Antec Kühler Vs. Corsair Hydro: Sealed Liquid CPU Coolers Compared

[Omi3D]

It should be pointed out that a proper installation is key to getting good temps.
A small damp smear of thermal paste on the CPU itself along with the stock paste on the cold plate is the winning solution. Better's the temps by 5 degrees than w/o it.

 

[iceveiled]

Haters gonna hate. My i2500k OC's all day 24/7 to 4.2 (I haven't needed to push it beyond that so it stays there for now) never goes above 36C on my H80, even after hours of games (WoW at 1080p a known processor intensive game, as well as titles like Crysis 2, etc), I have the fan profile set to max but the coolant never gets to the threshold to where the fans actually have to ramp up to be heard. The intake fan on my Corsair 600T is still the loudest component in my system, other than my GPU which I crank the fan up during heavy gaming sessions - can't hear anything during gaming anyway. My point is that cooler keeps my OC'd processor chilly and can never be heard above other components!

I understand all the air cooling guys getting all worked up (which always amusingly comes off as defensive and insecure), but if you can spare a little extra cash these liquid cooling kits really are nice. I was skeptical but after trying them for myself I don't know if I'll ever go back to air! My H80 is sleek and lightweight - it isn't putting all that extra stress on my expensive mainboard and best of all, isn't a towering heatsink obscuring all that beautiful hardware. No more guessing if your memory will fit with a massive heatsink, etc.

 

[panderaamon]

Ok so i've got a question?

What would happen if you got one H100 with a pair of 12 cm peltiers(TEC'S), 4 case fans, configuring the peltiers so they won't go below 5 C (or make 10 i dont know) and making a semi-custom TEC cooler system? would it be that much of a difference?

 

[JackNaylorPE]

What an odd comparison....a $60 cooler and a $100 cooler....a $95 cooler and a $125 cooler. The 620 / H50 works......the 920/H70 works. And I agree w/ others who've said the better air coolers are left out and instead a cheapo used.

here I see a review of various coolers, most of which are in a similar price range.

http://benchmarkreviews.com/index.php?option=com_content&task=view&id=721&Itemid=62&limit=1&limitstart=5

What's next, a graphics comparison of the 550 and 6970 ?

These self contained air coolers are IMO a viable alternative but I couldn't help thinking this is the article Corsair's sales department would have wrote after having released their new things. The 620 and 920 were Antec's versions of the H50 and H70.

I wouldn't put the Hyper 212 up against the Silver Arrow ..... and it just doesn't seem to be worthwhile reading the article which compares a mod level air cooler with the highest performing self contained units.

 

[Crashman]

[citation][nom]ubercake[/nom]Why put liquid anywhere near your system when an air cooling solution is just as effective?[/citation]That question has been answered many times in the article, so why ask it again? In case you somehow missed it over and over again, the answer is to get the big heavy thing off of your CPU socket.[citation][nom]Chesteracorgi[/nom]First: a criticism of your methodology. In comparing the cooling you only give CPU temperatures, and fail to give any acount of the ambient temperatures in the case.[/citation]The air cooler's installation was "optimized" by having the exhaust fan behind it and the top fan above the front of it. And then the temperature chart showed the difference between CPU temperature and the air outside the front of the case. Other components such as the graphics card should see minimal impact from CPU heat, and the VRM temperatures are also shown because that's the biggest secondary concern. But I appreciate the rest of the comment![citation][nom]joeladam02[/nom]The author is not attempting to pit the AIO WC setups against air coolers. The article is rather a synopsis of a broad range of the sealed water systems in comparison to one another, whilst using the air cooler as a reference.[/citation]Bravo and thanks for noticing. The biggest argument I've been trying to counter is from those that believe the "reference" cooler sets an invalid "reference point". Of course that belief is a little lopsided [citation][nom]JackNaylorPE[/nom]What an odd comparison....a $60 cooler and a $100 cooler....a $95 cooler and a $125 cooler. The 620 / H50 works......the 920/H70 works. And I agree w/ others who've said the better air coolers are left out and instead a cheapo used. here I see a review of various coolers, most of which are in a similar price range.http://benchmarkreviews.com/index. [...] mitstart=5What's next, a graphics comparison of the 550 and 6970 ?These self contained air coolers are IMO a viable alternative but I couldn't help thinking this is the article Corsair's sales department would have wrote after having released their new things. The 620 and 920 were Antec's versions of the H50 and H70. I wouldn't put the Hyper 212 up against the Silver Arrow ..... and it just doesn't seem to be worthwhile reading the article which compares a mod level air cooler with the highest performing self contained units.[/citation]It doesn't say this in the article, but much of the spark that started the article was when AMD and Intel started pairing the 920 and 620 with their CPUs. Corsair leading was just an "accident".

