Overclock.net banner

High End Air Cooling vs Closed Loop Water Cooling

79K views 164 replies 53 participants last post by  doyll 
#1 ·
I finally got tired of wondering which is better so I just ordered a Corsair H100 and 2 Scythe UK 3000s. Watch this space for comparisons between it, Phanteks PH-TC14PE and Silver Arrow SB-E.
______________________________________________________________________________________
System specs:
Intel 3770k @ 4.6GHz (load voltage 1.202v)
Asrock Fatal1ty z77 Pro
Gigabyte Radeon HD 6950 2GB (unlocked)
G.skill Ripjaws Series 4GB (2 x 2GB) @ 1866MHz
Crucial SSD CT128M4SSD2 128GB M4
Microsoft Windows 7 Ultimate x64 SP1
Corsair 800DW case
Antec Formula 7 Thermal Paste (I applied a small amount to the centre of the CPU)
I used Aida64 for 30 minutes to stress the CPU and monitor temps

Computer was left idle for 15 minutes prior to each test. There is a temp sensor on the case which records the room temp. It also has a probe which is inside the case (next to the motherboard) The fan controller also has 4 temp sensors placed at various points around the inside of the case. Ambient temperature and temperature inside the case were taken at the start, after 15 minutes and again at completion. The average temperature was recorded.
Delta temperature is the average core temperature at load minus the average temperature inside the case

Stock
393

UK-3000s
393
 
See less See more
2
#2 ·
Nice I am curious to know myself. I had a zalman cnpsx10 and it would hands down beat my h60 at idle and full load but noise was the reason i downgraded to a h60. I am one of the lucky few who has completely silent hydro series cooler. I also went with a closed loop because a monster air cooler looks terrible in my case hiding all my nice components.
biggrin.gif
 
#3 ·
I agree that the closed loop set up looks a lot better. I can't wait to test it out. It'll look great in my 800DW
 
#5 ·
Subb'd
 
#7 ·
I tell you what i had a H60 cooler and my Cooler Master v8 does much better job then that H60 some how the H60 would not seat right on Cpu was getting like high 60-70 Celsius for temps now down to high 20's to very low's 30 ' Celsius so very Happy with Air cool could not understand this WATER COOL ANY WAY in A case with the Heat generated from Motherboard,Hard drive, graphic card and Cpu i figure it's not enough room to cool off in that small space this is my tought any way i swear by my cooler Master .Ray

this is My overclock 410
 
#9 ·
There are any number of HSF vs. CLC tests showing that highend HSFs are superior, cheaper, quieter compared to CLCs and HSFs don't leak water and damage PC hardware like CLCs can.

If you have room for a highend HSF, on technical merit that is the best choice for most PC users. I suggest people educate themselves on the Pros and Cons of each type of cooling and then Buy What Makes YOU Happy!

http://www.anandtech.com/print/5054

http://www.tomshardware.com/reviews/h50-fort120-cogage,2401.html

http://www.frostytech.com/articleview.cfm?articleid=2665&page=7

http://www.frostytech.com/articleview.cfm?articleid=2665&page=6

As an example testing has shown the Xigmatek Aegir SD128264 is extremely quiet (only requires one fan but you can install a second one if desired) and capable up to 345w so there is no technical merit in a CLC for most people but people will try to sway you in one direction or the other. The Mugen 3 should also be fine with a second fan if you're running real high OC's and CPU voltage.

http://www.thelab.gr/heatsinks-coolers-watercooling-reviews/cpu-cooler-review-database-89014.html

Contrary to the FUD some people like to post, no one is having issues with the weight of the highend HSFs and mobos. If you prefer a CLC then that's what you should buy. Attempts at twisting the truth or posting FUD doesn't change the test results. Buy What Makes you Happy but don't post misleading info. as this is a disservice to other forum members.
 
#11 ·
Quote:
Originally Posted by sandy cheeks View Post

Nice I am curious to know myself. I had a zalman cnpsx10 and it would hands down beat my h60 at idle and full load but noise was the reason i downgraded to a h60. I am one of the lucky few who has completely silent hydro series cooler. I also went with a closed loop because a monster air cooler looks terrible in my case hiding all my nice components.
biggrin.gif
I hear ya..... while my Mugen-II was quiet it couldn't keep up with my overclocking, even with 2 fans. I like that the blocks are good-looking and functional pieces of hardware that does indeed let you show off your ram or motherboard heatsink's and components. Why get them if you're going to cover them up?

High end air coolers are not cheap either. I'm using a $140 loop that outperforms my Mugen-II and my Megahalem. Besides, I can't imagine hanging a couple pounds of copper and aluminum off my motherboard and listening to it creak from the pressure!
biggrin.gif
 
#12 ·
Quote:
Originally Posted by hollywood406 View Post

I hear ya..... while my Mugen-II was quiet it couldn't keep up with my overclocking, even with 2 fans. I like that the blocks are good-looking and functional pieces of hardware that does indeed let you show off your ram or motherboard heatsink's and components. Why get them if you're going to cover them up?
High end air coolers are not cheap either. I'm using a $140 loop that outperforms my Mugen-II and my Megahalem. Besides, I can't imagine hanging a couple pounds of copper and aluminum off my motherboard and listening to it creak from the pressure!
biggrin.gif
While cracking the motherboard is not impossible, I'd say the risk of the closed loop leaking all over your system is much higher than the risk of a normal large heatsink like a Megahalems/D14 etc. etc. cracking your board.

