How come a custom waterloop is better than an AIO? (in terms of cooling)

[Hi P]

I never understood this... since both of them use water and radiators

[Raudi_]

Block design 

[RONOTHAN##]

Custom loops tend to have better pumps with higher flow rates, better designed and more substantial water blocks, and larger, better designed radiators made out of better materials (copper rather than aluminum). 

 

There's nothing stopping an AIO from being as good as a custom loop, its just that it's really expensive and not really practical to do it properly. 

[TrigrH]

4 minutes ago, Hi P said:

I never understood this... since both of them use water and radiators

Better Blocks, significantly bigger and better pumps, higher water % in the coolant, larger tubing and inlet/outlets, typically larger radiators and a radiator/loop per device.

[shoutingsteve]

A big part of it is how much of the loop is in the radiator vs in the tubes.  When you custom cut tubing you make it the shortest length.  So the water is spending more time absorbing heat from the CPU or dropping heat at the radiator.  When you add in a bunch of extra tubing that is on the AIO setup the water hangs out in transport longer and therefor isn't the most efficient trip to or from your radiator.

[TrigrH]

10 minutes ago, shoutingsteve said:

A big part of it is how much of the loop is in the radiator vs in the tubes.  When you custom cut tubing you make it the shortest length.  So the water is spending more time absorbing heat from the CPU or dropping heat at the radiator.  When you add in a bunch of extra tubing that is on the AIO setup the water hangs out in transport longer and therefor isn't the most efficient trip to or from your radiator.

Why would the length of the tubing be a factor? its not like the cooled water is heating up on its journey back to the block? The time water is spent absorbing heat from the CPU would be determined by the strength of the pump no?

[Blue4130]

16 minutes ago, shoutingsteve said:

A big part of it is how much of the loop is in the radiator vs in the tubes.  When you custom cut tubing you make it the shortest length.  So the water is spending more time absorbing heat from the CPU or dropping heat at the radiator.  When you add in a bunch of extra tubing that is on the AIO setup the water hangs out in transport longer and therefor isn't the most efficient trip to or from your radiator.

Most custom loops that I have seen have much longer tubes than an aio when you add up all the runs.

[RONOTHAN##]

3 minutes ago, TrigrH said:

Why would the length of the tubing be a factor? its not like the cooled water is heating up on its journey back to the block? The time water is spent absorbing heat from the CPU would be determined by the strength of the pump no?

It technically gives slightly higher resistance to the pump, though unless you're using an entire spool for every run it doesn't matter. 

[tkitch]

21 minutes ago, shoutingsteve said:

A big part of it is how much of the loop is in the radiator vs in the tubes.  When you custom cut tubing you make it the shortest length.  So the water is spending more time absorbing heat from the CPU or dropping heat at the radiator.  When you add in a bunch of extra tubing that is on the AIO setup the water hangs out in transport longer and therefor isn't the most efficient trip to or from your radiator.

uhhhh no, this is entirely wrong.

 

Everyone else's comments about blocks, pumps, etc. are the correct answer.

[Somerandomtechyboi]

Usually in the radiators, aios usually come with slim rads likely for cost but also helps compatibility (liquid freezer ii literally the best aio on the market may have fitting issues with some setups due to the thiccness of the rad + fan)

 

And aios also have quite abit less water in em, less water = less thermal mass to absorb heat and cold which means when you load the cpu temps will spike (especially bad on the highend cpus) but with custom water theres alot more thermal mass depending on the res you use so more water to absorb the heat and cold so less spikey more stable temps. Kinda like flywheels, lighter make car go faster but when you lift the pedal itll slow down faster

 

 

A garbage custom loop with like no rad capacity will perform like trash but the normie ones that most ppl think of (at worst 360 20mm thiccness) will still likely perform better than even the liquid freezer ii. And theres also the old way of custom looping, rigging a god damn car radiator into a watercooling system though that isnt done anymore for obvious reasons (absolutely massive, pain to build, run a 20% antifreeze mix to prevent galvanic corrosion, and ppl prefer custom water mostly for aesthetics) great for pure performance and price/performance though which is why im building one

[shoutingsteve]

1 hour ago, tkitch said:

uhhhh no, this is entirely wrong.

 

Everyone else's comments about blocks, pumps, etc. are the correct answer.

I love when people use absolutes in these discussions, they are usually so excited to be presented with information beyond what they already believe. 

Wrap your hand around the return loop and tell me if it feels hot.  1/2 inch OD PVC tubing has a surface area of about 18 square inches per foot.  That is 18 square inches of heat source emanating inside your case rather than going out the radiator.  The high surface area copper drawplates are the largest factor and move heat away much faster than a small aluminum sink would, I would never say otherwise, but unless you plan on insulating your lines, you still need to make the shortest (and fastest, which means a speedy flow rate is important) run possible if you are serious about getting the heat out of the box.  Once out of the return pipes, it needs to spend as much time in the radiating heat into the air as possible before returning to the loop.  If you have a  half liter of coolant in your system, but 200 ml of that is in the transport lines, the other 300 ml isn't getting much time to cool off.

 

So, to repeat myself (an exact quote of my post): A big part of it is how much of the loop is in the radiator vs in the tubes.

[TrigrH]

3 minutes ago, shoutingsteve said:

I love when people use absolutes in these discussions, they are usually so excited to be presented with information beyond what they already believe. 

