to OC higher) then your current cooler is fine.

If you want to OC more then you should probably consider a new one.

Stability testing. Buffer zone. Peace of mind. Any of these ringing a bell? :rolleyes:
 
Top