top of page
Writer's pictureNick

"Top 6 Low-Profile CPU Air Coolers for Ryzen 7800X3D: A Comparison"


6 best cpu coolers for ryzen 7800X3D

Those who want to accommodate the currently best gaming CPU, a Ryzen 7800X3D, and the currently best gaming GPU, an RTX 4090, in their FormD T1, Fractal Terra, or A4 H2O, must pay attention to the compatibility of the CPU cooler. This is especially true if you want to rely on air cooling, be it for reliability, easier maintenance, or aesthetic reasons. Let’s take a look together at the most popular coolers in the 47mm segment.


I have to relativise this right away, there are two outliers in terms of height. Unfortunately, Noctua currently does not manufacture a cooler with exactly 47mm height. The Noctua NH-L9a is 10mm smaller, but still so interesting that I did not want to exclude it from this test.

The second outlier is the Thermalright AXP90-X53, which measures 53mm. I was particularly interested in how much better its cooling performance really is compared to the 6mm smaller models.


There are few comparisons online, that focus exclusively on 47mm models and what is close to them. Many comparisons bring in coolers with 70mm or more and leave out other popular options from the 47mm range. But I am mainly interested in those small form factor air coolers, which I can install in the following cases to get the best performance.

Case and Configuration

Maximum Cooler Height

FormD / Ncase T1 3,25-Slot Mode

48mm

FormD / Ncase T1 3-Slot Mode

53mm

Lian Li Dan A4 H2O

55mm

Fractal Terra Position 6

53mm

Fractal Terra Position 7

48mm


Test System and Test Methodology

Case

Ncase T1 2.5 (3-slot mode)

CPU

Ryzen 7800X3D @ 76,7W | CO -20

Case Fans

2 x Phanteks T30 @ 30% (600 rpm)

Graphics Card

Nvidia RTX 4070 Super FE

Motherboard

AsRock PG B650I Lightning WiFi

Power Supply Unit

Corsair SF 750

For testing, I use my own PC, which is housed in an Ncase T1. I deliberately refrained from using an open test bench setup because I specifically want to answer the question of which cooler works best and quietest in a sandwich case with limited space. The multiple conversions between coolers were quite annoying, especially since I tested all coolers at least twice, some even three times, each time with a new application of thermal paste.


For the synthetic benchmark I chose Cinebench R23 with a 30-minute runtime per test. I considered the following three scenarios:

“32 dB(A) Noise-Normalised” to find out what the coolers are capable of delivering with a really tolerable operating noise. Then “38 dB(A) Noise-Normalised”, which corresponds pretty much exactly to the noise level of my GPU and is still bearable. And last but not least “100% fan speed” to understand what the coolers can deliver when they are given free rein.

After each test, the otherwise 30% fixed Phanteks T30 fans ran for a few minutes at 100%, until the CPU had settled back at the initially measured idle temperature, before I started the next run.


For the games, I kept it simple and chose three titles. With Cities Skylines 2 (3440x1440) a somewhat CPU-intensive game, Cyberpunk 2077 (3440x1440) stands for rather GPU-intensive titles and Dark Souls 3 (2560x1440) I chose because I wanted to know where the temperatures land in an older game where the GPU doesn’t have to work much. For the temperature comparison in games, I set the CPU fans to the “38 dB(A) Noise-Normalised” profile before each run, because this, as mentioned, fits well with the noise level of my RTX 4070S FE.


To make the temperatures comparable, I used two different thermometers and measured the room temperature before each test. I subtracted the respective value from the result at the end and then added the result back with 23.5°C to be able to represent all values based on the same room temperature in a diagram.



Noise Normalised Testing explained


I’ve been mentioning Noise-Normalised tests for a while now, but what does that actually mean?

In this type of test, all fans are set to the same measured loudness. This is to prevent fan A from screaming like a jet turbine and thus achieving the same temperature as, for example, fan B on another cooler, which is virtually inaudibly quiet, but also gets that same temperature. Personally, I prefer it a bit quieter and don’t want to hear a screeching fan through my headset, while playing a game, so this type of test is always very important to me. Maybe you feel the same way.