Wrap your hand around the return loop and tell me if it feels hot.  1/2 inch OD PVC tubing has a surface area of about 18 square inches per foot.  That is 18 square inches of heat source emanating inside your case rather than going out the radiator.  The high surface area copper drawplates are the largest factor and move heat away much faster than a small aluminum sink would, I would never say otherwise, but unless you plan on insulating your lines, you still need to make the shortest (and fastest, which means a speedy flow rate is important) run possible if you are serious about getting the heat out of the box.  Once out of the return pipes, it needs to spend as much time in the radiating heat into the air as possible before returning to the loop.  If you have a  half liter of coolant in your system, but 200 ml of that is in the transport lines, the other 300 ml isn't getting much time to cool off.

 

So, to repeat myself (an exact quote of my post): A big part of it is how much of the loop is in the radiator vs in the tubes.

The amount of heat that radiates out of the tubing is irrelevant, longer tubing will only increase the heat released which would lower temps not increase it??? Either way Case fans deal with that? or am I missing something here? 

[tkitch]

9 hours ago, shoutingsteve said:

I love when people use absolutes in these discussions, they are usually so excited to be presented with information beyond what they already believe. 

Wrap your hand around the return loop and tell me if it feels hot.  1/2 inch OD PVC tubing has a surface area of about 18 square inches per foot.  That is 18 square inches of heat source emanating inside your case rather than going out the radiator.  The high surface area copper drawplates are the largest factor and move heat away much faster than a small aluminum sink would, I would never say otherwise, but unless you plan on insulating your lines, you still need to make the shortest (and fastest, which means a speedy flow rate is important) run possible if you are serious about getting the heat out of the box.  Once out of the return pipes, it needs to spend as much time in the radiating heat into the air as possible before returning to the loop.  If you have a  half liter of coolant in your system, but 200 ml of that is in the transport lines, the other 300 ml isn't getting much time to cool off.

 

So, to repeat myself (an exact quote of my post): A big part of it is how much of the loop is in the radiator vs in the tubes.

Only a sith deals in absolutes...  So apparently in this convo I'm a motherfuckin' sith, and I'm okay with that.

 

Okay, if I wrap my hand around an AIO's tube and it's warm, what's that supposed to tell me, other than it's doing it's job correctly?  Liquid in a cooling solution is supposed to heat up.  That's literally what it's there for.

 

Surface Area of the tubing in the case doesn't mean shit.  Why?  A few reasons:

1) PVC / PE / Whatever rubberized tubing they're using is a good insulator.  Yes, some heat escapes, but that's not the intended use, I imagine in AIOs it's selected for durability and reliability over anything else.

 

2) So despite being a decent insulator, yes some heat gets out, because physics says it will.  Regardless, that heat is in the air of the case, not in your processor, therefore it's effectively irrelevant, at least there.  (There's a possible small change to GPU temps if you warm up the air in the case, but if you have shit for airflow to allow that, that's on you.)

 

Mostly what you need to get better temps out of liquid cooling is better Blocks and Better Rads.  Custom Loop has both of those.  The pump also helps, but past a certain point, it's not a significant change.  (But yes, that point is well above what an AIO Pump can push.)

[shoutingsteve]

4 minutes ago, tkitch said:

Only a sith deals in absolutes...  So apparently in this convo I'm a motherfuckin' sith, and I'm okay with that.

 

Okay, if I wrap my hand around an AIO's tube and it's warm, what's that supposed to tell me, other than it's doing it's job correctly?  Liquid in a cooling solution is supposed to heat up.  That's literally what it's there for.

 

Surface Area of the tubing in the case doesn't mean shit.  Why?  A few reasons:

1) PVC / PE / Whatever rubberized tubing they're using is a good insulator.  Yes, some heat escapes, but that's not the intended use, I imagine in AIOs it's selected for durability and reliability over anything else.

 

2) So despite being a decent insulator, yes some heat gets out, because physics says it will.  Regardless, that heat is in the air of the case, not in your processor, therefore it's effectively irrelevant, at least there.  (There's a possible small change to GPU temps if you warm up the air in the case, but if you have shit for airflow to allow that, that's on you.)

 

Mostly what you need to get better temps out of liquid cooling is better Blocks and Better Rads.  Custom Loop has both of those.  The pump also helps, but past a certain point, it's not a significant change.  (But yes, that point is well above what an AIO Pump can push.)

You seem to think I'm disagreeing with you about the blocks and radiators...
What I have said (multiple times) is that running the shortest (read that as CUSTOM CUT TO LENGTH) tubes is a good practice for keeping temps down.

[tkitch]

7 hours ago, shoutingsteve said:

You seem to think I'm disagreeing with you about the blocks and radiators...
What I have said (multiple times) is that running the shortest (read that as CUSTOM CUT TO LENGTH) tubes is a good practice for keeping temps down.

And I'm saying nothing in that statement is correct.

 

If anything, longer tubes give more surface area for heat (even if just a little at a time) to escape into the air from the water.  

[shoutingsteve]

5 hours ago, tkitch said:

And I'm saying nothing in that statement is correct.

 

If anything, longer tubes give more surface area for heat (even if just a little at a time) to escape into the air from the water.  

Well, you can think whatever you want; that's the beauty of being close minded and ignoring physics...

[freeagent]

AIOs are throw away units, they have a finite lifespan, usually 2-3 years. They are e-waste imo.

[fonzz1e]

custom loop standard d5 pumps are rated for 50k hours at max speeds, that more than 5 years

 

in terms of flowrate my d5 at 40% speeds is still more than double of what most aio pump do at 100%

 

for temps custom loop usually has 2 rads or more, that's a lot more surface area for heat exchange