Testing Configuration and Benchmarks Used
System mounted NH-D15S CPU Cooler
System mounted NH-C14S CPU Cooler
To verify that the cooler kit operates up to specs, both the Noctua NH-C15S CPU cooler and the NH-C14S CPU cooler were tested with the CPU running in both stock and at a known stable overclocking speed using both Z87-based and X99-based motherboards, so that comparative results could be provided for both Haswell and Hawell-E based systems. The results are presented for the cooling kits under review as well as the Noctua NH-D15 CPU air cooler, the XSPC Raystorm D5 Photon RX480 water cooling kit, and the Corsair H100i AIO CPU cooler for performance comparison purposes. The benchmark tests used should give you a good understanding of the Noctua CPU coolers capabilities so that you, the reader, can make a more informed purchasing decision. The comparison coolers were selected based on their superior performance capabilities.
| Test System Setup | |
| CPU | Intel Z87-based system Intel Core i7-4770K Stock – 3.5GHz, 35 x 100MHz Base Clock Overclocked – 4.68GHz, 28 x 167MHz Base Clock Intel X99-based system Intel Core i7-5960X Stock – 3.0GHz, 30 x 100MHz Base Clock Max Air Overclock – 4.25GHz, 34 x 125MHz Base Clock |
| Motherboard | MSI Z87 MPOWER ASUS X99-A |
| Memory | Intel Z87-based system Corsair Vengeance Pro 16GB (2 x 8GB) DDR3-1866 modules Stock – 1600MHz, 9-10-9-27-1T Overclocked – 1780MHz, 9-10-9-27-1T Intel X99-based system Corsair Vengeance LPX 32GB (4 x 8GB) DDR4-2666 modules Stock – 2133MHz, 15-17-17-35-2T Overclocked – 2666MHz, 15-17-17-35-2T |
| Hard Drive | Intel 730 240GB SSD Western Digital Caviar Black 1TB SATA III HD |
| Sound Card | On-board sound |
| Video Card | NVIDIA GTX 570 1.25GB |
| CPU Cooling | Noctua NH-D15S CPU Air Cooler Noctua NH-C14S CPU Air Cooler Noctua NH-D15 CPU Air Cooler Corsair Hydro Series™ H100i Extreme Performance CPU Cooler XSPC Raystorm D5 Photon RX480 V3 WaterCooling Kit |
| Video Drivers | NVIDIA 335.23 |
| Power Supply | Corsair HX750 |
| Operating System | Windows 7 Ultimate x64 Windows 10 Pro x64 |
| Thermal Paste | MG Chemicals 860-60G Silicone Heat Transfer Compound |
| Coolant | XSPC Raystorm D5 Photon RX480 V3 kit Koolance LIQ-702 High Performance Liquid Coolant |
The 64-bit Windows 7 based test bench used for Intel Z87 LGA1150 board testing includes an Intel Core i7-4770K CPU, 16GB of DDR3-1866 memory, an NVIDIA GTX 570 1.25GB video card, and an Intel 730 240GB SSD drive. Using the selected components gives us the ability to demonstrate the cooling system's capabilities rather than that of the system components.
The 64-bit Windows 10 based test bench used for Intel X99 LGA2011-v3 board testing includes an Intel Core i7-5960X CPU, 32GB of DDR4-2666 memory, an NVIDIA GTX 570 1.25GB video card, and an Intel 730 240GB SSD drive. Using the selected components gives us the ability to demonstrate the cooling system's capabilities rather than that of the system components.
Benchmark Tests used for evaluation:
- AIDA64 Extreme Edition v4.30.2900 (Intel Z87-based testing)
- AIDA64 Extreme Edition v5.30.3500 (Intel X99-based testing)
- EVGA OC Scanner X v3.6.1.2
Morry,
As this article
Morry,
As this article appears on the Front Page, the “read more” link at the bottom of the snippet isn’t a link to this article, it’s a link to the Front Page. (On this page, the full article, the link is correct.)
Just tryin to be helpful. 🙂
Thanks for catching that,
Thanks for catching that, fixed up now.
Simply a message to
Simply a message to congratulate your effort, intriguing details in this posting..:)
My website … upgrade games
It’s unfortunately that I
It’s unfortunately that I didn’t known about this cooler a few weeks ago. Apparently some AM3+ boards only allow the CPU cooler to mounted in one direction. The CoolerMaster T4 overhangs the RAM slots making it impossible to remove the memory without removing the cooler.
Why didn’t you test the
Why didn’t you test the DH-15S with an additional fan? I have the DH-15 which has two fans and it blows the newer version out of the water other than at idle. It seems moronic for Noctua to design such a massive cooler for a single fan.