It’s also important to mention that my results, e.g. at 38 dB(A), are not comparable to other test results that you can find from other sources. The 38 dB(A) are measured in my specific setup. So someone would have to use the same measuring device, the same case, the same table, the same positioning, the same reflection surfaces, the same room, etc. (you get the point), to get comparable results. Therefore, please note that 38 dB(A) at your setup might be completely different than 38 dB(A) at mine. For the sake of completeness, I still want to mention that I measured from a distance of 50cm from the front right, so quite close to where I sit at my desk.


To get really accurate measurements, I wrote down the fan speeds corresponding to the loudness before each test and also checked again and again that the ratio of loudness and fan revolutions had not changed in the meantime, despite numerous restarts of the PC for the cooler exchanges.



The Gladiators Enter The Arena



Before we get into the details of the coolers, here is a quick run-through of the competitors.

Cooler Name

Height

Material

Price

Noctua NH-L9a-AM5 chromax.black

37mm

Aluminium

ca. 60€ / £50 / $55

Alpenfoehn Panorama 2

47mm

Aluminium

ca. 45€ / £39 / $118

Alpenfoehn Black Ridge

47mm

Aluminium

ca. 48€ / £44 / $103

Thermalright AXP90-X47 Black

47mm

Aluminium

ca. 24€ / £20 / $22

Thermalright AXP90-X47 Full

47mm

Copper

ca. 38€ / £41 / $34

Thermalright AXP90-X53

53mm

Aluminium

ca. 29€ / £26 / $22


Noctua NH-L9a-AM5 chromax.black



First up is the Noctua NH-L9a-AM5 chromax.black. This cooler exists in various versions, with the ‘a’ in the name of our model standing for the variant compatible with AMD processors, and the addition of ‘AM5’ means that the mounting kit for the AM5 socket is included. The corresponding version for Intel processors carries a small ‘i’ in the name, but is otherwise almost identical in construction. With “chromax.black”, Noctua designates their black-coated heat sinks with black fans and charges a premium for this compared to the coffee-brown-beige fan variant with an aluminum-colored heat sink. The technical data are as follows.

Height including Fan

37mm

Weight incl. Fan according to Spec

465g

Weight incl. Fan (own scale)

470g

Fan Radius

92mm

Fan model

Noctua NF-A9x14 chromax.black

Fan Maximum RPM (Spec)

2500 (±10%)

Fan Maximum RPM actual

2500

Fan Noise @ Maximum RPM

40,5 dB(A)

Price

Link to manufacturer

The Noctua, with a height of just 37mm, is the smallest cooler in the comparison, but with the NF-A9x14, it also has one of the best fans. This fan is both quiet and powerful. The Noctua scores additional points for the quality of the materials and the assembly. Although the cooler, like most others, is screwed in from the back through the backplate, the attachment is easy to handle and, unlike some of the other candidates, is compatible with all AM5 motherboards across the board.



Alpenfoehn Panorama 2



The Panorama 2 is Alpenfoehn’s attempt to continue the now 13-year legacy of the Panorama. The predecessor was already a popular option among the relatively large SFF coolers of the early 2010s. The Panorama 2 improves on this with a height of only 47mm. Alpenfoehn simply names the fan “Jetstream” after the larger 120mm model. Unlike the Noctua, it is not available for purchase individually. It is only available in conjunction with Alpenfoehn’s SFF CPU coolers. Let’s take a look at the data.

