Finding the Best CPU Cooler: 10 units reviewed and tested

[cliffordcooley]

I for one have not built up the nerve to go water-cooling, not even closed-loop. And if I did decide to get the best air-cooler, I'd like to know if changing the fan would make it more efficient. I understand testing what comes in the box, I feel that is more important to most.

Why would you buy an expensive cooler, then remove the fans and throw away the thermal paste?

Whether we would or not is our decision to make. And yes, I have thrown away the thermal paste that comes with the cooler, because I already have a specific thermal paste I use. If I were to pull the heat-sink and reapply, I'd use it anyway so might as well start with the same paste.

As for the fans, if you tell me that changing fans will not make a difference in thermals, I will leave it at that. But if the fans do make a difference your review leaves it open to speculation as to whether the weakness is heat-sink or fan. I was only suggesting, I'd like to know which.

 

[Steve]

As I said you need to read a heatsink review and not a cooler review, end of story. Why would I argue about the cooling performance different fans have? That’s not even up for discussion here, we are reviewing cooler packages, shouldn't be hard to understand that.

When we review a laptop for example we review the entire package which helps determine its value. We don’t just format the drive and set it up the way we like, we don’t use an external keyboard because we might prefer that and we don’t remove the battery and go with a bigger one because it lasts longer.

Good quality fans and even thermal paste for that matter aren’t cheap. So if you were to buy the Xigmatek Dark Knight for example and then throw the fan away and replace it with something you deem better along with some thermal paste you might be adding $20 to the cost. So how do we review these products according to value? The most important criteria that we go off for any review.

Furthermore you can’t even change the fans on the Thermaltake coolers and the Thermalright Archon SB-E X2 uses a custom 140mm fan. It might be possible to mount other 140mm fans though they wouldn’t fit as well.

 

[GhostRyder]

I for one have not built up the nerve to go water-cooling, not even closed-loop. And if I did decide to get the best air-cooler, I'd like to know if changing the fan would make it more efficient. I understand testing what comes in the box, I feel that is more important to most.
Whether we would or not is our decision to make. And yes, I have thrown away the thermal paste that comes with the cooler, because I already have a specific thermal paste I use. If I were to pull the heat-sink and reapply, I'd use it anyway so might as well start with the same paste.

As for the fans, if you tell me that changing fans will not make a difference in thermals, I will leave it at that. But if the fans do make a difference your review leaves it open to speculation as to whether the weakness is heat-sink or fan. I was only suggesting, I'd like to know which.

Indeed, I know many people who swap fans at times to improve certain areas, but the point of this review was what comes with the original kit.

I personally do the same thing Clifford, I use my own special Thermal paste I like over what comes in the kits. Makes the performance consistent at least in that respect and its going to be better than what comes with it.

 

[dividebyzero]

As for the fans, if you tell me that changing fans will not make a difference in thermals, I will leave it at that. But if the fans do make a difference your review leaves it open to speculation as to whether the weakness is heat-sink or fan.

That tends to straddle a number of possible test issues. The only real foolproof way to test comparative heatsink design efficiency is to measure the coolers in passive mode (sans fan(s)) which represents its own problems with coolers with integral fans.

When it comes to fans, you also have the problematic issue of changing specifications without documentation- either by the ODM/OEM, or by the batch assigned to various outlets. The classic case in point here is the Yate Loon D12 - a ubiquitous, cheap, readily available sleeve bearing fan sold under its own name or rebranded in various PSUs and coolers. The Yate Loon site lists a single specification for each RPM gradient (88 CFM for the SH high RPM variant), but it's fairly well known that these fans aren't created equal and a lot depends upon in what form these fans are purchased (or in this case, retail vendors batches).

