Which cooler for ryzen

[CiroConsentino]

Liquid metal... T-1000 "heat, you are terminated"
But seriously, if anyone can recommend me a "not so expensive" water cooler with 240mm radiator or even if the new one from Cooler Master (240mm radiator) is good, it's appreciated.

 

[Caring1]

I use a CM 240mm AIO on my main system and although it's not the latest is keeps it cool, and runs quietly.
And yes my current system is Intel, but the cooler has been used on various builds including Ryzen.

 

[mtcn77]

I agree that liquid metal be paired with a water cooler since that would increase the heatsink's efficiency to unbelievable proportions. Better to have the heat expelled outside the case, than within.

 

[CiroConsentino]

I use a CM 240mm AIO on my main system and although it's not the latest is keeps it cool, and runs quietly.
And yes my current system is Intel, but the cooler has been used on various builds including Ryzen.

May I ask for how long ? You probably have the 1st version with the "Cooler Master" logo on top of the pump. I've read that on the 2nd version the sealing is reinforced to prevent leaking.
Ryzen 3800X has a TDP of 105W, don't know what the max wattage capacity "Cooler Master" can handle.

 

[Caring1]

Actually they have had a few revisions of their 240.
Mine is the second iteration (still has the logo) and i've had it for over a year, maybe two, with no issues.
Don't worry about 105W, Intel can be a lot higher.

 

[CiroConsentino]

I'm beginning to regret getting the 3800X, should I change it to 3700X ? Is game performance really worse ? Playing at 4K res (or 1440p when needed). I think with 3700X I could actually get the Scythe TUF Mugen 5 - model SCMG-5100TUF and then add a second fan for a push-pull configuration, for extra cooling. I think 3700X is a lot easier to cool than 3800X. ?
I'm not sure if I can use any fan as the 2nd one. There is a "rev 2" model but I think it's the same fan, out of stock though.
I'm struggling here.

 

[Mr McC]

I'm beginning to regret getting the 3800X, should I change it to 3700X ? Is game performance really worse ? Playing at 4K res (or 1440p when needed). I think with 3700X I could actually get the Scythe TUF Mugen 5 - model SCMG-5100TUF and then add a second fan for a push-pull configuration, for extra cooling. I think 3700X is a lot easier to cool than 3800X. ?
I'm not sure if I can use any fan as the 2nd one. There is a "rev 2" model but I think it's the same fan, out of stock though.
I'm struggling here.

If you look back through the thread, on the first page, fourth post, Mussels states that a Dark Rock Slim provides adequate cooling for a 3700x. That would suggest to me that the Mugen 5, with a single fan, will be able to handle a 3800x without any problem. The Rev. B Mugen simply refers to the fact that the mounting system now works with the AM4 socket, which appears to also be the case with the TUF model.

 

[CiroConsentino]

Ok, so I'll buy another build with the 3700X and the MUGEN 5 TUF (haven't actually payed the other purchase yet). I'll probably add a 2nd fan for the push-pull config. My only concern is the speed throttling when gaming. The clock throttle is caused by the CPU temperature, right ?

 

[Mr McC]

Ok, so I'll buy another build with the 3700X and the MUGEN 5 TUF (haven't actually payed the other purchase yet). I'll probably add a 2nd fan for the push-pull config. My only concern is the speed throttling when gaming. The clock throttle is caused by the CPU temperature, right ?

I think you would be fine with the Mugen and a single fan. Boost clocks are limited by CPU temperature, but I don't think that is an issue with the Mugen you are considering: if a Dark Rock Slim can handle the processor, a Mugen most certainly can.

 

[Lindatje]

In my opinion you are downgrading your cooling, most air coolers are either going to be louder under load or perform worse. Are you sure the hum doesn't come from something else ?

Most watercoolers are louder then air coolers. Also for the 3600 is a watercooler overkill, a watercooler is only useful with CPUs that consume a lot, like the TRs or the new i7 / i9s. just look the reviews.

 

[Vya Domus]

Most watercoolers are louder then air coolers.

I find that hard to believe, I can't see how an AIO with fans running <1000 RPM can be louder than most air coolers which have to run at higher RPM to achieve the same temperatures.

Also for the 3600 is a watercooler overkill, a watercooler is only useful with CPUs that consume a lot, like the TRs or the new i7 / i9s.

That's for anyone to decide individually if it's overkill or not, there is definitely value in running an AIO for a 95W+ CPU.

