Liquid Cooler Lineup: Intel, Corsair, MainGear Tested

Article Index:   

Over the last few years, we've seen an increasing number of liquid coolers positioned as high-end alternatives to traditional heatsink+fan combinations. This has been particularly true in the boutique market, where a number of manufacturers now offer liquid coolers in one form or another.

Corsair has a full line of liquid coolers; we've reviewed the company's H80 and H100 here and compared them to Intel's own liquid cooling solution as well as an ultra-high end model Maingear sent over. This last entry is specifically designed for integration into the company's systems rather than as an aftermarket purchase, but it's an interesting data point to have in the charts nonetheless.

Intel's LGA2011 and hexa-core Sandy Bridge-E processors were the logical choice for evaluating maximum thermal dissipation, particularly considering that all of the coolers we tested retail for $75 and up. Unfortunately, this limited our choice of comparison hardware as far as what we had on hand.As a result, we don't have any results from older / lower-cost CPU coolers that aren't compatible with Intel's new socket.

How Liquid Cooling Works:

The image above is from Asetek's website, but all of the liquid coolers we've tested function in the same basic manner. Water is pumped through the CPU cold plate and up to the radiator where a fan blows cool air over the fins and lowers the coolant's temperature. The now-cooled water flows down the return pipe and is again pushed over the coldplate.

Water is a far better conductor of heat than air, which is why you'll often see water coolers promoted as being more efficient solutions for high-end enthusiasts. Simply comparing the heat transfer characteristics of air and water, however, grossly oversimplifies the situation. Water coolers still rely on fans (in some cases, more than one) and they incorporate an electric pump that generates additional noise and an additional point of failure.

The kits we're reviewing today are a far cry from the water coolers enthusiasts have been building for years. DIY water coolers typically involve separate reservoirs and external pumps. The systems we've tested contain significantly less fluid and use small pumps directly integrated into the cooling block.

Integrated all-in-one kits may not offer the theoretical performance of a high-end home-built system, but they're vastly easier to install and require virtually no maintenance. The tradeoffs are more than fair, provided that the coolers perform as advertised.

Image gallery

Related content


LLeCompte 3 years ago

Good review. I almost bought the H50 last year when i was building my new rig, the only thing that stopped me was that i read a review in which a H50 busted and leaked out into a system. That scared me off, so i went with a corsair A50 air cooler. It keeps my Phenom II x6 around 30C under full load.

realneil 3 years ago

These factory sealed water coolers do a great job of cooling. They still have fans attached to them, so the noise is not much different than an Air Cooler is.

I have an H50 and an H60 here now. They do a great job of cooling my i7-870 (H50) and my i7-2600K (H60).

I've read about problems that have cropped up with some lot numbers of the H100 coolers, but I suspect that Corsair is getting that fixed and under control by now.

I really like the H100 a lot, and would certainly appreciate having one to use on my next build, if that ever happens. If I can't afford the H100 at the time, I'll go for the H80 because it's a fine cooler too.

CMachicote 3 years ago

I like the closed system but I think should give options for video cards also a tee caped off in line and a water block for card would work out nice but then maybe you would need a better pump?? so maybe just a small closed system for cards

AKnudson 3 years ago

i really like the look of the H-80, but  am with Compte if there is a chance that it will leak into my system i would rather go with an air cooler.


What i the cost of air coolers, and how well do they cool?

AKnudson 3 years ago

The H-100 is alot bulkier than the H-80 where does the rest of sit? is it an external cooler?


marco c 3 years ago

H100 is only compatible with certain cases that have mounting locations at the top for a large radiator.

Drake_McNasty 3 years ago

Thanks for the review Joel. These coolers are definitely the way to go if you are building a new system. I really wish I would have gotten the epic 180 with my system! That thing is awesome.

What thermal paste were you guys using? I know maingear uses their own blend, it would interesting to see if the corsair stock paste or the maingear blend worked better for these coolers.

Also when are these companies going to start testing stock shrouds, most of the forums I've read see decreased temps with a push/pull configuration with a shroud.

dvsman 3 years ago

Good read! I have a Corsair H50 on my i7 920 and a Antec Kuhler on my 2600k. Both work great, low temps and volumes, but my Antec has developed a dripping sound since I have my Lanboy Air standing up "tower" style (no leaks just the sound).

der meister 3 years ago

The H100 looks quite nice. I will be defiantly looking at these enclose systems on the next rig I build.

Joel H 3 years ago

There's no reason to think these systems have a tendency to leak. With the H50 you refer to, I'm guessing it was either bad luck or a bad installation--possibly a configuration that put too much tension on one of the coolant lines.

I used the thermal paste that shipped pre-installed on all of the coolers in question. What, exactly that is, I don't know. Paste matters less than it used to, and most of the advantage comes from having it vs. *not* having it. Apart from the tendency of some low-quality pastes to dry out, you'll see similar results from using the silicon compound from radio shack vs. the high-end AS.

AKwyn 3 years ago

Good read; though shouldn't this be CoolIt vs. Asetek (since all they are is branded coolers)? In that case, it sure makes my Asekek cooler feel lonely if it can't handle the 3960X's high heat consumption at 4.5GHz.

Anyways, I think the tests you did were very detailed, so are the pictures... I personally think that thermal paste does matter. (currently using IC Diamond 7 in my build.) but to every opinion his own.

Keep up the good work Joel.

realneil 3 years ago

[quote user="TaylorKarras"]I personally think that thermal paste does matter. (currently using IC Diamond 7 in my build.) but to every opinion his own.[/quote]

I have some of that IC Diamond 7 and have recently tried it out twice. Seems to work well.

