This big wall in front of you consists out of 11 different low profile cooler models which are this time going to be pinned against each other. So, the reason why I decided to make this video is, well, before all to help you out in the process of picking out what’s the best low to lower profile coolers, I say lower because some of them are above 10 cm in height, so not that low as for example sub 4 centimeter tall Noctua’s L9, while also helping myself out in seeing what will the best choice for my personal ITX build project that I’m planning to start just as I finish off with this one, using the Streacom’s DA2 chassis, so be sure to subscribe for that one. As you can see, the whole premise is pretty simple and straightforward, so let’s just jump over to the matter at hand.
Since I’m doing a comparison test, I’ve decided to make everything as fixed as possible in terms of factors that could end up fluctuating the results, in particular frequency and voltage. The testing was done using two different platforms, or CPU’s to be precise, in the context of small air CPU coolers which I have here the first one is going to be a bit more casual Intel’s Core i5 8600K, and on the opposite end of the spectrum, for those same coolers a bit more demanding Ryzen 7 3700X.
Both of them were tested in a few different scenarios, differing in frequency, voltage, and fan settings. In particular, the 8600K had two setups, one with the fixed voltage at 1.1V and at its stock turbo frequency of 4,3 GHz on all cores, and the other one where it’s overclocked to 5,1 GHz at 1,310V. The Ryzen 7 3700X had three setups, one where it’s set to its base clock of 3,6 GHz and 1V, the other at 4,1 GHz and 1,2V, at last one, 4,25 GHz and 1,320V. The PBO feature and everything else related to frequency boosting were off both on AMD’s and Intel’s platforms, with the core voltage load line calibration set to be as flat as possible, to eliminate any variation in testing.
That’s also one of the reasons why I did my testing on an open testbed, so all of them have a level playing field, which will show you the actual cooling potential of the coolers themselves. More so since we’re talking about lower to low profile air CPU coolers which are most likely going to be used in an ITX chassis or other small enclosure, which are generally coming in numerous different sizes and build solutions, so the cooling performance can differ greatly from model to model, and that is only one factor to count in, while there are plenty of others that also have their own share of how they impact on cooling performance. One of which is the thermal paste, where I’ve dedicated to clean off the pre-applied ones on coolers which came with it, and use the same Thermal Grizzly’s Kryonaut thermal past in an X pattern for the best coverage with all coolers.
Low profile coolers round-up
So, who are our contenders? Starting with the alphabet order, we kick off with the to enthusiast all well familiar Alpenfoehn Black Ridge. Besides flipping its stock fan and creating another testing scenario out of it, because Alpenfoehn for some reason configures it to pull the air from the bottom, my guess that is because it produces less noise as in this alternative setup the blades are really close against the heatsink, while I’ve also used Noctua’s low profile NF-A9x14 and NF-A12x15 fans in few different testing scenarios.
Unfortunately, I didn’t have low enough RAM and motherboard with good socket clearance, I couldn’t test it with two NF-A12x15, although I want to check that out so badly. It is what it is.
Next in-line are BeQuiet’s pretty bulky duo, creating another Shadow Rock TF2 vs Dark Rock TF comparison, Noctua’s tried and true L9, L9x65, NH-D9L, NH-U9S models, and their bit bigger representatives, but still, small chassis approved NH-C14S and NH-L12 finished of with Raijintek Pallas and Scythe’s Shuriken 3.
Some you will probably notice that Cryorig’s C7 is missing. That’s because I just couldn’t get my hands onto one, I didn’t manage to get to them, although I worked with them before, couldn’t either buy it at the time, generally, I feel like they’re not that present as they were before. Anyhow, I did test their C7 Cu and their regular model a long time before, so I know how it performs. It’s alright if you were to cool a lower TDP CPU’s, but the fan was always something that bothers me, it’s was just too loud for my personal liking. With that said I’m pretty sure it would end being placed somewhere in the middle to low range in relation to the rest of the cooler offering that I have here based on the CPU’s, I’ve chosen to test them out on.
Speaking of loud, as you’ll see soon, I’ve used both the motherboard’s default fan speed and full fan speed in my testing, mixed and match it across the board, so you get a better understanding of what’s the cooler actually capable of doing performance-wise at its peak. I’ve done my noise testing on Intel’s platform since It had a larger range of noise samples so to speak, as with AMD’s platform after you would bump up the core voltage and frequency, they would all pretty much be max out by motherboards default fan profile, while with Intel it had a more gradual representation.
I won’t spend too much time talking about the installation process of each cooler, for the majority of users it’s a one-time thing after all, so how long does take do it is not that important in the great scheme of things, although it can definitely make the whole process more convenient, that’s for sure. Here Noctua takes the cake with its SecuFirm2 installation solution, while their L9 series only needs four screws, similar to Black Ridge model, while the rest of them, take an equal amount of time to be installed, some do it in a more practical way with fewer parts, some not. I can say one thing though, I would have loved to have few popular ITX motherboards in my hands at that time to also confirm that they are 100% compatible with these coolers, but that, unfortunately, wasn’t possible, so be sure to double-check by going to the website of coolers manufacturer, they tend to have listings of compatible motherboards for their products. The same goes for the RAM cleaner around the cooler, triple-check that your fits before you go and finish your purchase.
So, that’s a total of 10 different CPU coolers, or 11 since Noctua has an AMD and an Intel version of its L9, and 18 different cooler setups, four of the different testing scenario for each platform, so 144 test runs, and 72 two more for the noise testing. Yeah, that’s a lot, so let’s just dig into it.