It's probably a given that higher RPM results in higher static pressure which yields higher airflow over the fins/heatpipes, which lowers temperatures- at least until the point is reached where the wick inside the heatpipes becomes overwhelmed (conduction) and/or the cooling area is simply not sufficient (convection). So, for individual testing it's probably more a case of doing some detective work on what fan you actually have (if it can be ascertained) and measuring its static pressure specification against those of other vendors...assuming they publicise the information- many noted fan manufacturers such as Scythe do not list static pressure....and a number of those that do are rather economical with the truth regarding pressure and airflow claims.

For comparative testing, issues with sizes are going to come into play (and possibly proprietary designs- EDIT Ninja'ed by Steve)- do you test a cooler designed for a 92 or 120mm fan with the same 140/150mm fan(s) that come packaged with some of the larger coolers even though the smaller cooler wouldn't likely benefit too much? You certainly couldn't run the reverse scenario since a buyer of an NH-D14 wouldn't turf out the bundled fans to use 120mm models.

 

[Steve]

Your effort and general ‘can be botheredness’ really puts me to shame, well done once again sir.

 

[dividebyzero]

Your effort and general ‘can be botheredness’ really puts me to shame, well done once again sir.

Thanks! I remember starting out with air coolers and while it was fun to debate which cooler was 1°C better than its competitor (at least for the few weeks it stayed unclogged with dust), that pretty much wore off along with the novelty of tower coolers. After having to fit, replace, and clean these damn things the overriding considerations are now how easily it can be installed, and do I have to take the motherboard out.
Any of these aftermarket coolers are going to do the job on a stock/mild OC CPU, and to my way of thinking, if you're getting more aggressive with clockspeed bumps then a certain onus falls on the user to research the finer points of the particular hardware. No review I don't think can be a one-size-fits-all scenario, any more than you can say that X MHz overclock is guaranteed using Y extra voltage producing Z amount of heat.

 

[cliffordcooley]

As I said you need to read a heatsink review.

TS link please, comparing all these coolers. Am I mistaken by thinking there was a line suggesting it has been a long time since a review has been done?

Wait, I know what it is. You are telling me to go somewhere else and read my technical news.

That’s not even up for discussion here

In other words you were in a hurry and personally feel the review is closed.

we are reviewing cooler packages, shouldn't be hard to understand that.

I didn't realize you could buy them any other way. Forgive me for suggesting a more complete review of the packages, which packages seems to be the only way you can get them. I'm not downing your review of the packages. I'm not suggesting you change anything that was already done.

Side note: The coolers that have custom fans are coolers I wouldn't be interested in. I hate proprietary crap, that can never be serviced. I wouldn't want to replace the entire cooler because the fan issues. I wish everyone would quit buying things that can not be serviced.

 

[Steve]

TS link please, comparing all these coolers. Am I mistaken by thinking there was a line suggesting it has been a long time since a review has been done?

Wait, I know what it is. You are telling me to go somewhere else and read my technical news.

If I have to be the one to break the bad news that we cannot possibly cover every last topic at TS then so be it. You need to find an article that takes the very best air-cooled heatsinks and compares them with the same paste and fan(s). We don’t have such an article and I am not sure if and when we will.

In other words you were in a hurry and personally feel the review is closed.

If I had to put it into other words I would say it’s a stupid comment to make given at no point did I make such a claim and nor would I. If by now you don’t get it, given the explanations myself and dividebyzero have given, then I tap out.

I didn't realize you could buy them any other way. Forgive me for suggesting a more complete review of the packages, which packages seems to be the only way you can get them. I'm not downing your review of the packages. I'm not suggesting you change anything that was already done.

If you believe that then you really need to do some more research. There are plenty of high-end heatsinks that come as a stand-alone item, there have been for at least the last decade anyway.

 

[Guest]

Please update it with coolermaster hyper 212 evo...It's better than many of the costly CPU coolers in the market

 

[Guest]

Thank you for the article - I just bought a Intel 4770K and even running at stock, the boxed fan idles at 80-85C and I can't even reboot without receiving a CPU over temperature warning from my BIOS!