 

[milewski1015]

I find that hard to believe, I can't see how an AIO with fans running <1000 RPM can be louder than most air coolers which have to run at higher RPM to achieve the same temperatures.

I think Lindatje was referencing the noise from the pump. I've only used air cooling so I can't say I have any personal experience to go by, but the complaints I generally see about AIO noise is in regards to the whir of the pump impeller.

 

[mtcn77]

If you look back through the thread, on the first page, fourth post, Mussels states that a Dark Rock Slim provides adequate cooling for a 3700x. That would suggest to me that the Mugen 5, with a single fan, will be able to handle a 3800x without any problem. The Rev. B Mugen simply refers to the fact that the mounting system now works with the AM4 socket, which appears to also be the case with the TUF model.

If you need help, there is a seperate list for mid-range coolers. Basically, Kotetsu 2 is a better candidate if your heat load isn't so extreme.

62 CPU-luchtkoelers vergeleken: hoe blijf je deze zomer koel?

Waterkoeling wordt steeds populairder. Niet vanwege de prestaties danwel de looks die de doorslag geven. Toch zijn luchtkoelers nog altijd de top drie i...

nl.hardware.info

If you need seperate filters, hardware.info can tabulate seperate heat loads. I can prove it, if you don't take my word for it that mid range coolers do better in this regard than AIOs and big towers.
I tried telling before, these cpus get better with an AIO solely because of the better cold plate thermal interface. That is where your effort should be focused.

I find that hard to believe, I can't see how an AIO with fans running <1000 RPM can be louder than most air coolers which have to run at higher RPM to achieve the same temperatures.

AIO's don't run quiet, they start at 50dBA. I'm not being picky with reviews, it is just the general trend.

I think Lindatje was referencing the noise from the pump. I've only used air cooling so I can't say I have any personal experience to go by, but the complaints I generally see about AIO noise is in regards to the whir of the pump impeller.

AIO's offer the liquid metal advantage.
Recent reviews state cold plate interface accounts for 30%10% of the cooler performance, the fin array and forced conduction is only 45%, 10%30% is the tim.

 

[Vya Domus]

AIO's don't run quiet, they start at 50dBA. I'm not being picky with reviews

I don't get it how you can simultaneously say you're not picky with reviews yet you throw some random number out there like "50dba" as if it even meant something. You are aware noise measurements are essentially worthless without a reference point or context, right ? If you are not picky then surely you can look at any review from Gamer Nexus for example where they normalize tests based on noise and AIOs always come on top.

I am still kind of staggered that you guys are seriously arguing that AIOs are quieter than air coolers in general. A radiator filled with water has more thermal capacity, that fact by it self implies you need less airflow and therefore less fan RPM in order to cool something to the same level compared to using an aluminium heatsink with copper heatpipes.

I think Lindatje was referencing the noise from the pump.

Yes, the pump can be audible, I would know I have two of them in my PC, if I turn every other fan off in my system I can hear them. However once the fans ramp up it makes no difference and since AIOs need to ramp up the fans to a lesser extent than your average air cooler the total noise output will still be lower.

 

[CiroConsentino]

If you need help, there is a seperate list for mid-range coolers. Basically, Kotetsu 2 is a better candidate if your heat load isn't so extreme.

62 CPU-luchtkoelers vergeleken: hoe blijf je deze zomer koel?

Waterkoeling wordt steeds populairder. Niet vanwege de prestaties danwel de looks die de doorslag geven. Toch zijn luchtkoelers nog altijd de top drie i...

nl.hardware.info

Thanks, I bought Scythe Mugen 5 TUF in the end. I would ask for some advice on how to apply the thermal paste on AMD CPUs. Does the paste need to cover the entire CPU ? For Intel CPUs, I've always applied a straight line in the center from top to bottom as Intel recommends. On placing the heatsink, the paste would spread on its own.

 

[mtcn77]

I am still kind of staggered that you are seriously arguing that an AIO are quieter than air coolers in general.

I just discovered multi-quotes, so stick with me.