Single fan helps with
Single fan helps with compatibility and noise (just look at the noise charts in this review: the D15S and C14S are in a league of their own), but it still doesn’t make sense for it to cost so much; you may as well get the regular D15 and remove one of the fans. At $70, the single fan coolers would be a good option.
Sorry I cannot hear my twin
Sorry I cannot hear my twin fan DH-15 at all, so who cares if the DH-15s is less noisy if it doesn’t cool and it’s clearly an inferior product as tested. Have you asked yourself that the reason it’s so quiet is the reason it sucks as a cooler. No free lunch.
The D15 may be quiet, but it
The D15 may be quiet, but it is not silent. Maybe you can’t hear it because your PC (or room) has other, louder components, but for people who aim for exceptionally quiet PCs, that difference does matter. And saying that the cooling performance “sucks” is a bit of a stretch – this review tested it against a flagship AIO and custom watercooling, both of which used loud, high speed fans. Any single fan air cooler will lose against those unless it uses a 8000rpm Delta fan. It should have been compared to other quiet air coolers, not only to coolers intended for the last word in overclocking.
Still, I agree that as it is, these are not good coolers. Just too expensive, but with a solid price drop, they could be reasonable.
I wonder how the single fan D15 compares to dual fan with LNA. You might be able to get a better temp:noise balance with two slower fans.
I’d like to see it tested
I’d like to see it tested with an extra fan too. This seemed like it would be the perfect air cooler being an offset NH-D15 for PCIe slot compatibility but those Haswell-E numbers are ugly.
I’m sorry, but these tests
I’m sorry, but these tests doens’t make any sense to me. According to other reviewers the D15S beats the C14S with quite a margin.
Since noise and size is more
Since noise and size is more important to me than getting that last couple hundred MHz, the C14S will be on the shortlist for my next build. Thanks for the review!
Wish I knew about this
Wish I knew about this Heatskin before purchasing the Cooler Master Hyper 212 EVO. Having the same Corsair Vengeance Pro Memory with the Cooler Master Hyper 212 EVO makes the first DIMM 1 Slot unusable and even more with Corsair’s Memory Cooler.
PhoneyVirus
http://twitter.com/PhoneyVirus
http://phoneyvirus.wordpress.com/
Check out the CRYORIG H7. It
Check out the CRYORIG H7. It cools better and is quieter than the 212 EVO while also having full RAM clearance for the same price.
I just bought a D15 and I
I just bought a D15 and I have to admit whilst its thermal performance and acoustic performance under load is admirable (PWN/Silent Mode — the deep tonality helps), the fact that its idle/low-power noise is audible in any capacity is somewhat disappointing (and something I’m hoping to address in software).
Keep in mind a low power state is the one I do most of my work in (Word, Spreadsheets, emails) and where I appreciate absolute silence.
No doubt it will be quiet enough for most, but if you are the kind of person paying the premium price for these kinds of coolers then you are also, in all liklihood, a complete silence freak, right down the last decibel.
My machine is pretty much made from every conceiveble angle with silence in mind, so it’s safe for say the D15 is actually the loudest component in the machine in the lower power scenario.
In the high power scenario the graphics card kicks in somewhat but to be honest if I’m in that power state it means I’m playing a game. And if I’m playing a game the audio from the speakers handily drowns out the sound of the machine.
From my perspective, idle acoustic performance is actually more important than acoustic performance under load.
I’m also not shouting in
I’m also not shouting in silence but do want silence when I’m quiet. Nice perspective, like it. 😀
Anyhow, looks like NH-D15S failed on X99 Broadwell-E. Good to know, guessing I’ll go with Corsair H100i GTX when 6900k comes out.
We received new samples to
We received new samples to test with from Noctua on the Hawell-E system and the performance was improves dramatically over the orginal review samples. You may want to give the Haswell-E performance numbers another look…
So what went wrong if both
So what went wrong if both units were faulty should owners of the coolers be looking for something.
Or were they not installed properly or something went wrong in the testing of them.
Cheers
Not sure what the issue with
Not sure what the issue with the units. The old and new units were tested in the same method each time. Could have been something with the shipping of the units.
However, if you do run into similar issues, contact Noctua and they should replace them for you without issue.
(NH-C14S – Single Fan) My
(NH-C14S – Single Fan) My 4790k @ 1.25 volts, 4.6ghz on all cores (override mode for testing) OCCT peaks around 80 C, I mean its not amazing but it works and is pretty quiet. I considered an NHD15 but its massive. Hit 85C @ 4.7ghz @ 1.275 volts. Basically what I am saying is this cooler seems good for around 1.25 volts for nice temps during normal use.