 

[sempifi99]

Flow the most important? Not really, head pressure is just as important, if not more so. If flow was the only consideration you'd see all the modders using aquarium pumps in there systems. However, that's not the case because these loops have a bit of restriction to them and those pumps wouldn't really cool anything because they offer virtually no pressure, just high flow rate.

And I have seen people actually use these systems over at XS as they have a use in certain situations.

I would like to argue that in fact fluid flow is more important than head pressure when comparing systems like this. If you took physics in high school you should be familiar with Poiseuille’s Law, which is used to calculate flow for viscous fluids. In which pressure is equated to flow rate.

For the systems tested here, pump head pressure would yield an arbitrary number, since each system (radiator, tubing) would impart different resistances on the fluid. And when buying one of these systems one would not replace the pump/heat sink unit. So taking head pressure values would be useless unless you would be including the entire system in the return loop. And if that is what you meant then it would be the same as measuring system flow.

 

[sempifi99]

Pump flow is very questionable- even Corsair and Antec do not list this on their own sites. Pump flow is one of the most important selling points for every other pump used in watercooling. I've looked for flow stats on these before and have only seen 1-2 articles from people who have broken them down and done some independent testing...nothing from the actual manufacturers themselves. However, most of us full-watercooling guys don't consider these coolers as watercooling...they are simply an out-of-the-box gimmicky solution that basically 'competes' with good air cooling.

The reason it would be good to know pump flow is because you can overclock these pumps. Though I do understand why manufacturers would not want to publish pump flow on prebuilt systems like these.

It goes without saying but I do agree that these do not compete with a completely custom built water cooling systems, and that, that is not their intended market.

 

[triclops41]

As someone who has owned the H50 and currently has the Noctua D14, I cannot understand the gate for the closed loop CPU cooling systems here. Calling it "fake WC" is technically false and needlessly disparaging. Maybe calling it "little water" would make more sense, as these are less expensive and smaller than big boy WC.
But these are typically the best solution for cooling OCed CPU in small form factor cases. You can use them effectively as an intake or exhaust, and are much smaller than big boy air cooling.

Again, just because you can get better cooling, value or acoustics with other coolers doesn't mean that these have no value, or are just for people you consider "newbs". My old system was an i7 875k in a nice micro atx case, and the H50 with a couple of upgraded fans was by far the best solution for cooling my OCed 875k. As an Exhaust, it allowed me to fit a GTX 480 with an aftermarket cooler, all of my hard drives, and giant Blue-Ray burner, etc, with low temps, low noise, and a compact case.

 

[triclops41]

*understand the hate*

 

[rgd-5]

Maybe I'm seeing it wrong here, but it looks like you installed the Corsair H80 with the fans exhausting air out of the back of the case. Doesn't Corsair recommend using the fans to pull outside air in?

 

[clonazepam]

The idea of these closed loop systems is to simplify things with good to great performance. I don't mind that they used the EVGA cooler. The noctua requires consideration what ram you choose, what motherboard you use, whether or not you want to use a side case fan, etc, and that's not exactly simple for many. How many forums are filled with the questions, "Will noctua fit in this system?"... Many of those considerations are eliminated by choosing closed loop systems, and simple design air coolers like the evga product.

I would like to know how many mm of space there was between the evga air cooler and the reference gtx 580 in the one pic. It looks pretty close based on the angle of the camera, so maybe its just an illusion. Jus' curious...

 

[clonazepam]

Even though I've since moved on, I still believe firmly that the H50 is a decent product, but gets a little expensive when you swap out the factory fan for two good ones. Any of the closed loop systems do have that weakness as well, as far as many owners being compelled to add more or replace the existing fans with either something that's just more exotic, or quieter, or has more pressure.

 

[10tacle]

I paid ~$70 for my NH-D14 and am perfectly happy running my i2500k at 4.83GHz never hitting over 65C except in summer ambient heat. My PC goes places, and I just don't have a high confidence of moving it around with a water solution like these.

 

[clonazepam]

[citation][nom]10tacle[/nom]I paid ~$70 for my NH-D14 and am perfectly happy running my i2500k at 4.83GHz never hitting over 65C except in summer ambient heat. My PC goes places, and I just don't have a high confidence of moving it around with a water solution like these.[/citation]

One of the main points of the article is that these closed loop systems put very little stress on the motherboard / socket. The Noctua NH-D14 (I own one) could be considered to be a lot more dangerous when transporting. "A lot more" in relation to the closed loop system, but still pretty safe if the proper precautions are taken.

 

[Crashman]

[citation][nom]clonazepam[/nom]I would like to know how many mm of space there was between the evga air cooler and the reference gtx 580 in the one pic. It looks pretty close based on the angle of the camera, so maybe its just an illusion. Jus' curious...[/citation]Probably around 10mm, you can't even get the fan clip open all the way with the cooler installed. Hard to imagine a bigger cooler in its place!