I normally use Arctic Cooling Ceramic TIM and have been using it for years.

Joel H 3 years ago


In this case, opinions are irrelevant--and by that, I mean both yours and mine. Data matters. Facts matter. On that note:

Note that Vegemite and toothpaste produce better short-term results than thermal paste--and that all the thermal pastes cluster together at the 0.50C/W mark.

But hey -- that's a very old article. Couldn't blame you if you wanted something newer. So let's check here:

Note that on Page 4, the article states that differences of less than 2'C are considered to be equal. Here, we see *nine* thermal pastes within a 2'C difference. The single outlier, ProlimaTech, might be better--or it may be that the ambient air temperature was slightly lower when that compound was tested.

Now, it's absolutely true that all of the compounds in question turned in results between 32-37 degrees, while the score for having no thermal compound at all was 62 degrees. That's because you really, really need to have *something* in between your CPU and your heatsink. But the question of what that should be?  Just about anything that won't evaporate or turn powdery works.

The only exception is if you plan to OC below the freezing point of water, as I've previously stated. It may also be true that certain compounds dry out more quickly if you plan to run at extremely high temperatures, but again, this is more about choosing the proper compound for a fringe scenario than any normal benefit of one paste over another.

Hope this clears some things up.

realneil 3 years ago

[quote user="Joel H"]Data matters. Facts matter.[/quote]

You're right, but I'm not gonna be spreading Cheese or Vegemite on any CPU's anytime soon. Big Smile

Selling the "best" TIM material has turned into a huge business over the past few years. There are so many studies that crow about this or that working better than all of the rest that you just can't keep up with it,......even ~if you really care~ about it.

In my circumstance, I don't do a lot of overclocking, and I never radically OC any system because I can't afford to keep re-buying parts all of the time. I have to be careful with this kind of activity.

The Arctic Cooling Paste that Someone gave to me (a lot of it) a year and a half ago works well. It establishes that solid, continual, interface between the CPU and the surface of the cooler that helps to mitigate almost all of the heat being produced.

The Ceramic material that's in it has a high degree of thermal conductivity compared to other pastes. And my newest stuff Is IC Diamond (won it in a contest) and that stuff is supposed to be the most conductive there is, because of the super fine Diamond dust that's in it. The theory here is that since Diamond is the most conductive crap substance around, having ~super-tiny~ ground up bits of it in your paste will work the best for you.

The actual difference between the two supposedly premium Pastes? (as measured by my unscientific methodology)   

1 degree Celsius most of the time, and sometimes 2 degrees. This is while running without stressing the system, or doing everything I have to throw at it while my mild overclock is going. (2600K @4.4GHz.) (i7-870 @3.8GHz.) My water coolers are all overkill for the amount of heat that I'm churning out here. My temps stay low all of the time.

So I understand your point about using any old thing that will work. But I can't bring myself to do it like that, especially since the good shtuff isn't all that much more.

Joel H 3 years ago


Sure. I'm not seriously suggesting someone start using toothpaste. FIY on the diamond paste -- don't expect to see a difference. Diamond *is* a fabulously conductive substance, but the entire point of thermal paste is that it fills tiny, microscopic air gaps between CPU and heatsink. Diamond, being the hardest substance on earth, doesn't compress worth a darn (it's already extremely compressed carbon) and it doesn't exactly flow, either.

What you end up with are little bits of diamond with paste around them. You'd need to lay down a layer of diamond just a few atoms thick (and grown so as to align correctly) to realize the potential benefit.

realneil 3 years ago

[quote user="Joel H"]I'm not seriously suggesting someone start using toothpaste. [/quote]

I know this. I never thought that for a second. Surprise

As for the diamond, it works as good, and a tiny bit better than the other high-end stuff I have on hand.

Here is a page with the specs on it. It is made up of Really-Illy tiny-azz particles of diamond when they're done with it, and there is a crapload of it in the paste so you get a very thin film of diamond material that mitigates the heat better. (so they say)

It works well for me and I got it for free, I'll use it.

EDIT: Maybe you should try it out?

MayhemMatthew 3 years ago

These things leak when its a picnic. Not the products fault. I much prefer custom setups, just because you can mount outside of the case rather than all internal. They work, but air coolers are often as efficient while being cheaper.

ajm135 3 years ago

Yeah i would really be concerned about leakage. One small crack or lemon product and bye bye computer

ThunderBird 3 years ago

Good review but I wonder if better temps might be achieved if one were to clean off the heat sink apply your own thermal compound?

Joel H 3 years ago

Doubtful. Most of the advantage of thermal paste lies in having it vs. not having it. Once you've got some properly applied, the difference between the $5 paste at radio shack and $15 stuff is extremely small.*


* - If you plan on using dry ice or phase changers, this isn't true--you need something that won't freeze to the CPU socket or change consistency after it melts.

Ryan H. Cowley 3 years ago

My liquid coolers broke i need a new one =(, I'll probly be getting an Intel one over the Corsair.

cowboyspace 2 years ago

H100 for the win. Just one syupid Question do you have to put liquid in it O_O

realneil 2 years ago

[quote user="cowboyspace"]H100 for the win. Just one syupid Question do you have to put liquid in it O_O[/quote]

H100 is closed loop. (sealed unit) Choose something else if you're looking. H100 has problems. Go to the Corsair forums and read about it there. You don't want it until they fix what's wrong.


Post a Comment
or Register to comment