Height including Fan

47mm

Weight incl. Fan according to Spec

350g

Weight incl. Fan (own scale)

350g

Fan Radius

92mm

Fan model

Alpenfoehn Jetstream (92mm)

Fan Maximum RPM (Spec)

3000 U/min (±10%)

Fan Maximum RPM actual

2950 U/min

Fan Noise @ Maximum RPM

49,5 dB(A)

Price

Link to manufacturer

What stands out here is the child’s play assembly. It is the only cooler in the test that does not require a backplate or screwing through a backplate. Instead, it can be conveniently screwed into the AM5 stock backplate. Like the Noctua, it also has a shroud on its sides, which is supposed to ensure that the exhaust air is not discharged to the sides, but ideally upwards. The cooler feels quite hefty and, at 350g, is exactly within the specification.



Alpenfoehn Black Ridge



Optimum Tech has set the example and used the Alpenfoehn Black Ridge in most of its FormD T1 PCs. Many have followed suit, making the Black Ridge one of the most popular coolers in the 47mm segment. Known disadvantages are its limited compatibility with the motherboard and RAM (max. 33mm high), as well as the lack of ability to mount it on the AM5 socket. At least until now, because the new revision d finally has an AM5 mounting kit in the package. Here’s a look at the data.

Height including Fan

47mm

Weight incl. Fan according to Spec

438g

Weight incl. Fan (own scale)

434g

Fan Radius

92mm

Fan model

Alpenföhn Jetstream (92mm)

Fan Maximum RPM (Spec)

2800 U/min (±10%)

Fan Maximum RPM actual

3000 U/min

Fan Noise @ Maximum RPM

52,5 dB(A)

Price

Link to manufacturer

The assembly, like with the Noctua, also happens from the back here and was even a little easier to accomplish than that of the Noctua. What’s bothersome here is that for a fan change, due to the design, the entire cooler must be removed from the socket. Alpenfoehn does not provide a mean time to failure (MTTF) for the Jetstream fan (92mm).


To optimally use the Black Ridge in the airflow intended for the FormD T1, the fan should be turned around once before assembly, so that it sucks in the air from outside through the side panel. The major disadvantage of this configuration is, that the impeller blades of the fan are positioned so close to the fins of the cooler, that a very unpleasant noise is created by turbulence at medium to high speeds.


To somewhat mitigate this high-frequency sound, the spacer mod has been established in the community. A spacer is positioned between the fan and the heat sink so that the impeller gets just enough clearance to minimize the resulting turbulence. Popular here, for example, is the 2mm foam pad of the Noctua NA-FD1.


I myself, true to the motto “a lot helps a lot”, used a 3D printed 4mm spacer made of flexible TPU. The included 16mm screws for attaching the fan then no longer fit, but in any well-sorted PC builder inventory, there are certainly a few M3x20 to use the fan including a 4mm spacer. More on the mod later.



Thermalright AXP90-X47 Black



The AXP90-X47 from Thermalright exists in numerous different colour and material variants. In this naming, it is the standard version in silver, black, or white with an aluminium heatsink. The fan here, unlike with the previous options, is not screwed on, but only clamped on. These and other subtleties probably ensure that Thermalright can offer the cooler at such an attractive price. Let’s take a look at the facts here.

Height including Fan

47mm

Weight incl. Fan according to Spec

310g

Weight incl. Fan (own scale)

301g

Fan Radius

92mm

Fan model

Thermalright TL-9015B

Fan Maximum RPM (Spec)

2700 U/min (±10%)

Fan Maximum RPM actual

2700 U/min

Fan Noise @ Maximum RPM

44,5 dB(A)

Price

Link to manufacturer

You might be wondering what’s so special about this cooler, that I included it in the comparison. As mentioned, it’s cheap, almost outrageously so. With just over a third of the price called for the Noctua and roughly half of the rest of the non-brand competitors, the little one is already a price tip. Will the performance be decent? More on this later in the load tests.



Thermalright AXP90-X47 Full Copper



The AXP90-X47 Full (Copper) is characterized by its heat sink, which is made entirely of copper. Copper conducts heat about 70% better than aluminum, but it is also more expensive. Now, a 70% better thermal conductivity on paper does not necessarily mean a 70% better performance.


