Is dual tower always better?

[Knight47]

I would like to know if dual tower coolers are always better than single tower?

I see some review that says Arctic Freezer 36 is just as good as Noctua D15.

 

[P4-630]

Also check the diameter and amount of heatpipes used.

My single tower Noctua NH-U12A is a 120mm cooler but with 7 heatpipes and 2 fans.

 

[dirtyferret]

I would like to know if dual tower coolers are always better than single tower?

I see some review that says Arctic Freezer 36 is just as good as Noctua D15.

way too generic of a question, as @P4-630 stated fans, heat pipes, how well the tower is built and tower specs all play a role

 

[phanbuey]

the best 2 aircoolers are dual tower, so if you're getting one of those than yes it's better

 

[Knight47]

way too generic of a question, as @P4-630 stated fans, heat pipes, how well the tower is built and tower specs all play a role

https://gelidsolutions.com/product/glacier-black/ or https://www.arctic.de/en/Freezer-36-Black/ACFRE00123A

 

[Frick]

https://gelidsolutions.com/product/glacier-black/ or https://www.arctic.de/en/Freezer-36-Black/ACFRE00123A

Depends on how much they cost where you live, what case and other cooling you have and what you're planning to cool. Between those two I'd probably go for whichever is the cheapest.

 

[Noci]

way too generic of a question, as @P4-630 stated fans, heat pipes, how well the tower is built and tower specs all play a role

All above is thru, but with these aspects in mind, the bigger the heat dissipiating surface is (cooling fins), the more efficient (not better) a air cooler can function. So, yeah a dual tower has more fin surface to dissipiate heat to the air so in theory it should be more efficient as a single tower cooler (but only if all other specs are the same).

Regarding the question in the title; for example, the answer is no, a well build single tower cooler can perform better than a poorly build dual tower model.

Best is to know how much heat you want to dissipiate (which CPU and OC or not) and check the reviews to be able to that cooling capacity than you should be OK.

Depends on how much they cost where you live, what case and other cooling you have and what you're planning to cool. Between those two I'd go for whichever is the cheapest.

As above, this is unique for every situation; available budget, does it fit in your case, noise levels, are the 'looks' are important? and so on.
As long as you just make sure the cooling capacity is suited for what you are going to cool.

It takes some research, but sometimes it's all about the journey, not the destination.
Have fun.

 

[dirtyferret]

https://gelidsolutions.com/product/glacier-black/ or https://www.arctic.de/en/Freezer-36-Black/ACFRE00123A

assuming the glacier black performs the same as the glacier rgb (they look identical and have identical specs) than the glacial black is the better cooler

I see some review that says Arctic Freezer 36 is just as good as Noctua D15

It's a basic budget cooler, it may do the job depending on your needs but it's not as good as the Noctua D15

 

[silentbogo]

They are not "better" per-se, they just have space for 3 fans to push more air through. With a decent-enough single-tower, more case fans and enough airflow you can do just as good as D15. Heck, my old Thermalright Ultra 120 did miracles on my ancient LGA1366 rig. I think it was able to cool around 300W of heat when I was going for my 4.7+GHz runs of Cinebench and 5GHz OC just for fun and bragging screens.
But under normal circumstances - yes, I'd likely go for Glacier Black as well.

 

[#22]

Cooler performance is mostly about fans, so there exist dual towers being worse than some single tower. That's why if we want to form some universal rules, let's talk about the best ones - it only makes sense.

Dual tower coolers with two fans tend to be the best. They are the most powerful, but also the lower you normalize the noise, the more they tend to win cooling wise, especially bigger ones. If we talk the best ones on the market, these tendencies transform into facts. One exception are dual tower coolers with only one fan and here it depends. E.g. Noctua D15S performs totally similar to U14S being like half of that - no matter if we talk low noise or high wattages, whatever man would trow this pair into, they will score similarly.

 

[Knight47]

Depends on how much they cost where you live, what case and other cooling you have and what you're planning to cool. Between those two I'd probably go for whichever is the cheapest.

Gelid costs like 24 euro and Arctic 28 euro. Phanteks P400A case with two P14 intake and one P12 exhaust, It will be used on a Ryzen 5 5600.

 

[Calenhad]

Gelid costs like 24 euro and Arctic 28 euro. Phanteks P400A case with two P14 intake and one P12 exhaust, It will be used on a Ryzen 5 5600.

