Another Q6600 first overclock

[jhick]

I just got my Q6600 G0. It is running a arctic cooling 7 pro heat sink.

I have it running at:

FSB = 2700MHz
multiplier = 9x
core voltage = 1.208

Using core temp 0.95.4 my idle temps are:

42 - 42 - 38 - 36

I was just wondering if these idle temps are OK? They seem a bit high to me for such a small overclock. Thanks.

 

[gamerxx13]

totally fine, can push it till temp is around 60, please read forums about overclocking before you do. i can see you are already overclocking.

 

[jhick]

I have read about 10 guides on overclocking and have read through many threads of people overclocking the Q6600. I just cant seem to find any again regarding tempts. I was just wordering about my temps because there are people reporting going up to 3000MHz with a stock cooler. Could I lower my vcore a bit more to get lower tempts? Seems my vcore is already pretty low.

 

[sutahz]

Idle temp is also Heavily dependant on ambient temp. If the heater's a blasting... your core will feel it.
You may be able to drop your voltage, whats your VID? 1.3? 1.2? either way, you may be at your max GHz for that Vcore or not, every cpu is different.

 

[jhick]

My room temp is 21C. VID is 1.325 according to crystal CPUID

 

[BonzaiDuck]

I commend you on your common sense, even though you seem new to the C2D/C2Q Intel technology. And I only say "seem."

You say "FSB = 2700 Mhz," but I assume you mean the CPU is running at 2.7 Ghz -- a 300 Mhz over-clock.

Your core temperatures seem a tad warm for that processor, that voltage setting, and that over-clock level.

My best sense of VCORE adjustments is this: You should not fear damage to the processor if you set the VCORE as high as 1.38V. But see my bolded remark, below. The nominal idle vcore for both G0 and B3 steppings is around 1.372V. On air-cooling, my highest over-clock setting is 1.4183V. But I want to keep it closer to 1.37. You should be able to over-clock farther than I can on a setting of 1.31 or 1.32V. My sense of it is that both G0 and B3 steppings can use about that much voltage when running the processor at 3.0 Ghz and FSB = 1,333 Mhz. Some G0's will be rock-stable with less than that, or VCORE between 1.25 and 1.29V (but assume the worst.)

By comparison, my B3 at room ambient 66.5F(ahrenheit) is showing idle CoreTemp values of 35,35, 31, 33 with the Vcore at 1.3625V and processor running at about 3.08 ghz.

I'm guessing that you're using the stock INtel heatsink-fan. I'd be interested in a description of your computer case and fan-deployment.

With my custom provision for air-cooling enhancement, I can expect my load temperatures (average for all four cores) to top out at around 59C when room ambient is 76F. If I use an over-clock profile for running the C2Q @ 3.2 Ghz and vcore = 1.4183V, my load temperatures may be as high as 63C at that room ambient.

 

[BonzaiDuck]

Computronix' C2D-C2Q Temperature Guide at Tom's Hardware

 

[jhick]

Originally posted by: BonzaiDuck
Computronix' C2D-C2Q Temperature Guide at Tom's Hardware

I have read every guide out there about quad core overclocking and a couple specific to my motherboard (Asus P5K-E and Q6600).

 

[jhick]

Originally posted by: BonzaiDuck
I'm guessing that you're using the stock INtel heatsink-fan. I'd be interested in a description of your computer case and fan-deployment.

I am running a arctic cooling 7 pro heat sink. I have a coolermaster 690 with a 120 mm intake, 120mm exhaust, 2 120mm blow hole exhaust, 1 120mm bottom intake, and 2 120mm side intake. Most are scythe or yate loon fans rated at around 45cfm. So I think the cooling is OK? I also have re-done the cabling so there is nothing obstructing airflow.

Yes, I meant fsb = 300MHz. haha, not 2700.

I just left the computer on all night running and temps have dropped down to 37-37-35-33 idle. I think everything just needs to settle and everything and warm up.

 

[BonzaiDuck]

The cooler is "sub-stellar," but if those are your idle core temperatures at 2.7 Ghz, and if going up to 3.0 ghz doesn't raise them much above 40C, you should be able to stay under 70C at load -- I would think. I also think the very minor VCORE adjustments you might need to get there may not have an appreciable impact on the temperatures. I think you said your ambient room temp was about 21C or 70F?

Looking at the CoolerMaster case, I can see how cooling might be improved just a bit without spending a lot or cutting on the case. But if I were going to do it myself, I'd swap out the cooler for a ThermalRight Ultima 90, first, and that's about ten times the cost I had in mind.