The heat pipes, i.e., the part of the cooler that takes on the lion’s share of heat transport, are already made of copper in all reputable coolers. The use of copper as a material for the cooling fins offers only one relevant advantage for our test: more weight, thus more thermal mass. So let’s take a look at the data here as well.

Height including Fan

47mm

Weight incl. Fan according to Spec

520g

Weight incl. Fan (own scale)

504g

Fan Radius

92mm

Fan model

Thermalright TL-9015R

Fan Maximum RPM (Spec)

2700 U/min (±10%)

Fan Maximum RPM actual

2900 U/min

Fan Noise @ Maximum RPM

44 dB(A)

Price

Link to manufacturer

The assembly of the Full Copper, just like its two siblings, is a bit annoying. Although a dual-purpose backplate for AMD and Intel is included, this cannot simply be used on AM5 instead of the stock AMD backplate. Because this secures the socket retaining mechanism. So it has to be attached in addition to the stock backplate, which then requires quite a lot of space behind the motherboard.


I decided on a different solution and simply used the holes of the stock backplate and countered on the back with four M2.5 nuts. This makes the cooler sit very well, even though the initial assembly was quite fiddly and you need a bit of finesse when tightening the nuts. You should not tighten these more than hand-tight to avoid damaging anything.



Thermalright AXP90-X53



Everything that applies to the AXP90-X47 also applies here, except for the height, of course. With 53mm, the X53 is a whole 6mm higher and should theoretically offer better cooling due to its larger thermal mass. Otherwise, like its two siblings, it relies on the same fan in a different color and the same fiddly clamp mechanism for its attachment. Let’s take another look at the fact sheet.

Height including Fan

53mm

Weight incl. Fan according to Spec

320g

Weight incl. Fan (own scale)

323g

Fan Radius

92mm

Fan model

Thermalright TL-9015

Fan Maximum RPM (Spec)

2700 U/min (±10%)

Fan Maximum RPM actual

2800 U/min

Fan Noise @ Maximum RPM

48 dB(A)

Price

Link to manufacturer

In our specific test case, it should be noted that due to its height, the cooler will sit very close to the side panel of the Ncase T1, causing air turbulence. This turbulence usually leads to higher noise levels. We already observed the same with the Alpenfoehn Black Ridge. This effect will be significantly more noticeable in the Fractal Terra. Gamers Nexus has covered this topic extensively in their review of the case.



The Best Small Form Factor CPU Cooler for the AMD Ryzen 7800X3D - The Results!


Synthetic Benchmark: Cinebench R23


Synthetic Load Scenario in Cinebench R23 @ 32 dB(A)

Let’s start with Cinebench R23 and the temperatures and clock speeds at 32 dB(A). Here, both the Noctua NH-L9a and the Alpenföhn Panorama 2 are struggling to keep the 7800X3D from climbing over the 89.9°C mark. With the Noctua, it’s very clear from the clock speed in the third image that the 7800X3D is already downclocking a bit here to stay below the TjMax. Meanwhile, the Panorama 2 just manages to maintain its clock speed of 4480 MHz at 32 dB(A).


The Alpenföhn Black Ridge fares a little more than 2 degrees better. It can also boast a stable clock speed of 4480 MHz.


Next up are the two Thermalright coolers, AXP90-X47 Black and AXP90-X53, which deliver virtually the same result, with the 6mm flatter X47 leading the way. Why is that? The keyword here is again “noise-normalised”. The fan of the X53, due to its proximity to the side panel, already causes such strong turbulence at the ~1530 rpm here that the identical fan of the smaller brother X47 can spin about 140 rpm faster at the same volume. If we also take a look at the weight difference of the two models a bit further up, it’s not even 10% with my two specimens.


So it’s not surprising that the AXP90-X47 can outpace its big brother for now.