Single tower coolers are more than enough for that cpu. The stock cooler is basically enough for that cpu. Since it is the non-X version.

 

[Shrek]

space for 3 fans to push more air through.

How advantageous are 3 fans in the sense of diminishing returns?

 

[Slyr7.62]

How advantageous are 3 fans in the sense of diminishing returns?

The 3rd fan is definitely not very useful, and even a 2nd fan doesn't do a lot depending on the cpu setup. As a quick example w/ my 12400 under a Frost Spirit 140, ran it passively for a few tests, and stayed under 75c through a 10minute loop of cinebench. Of course a 65w cpu doesn't take much, but I still won't bother w/ a 3rd fan if/when I use a hotter cpu.

 

[tabascosauz]

I would like to know if dual tower coolers are always better than single tower?

I see some review that says Arctic Freezer 36 is just as good as Noctua D15.

It's clearly a U12S-level product, so no? Not even close. Put it on top of a part that can actually get heat out of the IHS and then compare.

Dual towers aren't automatically better, but once the finstack gets fat and heavy enough then splitting it off into two finstacks with a middle fan will be more effective. TR Macho and the like didn't last, for good reason.

If you are using it on a Ryzen then none of this matters, and it says nothing about cooler performance. Once you get above a certain level of performance, all coolers pretty much perform within margin of error of each other, when tested on a Ryzen that is density-limited more than it is power-limited (ie. 6- and 8-cores).

The serious cooler performance testing is always done on Intel or on a custom testbed where you can scale the wattage at will. And even then the medium-high end dual towers just all perform in the same clump, boring as hell to try and argue which is better than the other.

Aren't many CPUs easier to cool than the Ryzen 5600.

 

[Knight47]

Single tower coolers are more than enough for that cpu. The stock cooler is basically enough for that cpu. Since it is the non-X version.

Stock fan is too noisy, so it needs to go.

It's clearly a U12S-level product, so no? Not even close. Put it on top of a part that can actually get heat out of the IHS and then compare.

Dual towers aren't automatically better, but once the finstack gets fat and heavy enough then splitting it off into two finstacks with a middle fan will be more effective. TR Macho and the like didn't last, for good reason.

If you are using it on a Ryzen then none of this matters, and it says nothing about cooler performance. Once you get above a certain level of performance, all coolers pretty much perform within margin of error of each other, when tested on a Ryzen that is density-limited more than it is power-limited (ie. 6- and 8-cores).

The serious cooler performance testing is always done on Intel or on a custom testbed where you can scale the wattage at will. And even then the medium-high end dual towers just all perform in the same clump, boring as hell to try and argue which is better than the other.

Aren't many CPUs easier to cool than the Ryzen 5600.

GN tested the Freezer 36 on 200W 3950X, it was 56.7 C while D15 was 55.2 C.
People told me that 7800X3D is easy to cool too, still reaches 75+ C on high single core load.

 

[tabascosauz]

Stock fan is too noisy, so it needs to go.

GN tested the Freezer 36 on 200W 3950X, it was 56.7 C while D15 was 55.2 C.
People told me that 7800X3D is easy to cool too, still reaches 75+ C on high single core load.

Like I just said, density considerations are not equal between 1CCD and 2CCD. 2CCD gives direct touch heatpipe coolers a better chance, than literally all cores under 1 heatpipe.

7800X3D is relatively easy to cool. 70+ on punishing single core loads for air isn't news at all, even for AM4 CPUs - if you aren't hitting near it just means your CPU doesn't have a high enough Fmax to require enough Vcore or generate enough per-core watts to matter (ie. anything below 5800X), or it's not being pushed hard enough. The single thread temps have zero bearing on Ryzen thermal performance under multi-thread. I don't know why you are bringing the 7800X3D into this, AM4 and AM5 are not the same.

In any case, the Freezer is way more cooling than your 5600 needs. But your question is whether dual tower is always better. Answer is no one makes heavy monolithic single towers anymore for good reason.

Also, why are we comparing to the D15? It hasn't really been top dog for a long time. Go check out the Gen2 review to see where it stands compared to even the PA120s and PS120s.

 

[silentbogo]

How advantageous are 3 fans in the sense of diminishing returns?