Another thing you could do is to try and reverse the blow-hole fans to intake so that all air is forced out of the rear exhaust fan-port, and per my last paragraph, build a foam-art-board duct box between the cooler and the rear-exhaust fan. I'm only guessing that airflow through the PSU is more or less independent of case air, but if it's an efficient power supply, it is not an important consideration. With the PSU in the bottom of the case, I'd guess that internal case air would be warmer, regardless.

Others using that cooler with the q6600 report over-clocks to 3.2. But first, run PRIME95 and CoreTemp to get two load temperature baselines: at stock settings, and your current setting of 2.7 Ghz.

If you're not compulsive-obsessive about over-clocking, I still think this should run fairly cool at 3 Ghz, and sufficiently so with the cooler you cite.

 

[MadScientist]

Here's my Core Temp load temps (Prime95 blend setting after 6 hrs) on my Q6600 GO with another "sub stellar" HS, the Cooler Master Hyper TX2, which is close in performance to your AC Freezer Pro. I have a "stellar" case though, the Antec 900, that has excellent airflow, and my ambient (room) temp is 68 F. Prime95 Small FFts setting gives me about the same temps.
Cpu Speed(Ghz)----FSB----Bios V----Temp (avg 4 cores)load
2.448-----------------272----1.2375------------43 C (stock default on my MB)
3.303-----------------367----1.2575------------46 C
3.402-----------------378----1.2775------------50 C
3.510-----------------390----1.3175------------53 C
3.600-----------------400----1.3575------------58 C
My idle temps range from 29 C at stock to 36 C at 3.6 Ghz. The Duck is right. Your idle temps at that vcore are high. Do you have your AC Freezer Pro HS aligned so the fan is facing towards the back, not up or down, in your case? Did you check if your heatsink's base and the IHS of your CPU is flat? What TIM (thermal paste) did you use with your heatsink, the one that came with it or something else? If something else did you apply it correctly? Did you install your HS correctly? Those plastic push pins are sometimes a PITA to get in and can pop out if not fully seated.

Sorry Duck, I couldn't resist. Those were "stellar posts".

"Happy Holidays" I'm trying to be politically correct.

 

[BonzaiDuck]

Feliz Navidad.

I'm still "pushing" IC Diamond or similar thermal compound, but not enough to take time linking it in here from HeatSinkFactory.

That's pretty darn good with that G0 and the "sub-stellar" HSF. Lessee -- 3.6 Ghz will give you about 64C core average at load when ambient is 78F.

It looks as though G0 Q6600 is still an option for me, if I don't want to replace my mobo or wait for Penryn prices to drop. I may not be able to use the Yorkfield with the mobo, either.

[Darn!! That's good -- 3.6 and 1.36V? You got a GOOD one! The bar-chart distribution of VIDs and VCORE settings I've seen -- if I recall -- puts your stock VCORE below average. I'll want to look again, but . . . . . GEEZ!!]

 

[MadScientist]

It's an OEM chip I got from ChiefValue. It surprised me too that I could reach 3.6 Ghz with only 1.3575V. It will do 3.285 Ghz at stock 1.2375 vcore Prime95 stable. I've got the fans on the Antec 900 on mid setting and keep my thermostat at 68 F. I'm going to set it at 3.3 or 3.4 Ghz. My son, the next owner, likes to keep his apartment toasty.

 

[amheck]

Is the arctic pro 7 cooler really that bad? I did some reading, and it seems that it was a pretty good bang for the buck. It seemed to be half the price of some of the other big names and people seemed to be pretty fond of it, from what I gathered.

Oh well, mines out for delivery today.

 

[BonzaiDuck]

Not THAT bad. Consider that I, myself, am compulsive-obsessive about "best" air-cooling. I had order the TR-Ultra-120 (original) for this system, getting ready to make the installation when the news about the "Extreme" version appeared. Then there was the news about the "Extreme's" slightly convex shape, and I realized I'd have to lap the heatsink base. Then I found a reseller selling "custom-lapped" Extremes. I paid the extra $30 to save myself the trouble.

Reviews often mask "performance" in two ways. They either don't show comparisons with other coolers, or they exclude the best coolers from the comparison. Beyond that, they usually show the evaluation with a controlled room-ambient temperature, the idle temperature, a load temperature, and a processor spec. The processor spec, in turn, gives a TDP or thermal-design-power in thermal watts. They may compare the same processor at different over-clock settings, which then implies higher TDP's for those over-clocks.

But they often do not show "thermal resistance." TR folds into a single number the information about idle and load temperature values with the TDP, and excludes a need to adjust for room-ambient, because it measures a change in temperature given a certain processor thermal wattage. It is THE accurate measure of heatsink effectiveness and performance, provided that a range of fan-speeds that may change thermal resistance are also reported.

