My next cooler will probably be AIO, mostly because the bulkyness I think is annoying.
And a pump, radiator and two stiff tubes are no hassle? A big tower cooler is much easier to mount/dismount than an AIO.
Depending on what your needs will be by then, and whether you want to overclock or not. My go-to recommendation for non-OC is the often overlooked NH-U14S(~$80) which is performing a little lower than the newer NH-U12A(~$130), but is fairly compact and surprisingly capable. It should also be close to the performance of NH-D15S which you list in your specs. My biggest problem with AIOs is the lack of advantages for all the disadvantages, if you need extreme cooling then a custom loop is the way to go, otherwise a good air cooler is going to give the best cooling and noise levels, unless you have a very special case where only an AIO makes sense.
It is worth paying attention to what comes next from Noctua. We've been waiting two(or three?) years for their next generation "NH-D15" now, and currently it's scheduled for Q2, but who knows. But I'm curious to see if it will still be a dual tower, or if they basically make a larger version of the very impressive NH-U12A (and sell it in single and dual fan models). Those improvements with 7-8 heat pipes and their new fan will get very interesting.
One thing I do find interesting, is that Noctua's HEDT/server versions of their lineup, the NH-U14S with one or two fans, is their top product (not the NH-D15), and this is capable of cooling 450W HEDT CPUs, while the normal desktop version is capable of ~250+W. This sheds light on one important aspect of any kind of cooler; modern desktop CPUs have very concentrated heat, and the test results may actually vary depending on testing on a hot plate vs. actual real world CPUs. There is also the endless debate over open bench vs. case, especially since tower coolers generally does well in cases with a decent air flow. With respect to this review and others, many such variances in testing methodology makes may have significant affects on the results, and sometimes it's not necessarily
wrong either. But regardless, I have no doubt that either this reviewed cooler or counterparts from Noctua are more than good enough for any real world usage (non-OC) for LGA1700 and AM5.
And there are probably other worthwhile alternatives too. (Not all alternatives are widely available in my region, so my perspective is incomplete.)
Paying that much for an air cooler is insane, but at the very least the performance backs up the price.
But it would have been OK if it was an AIO with less cooling performance?
I never really got the point of these insanely expensive air coolers. I'd rather buy an AIO if I wanted to spend that much. But I guess there are cases where an AIO won't fit.
But the good air coolers beats most AIOs up to 280mm at least, and usually is cheaper than the AIOs they beat.
There's clearly some issue with these coolers - looking at a quick sample of google results for "Dark Rock Elite Review" there's a
glaring lack of consistency.
- TPU - Editor's choice, best air cooler to date, vastly outperforming the Noctua D15
- Anandtech - Decent, but no better than a Noctua D15
- THG - Decent, ties a Noctua D15 but only tested on Intel LGA1700. Marginally better than the $33 Peerless assassin but worse than the Frost Commander 140 and Assassin IV.
- eTeknix - Good, but found zero difference between the Dark Rock Elite and Dark Rock 5, which is in stark contrast to most other reviews of both and an inconsistency of its own.
- OC3D - Decent, but inferior at both fixed fan speed and max fan speed to both the AK620 and Assassin IV
- Guru3D - Poor, barely creeping ahead of the single-tower 120mm AK500
- Hardware Canucks - disaster, despite restesting and high expectations.
- Hardware Busters - poor, matches an old Assassin III and way behind the Noctua D15.
I'm sure there are more reviews, but those are the ones I found first and skim-read. What's obvious is that performance is all over the place, which is fine if you're gambling at under $50, but this is the single most expensive air cooler ever released.
It may very well be related to testing methodology.
Did you notice any correlation between testing and results? I just skimmed through a few of them and failed to get a good understanding of whether it was open bench or case, and how room tempature is regulated, humidity, etc, and whether other coolers were re-tested under the same environmental conditions or if they reused old results. (all of which may actually lead to a different conclusion)
This must be an honest mistake.
