[POST IDEA] Best Current CPU Coolers

[jordan1794]

Hello all!
So I had an idea for a post, was wondering if others would be interested in helping out.

Seems like there are tons of reviews of coolers out there, but obviously you can't test EVERY cooler at once on the same test bench.

For this reason, I was thinking about trying to create a list of the best CPU coolers, to be continually updated. By researching and referencing other reviews to make a "master list".
(Talking PURELY in cooling performance, NOT taking into consideration value, durability, etc.)

I.E.

Review 1 shows Cooler A is better than Cooler B
Review 2 shows Cooler C is better than Cooler A

This would (theoretically) put the "pecking order" as
#1 Cooler C
#2 Cooler A
#3 Cooler B

The replies to the thread could be used for discussion for fine tuning the list, as well as suggesting new entries.


What do you guys think, good idea or terrible?

 

[jgunning]

I think the Cooler Master Hyper 212 Evo should be in that list!



JG

 

[jordan1794]

Alright, so here is what I have so far, just to give a rough idea of how I'd do it...
Base on just googling a couple reviews and using the numbers listed, here is the "pecking order". Using a Hyper 212 Evo as a baseline, and comparing the temperature difference each cooler has from it.
(Important to note that some of this is inferred, as explained in the first post!)

#1 Corsair H110i GT -28.4
#2 Glacier 240L -27.9
#3 Corsair H100i -27.2
#4 Corsair H110 -26
#5 Phanteks PH-TC14PE -25
#6 LEPA Aquachanger 240 -23.5
#7 Silverstone Heligon HE01 -23
#8 Noctua NH-C14 -4
#9 Cooler Master Hyper 212 EVO (BASELINE) 0

 

[jordan1794]

Also, list would be further broken down into Air/ Water categories, as well as 120mm rads, 240mm, and any rads over 240mm.

 

[MIRTAZAPINE]

The problem with cpu cooler reviews is they are reviewing a "cooler" that means the heatsink with the included fan for air cooler or for water cooling the radiator, pump and fan combo. Sometimes the thermal compound bundled is factor in. This carry with it many variables. I see you want to do some kind of meta-analysis or some sort. You would have to take to account the testing methodology for different places. Cooler have variation between the even for the same model due to manufacturing tolerances. There is also the factor of how you mount it and how the thermal paste spread and form.

If you want a controlled review for just air coolers testing it with no fans at a start is one way. And testing the air coolers with the same for 120mm fan. Even with this it is not that straight forward and you need a few baseline fans as reference. Some cooler works well for low rpm fan, some don't. Some need high pressure fan and some less. In the test you would need to at the very least remount 5 times and do the test 5 times to see the variations.

 

[jordan1794]

The problem with cpu cooler reviews is they are reviewing a "cooler" that means the heatsink with the included fan for air cooler or for water cooling the radiator, pump and fan combo. Sometimes the thermal compound bundled is factor in. This carry with it many variables. I see you want to do some kind of meta-analysis or some sort. You would have to take to account the testing methodology for different places. Cooler have variation between the even for the same model due to manufacturing tolerances. There is also the factor of how you mount it and how the thermal paste spread and form.

If you want a controlled review for just air coolers testing it with no fans at a start is one way. And testing the air coolers with the same for 120mm fan. Even with this it is not that straight forward and you need a few baseline fans as reference. Some cooler works well for low rpm fan, some don't. Some need high pressure fan and some less. In the test you would need to at the very least remount 5 times and do the test 5 times to see the variations.

Yeah, it's going to be far from perfect, but all the variations combined aren't going to skew results to the point where the list is pointless.

I'd also probably just go with the coolers as they ship out, with their stock fans.

 

[sneekypeet]

Been able to run the same test bench for the last two years for CPU coolers.
I also feel the only major change for a lot of reviews is the test used to load the processor, and the included paste can play a degree or two in the results, and is why I use the same paste on every cooler. As far as fans go, you have two options really. Use the same fan on all coolers, or use the one that comes with it and show what potential customers will get out of the box, not best case scenario. As far as running coolers without a fan, a lot of coolers will fail right out of the box with stock CPU loads applied.

As I say testing is the toughest hurdle to look over when trying to compare coolers from site to site.
For a thread such as you want to compile, I would go to sites and find the results, such as this: http://www.techpowerup.com/reviews/Silverstone/Tundra_TD02-LITE/6.html or http://www.tweaktown.com/reviews/73...formance-liquid-cpu-cooler-review/index7.html where you can digest a ton of coolers at one time. Find a few that run the same testing methods, and you could create a spreadsheet pretty easily and combine them into your own easy to read chart.

Just a few pennies from someone in the game.

I would also assume @crazyeyesreaper may have some tips or points to consider when trying to assemble this sort of a list.

 

[trog100]

noise and the speed the fans have to run at plays a big part.. its not rocket science.. the bigger and noisier the cooler is the better its going cool..

to be honest they are all hitting the law of diminishing returns.. double the size and weight they need to be just just to win out by a couple of degrees on some review site..

the whole cooling game (including case fans) is turning into over hyped nonsense..

trog

 

[crazyeyesreaper]

All i can say is good luck sites that use similar testing methods would be easy to take data from. However lesser known sites. Hardware configuration differences. say x79 / x99 vs z87/ Z77 / Z97 / 170 etc etc.

THe other problem is do to cooler design aspects example: convex or concave base can favor certain CPUs. Even then the disparity between voltages on the CPU / Clock speeds / Intel's TIM under the IHS on mainstream chips making temps variable between processors etc.

