Now that CPU cooler reviews have begun in earnest here at AnandTech, it's been interesting to see just how conventional wisdom plays out in practice. There's been a pervasive attitude that closed loop coolers are only really competitive with the highest end air coolers, and there may be some truth to that, but we have at least one of those flagship coolers on hand today along with parts from SilverStone, be quiet!, and Cooler Master.

Once we got in touch with Noctua and let them know we were doing cooler reviews, they gave us the opportunity to correct what I'd consider to be a sizable omission in terms of coverage in general: no review of the flagship NH-D14 CPU cooler. The NH-D14 is big, beefy, expensive, and typically regarded in enthusiast circles as one of the finest air coolers available. Alongside the NH-D14, Noctua also sent us their NH-L12 and NH-L9i low-profile CPU coolers; while the NH-L9i is potentially underwhelming, the NH-L12 stands to impress as potentially the most powerful downward-flow cooler on the market.

In the interests of making it a full-on roundup, three additional coolers were brought in for review. First is the flagship SilverStone Heligon HE01, a substantial dual tower cooler with a massive 140mm (38mm thick!) fan in the center and rated to cool a staggering 300W. Next up were two coolers I've had in house for a little while that are going to get to see sunlight and scrutiny: the be quiet! Dark Rock Pro 2 (rated for 220W) and the cooler from my case testing bed, the Cooler Master Hyper 212 EVO. The EVO can typically be found for under $40 (and usually much closer to $30) and is regarded as one of the best budget coolers on the market.

  Noctua NH-D14 Noctua NH-L12 Noctua NH-L9i
Dimensions (in mm) 158x126x120 93x128x150 95x95x37
Fans (Supported) 1x 140mm & 1x 120mm (3) 1x 120mm & 1x 90mm (2) 92mm (1)
Weight 1240g 680g 420g
Rated Noise in dB(A) 13.2~19.8 13.1~22.4 14.8~23.6
Price at NewEgg $81 $69 $48

  SilverStone Heligon HE01 be quiet! Dark Rock Pro 2 CM Hyper 212 EVO
Dimensions (in mm) 140x119x160 147x138x166 120x80x159
Fans (Supported) 140mm (3) 1x 120mm & 1x 135mm (2) 120mm (2)
Weight 926g (w/o fan) 1250g 580g
Rated Noise in dB(A) 18~41 13.5~26.4 9~36
Price at NewEgg $75 $99 $33

Before we get started with testing, some notes. First, the NH-D14 that Noctua sent is their Socket 2011 edition, but there's no appreciable difference between that one and the standard version; the mounting brackets from the NH-L12 were used for the NH-D14 and worked like a charm.

The be quiet! Dark Rock Pro 2 is unfortunately, like the rest of be quiet!'s line, still a bit rarefied stateside. That's unfortunate, because this little company has a lot to offer (as you'll see later). Of all the coolers tested, the Dark Rock Pro 2 is the most intimidatingly large, but be quiet!'s products are designed for silence first, so we'll see how it works out.

Finally, having the Cooler Master Hyper 212 EVO in this lineup almost seems unfair; it's smaller than the other coolers, only really benefits from one fan, and is the least expensive by a longshot. Looks can be deceiving, though. I used to run a Hyper 212 Plus and can attest to that cooler being both remarkably inexpensive and efficient, and the EVO's fan is both more powerful and quieter than its predecessor's.

Ease of Installation
Comments Locked


View All Comments

  • stennan - Thursday, March 14, 2013 - link

    Hey Dustin, great review!

    Some friendly feedback: could you show some more pictures of the coolers tested earlier in review/roundup? It wasn't until the last page that the coolers were shown, also pictures of the cooler on the motherboard or in the case would be nice.

