120mm AIO Roundup: Testing Be Quiet, Corsair, Cooler Master, and Enermax models

I had wondered about something, after seeing news about EKWB taking over Intel's Cryo Cooling:

Are there any AIO coolers that incorporate TEC to reduce water temps to potentially below ambient? Putting it in the radiator seems like a way you could utilize Peltier cooling with less waste heat, and activate it only when really needed.

I've seen external chillers, but wondered whether anything like that was incorporated into radiators intended for internal mounting.
 
  • Like
Reactions: Albert.Thomas
Corsair iCue has been buggy the last few releases if using a custom fan curve, as it'd lock the fans to the lowest speed no matter the temperature, which was just fixed in the latest release of 5.11.96.
This AIO isn't controlled by iCUE, it's controlled by PWM
Enermax's website doesn't show the LiqMaxFlow 120, only the LiqMaxFlo SR 120. Can you verify this is the LiqMaxFlo SR 120??
Good catch. Indeed, it is the SR 120.
I had wondered about something, after seeing news about EKWB taking over Intel's Cryo Cooling:

Are there any AIO coolers that incorporate TEC to reduce water temps to potentially below ambient? Putting it in the radiator seems like a way you could utilize Peltier cooling with less waste heat, and activate it only when really needed.

I've seen external chillers, but wondered whether anything like that was incorporated into radiators intended for internal mounting.
I've had similar thoughts. In theory, you'd want it to cool temperatures to just above ambient to avoid issues with condensation.
 
I've had similar thoughts. In theory, you'd want it to cool temperatures to just above ambient to avoid issues with condensation.
I think you can afford to go a little below ambient, if the feed hose is insulated. I expect the waterblock, itself, will stay at least at ambient.

Whether it's just below or just above ambient, I think the main point is that there'd need to be a thermostat controlling the TEC, so the water doesn't get cooled too much (and the flipside is you don't want to needlessly burn power chilling the water when it's not necessary). If TEC kicks in pretty much only when the CPU is boosting, then I think we won't have to worry too much about below-ambient water temperatures.

Heh, the thermostat feature could be implied by calling it Variable TEC, or "VTEC", for short. Then, we could revive a whole generation of memes.

vtec-just-kicked-in-yo-1.jpg

 
I always wonder, why is three no comparison to a stock air cooler? They are supposed to have capacity to cool at max power, sustained (assume that pc case itself has adequate cooling). What willa one gain for using liquid cooling - higher (meaningfull for gaming, encoding etc) performance? OC capacity?
 
Corsair iCue has been buggy the last few releases if using a custom fan curve, as it'd lock the fans to the lowest speed no matter the temperature, which was just fixed in the latest release of 5.11.96.

Do you have a link to a different, perhaps non-spoofed page? Something smells phishy..

The typo belows doesn't happen much (or for long) on (properly editor-reviewed) published sites for .com based companies. From the 5.12.97 release notes:

"• Background image will now display properly when using the 640x48 px window size".
 
Not sure that four products constitute a "roundup". The Arctic Liquid Freezer line has consistently trounced the competition for many years, at really competitive prices...

Full disclosure: yes, I do own an Arctic cooler, but spent many hours reading reviews. That's why I went with their product.
 
Not sure that four products constitute a "roundup".
*Five Coolers, Four Manufacturers 😉
The Arctic Liquid Freezer line has consistently trounced the competition for many years, at really competitive prices...
The Arctic Liquid Freezer II 120 was discontinued last year and I don't see a Liquid Freezer III 120 available anywhere yet
 
  • Like
Reactions: dave.rara66
To what end?

Unless you are on the hairy edge of overclocking and are actually thermal throttling, "cooler" does not mean better performance.
These days, you don't need to overclock the higher-end CPUs, for cooling to be your limiting factor. Since we're talking about 120 mm AIO's, I wondered how much more mileage TEC could get from them. Perhaps it could even be a more cost-effective option than going up to 240 mm, or at least allow the fans to run a little slower for the same amount of cooling.

Note that this article found these 120 mm AIOs delivered CPU temperatures of 78 to 84 C, at just a 95 W workload. It's not at all hard to see why you might want more cooling capacity, if you do anything CPU-intensive and have a more power-hungry CPU than the 7700X used in these tests.

For instance, a Noctua NH-U14S (6, double-ended heatpipe cooler, with a 140 mm fan) can only cool a stock i7-14700K to 98.3 C.

