What to expect from Noctua’s new top-end CPU coolers (and fans)?

Noctua NH-D15 G2 and Noctua NF-A14x25r G2 PWM

Well, after all! The NH-D15 G2 cooler will no longer be on Noctua’s roadmaps. Neither will the separate pack of 140mm NF-A14x25r G2 PWM fans it uses. Both Noctua’s new flagship cooler and the first fans from the new generation are on their way to stores. This comes after years of development and detailed tuning. We can now take a look at the key elements involved with these cooling components.

The Noctua NH-D15 G2 cooler…

The popular Noctua NH-D15 cooler has its second generation after ten years. This cooler carries the same designation (NH-D15) with the additional designation G2 (NH-D15 G2). There are actually more of those additional designations as there are more variants of these coolers, but more on that in a moment. Before that, some “basic characteristics” are important.

So, this is a cooler with two towers, where each is cooled by a 140-millimeter fan, two in total. Compared to its predecessor (NH-D15 G1), the NH-D15 G2’s heatsink is more robust, has two more heatpipes (eight in total) and, in addition to the higher weight (1180g without fans, 1525g with fans), it also has denser finning. Instead of 1.9 mm (NH-D15 G1), the NH-D15 G2’s fin spacing is “only” 1.6 mm.

The NH-D15 G2’s heatsink is diverted away from the PCI Express slots. This used to be the case in the previous generation as well, but only with the NH-D15S cooler, which is a single-fan variant. Now, better compatibility on motherboards typically with graphics cards will apply to all variants of the NH-D15 G2 cooler. The cooler is 168mm tall, 150mm wide and 127mm deep, or 152mm with the fans mounted.

The cooler comes in three variants, each with a different designation (without end marking, with the end marking “LBC” and in the case of the third variant the end marking is “HBC”). The only difference is in the base modification, the area of which varies with respect to the needs of a particular platform. In the table below Noctua indicates how each variant will work on which platform. It is all about the quality of contact and the potential to achieve the highest cooling performance.

Given that on each platform’s processors, the critical temperatures are in different places, Noctua responds with coolers with different bases that react to this. They do this by better contact/higher pressure in the area with the highest temperatures. See the Computex 2024 article for imprints with mounting pressure distributions.

The cooler comes in three variants, where each has a different designation (without an end marking, with the end marking “LBC” and in the case of the third option, the end marking is “HBC”). The only difference is in the base modification, which varies in area/convexity with respect to the needs of the specific platform. In the table below Noctua indicates how each variant will work on which platform. It is all about the quality of contact and the potential to achieve the highest cooling performance.

The “standard” base is like the one Nocua has used on its coolers until now, LBC (optimized primarily for AMD platforms) stands for Low Base Convexity and HBC (suitable for Intel LGA 1700) stands for High Base Convexity. Noctua responds to different situations with different bases. This is with a view to achieving better contact/higher pressure with respect to the best possible heat transfer from the CPU to the cooler. For imprints with mounting pressure distributions, see our Computex 2024 article.

However, this does not mean that you cannot install any of the variants on any supported platform (AMD AM5, AM4 and Intel LGA1851, LGA1700, LGA1200, LGA115x). You could, but optimum cooling performance will not be achieved in any given combination. Noctua ranks this within its system (NSPR) at 228. The second highest NSPR (183) is held by the first-generation NH-D15. NSPR/TDP naturally relates to maximum performance with the highest achievable fan speeds. These are up to 1500 rpm for the NF-A14x25e G2 PWM fans used.

One of the new psychoacoustic optimizations to achieve the highest possible cooling performance at the lowest possible noise is the speed offset of the second fan, which is slightly slower. For a detailed physical explanation with practical examples, you can go to Noctua’s topical website. If anything is unclear, there is always room for discussion (below the article) and for interpreting things “in different terms”.

The manufacturer’s suggested price for all variants of Noctua NH-D15 G2 coolers is 150 EUR. The coolers are available to order globally on Amazon, for example.

… and Noctua NF-A14x25r G2 PWM fans

Fans from the NH-D15 G2 cooler are also immediately hitting the shelves separately. However, it should be noted that this is a variant with a mounting hole spacing of 105 × 105 mm, which is atypical for 140 mm fans (and standard for 120 mm fans). It doesn’t matter on the NH-D15 G2 cooler, its clips are well suited for such spacing, but using these fans in other situations may not be possible.

You won’t be able to install them on liquid cooler radiators, and similar complications could arise inside cases. Even if you do mount the fans, the impeller or frame might be too big for the 120mm position. In 140mm positions, you will only be able to install the square-framed variants of these fans later on, which will not have the letter “r” in the designation (NF-A14x25 G2 PWM). Their planned release for September this year was recently confirmed in our interview with a Noctua representative.

The most noticeable change of the new 140mm Noctua fans is in the completely different geometry with humped blades (we already discussed why this modification is good in an earlier article) or the Centrifugal Turbulator (we also wrote something about this…).

The Noctua NF-A14x25 G2 PWM fans will be available individually and also in a two-pack with the additional designation Sx2-PP (NF-A14x25r G2 PWM Sx2-PP). One fan (PPA) is 50 rpm slower than the other (PPB). Noctua’s second-generation 140mm fans also use a new PWM controller – the NE-FD6.