Look for reviews on the Arctic Cooling 7 Pro, see if you can assess the thermal wattage of the test-configuration, make a note of the room-ambient test-condition, and see if you can't compare the temperature data with the data shown in the Anandtech review comparisons (some 20 coolers) with the ThermalRight Ultra-120-Extreme (May 5, 07) and the Ultima 90 (August-something, 07). That'll give you a better idea . . . .

 

[BonzaiDuck]

FN: In event people find reviews that feature measures of thermal resistance, pay attention to the following:

"As performance and effectiveness increase, THERMAL RESISTANCE DECREASES."

We estimate the thermal resistance of the TR -ULTRA_120-Extreme to be about 0.097 C/W, based on data per the unlapped heatsink sample. By comparison, a high-end $200 to $300 Swiftech water-cooling kit shows effective thermal resistance closer to 0.12 C/W.

 

[plewdawg]

Originally posted by: MadScientist
It's an OEM chip I got from ChiefValue. It surprised me too that I could reach 3.6 Ghz with only 1.3575V. It will do 3.285 Ghz at stock 1.2375 vcore Prime95 stable. I've got the fans on the Antec 900 on mid setting and keep my thermostat at 68 F. I'm going to set it at 3.3 or 3.4 Ghz. My son, the next owner, likes to keep his apartment toasty.

Do you know what batch # you got?

 

[jhick]

I figured out what the problem is. I had set the motherboard to control the HSF speed. It ended up that the RPMs would not go about 1300. Which is about 50% of max speed for the arctic cooling pro fan. I now set up a profile with speedfan and my idle temps with 2600rpm fan speed are now:

34--34--32--30--

That is with no overclock. So fsb=267 and vcore=1.3

I will prob keep my the highest temp for a core around 37 since at that temp the fan only needs to operate at 40% and is very quiet.

When I overclocked...

fsb = 333MHz
Vcore = 1.325V
load temps = 53-53-47-46 after 15 minutes of prime95

I could lower the vcore at that fsb, I know it can run at 1.265 at 333MHz and was stable for 2 hours. So I will probably lower the vcore or if I am feeling ambitious later do some more overclocking.

I am happy with that now. The 5+ degree drop makes a big difference.

 

[TwYsTeD]

Not to hijack this thread or anything but I'm kinda new to intel cpu's and overclocking...

I just got my G0 q6600 from newegg - got it to 3.5 @1.41 vcore. I cant seem to make it to 3.6 despite pushing to 1.48v. My temps are fine, at 3.5 im idling at 35 - 38 - 32 - 36 (coretemp). At 1.48v i can boot into windows at 3.6 but I'm having stability issues. I think my temps are still ok but im hesitant to push my vcore to 1.5+. Should I stay away from the 1.5ish vcore range?

 

[jhick]

the key is what your temps are at load. You should be running prime95 for a period of time to check for stability and temps after every adjustment. If your temps are below 70 and you have stability at 1.48 vcore. I can't see why you shouldnt increase. Every chip is different, so as much as you can get advice online, it is a lot about just testing and retesting.

 

[Idontcare]

Originally posted by: TwYsTeD
Not to hijack this thread or anything but I'm kinda new to intel cpu's and overclocking...

I just got my G0 q6600 from newegg - got it to 3.5 @1.41 vcore. I cant seem to make it to 3.6 despite pushing to 1.48v. My temps are fine, at 3.5 im idling at 35 - 38 - 32 - 36 (coretemp). At 1.48v i can boot into windows at 3.6 but I'm having stability issues. I think my temps are still ok but im hesitant to push my vcore to 1.5+. Should I stay away from the 1.5ish vcore range?

Are you volting up your NB to handle the increasing FSB frequency? What about your NB temps?

Juicing the quad will only take your overclock to the point where the other components in your system can start to become the limiting factors.

 

[TwYsTeD]

Originally posted by: Idontcare

Originally posted by: TwYsTeD
Not to hijack this thread or anything but I'm kinda new to intel cpu's and overclocking...

I just got my G0 q6600 from newegg - got it to 3.5 @1.41 vcore. I cant seem to make it to 3.6 despite pushing to 1.48v. My temps are fine, at 3.5 im idling at 35 - 38 - 32 - 36 (coretemp). At 1.48v i can boot into windows at 3.6 but I'm having stability issues. I think my temps are still ok but im hesitant to push my vcore to 1.5+. Should I stay away from the 1.5ish vcore range?

Are you volting up your NB to handle the increasing FSB frequency? What about your NB temps?

Juicing the quad will only take your overclock to the point where the other components in your system can start to become the limiting factors.

No I havent upped the voltage on the NB at all