The AXP90-X47 Full can also be a big brother with its 504g, and how. It easily beats its two siblings by more than 1°C and all other competitors by 3-5°C. Not bad, Mr. Woodpecker. However, in terms of clock speed, it can’t set itself apart from the aluminum AXP90-X47. As a reference, I also tested my EK Nucleus CR 240 AiO. The last diagram shows what clock speed is still possible with this cooling performance. Of course, no 47mm SFF air cooler will ever come close to the performance of a 240mm AiO.


Synthetic Load Scenario in Cinebench R23 @ 38 dB(A)

At 38dB(A), the Noctua NH-L9a is the first to fight the 7800X3D out of the thermal limit. It does this confidently and with a quite pleasant sound profile. It’s very impressive when you consider that the cooler is 10mm, or 22% flatter than most others in the test field. It now also maintains a stable 4480 MHz, the clock speed of the Panorama 2 from the first test. This one, unlike the Noctua, could not improve and remains at this clock speed. The temperature, as expected, goes down a bit and is now at a somewhat milder 86.2°C.


What is the Alpenföhn Black Ridge doing in the meantime? Its thing quite well, I would say. The temperature at 38 dB(A) is now just under the 85-degree line, which many motherboards preset. However, the sound profile is somewhat more unpleasant than that of the Noctua at the same volume, which is certainly due to the orientation of the fan.


The three Thermalrights also don’t differ too much here. The AXP90-X47 Full Copper cools the Ryzen 7800X3D about 1°C better than its taller aluminum counterpart, the AXP90-X53. The AXP90-X47 is pretty much in the middle. Even at this volume, the X53 can’t fully unleash its potential. Interesting: The black fan of the X47 can spin minimally faster, about 70 rpm than the identical but differently colored counterpart of the X47 Full, which rotates here at about 2090 rpm.


Synthetic Load Scenario in Cinebench R23 @ Full Speed (100%)

At full speed, all fans except for the Noctua are unpleasantly loud. This one, in contrast to the rest of the field, only does “only” 2500 rpm, and therefore does not make such a big leap down the temperature ladder as the other coolers. At 87.6°C and comparatively pleasant 40.5 dB(A), it’s the end for the Noctua NH-L9a.


The Alpenföhn Panorama 2 can now clearly distance itself from the NH-L9a and moves a tiny bit closer to its brand mate. However, at 49.5 dB(A), it is noticeably quieter than the Black Ridge, which can no longer hold anything back and, with a deafening screech and 52.5 dB(A), also makes it clear to me that it is not made to suck air through its heat sink. We’ll take a look later to see if a bit of modding can help.


Meanwhile, in the Thermalright trio, the AXP90-X53 manages to break out for the first time and overtakes its little brother, the X47, by just 0.5°C. For this, however, it is also significantly louder at 48 dB(A) than the AXP90-X47 Black with only 44.5 decibels. Both coolers, however, keep the 7800X3D stable below 80°C. Really good.


The clear winner of the performance test, however, is the Thermalright AXP90-X47 Full. Despite half a decibel less (44 dB(A)) than the X47 Black, the copper-colored fan manages to perform almost 200 revolutions per minute more than the black fan of the X47. Thunderstorm. What it lacked in performance at 38 decibels, it makes up for at the back end. The gap has increased from not even 0.5 to exactly 2°C, and the X47 Full lines up at the top of the field with 76.7°C on the 7800X3D.



Gaming Performance: Cities Skylines 2, Cyberpunk 2077, Dark Souls 3



Gaming Load in Cities Skylines 2 3440x1440 High Settings @ 38dB(A)

In Cities Skylines 2, all coolers show temperatures in the mid to high 70s. We shouldn’t interpret too much into these numbers, as I loaded the same city and positioned the camera at the same spot each time. As long as all coolers are roughly in the same temperature range, I think we can speak of a passed test.


However, a clear trend is emerging. The Noctua and the two Alpenföhns are on average a bit worse at keeping the 7800X3D in check. All three Thermalright coolers are about 1.5°C - 3°C ahead of the group of Central Europeans. All in all, as already written, all coolers are completely sufficient and keep the 7800X3D at least 10°C below TjMax in this demanding scenario.


