The Official Cooling Thread!

Remember they had the Freezer 7 Pro posting better temps than a TRUE. They both had fans BTW, so in my mind they are just unreliable from then on.

 

Thermalright Ultra-120A (not extreme)
3.15Ghz CPU (75% overclock) at 1.41V
All case fans lower than 750rpm
An hour of Supreme commander:Forged alliance
CPU temp? 39C

No I'm not!

I'm going to passively cool my graphics card next. I'm thinking of waiting for the new generation cards before I do that, since if I can get the same level of performance with a single card as I do with crossfire now, and in every game, then spending all that on getting a pair of GPU heatsinks is madness.

 

Uhuh, now try doing that WITH NO FANS on an overclocked Q6600.

You still have case fans.

 

That's what I'm going to do once I buy a Q6600. I'm not anticipating any problems.

 

Yeah, well I am, Unless you live in England or something where it's always cold lol.

Seriously, I know how horrible it is trying to keep a Q6600 cool. With a freezer 7 Pro at FULL SPEED it still would go OVER 70C easily in about two minutes of running Prime 95 at 3.6 Ghz.

 

With a freezer 7 Pro at full speed my CPU ran at nearly 50 degrees. I think you underestimate just how much more powerful the Ultra120s are. Not only that, I'm not going to push the Q6600 to its overclocking limit, nowhere near it. This is for several reasons really. Firstly so it doesn't dump so much heat into the case I have to turn my case fans up, secondly so it doesn't get so hot I need to worry about having a fan on my CPU cooler, and thirdly so my power consumption remains relatively low. Electricity's not cheap these days. My plan is to run it at pretty much the stock vcore, I'm aiming for a CPU speed around 3.1Ghz or so, in continuity with the speed I have now, but just more cores of it.

Additionally, Prime95 isn't your average program. I've no doubt that my CPU will get hot running it for extended lengths of time, but if it can run that without throttling, then temps running a normal application will be fine. I assume Prime is multi-threaded?

 

Yeah I guess it will do fine then. Yes Prime95 uses all four cores. Next time I get a Q6600 I am getting a much better CPU cooler and KEEPING it at 3.6Ghz.

EDIT: Sam, also if you want to save electricity why do you keep your voltage on auto?

 

Because as nice as saving electricity is, I'm willing to sacrifice a small amount to guarantee my CPU's stability - as you know I like overclocks to be plain and simple, and even if it costs me a few watts, that's fair enough. Changing the brightness on my monitor by one notch is enough to outweigh the gains by optimising my CPU voltage.

 

don't you save elec if the VCore is on auto? why is that so?

my E2160 @2.2Ghz temp is very low, 17 degrees at idle and about 40 degrees on load.

Why some CPU is hotter than others?

 

If the voltage is set to auto the motherboars selects a voltage that it knows should be stable for the CPU. Most of the time you can run the CPU a fair amount lower than this if you're willing to spend the time finding out how low you can go before stability problems occur.

Your CPU will run cool because it isn't overclocked very far, remember we effectively have the same CPU, just yours has less cache, and mine's running 950mhz higher than yours. However, to have an idle temp of 17C your room must be very cold, a lot colder than most.

 

@sam
It doesn't take very long to figure out the lowest voltage stable for a certain overclock, and then if you leave it running like that till your next upgrade(6-7 months) w/o any tweaks, it would be worth keeping the cpu at a lower temperature.
For example, if I "auto" the voltage on my e6300, the mobo selects 1.45 volts whereas if I select it myself, I can go down to 1.15 which is awesome cause I can't afford to run the A/C in my house ON 24/7 and if the A/C is not on, the cpu is not stable with 1.45volts running through it. Now, when I drop the volts to 1.15, it is completely stable 24/7 with orthos, prime95, etc w/o needing the A/C.
-im1992

 

Fair enough - I might give that a try, certainly will with my quad core when I get it.

6-7 months per upgrade? Wow, money runs well in your household! I've had my E4300 for 16 months, and it'll be another couple of months at least before I get a Quad.

 

from anand tech




i see wat you are saying here....

but now look (from the same article)

THIS IS WHY YOU BUY CPU COOLERS:






boozer, you cant use just oen screenie.

the last two are the money shot. they are the ones too look at.

 

Would this be the same anandtech review that we used as why you shouldn't trust anandtech?

 

lol, i was guessing other ppl who actually have a job probably upgrade their computers every 6-7 months
I am 15 and don't have a job so I tend to upgrade(more ram, new gpu, new cpu) my computer every 10-11 months and if i find a great deal, even earlier.
My experience with upgrading CPUs is that every time you upgrade the CPU, you need a new mobo at least and thats where my dads credit card comes in, hehe.
-im1992

 

no its the one that proves that they are not crap.