Note that they increased the default thermal throttling limit to 115 C, for that test. Otherwise, the temperature ceiling would be 100 C. When you consider that a lot of motherboards will run that CPU with power limits removed, by default, it would be well into throttling territory!
 
Are there any AIO coolers that incorporate TEC to reduce water temps to potentially below ambient? Putting it in the radiator seems like a way you could utilize Peltier cooling with less waste heat, and activate it only when really needed.
I think no one adds a TEC to the radiator side, because it's going to be very clunky and expensive.
TECs are usually shaped like a square plate with a small surface area. If you want to increase its surface area, it's going to need its own water block on the water side, and maybe a regular heatpipe tower on the air side.
Using some very rough maths a contraption that has a water block attached to a heatpipe air tower is going to be some 60mm x 120mm x 180mm in size for a 120mm fan.

I would think it's cheaper and more compact to increase surface area of the radiator, and add the TEC to the CPU waterblock side.
 
  • Like
Reactions: bit_user
Then you shouldn't be using low end 120mm AIOs.
What if TEC would add enough cooling performance that you could? There are advantages to the 120 mm form factor, other than cost. Their small size could make them more viable for some mini-ITX cases. Or what about boosting the performance of 240 mm radiators, so that you don't need to go even bigger for i9 and R9 cpu models?

Maybe you disagree, but I just thought it might be an interesting idea to explore. Anyway, you asked and so there's my explanation.
 
What if TEC would add enough cooling performance that you could? There are advantages to the 120 mm form factor, other than cost. Their small size could make them more viable for some mini-ITX cases. Or what about boosting the performance of 240 mm radiators, so that you don't need to go even bigger for i9 and R9 cpu models?

Maybe you disagree, but I just thought it might be an interesting idea to explore. Anyway, you asked and so there's my explanation.
Yeah, I get what you're saying.

Its obviously not an impossible design. But as we don't have any yet....
 
TECs are usually shaped like a square plate with a small surface area.
I wonder what the real restrictions are on its size and shape.

I would think it's cheaper and more compact to increase surface area of the radiator, and add the TEC to the CPU waterblock side.
As far as I know, atop the CPU is where people normally put the Peltier device(s). The point was to consider putting it on the other end of the system. I think it could be more efficient there, since you wouldn't need such a large temperature gradient for it to provide a benefit in the radiator. I'd put it only in the latter portion of the radiator, which is naturally less efficient than when the warm water first enters.
 
@bit_user: Dude, I'm a fan and always read your comments with great interest. But you have completely misunderstood or circumvented USAFRets question here imho. Total cooling capacity is a completely different issue to minimal reachable temperature. You would have to explain how my cpu would profit from 25°c as apposed to 30°c summer ambient.

I know that I would definitely perform better; my personal operation specs require throttling or cold beer as of 28°c ... But that's not the issue here. 😉
 
So, I found this discussion of the approach:

Just reading through it now, but I see two compelling arguments against it.
  1. You need a lot of peltier devices to cool the water by a relevant amount to affect load temps.
  2. If you don't adequately cool the "hot" side of the peltier, I guess the "cool" side doesn't get very cold.

As @Notton said, you'd probably need to attach the peltier to a waterblock with heat pipes, and then run those heat pipes through part of the radiator. Perhaps an easier way to think about it would be like bolting a tower-style air cooler to the second water block (this one being connected downstream of the normal radiator), but with the peltier in between. Would add cost, bulk, and expense.

I haven't sought out the referenced discussion at overclock.net - I think finding the reddit thread satisfied enough of my curiosity. Thanks for the replies, everyone.
 
Last edited:
I wonder what the real restrictions are on its size and shape.


As far as I know, atop the CPU is where people normally put the Peltier device(s). The point was to consider putting it on the other end of the system. I think it could be more efficient there, since you wouldn't need such a large temperature gradient for it to provide a benefit in the radiator. I'd put it only in the latter portion of the radiator, which is naturally less efficient than when the warm water first enters.
Alternatively, I suppose the TEC could be added between the inlet and outlet. It would still require water blocks on both sides to get the most surface area, but at least it won't take up that much room.

As for size constrains, it seems like it can be rectangular, but too large in size could cause de-lamination/failure of the TEC.
https://americas.kyocera.com/news/2020/07/22184006.html
 
  • Like
Reactions: bit_user