Best of the best low profile CPU coolers
Taking a look at the results, stated in hottest core temperature and average core temperature during the half-hour stability test in AIDA64, starting with Intel’s platform, you can see that the bigger examples of the group, the Dark Rock TF and NH-C14S with their bulky design and a lot of heat pipes, have the best CPU temperatures at stock figures, while that is also translated on the overclocking scenario, where they more than successfully sustained CPU’s core temperatures, while the smaller ones like the L9 and L9x65 expectedly didn’t manage to handle that, as well as the Alpenfoehn in its stock fan configuration. In its example, it was also really fascinating to see that just by flipping the fan or chaining fan configuration you’ll receive significant temperature drop, even over 10°C difference in some cases and it actually had respectable results compared to those two bigger cooler examples, which is even more impressive knowing how compact this cooler is. Other more compact coolers like the Pallas 120 managed to handle the overclocking scenario, although it had one of the higher recorded temperatures, while the Shuriken 3 was next to it basically within a margin of error. It was also interesting to see the competition between the two pretty similar brothers, the U9S vs D9L. Another interesting observation was how there wasn’t any bigger difference with Noctua’s L12S when it comes to its different fan configurations testing, and C14S for that matter, but as you’ll soon see this will change with higher TDP CPU. You’ve probably already noticed it by now, but since the height of the cooler is a limiting factor when it comes small enclosures, being it you buying the cooler around the chassis, or chassis around the cooler, I dedicated to also put the height specification for each cooler right next to its name, just so you can know what you’re dealing with.
Crossing over to AMD’s platform and Ryzen 7 3700X, you can see that the results are pretty reflective of what I’ve experienced with Intel’s Core i5 8600k in terms of the performance differences and relations between each cooler. At base settings, they were all pretty much holding onto the 60°C mark, expect the smaller ones and the Black Ridge with its stock fan configuration. What’s interesting to see is that rarely any of the coolers managed to handle the 4,25 GHz frequency overclock, which is understandable everything considered, from voltages to this being an 8 core 16 threaded CPU. Only the Dark Rock TF and NH-C14S managed to handle it and keep it cool enough so the test won’t crash, with a slight performance advantage for BeQuiet, but basically within a margin of error.
The more conservative 4,1 GHz overclock was somewhat of an easy hurdle for most of them, and it was interesting to see how the tables have turned here, where the Pallas 120 performed better the Shuriken 3, while they were constantly going back and forward before, although the Pallas 120 did performer better from start to finish with AMD’s CPU. For your i3’s and lower end Ryzen CPU’s, the L9a and L9x65 will be ideal, but anything above that is a challenge, especially if you want to keep it really cool and quiet. Now we can also more distinctly see how Noctua’s horizontal heat-sink cooler design performs depending on the fan setup.
All in all, a lot of decent contestants, some expected results, and some surprises, we got a pretty clear overview of what, and depending on what can you put in your chassis size-wise, the final choice can be really easy. But, of course, I’m not done yet, let’s now hop over to my personal favorite part of every testing and a pretty crucial part of every ITX build – noise.
Yes, I know that this looks a bit overwhelming right now, but bear with me. I purposely aimed to round up all the noise testing results in one place so you can see how they perform throughout the whole benchmarking range. I’ve highlighted the top 5 quietest performs with green color for each scenario. As you can see here, BeQuiet is pretty much leading the pack in every regard, from idle to full fan speed. Noctua is keeping up with the pace, right behind them, until we get to the top end, where their fans end up being louder, but then again, you have to consider that you won’t be using the fans at their top fan speed, but more somewhere in between that, where they’re still overall among quietest in comparison to others. The Black Ridge’s stock 120 mm fan was the loudest one in its stock setup, while the Big Shuriken 3 was pretty quiet at the top end. Of course, this was an as-is situation, so there’s always room for improvement in terms of the performance to noise ratio, where, depending on your build, chassis and fan setup, you can do some manual adjustment of the fan curve speed in motherboards BIOS.
I’ve recorded few sound bites just so you can get a better grasp of the difference between quieter and louder noise levels which were shown in the results.
Bigger is better?
With all of the testing done, you are probably most interested in the top picks when it comes to value for your money. To establish a baseline, I’ve used the lowest price I could find on these coolers at the time of doing this video on Geizhals.eu, one of the OG price search comparison websites, since in the US most of the products were out of stock. Taking that and performance into consideration, I would say that the Shadow Rock TF 2 represent a really good value, as well as the NH-L12S, especially since most of the coolers here are priced around 50 to 60$ range, expect the big two, and of course the Shuriken 3 and Pallas 120 which are really affordable, and definitely choices to consider for lower TDP CPU’s, but that’s not a surprise since both brands follow that value for your money philosophy. You can also cross-reference the price with the noise performance of each cooler, and see what would be the best value pick that’s also among the quietest ones. Generally speaking, It’s a bit hard to draw any definitive conclusion as all of this comes with the premise of not having any height restriction because if it was otherwise your options would be limited. Especially if you have a budget for something more expensive, but you’re aiming for a more compact chassis, you probably won’t be able to get both the quietest and best-performing air CPU cooler out of this line-up, which leaves a bit of salty taste knowing you could have it, but you just can’t fit it in. As always, doing an ITX build often ends in being a game of finding a good balance between pretty much every aspect of that whole process.
Well, this was exhausting, but in my opinion well worthy of the time, hope this will at least bring some level of clarity when you make your own purchasing decision. As for me, with everything shown here, since the Streacom DA2 supports up to 145 mm tall CPU coolers, I think I’ll go for either the C14S or the BeQuiet’s Dark Rock TF2, because I need as much as performance possible. Be sure to subscribe for that build, so you find out which one I actually end up using.
That’s it for this time, thanks for checking out my best low profile CPU cooler comparison, if you have any question feel free to hit me in the comments section of my YouTube video listed above, you can contact me via my social media channels!