 

[-1nf1n1ty-]

I think this would be a great idea, I was just thinking about making a thread asking for CPU cooling advice as my Antec Kuhler is getting old and it's bulky, I hope its fine if I ask my question here?
http://www.ncixus.com/products/?usa...&vpn=EN6619&manufacture=Xigmatek&promoid=1428

I want this for my i7 4770k but I haven't seen any reviews last time I got a great CPU cooler Noctua and Xig were the "go to" coolers and they were awesome.
don't mean to hijack the thread if I am then just ignore me and Ill keep on looking!

 

[Hnykill22]

NZXT Kraken X61 is an excellent choice for high end cooling

 

[bonehead123]

I think the Cooler Master Hyper 212 Evo should be in that list!



JG

^^^ +1

 

[dorsetknob]

To many variables to take into consideration
Eg
ambient air Temp
ambient Humidity
Software testing load need i go on

still if you want to do it then i am sure your have the support of the mods and lots of replies

Also lots of trolls and flamebaiters ing just waiting to lock and load.............

 

[zo0lykas]

Alright, so here is what I have so far, just to give a rough idea of how I'd do it...
Base on just googling a couple reviews and using the numbers listed, here is the "pecking order". Using a Hyper 212 Evo as a baseline, and comparing the temperature difference each cooler has from it.
(Important to note that some of this is inferred, as explained in the first post!)

#1 Corsair H110i GT -28.4
#2 Glacier 240L -27.9
#3 Corsair H100i -27.2
#4 Corsair H110 -26
#5 Phanteks PH-TC14PE -25
#6 LEPA Aquachanger 240 -23.5
#7 Silverstone Heligon HE01 -23
#8 Noctua NH-C14 -4
#9 Cooler Master Hyper 212 EVO (BASELINE) 0

from this list i would pick CM hyper 212, if you wanna air cooler
but for sure i would go for "zalman reserator max 3" if i got remember hes can cool down up to 300W cpu

 

[crazyeyesreaper]

Why do people love the CM Hyper 212? Its now more expensive then better performing competition. I can't quite understand the communities attachment to said cooler.

 

[Jetster]

It trendy. It goes on sale more often then most so there are allot of units out there. This idea of a list may work but really without a controlled environment it would be very subjective

 

[bonehead123]

Well, when I looked this weekend, it was $28 something at amazon......

Just exactly how much should the average user pay for a really good air cooler anyways ?

And most the most likely reason for people liking it is that it just works, is easy to install, looks good, and does it's job with no muss, no fuss.....PLUS it is very affordable too

 

[rtwjunkie]

Why do people love the CM Hyper 212? Its now more expensive then better performing competition. I can't quite understand the communities attachment to said cooler.

Indeed. It's decent, but there are ALOT of decently priced coolers better. Heck, although about $40 more, the one in my system specs is WAY better, but not even in the sample list.

In general, I think this is a noble idea, but coolers are one of those subjective (in part) pieces of equipment, because A may not test well here, but performs great for individual over there.

Also, doesn't (Edit) Frostytech already maintain a thorough list?

 

[jordan1794]

Well, no one seems to think it's an absolutely terrible idea, so I'm gonna go ahead and work on it a tab bit.
Basically I'm just going to google popular coolers and include them, if anyone has any that they think should be added please list it (If you can provide a link to a testing page, with several other coolers tested on the same bench, that would be INCREDIBLY helpful)

 

[rtwjunkie]

http://www.frostytech.com/articles.cfm

 

[n-ster]

I don't think you can create an objective list, and I don't think using only cooling data is going to be terribly relevant. Noise will be a factor for most, separating Water and air and subdividing water into single vs dual (ie 120/140 vs 240/280) and having a low-profile section for the air coolers is a start.

I think you can make a "recommendation" style article, with certain criteria, but it will be subjective. Best all-around cooler, Best noise per cooling, Quietest etc based on reviews though.

EDIT: sorry for my terrible wording

 

[Kanan]

The king is missing, the good old Noctua NH-D14 - and do I really need to post a review?

Though I like this review: http://www.tomshardware.com/reviews/nepton-280l-tundra-td02-water3.0-pro-reserator3-max,3607-13.html

 

[jordan1794]

Yeah, I guess to be more specific I'd be going for the best coolers based SOLEY on performance, for the people who don't mind noise.
I could then, rather easily actually, create sub-sections for coolers under 40 db, 30 db, 20 db, etc.

Time to build one hell of a Excel spreadsheet haha. Luckily I absolutely love making them

 

[Jetster]

Fill out the country in your user CP What is a db?

How about a poll? They you can make a spread sheet of the results maybe?

Or have people list there top three

1.
2.
3.

Because you cant just go by results people have. Its not controlled

 

[jordan1794]

Fill out the country in your user CP What is a db?

How about a poll? They you can make a spread sheet of the results maybe?

Or have people list there top three

1.
2.
3.

Because you cant just go by results people have. Its not controlled

Haha, sorry...was talking about noise.
Guess the correct abbreviation is dBA

And what I figure is, I'll make the list based on the findings that I can google. Then the community can help by challenging the list order, either by their own personal results or by linking reviews that provide contradictory data.
It will never be perfect, however it will still be helpful...obviously the top 5-10 coolers will all be nearly identical in real world usage, but having them listed together makes a decision much easier, as you can single it down to price and/or noise levels.

Further, just to specify, I would only go by reviews where several coolers were tested on the same bench, hopefully removing a good portion of variation.
Again, it will be VERY far from perfect, but I think it will still help.