    Keep up the work!
  • cbrownx88 - Thursday, March 14, 2013 - link

    I'd buy a Noctua cooler in a heartbeat if they didtn have those god awful colored fans.
  • cmdrdredd - Thursday, March 14, 2013 - link

    You can use any 140mm fans you want
  • mapesdhs - Friday, March 15, 2013 - link

    Thoroughly agree! I bought a Noctua NF-A15 140mm to replace the stock rear fan in a
    Coolermaster HAF 932 case. The Noctua fan performs fantastically well, but the colour
    scheme is horrible. Having said that, Noctua's packaging and presentation is very good.
    Opening the fan pack feels like a special event, it all has a certain luxury sheen which at
    least makes one feel the high price is going to be worth it.

  • K1wi - Thursday, March 14, 2013 - link

    I'd love the NH-L12 for a compact build as I'm a huge Noctua fan (best post-purchase support I've ever encountered), but it looks like it blocks the top PCI-e slot, so its a no-go for me. Its a pity because it looks like it only barely does... I wonder why they didn't choose to offset it to stay out of the PCI-e area and rather more over the top of the board where most ITX boards place their chipset stuff etc.

    That only leaves the NH-L9i, which is more about reducing noise than improving thermals :(
  • seven2nine - Thursday, March 14, 2013 - link

    Ease of installation
    first paragraph "surrounding all the coolers here but but...."
  • Lord 666 - Thursday, March 14, 2013 - link

    After happily winning a z77 Sabertooth on Anandtech, was bummed my existing Zalman would not fit over the armor.

    Without a doubt, will the NH-D14 fit?
  • garadante - Thursday, March 14, 2013 - link

    Honestly, these closed loop liquid cooler results aren't necessarily valid, as fifteen minutes of load testing probably isn't enough to thermally saturate the coolant. As soon as the coolant is thermally saturated, shouldn't the performance be based purely on the heat transfer of the radiators between the cooling medium, intake air, and the heat transfer medium, the coolant? At this point, I imagine a high end air cooler would perform more consistently in the long run.

    Liquid cooling really shines with actual cooling loops, built yourself, with high end components, large radiator surface area, and more coolant to thermally saturate than is present in these closed loop coolers. There's only so much heat a 120mm closed loop cooler can dissipate, once the coolant reaches it's thermal saturation point.

    These closed loop coolers are better for bursts of activity, but I have my doubts with them in extended torture testing. Anyway, I'd rather use a 480mm radiatior, or even a 480 paired with a 240/360, so I can run the fans at nothing/almost nothing overnight, with the heat capacity of the coolant taking care of the cooling, the only sound coming from the pump.
  • ShieTar - Friday, March 15, 2013 - link

    Potentially good point, but lets throw some numbers in to test it:

    Heat Capacity of water is about 4 J/g/K, so if the closed-loop system come with 1/4l of water (I doubt it is more than that), they can take 1 kJ/K. The overclocked 2700K can possibly touch 100W when overclocked and stressed, so it puts out enough power to heat up the water by 1K every 10 seconds, or 90K over 15 minutes if you ignore the cooling effect of its radiator.

    With this amount of power I would guess that this small, closed-loop systems are already relatively saturated and stable. Of course if you have a big system with 4 liters of water and a few Kg of radiator metal, you are looking at a much higher heat capacity of more than 20kJ/K, and if you just use normal loads (~50W) instead of a torture test, you could indeed cool for an hour or two on heat capacity alone. If you add an overclocked 680 or 7970 to the loop though, even this big system should hit saturation within much less than one hour.
  • mapesdhs - Friday, March 15, 2013 - link

    I talked to a German company which specialises in water cooling. They told me that to achieve
    cooling performance "significantly better" than a high-end air cooler would require one to spend
    at least 200 UKP ($300+). Otherwise, unless there are other considerations, a water cooler is
    not going to beat a good air cooler.

    This is why I keep bagging used TRUEs, etc. off eBay when I can, typically get them with
    fans included for around 10 to 17 UKP. Also just bought a Venomous-X for only 25 UKP total.


Log in

Don't have an account? Sign up now