The suggested price for the new NF-A14x25 G2 fans is 40 EUR or 80 EUR for the two-pack. They will be available soon, again for example on Amazon.

English translation and edit by Jozef Dudáš


  •  
  •  
  •  
Flattr this!

Scythe Mugen 6 (Black Ed.): Does a second fan finally make sense?

I give Scythe coolers less attention here than they really deserve. Today I’d like to remedy that, so I’m bringing you a double dose of Mugen 6 and Mugen 6 Black Edition cooler tests. The only difference between these coolers other than the color scheme is the presence of a second fan in the case of the black version. Given their design, I have high expectations for efficiency and hopefully I will be satisfied with the acoustic performance as well. Read more “Scythe Mugen 6 (Black Ed.): Does a second fan finally make sense?” »

  •  
  •  
  •  

PCCooler CPS RZ820: Against the elite, weighing over 1.8 kg

They were so emphatic about their cooler being optimized for Intel LGA 1700 processors that we had to try it out. And given that the CPS RZ820 is a “hefty chunk” of material, the reference sample selection included the kind of competitive coolers that are at the top of the tower model database. These should give the PCCooler CPS RZ820 a good run for its money. At the same time, this cooler should not be more expensive. Read more “PCCooler CPS RZ820: Against the elite, weighing over 1.8 kg” »

  •  
  •  
  •  

DeepCool Assassin 4S – The stealthy tamer of high temps

A year ago, the excellent DeepCool Assassin IV cooler was tested here and eventually took our top award. Today we take a closer look at a modified version of this cooler called Assassin 4S. This variant may have lost the rear fan, but the middle one has been completely redesigned to compensate for this loss. So today’s article logically aims to find out how much of an impact this major change has on efficiency and acoustics. Read more “DeepCool Assassin 4S – The stealthy tamer of high temps” »

  •  
  •  
  •  

Comments (8) Add comment

  1. PPA is actually 50 RPM slower than PPB; one is 1475, the other 1525.

    Seeing how they’re talking about power per torque in their fan motor article, I wonder if max motor torque is something measurable (in addition to motor power).

    1. Very nice, illustrative video. I’ve been meaning to do a similar test for a long time, albeit demonstrated in a slightly different way. So far, unfortunately, there was no time for it.

  2. There’s some seriously impressive claims from Noctua on the A14 G2 (https://noctua.at/en/nf-a14x25-g2-performance-improvements), in particular the sound pressure level vs. airflow on 38 mm radiator chart.

    In this environment, the A14 G1 shows a maximum airflow of ~72 m3h, which closely matches your test data of 70.82 m3h on the 45 mm radiator. Thus, we can reasonably assume both radiators interacts with the fans in a similar fashion.

    If their chart is accurate, it would mean that @31 dBA, G2 will have a massive 50 m3h airflow (vs 23.61 of the G1 and 36.18 of the Toughfan 14 Pro), and @42 dBA, 92 m3h (vs G1=50.29 and TF14P=63.89), and it would have a maximum airflow of 95 m3h. Its maximum noise level on this radiator (~26 dBA) also barely increases from their measured value in free flow (24.8 dBA).

    To be honest, it sounds a bit too extreme to be true, but we’ll see🙂 Perhaps its acoustics on radiators is really that good, as suggested by the fact that when noise-normalized by Noctua, the G2 runs at 1500RPM but the G1 runs at 1200 in this environment.

    1. Radiators of liquid coolers do not really increase the noise level of fans, regardless of their orientation in relation to the fins. In some cases, even with the same fan speed, the noise level even decreases. This is because sometimes the radiator has the effect of mitigating turbulent airflow, which is noisier than laminar airflow. We will definitely write a detailed article on this one day, just have to find the time… 🙂

      1. Right, it’s a bit strange that Noctua’s data show A14 G1’s noise level increasing significantly on the radiator, but your tests found little to no differences. Maybe it’s down to the exact type of radiator used (for example, how close the fins are to the fan).

        1. Of course, many different measurements can contribute to the results of noise measurements. I don’t know how exactly Noctua measures the noise in these cases (if this is publicly known from well available materials, I apologize for my ignorance). I don’t think it’s exactly according to the standard from a distance of one meter. I assume that their goal will also be higher resolution gained by shorter distance. And yes, as you write, another thing is that each radiator causes different behavior.

          The noise level can also be significantly increased by details to which you may seemingly not attach much importance. I mean, for example, the adjustment of the edges of the radiator channels between the fins and so on. As you well know, different noise levels are achieved when installing a push fan, and another when installing a pull fan… no matter how precise and generally valid the results you strive for, it is practically impossible to achieve them, because the variability of the individual components that exist is really large. And the variability of measurement techniques is also wide (and in the end it’s about who follows more or less such a procedure in order to be able to draw the most relevant conclusion). If some tester claims to have the only correct results, that’s (quite) a bit ridiculous, because they don’t realise (or pretend otherwise, which is even worse…) a lot of things. 🙂

Leave a Reply

Your email address will not be published. Required fields are marked *