Gaming Load in Cyberpunk 2077 3440x1440 High Settings @ 38dB(A)

Cyberpunk 2077, unlike Cities Skylines 2, should be a very graphics-limited game, and we see this in the temperatures. The Noctua NH-L9a and the Alpenföhn Panorama 2 are suddenly ahead of their more potent competitors, unlike in the synthetic tests.


The slightly higher temperatures with the AXP90-X47 and Alpenföhn Black Ridge could have been triggered by a background process. Perhaps Windows Update wanted to check again at the moment the Cyberpunk benchmark was running to see what’s new.


Here too, the diagram shows a somewhat reduced meaningfulness for the actual performance of the coolers. Conclusion: All are perfectly fine for gaming. It’s interesting that here the AiO, which I also included as a reference, makes not 16°C, but less than 10°C difference to the air coolers on the 7800X3D, unlike in the previous test. This could be due to the higher heat output of the graphics card, which has to go through the radiator and thus makes CPU cooling a bit more difficult, but it doesn’t have to be. I did not conduct conclusive tests on this.


Gaming Load in Dark Souls 3 2560x1440 Ultra Settings @ 38dB(A)

In Dark Souls 3, which I chose for comparison to represent the “old games”, a similar trend is again evident as in Cities Skylines 2.


The coolers that performed best in the synthetic benchmarks also stand out here again with about 2-3 degrees less. Ultimately, however, it must also be stated here that all coolers are absolutely capable of handling the waste heat of the Ryzen 7800X3D without any problems.


So, what conclusion do we draw from the gaming benchmarks? There is no clear winner. All coolers are capable of dissipating the heat generated during gaming without any problems. The test with Cyberpunk 2077 in particular shows how irrelevant the choice of CPU cooler can be in some titles. In the other two games, there is a slight trend towards the Thermalrights, but we don’t want to attach too much weight to this here. So if you just want to play, any of the coolers from this comparison is a great choice!



Popular Optimizations and Mods from the Community tested


I was also burning to know whether it makes sense in terms of performance to invest an additional €28 when buying a Thermalright cooler like the AXP90-X47 Full and put this into a Noctua NF-A9x14 chromax.black to replace the standard fan with this one.


After all, the Noctua NF-A9x14 is supposed to be an excellent 92mm fan. Let’s see what it does to the Alpenföhn Panorama 2, which has so far delivered rather below average in my tests.


Furthermore, the Alpenföhn Black Ridge also gets a small upgrade, by attaching a 4mm thick, 3D-printed spacer made of thermoplastic polyurethane (TPU) between the fan and the cooler to eliminate the turbulence. TPU is a very flexible and heat-resistant material and is therefore excellent for this application.


So let’s first take a look at the temperatures and clock speeds using these popular mods compared to the stock configuration.



Here, the Alpenfoehn Panorama 2 with the Noctua NF-A9x14 screwed on just manages to free the 7800X3D from thermal throttling at 32 dB(A). At 38 dB(A), it’s so significant that the CPU can now also boost to 4550 MHz. With a volume of 38 dB(A), an improvement of almost 1°C can even be seen. If you let the Noctua fan spin up to its 2500 rpm (stock fan 3000 rpm), the volume of the Panorama 2 under full load is reduced from 49.5 decibels to only 44.5 dB(A). Whether this is worth the extra cost of €25 is something each Panorama 2 owner has to decide for themselves.


The Black Ridge, with the spacer mod, also becomes a bit quieter in both volume classes and can therefore also record better temperatures. The 7800X3D becomes about 1°C cooler. That doesn’t sound like much at first, until we look at the total volume at 100%. Here, the Black Ridge can only work its way down from a disastrous 52.5 dB(A) to a respectable 49 decibels through the spacer. This makes it even a bit quieter than the Panorama 2 in the stock configuration.


Given the other results, I can well imagine that the Noctua fan mod on the Black Ridge would also get a bit more performance at lower volume.