 

I still don't believe them. My fanless TRU-A gives me a temp 8C lower than I got with the freezer at max speed, and at a slightly higher voltage too. I now run the voltage lower, but that's not really relevant here. That review is bull.
Oh, and I didn't redo the arctic silver when I plonked this cooler on either...

im: Heh well, this is my second mobo for the same CPU currently, but I've no need to get another for the next CPU I'm planning on adding...

 

abuzar1,

And your point is? To me, your point seems to be that either you don't know what you are talking about, or you don't understand it! Not to pick on you, but you have a very bad habit of badmouthing products with little or no knowledge or experience with them!

My temps aren't going to meaningfully get any lower if I replace my AF7 Pro with a TRUE tomorrow. That doesn't mean that I'm saying that the AF7 Pro is better. All it means is that it does an excellent job for me in my given setup, so I don't need a TRUE! Some people right on this thread have had problems cooling, even with a TRUE! Did you think that Anandtech was trying to boost Arctic's sales by making false claims? That's not very likely as it's already a very popular and effective cooler to begin with, and one of the top sellers!

I know I've said this before, but Case Ventilation is everything to your CPU Cooler! Case Ventilation is rapidly becoming an art, and a Black Art at that! We've reached the point with all that goes on inside the case these days, that the addition or change of a single component can and does wreck havoc with the airflow in your case! I had to radically change my case's airflow after I installed the Enzotec Northbridge cooler! Why? Because by adding a simple Chipset cooler it changed where more of the heat was coming from after it was installed!

I turned the cover 80mm fan around to make it an intake instead of an exhaust, running it at 25 cfm at slightly above minimum speed. I had to increase the rear case 120mm to 1400 rpm and about 78 cfm. 1300 rpm would be ideal, but there's an annoying beat frequency between the rear fan and something else in the computer at that speed. one of those annoying R-R-R sounds as the sounds synchronize in and out of frequency! I still maintain about 100 cfm of case airflow, have enough negative case pressure to supply air to the static vents on the side cover and the back of the computer. Case manufacturer's are already starting to experiment with miniature air ducts to better direct the airflow to where it's needed inside the case. I personally hope that they work out as it's a PITA to have to try and control where the air goes the way I'm doing it now. What happens if I decide to upgrade my video card next week? It's back to the drawing board for the airflow, again!

Speaking of the Drawing Board, Right now I'm working on what I call SCCS (catchy, no? LOL!!) or Secondary Computer Cooling System. If I'm successful, it should put an end to most cooling problems that are almost always caused by the differences of "what's in the Box" to begin with! There are thousands of combinations of components and hardware, any of which can cause heat problems! Add to that the hundreds of different cases on the market and it makes for the problem I'm hoping to solve!

BTW! When you read the Anandtech article, did you notice what the Ambient temperature was in the test setup? It was 65F/18C, or down right chilly! Not at all a temperature you are likely to find in the real world home or office. Why is it so surprising that the AF7 Pro did a little better than the TRUE under those conditions? It takes up much less space and interferes with the case airflow less than the TRUE, so why shouldn't it do better? I'm sure if the Ambient temp was a more normal 24-25C that the clear winner would be the TRUE! I can see Anandtech's thinking! 15 or so computers running in the same room generate a lot of heat, "so we'll keep it cool"! In reality setting the thermostat to 75F or so wouldn't have caused any problems at all and would have made for a better, more comparitive test!

Fault Anandtech for not using better judgement on the room temps, yes! But to say that you don't trust them because of the results they simply reported just isn't very fair in this instance!

Respectfully,
Russ

 

yey russ. I <3 Russ!

xD

its true though boozer, you do slag off products very quickly, w.o personally testing them.

 

Turbulence or otherwise, my CPU uses a very similar amount of power to the one in that test (77W if Everest is to be believed), and yet by comparison the TRUA is vastly superior, let alone the TRUE. This is conducted with a room temperature of 21ºC, not too far off the test environment they used. I'm not saying the results are made up, but there's definitely something fishy going on.

As for your warbling problem Russ, the main check is to ensure that there aren't any fans that are spinning at the same speed. At the level of volume of my case fans, superposition isn't a problem, but once the fans become a bit more audible it can really get annoying.

At 1300rpm, the 9-blade Silverstone will be putting out at 195Hz, or a G note. If you've a 7-blade fan running at 1650-1700rpm or there abouts, that will also be a G note, and at that speed will be loud enough to noticeably interfere, even if it's an 80mm. That frequency is above that produced by even 10,000rpm hard disks, so you can absolve those.
It could be the PSU fan, but if it's a 120mm fan in a PSU, unless it's got a crazy number of blades it'll be a bit noisy, which doesn't sound like the case for yours - presumably due to mounting methods PSU fans get noisy around 1200rpm for 120mms, and about 1600rpm for 80mms.

Edit: oh, and why is Abuzar getting flak for this? He's not even been in this discussion this time round yet!