 

[sempifi99]

Anyone know where the H80 is for sale at $75? I see it listed for that on the article for its price, but can't seem to find it anywhere near there. I have seen it on newegg ~$95 but that is the closest I can find.

 

[Crashman]

[citation][nom]sempifi99[/nom]Anyone know where the H80 is for sale at $75? I see it listed for that on the article for its price, but can't seem to find it anywhere near there. I have seen it on newegg ~$95 but that is the closest I can find.[/citation]I'm pretty sure it was Newegg or Directron, and of course the price took a jump right after the article was published. You should send a nifty letter to Corsair to see if they can make a deal with one of these companies to resume the temporary discount until...Christmas?

 

[aaronstyle]

Great article. Plenty of useful information, however it says the test system is an i7 2600k, but on the page with temps above ambient, it says i7 870.

I think that using a closed loop system is nice because, it's got a great form factor, and looks sharp. To me, I'd rather save the space, at the expense of an arguable couple of degrees versus high end air coolers. I enjoy my h60, it does what it says it does, so I can't complain. Even with a modest overclock on my phenom II, I never see core temps over 43-45C even after gaming for several hours.

 

[ien2222]

[citation][nom]rubix_1011[/nom]I agree that head pressure takes precedence over flow rates; not arguing there. What has been noted by these teardowns is that there is very low flow and minimal head pressure (since the loop is minimal and predetermined by the tubing) there doesn't really need to be a lot of head, here. And since radiators are the least restrictive components, it comes down to flow rates in a simplistic 'loop' like this.People everywhere use these in certain situations, but the WC'ing guys I know from XS and on OCnet really don't have much use for them outside of simplistic installation and use. We have an argument that discusses this in the watercooling forum if you'd like to join.[/citation]

Heh, sorry, I was in a hurry when I read your post and misread it. I thought you were talking about WC in general, that's my mistake. Yes, in a simplistic solution like this flow rate is really the only thing that matters for the reasons you gave as far as the pump goes. Sorry about that.

Oddly, I've seen several builds on XS with these type of systems. But yeah, definitely very specific applications as I've only seen them in SFF's. But even Vapor (hmm..or maybe it was Martinm?) said that they aren't as bad as some people make them out to be and aren't a bad choice in certain applications.

 

[xyriin]

I see a lot of hate on the closed loop liquid coolers stemming from "not real liquid cooling" or "air is just as good". I fully agree with both sentiments, but the real analysis is the pros and cons.

It isn't real liquid cooling, therefore no upkeep and maintenance requirements. You don't get the full performance of true liquid cooling, but you avoid the hassle.

Air is just as good. Really it is, but you have to use a heatsink and fan that takes up 5 times the space or more and increases ambient board temperatures. Air is also more noisy. Let's also not forget that a massive and weight air cooler also places strain on the board itself. If you ever move from 5x120mm +CPU air cooler +high end video card to 5x120mm +high end video card you WILL hear the difference.

 

[eddieroolz]

Thanks Toms for this comparison. I was planning to build a machine sometime with water cooling.

I was initially leaning towards an Antec Kuhler but it seems the Corsair H80 is a much superior choice. Glad I didn't buy the hardware yet.

 

[Guest]

I have the H80 and love it. I leave it on the lowest setting and it just does its thing. Corsair should make the fans that came with it available for purchase as they are very high quality and I would like to have a couple in reserve to swap out for cleaning. This is an A1+ device for me and I highly recommend it.

 

[rantoc]

[citation][nom]ien2222[/nom]Flow the most important? Not really, head pressure is just as important, if not more so. If flow was the only consideration you'd see all the modders using aquarium pumps in there systems. However, that's not the case because these loops have a bit of restriction to them and those pumps wouldn't really cool anything because they offer virtually no pressure, just high flow rate. And I have seen people actually use these systems over at XS as they have a use in certain situations.[/citation]

Both are important, the water loop setup dictates what value is the most important however. If you got a small loop with parts that don't hinder the flow much a high flow pump is more beneficial than a high pressure one. Same goes for big loops with several blocks (like 2xgpu, cpu, mem ect) that hinders the flow then a high pressure pump is recommended - The right pump for the job in other words!

 

[nexus_storm]

I use a Hydro 80 in my system and would recoomend it to all looking for a closed looped water cooling solution I basically have the fan speeds set to lvl 2 can't hear it and the CPU stays damn cool compared to the heavier heatsinked coolers, just wish Corsair would bring out a closed looped system for GPU's, instead of having to build a water system as I have done. Living in a hot environment (Australia) up north cooling is important for me.

 

[barren]

Technical differences aside the weighting to success here seems to favour cost vs (Tom's) measured performance. As a UK based enthusiast using todays exchange rate list pricing from a reputable site the 620 comes in at $62.54 while the H80 is $114.51 (!). Presumably this would indicate the 620 to be the preferred option?