The only downside here is the price. The Alpenföhn BR already costs just under €50. If we add another €25 for the Noctua and €10 for a 3D printed spacer (if you don’t have a 3D printer at home), we’re already at €85. Personally, I would like to see Alpenföhn include a 4mm spacer and a second set of longer screws in a new revision to make the Black Ridge also well usable as an intake.


Let’s move on to the AXP90-X47 Full. Yes, the Noctua mod brings a bit better temperatures and admittedly also looks really good, but whether the €25 for the fan swap is justified here, everyone should judge for themselves. If you only consider the appearance, my clear conclusion honestly is, yes. The NF-A9x14 in the chromax variant is simply chic.


Where do the mods fit into the overall context? For this, we call up the very first diagrams again and add the curves and bars for the modified coolers.



The two coolers that underwent a fan swap impress with better performance from the Noctua, especially at lower volumes, and improve by 1 to 1.5°C. This effect slowly diminishes as the fan speed increases, which is due to the fact that the Noctua impeller only spins at 2500 rpm and therefore can shovel a bit less air over the fins at full load. However, it remains somewhat quieter on both test candidates. With the AXP90-X47, however, only by 0.5 dB(A), while the noise improvement with the Panorama 2 is a full 5 decibels under full load.


So is the Noctua worth it for a cooler like the Alpenföhn Panorama 2? From my point of view, no. The fan of the Panorama 2 fits in well visually and delivers sufficient performance. Therefore, it is not worth driving up the price of the cooler by over 50% again for this minimal improvement.


It looks a bit different with the AXP90-X47 Full, as mentioned above. The stock fan is, let’s call it “a matter of taste” in terms of color, and the cooler offers so much thermal mass compared to the others in the test field that a fan like the Noctua enables it to deliver good cooling performance even at low speeds. Mainly, I would make the fan swap here because of the looks. I’m happy to take the better volume and CPU temperature with me.

With a total price of around €60 compared to the almost €76 of the Panorama 2 with NF-A9x14, the AXP90-X47 Full is the better package here.



The Conclusion: Which Small Form Factor Cooler is the Best for the Ryzen 7800X3D?


When we take a close look at all the competitors in this comparison and weigh their pros and cons against each other, there can really only be one winner, the Thermalright AXP90-X47 Full (Copper). This cooler has consistently achieved the best temperatures on the Ryzen 7800X3D in almost all tests. It receives the Tiny Tech Tweaks award for the best SFF CPU cooler in this test. With a price around €35, it is also quite affordable.

Check out AXP90-X47 Full* on Amazon.co.uk!

Check out AXP90-X47 Full* on Amazon.com!


For those who find that too expensive, the AXP90-X47 Black is the price-performance winner of this comparison. It is not only over €10 cheaper than the full copper variant, but also only 1-2 degrees inferior in most tests. In case my EK Nucleus CR 240 AiO should ever quit service, I have the AXP90-X47 Black in the drawer and can continue my gaming activities without performance losses.

Check out AXP90-X47 Black* on Amazon.co.uk!

Check out AXP90-X47 Black* on Amazon.com!


Are the other coolers not recommendable? Yes, they are! All in all, each of these CPU coolers is recommendable for a gaming PC with a 7800X3D in an SFF case like the FormD T1 or Fractal Terra.


Especially the Cyberpunk benchmark has shown how little weight lies in the choice of the CPU cooler, as long as you choose a reasonably sensible one, and more than that are all the specimens represented in this comparison.


So just go for the one you like best or that visually appeals to you the most. To the quietest or to the one with the easiest assembly. If you mainly use your PC for gaming, any of these coolers is a good choice.


What are your experiences with the mentioned coolers?

Have you tried one of the modifications or another one yourself?

Would you like more of these comparisons?


Feel free to leave me a comment!



*Links marked with * (asterisk) are so-called affiliate links. If you make a purchase through one of these links, I receive a small commission. The price for you does not change as a result. Thank you! :)

bottom of page