Honestly - barely noticeable and not worth it. The only 3 times it helped were:
1) When I was rebuilding my office neighbor's PC into a portable ITX cube case before his vacation. Had to zip-tie a third 140mm fan onto his D14, maybe dropped temps by 2-3C. In a normal ATX case it may not make any difference at all, especially if you have a decent intake/exhaust.
2) open bench, which I assume does not apply to your use case.
3) Not too long ago I used to have an LGA2011 rig (along with my Z620 ghetto-rig). Cooled it with a Snowman HSF, probably the cheapest adequate cooling for this socket. Also added 3rd fan once I swapped CPU to a 115W Ivy SP. It was a shoestring-budget PC, so the case was awful(a 20 y.o. ASUS Vento or something like that), only added 1 intake fan and left the exhaust empty since CPU fans did a better job at blowing air through. I think the whole rig cost me less than $150 including a brand new Quadro P400.

 

[Darkholm]

Gelid costs like 24 euro and Arctic 28 euro. Phanteks P400A case with two P14 intake and one P12 exhaust, It will be used on a Ryzen 5 5600.

Single tower with good fan. I had 5600X + Thermalright Venomous X w/ Noctua F12 fan on it and it was cool.
On my working rig, there is 5600G and it is cooled by DeepCool AG400 (dirt cheap cooler) and I have not heard it ever (only fan in my work system), even when rendered some promo video in DaVinci Resolve.
R5 5600 is not heat monster like 5800X or 5900X so you have no need to bust your head which cooler.

 

[#22]

Stock fan is too noisy, so it needs to go.

GN tested the Freezer 36 on 200W 3950X, it was 56.7 C while D15 was 55.2 C.
People told me that 7800X3D is easy to cool too, still reaches 75+ C on high single core load.

Put Freezer 36 on some bigger load, on CPU with more sensible for testing purposes thermal dissipation or lower the noise of both coolers and you will see D15 beating this Arctic by large margin.

Why you even bother with this Gelid? Is there any decent review showing it's performance? Or you just find it pretty?

Today if your priority is cooler being cheap, search no further than Thermalright. If not available in you region, nice picks is Arctic you consider of Endorfy/SilentiumPC. ID Cooling is not bad too. Keep in mind that cheap cooler means bottom of the barrel anything, especially fans of random longevity and likelihood of acoustic anomalies which may irritate you on daily basis.

 

[JustBenching]

All else being equal yes, dual towers are better

 

[Knight47]

Why you even bother with this Gelid? Is there any decent review showing it's performance? Or you just find it pretty?

Today if your priority is cooler being cheap, search no further than Thermalright. If not available in you region, nice picks is Arctic you consider of Endorfy/SilentiumPC. ID Cooling is not bad too. Keep in mind that cheap cooler means bottom of the barrel anything, especially fans of random longevity and likelihood of acoustic anomalies which may irritate you on daily basis.

Because it's cheap, I can lower the middle fan to get some airflow over the VRM.

Cheapest TR coolers are Assassin Spirit/King/X for 25€. Endorfy Spartan 5 is 20€, Fera 5 30€.

 

[#22]

Because it's cheap, I can lower the middle fan to get some airflow over the VRM.

Cheapest TR coolers are Assassin Spirit/King/X for 25€. Endorfy Spartan 5 is 20€, Fera 5 30€.

Cooling CPU surroundings is great advantage of dual towers, but more theoretical until you have VRM shitty to the point that you need to rely on damn CPU cooler or powerhungry card (~300W) which would make CPU thermals noticeably hurt from not having this ventillation. The latter is this less obvious, reviewers rather don't point it out, but is likely to create bigger temperature differences.

I have never seen the review of this Gelid, so I wouldn't buy such thing. If your really need it to be cheap like 30 euros, fine - by being dual tower itself it's likely being better than tiny coolers from Endorfy or Thermalright. But if I were you, I would pay probably 10 euros more for Thermalright Phantom Spirit, one of their dual towers at least. Then you would have top performing cooler from the brand having tradition of providing mounting kits for future sockets, so saving you buying next cooler. Imo totally worth it instead of some unknown from the brand not counting anymore.

 

[sneekypeet]

Dark horse solution for bang-for-the-buck... ID-Cooling FROZN A620 PRO SE. $29.99 and performs well above its class.

 

[Knight47]

But if I were you, I would pay probably 10 euros more for Thermalright Phantom Spirit, one of their dual towers at least.

50€, basically every other dual tower cooler costs twice as much.

Dark horse solution for bang-for-the-buck... ID-Cooling FROZN A620 PRO SE. $29.99 and performs well above its class.

53€