The only way I can see a heatsink cracking your board is if you try to ship it with the heavy cooler installed. That or maybe the user bumps it while working inside the case.

A PCB is pretty strong. I watched a video on electronics being made not too long ago and the robots installing the components are just slamming them into the board and making it flex much more than my megahalems ever made my pcb flex. Although I also watched a video of the Gigabyte factory and the way they make their boards looks very gentle.
 
#13 ·
Quote:
Originally Posted by tw33k View Post

I finally got tired of wondering which is better so I just ordered a Corsair H100 and 2 Scythe UK 3000s. Watch this space for comparisons between it, Phanteks PH-TC14PE and Silver Arrow SB-E.
What are the Scythe's for? The H100 is loud enough already.

Anyway, people have already done this and they never describe/disclose their methods, so make sure you setup properly: keep the ambient under control (or at least log it), use a proper open-air test bench (no case), apply the same amount of the same paste the same way each time, use a fresh Windows install, run a variety of clocks and voltages to show scaling, give a very generous cooldown time while powered off. There is probably more to account for.
 
#17 ·
The high end air coolers are capable of nearly the same cooling potential as an H100 but, at least in my experience, require a huge amount of air displacement (high, loud fan speeds) to equal what the H100 can do at dead silent operation.

I'm doing push pull on my H100 and the fans are not audible to my ear in the slightest. In fact, my entire machine is dead silent. Still, the H100 keeps a 2700K @ 5ghz 24/7 plenty cool (real world loads, video rendering and what not, sometimes break 70C in a 24C ambient room.

My experience might not be typical, but I've found air coolers need a great deal more air flow through them to equal the same or similar performance I'm seeing out of the H100 silent.

I've also modified the H100 quite a bit (high pressure mount mod, cut holes in my case, lapped, etc.) but it was still pretty respectable at stock.

But above all, it's the cooling coupled with the insanely quiet operating volumes that made me stick with the H100 for now (will invest in a custom loop eventually, however).

You just better hope you get an H100 pump that doesn't knock and grind and rattle, though. LOL ... cause that was WAY worse than any fan noise I've ever heard. Sheez. Talk about a nightmare. If you get one of those you can either play the RMA game (and probably get another noisy one) or just run the pump on a fan controller and drop the RPM's by about 2/3% ... that shut mine up completely and didn't dent he temps in the least.
 
#19 ·
Highend HSFs do not typically require a high volume of airflow compared to CLCs because they are much more thermally efficient than transporting the CPU heat to water and then to a radiator where AIR is used to cool the radiator/water. CLCs are known for being noisy because of the fan static pressure required to actually push enough air thru the radiator to cool the water properly. Thus the CLC fans must typically run at high speed. This creates more noise than a typical highend HSF system. You can buy better fans for the CLCs but why bother when there is no technical merit to the LCLs over a highend HSF that is quiet by comparison.

Anyone who claims an H100 is "insanely quiet" either has their PC shut-off or needs a hearing aid as all reviewers note how loud the fans are on the H100 in particular, which many consider simply unacceptable.
wink.gif


BTW, ALL of these issues have been discussed countless times in this and other forums and nothing has changed the physics or test data results. Every 2-3 weeks we get a new thread with all the old beliefs...
 
#20 ·
Quote:
Originally Posted by PontiacGTX View Post

you could say 25 to 35% +/- in prices but NH D14 doesnt costs 55$ in mot of store is about 80s
Actually the Xigmatek Aegir is only $48 + shipping @ Newegg and it's every bit as good as any of the other top HSFs with just one fan and can equal or beat the H100 with two fans. With two fans the Aegir is at the top of the heap along with the Mugen 3 up to at least 345w. Prices obviously can vary globally but for this discussion the Aegir is basically 1/2 the price of an H100 (with rebate) and the Aegir can't leak water - which is a VERY good thing.
thumb.gif


http://www.newegg.com/Product/Product.aspx?Item=N82E16835181017&Tpk=Corsair%20H100

http://www.newegg.com/Product/Product.aspx?Item=N82E16835233087

http://www.thelab.gr/heatsinks-coolers-watercooling-reviews/cpu-cooler-review-database-89014.html
 
#23 ·
#24 ·
#25 ·
Just saying that you can easily cram in more fins and use extreme fans, but to find the balance is what matters, not to mention the payoff--the performance increase. There is no CLC that gets you more MHz for more FPS, or more life out of your CPU than $50-80 air cooler, but the CLC makers just use NOISE to try and do it.

And don't bring up RAM clearance or cramped case problems from "huge" coolers, or whatever.. no one needs giant RAMS sticking off their board and the H100 is the hardest cooler to fit anywhere. HELL, YOU NEED A $150 FULL TOWER CASE JUST TO MOUNT ONE!

lachen.gif


So let's see tw33k settle this argument once and for all.. with temp differences.
rolleyes.gif


(I'm joking, there is no real benefit to having ~67C load vs ~70C load, which is all we could possibly see).
 
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top