A radiator filled with water has more thermal capacity,

This is true, however I didn't say it per se. You did. I argue for its relative high performance which comes courtesy of more noise.
I have seen and listen to reviews which offer advice that the next best step AIOs should take is to increase the pump water flow. Well, if you would do that, you lower the temperature gradient at the radiator. See what I mean? AIO heat exchange is limited by the fundamentals without going too much into detail. If you are going to top the charts with the microarray cold plate, you give up some of the flow speed.

that fact by it self implies you need less airflow and therefore less fan RPM in order to cool something to the same level

This part is where you will not support me.
To imply something, is not to say 'it is', in other words naturalistic fallacy. "Is" is not what "ought to be".
AIOs tend 'not to be' quiet imo, they neither top the charts on air flow. They are quite restrictive, both at the pump and the fin-fan arrangement.
I am not saying they don't work at high heat loads, to the contrary, they work better than air coolers.
They skew the bounding issue from the radiator impedance towards impedance due to less clutter for the rest of the fans.
They tend towards open bench results since they feed from the outside without elevating case internal air temperature. They offer a great trade off between higher radiator impedance causing more noise while lowering case airflow impedance and turning it into an open bench suite with its constant cold air feed & sealed exhaust if you do it correctly.

compared to using an aluminium heatsink with copper heatpipes.

Copper heatpipes don't experience galvanic corrosion since they are fully sealed.

 

[Vya Domus]

I argue for its relative high performance which comes courtesy of more noise.

Which is untrue, you saw the chart I posted. Even when you normalize noise AIOs perform better, it's still beyond me how you can believe they are louder when nothing supports that idea.

Copper heatpipes don't experience galvanic corrosion since they are fully sealed.

I can't see why that matters to any significant extent. An AIO will fail due to other reasons long, loooooong before corrosion gets involved, unless there was something terribly wrong with it from day one.

AIOs tend 'not to be' quiet imo

I guess it's just your opinion then, unsupported by any kind of testing.

Again, trust me, I would know if they really were louder I have two of them running in one system.

 

[mtcn77]

Thanks, I bought Scythe Mugen 5 TUF in the end. I would ask for some advice on how to apply the thermal paste on AMD CPUs. Does the paste need to cover the entire CPU ? For Intel CPUs, I've always applied a straight line in the center from top to bottom as Intel recommends. On placing the heatsink, the paste would spread on its own.

I made some mistakes accounting for the percentages. Essentially, 10% is the latching mounting pressure. You have to tighten the nuts well. Be sure to fill the gap with some more extra because the ihs plate is concave. Requiring for you to account for it, either with a convex base cpu cooler, or good pressure.
Tim is 30%, I couldn't find the chart, but it is citable.

I guess it's just your opinion then, unsupported by any kind of testing.

I'll leave it at that, if you found so little.

 

[Mussels]

The noise is about idle vs load
Air is quieter at idle, water is quieter at load

pump noise is a total bitch, it all depends if you get a quiet one or not... out of maybe 6 AIO's i've had, i've only had one with a silent pump.

 

[CiroConsentino]

Ok, I think I'll be fine with the

I made some mistakes accounting for the percentages. Essentially, 10% is the latching mounting pressure. You have to tighten the nuts well. Be sure to fill the gap with some more extra because the ihs plate is concave. Requiring for you to account for it, either with a convex base cpu cooler, or good pressure.
Tim is 30%, I couldn't find the chart, but it is citable.

I could go for the X paste method, and maybe, a small dot on each of the "V" sides. The blob method might not reach the corners. Too much paste would just fall off onto the motherboard, right ?
Maybe I'm just overthinking... I only applied paste on 2 Intel CPUs, straight line in the middle.

 

[Mr McC]

Ok, I think I'll be fine with the

I could go for the X paste method, and maybe, a small dot on each of the "V" sides. The blob method might not reach the corners. Too much paste would just fall off onto the motherboard, right ?
Maybe I'm just overthinking... I only applied paste on 2 Intel CPUs, straight line in the middle.

I don't think there is a definitive answer, but this method seems as good as any:

 

[Ruru]

Liquid metal is pretty meh unless it's used on delidded CPUs (or laptops).

 

[CiroConsentino]

The 5 dots seem to be the one AMD recommends.

 

[xman2007]

Liquid metal is pretty meh unless it's used on delidded CPUs (or laptops).

I take it you have done your own thorough reseach and development in a controlled environment with various test systems at controlled ambient temps to come out with such a claim?

 

[Ruru]

I take it you have done your own thorough reseach and development in a controlled environment with various test systems at controlled ambient temps to come out with such a claim?

The hassle vs the little benefit isn't worth it. Also it likes to stick on the IHS/cooler/waterblock so goodbye warranty.