Four New Closed-Loop Liquid Coolers Versus Noctua's NH-D14

What Does It Take For A Closed-Loop Cooler To Beat Big Air?

Historically, price was the biggest competitive problem for closed-loop liquid coolers. But several manufacturers recently dropped the prices of previous-generation products. The baseline NH-D14 is also expensive by air cooling standards, and is now matched by the reduced-price H90, X40, and 240M.

Price reductions are how the X40 and 240M beat Noctua's NH-D14 in a cooling/cost comparison. The H90 can beat the NH-D14’s value, but only when it’s mounted backwards. NZXT's Kraken X40 remains the real leader here.

However, overall performance compares cooling to noise. Using the numbers from that chart to map performance to price, we again find the NH-D14 on top. Thermaltake’s older Water2.0 Extreme takes second place when its fans are slowed through its automatic controller, and Zalman’s old LQ320 matches Corsair’s even older H90 with its fans at full-force. Both the LQ320 and H90 can be slowed through motherboard control, and either of these probably could have won the value debate if we arbitrarily set the motherboard’s fan slope to something approximating ideal cooling-to-noise conditions.

And the winner is? Well, we’d recognize the NH-D14 again if not for several factors, including the fact that this is a liquid-cooling shootout. A primary reason to choose one of these closed-loop solutions is that they place much less stress on a motherboard compared to big heat sinks. We’ve already destroyed a few thin platforms using massive chunks of metal bolted to processor interfaces, and we once had a board destroyed in shipping by its heavy cooler. It’s consequently difficult to recommend a socket-mounted cooler to anyone who moves their machine around, making it far easier to pick from the closed-loop contenders listed above.

Extra cooling capacity makes Thermaltake’s Water2.0 Extreme appear the better choice for large enclosures. Great value makes Corsair’s H90 and Zalman’s LQ320 appear excellent choices when the build is limited to a single fan mount.

Thomas Soderstrom
Thomas Soderstrom is a Senior Staff Editor at Tom's Hardware US. He tests and reviews cases, cooling, memory and motherboards.
  • squirrelboy
    length, width and thickness in inches, fan size in mm, weight in ounces. can we just have mm everywhere, and preferably kg as well?
    Reply
  • razor512
    Seems they reduced their old BS claim of handling 400 watts of heat on the Reserator 3 MAX, it is now 350 watts, but even that seems unrealistic.

    Tomshardware should place some of these coolers on a resistive load of 300-400 watts and see if these coolers can actually handle the heat, or will the fluid boil and build up enough pressure to pop.
    Reply
  • xiinc37
    Isn't the whole point of watercooling to move the cooler far away from the cpu, so that there is more room to utilize a significantly larger radiator? The thermalright and zalman options look smaller than the noctua...
    Reply
  • rmpumper
    Grabbed a Dark Rock Pro 2 for my new setup. No regrets - looks a billion times better than D14, is less noisy, performs +-1C the same.
    Reply
  • ingtar33
    got an old corsair h100 for $50, which was less then the old heatsink i had on my phII x4 965 used to go for (thermalright ultra 120), Gained a solid 8C improvement in temps, lost some of the noise advantages (the ultra had two noctua's on it)... setting it to medium gave me a 5C improvement in temps, and about the same noise levels. overall i'm pretty happy with it. it looks cleaner, that's for sure.
    Reply
  • bigcyco1
    Thanks for the review.I always have said if someone is going to watercool they should go custom.IMO those all in one water cooling kits just are not worth it. They are no better than high-end air cooling like a Noctua D14, Thermalright Silver Arrow,Phanteks PH-TC14PE elect.If you want to go water cooling do a custom loop if not just go with top of the line air cooling.Better bang per buck!
    Reply
  • itzsnypah
    I wish there was some way to measure the DeltaT of the Ethylene Glycol used in these CLC's because truthfully I wouldn't be surprised to see >25c when using a overclocked 3960X.
    Reply
  • ChromeTusk
    I wonder how much the rear exhaust fan would have affected the Reserstor3’s results.
    In any case, I will wait for price drops or rebates.
    Reply
  • djorgji
    Is it so difficult to sort the results from best to worst? Lowest to highest etc?

    Like this it is impossible to read.
    Reply
  • nilfisktun
    Well, those stock fans all blows, in my optic. You might call it cheating, but try slamming in two noctua pwm fans in a H100i for instance. I run this at home, and those fans typically runs at 600-800 rpm, therefore being totally silent. My stock temps rest around 38C, with 22C ambient, and hit around 60C at full prime load. i7 2600K @ 4.2 ghz. 1.28v
    Reply