Where to mount the AIO radiator? .. my findings

NorKris · April 14

Hi guys, back with more test results

The goal: where is it best to place the AIO-radiator, in this testing i was assuming the person did not buy any additional case fans. more on this later.

About the tests:
- 10m runs of cine R23 multi + 720p Furmark

- Radiator Fans: 1500 RPM (fixed)

- case fans: 1250 RPM (fixed)

- Room Temp was measured about 30cm from the PC case with a probe

- System power was taken with a "in line power meter" ( power plug with read outs)

- the case is a "Corsair 4500D" (read the Q/A for more info)

- every config is run'd 3 times to eliminate strange spikes or other strange behaviors.

- the numbers in the result is a average of 3 runs for every config.

- Setup 7 was run without the top mesh due to how restrictive this was, more testing on this later ( but this is a case specific thing ofc)

the results:

Here is a fast Q/A of things you guys might wonder about

Spoiler

Why only “push” on the fans?

To make the only variables the rad location

Why the same fan speeds?
- to make the only variables the rad’s location (or in vs out)

Why only run push on the fans?

- to make the only variables the rad’s location (or in vs out)

Why buy and mod a 4000D instead of buying a 5000D?

The 4000D was avaiable in a local small pc store, so fast delivery + i wanted to support my local store,

What is the Mod?

- Side mount for radiator and mesh on the sidepanel, i named it a 4500D (dont sue me corsair )

Why run furmark (not on full resolution) on the graphics card?

To simulate a gaming load

Why only run 2 case fans and 4 in total?

- using the fans from the case and the fans from the AIO was a combo i found to be a realistic scenario for many if not most users.

What CPU was used?

- 7940x OC’d to 4,1 Ghz

What cooler was used?

- Cooler master ML240L

What GPU was used?

- A GTX 680, i didn't have any newer, sry

Why 10m Cine r23 and furmark test?

- i found that with a 240 AIO at this heat load that I'm reaching equilibrium after one cine run.

How is the intakes?

Every intake has a mesh of some kind

Why is the CPU watt just “approximately”?

- my mother board has never been available to report CPU watt/power

What's up with the arrows in the pictures?

- it's to indicate if are has touched/moved over a heat producing component. (blue for none, purple for one and red for two)

Why is the heatload on the CPU so high?

- To eliminate margin of error and to highlight the "extremes abit", and cpus are getting hotter and hotter...

here is some pics of the "4500D"

Spoiler

my toughts:

1. some of these are inline with what i have seen from other testers

2. the temps for top exhaust(8) is higher than i suspected. and this can be down to 2 things:

my test setup being limited in these runs to case fans + what u get with an AIO. i want ur guys inputs on how many fans u want me to test with next time is it another number u find to be more realistic?

or it could be how restrictive the top mesh is on the 4000D, cuz when i ran config (7) with the mesh i got 100c after 2m... so i had to run it without.

i hope you enjoy'd the reading and apricate the testing. plz stay on topic this time guys

IkeaGnome · April 14

33 minutes ago, NorKris said:

cuz when i ran config (7) with the mesh i got 100c after 2m...

In your chart it says the max temp was 93.3c though….

NorKris · April 14

13 minutes ago, IkeaGnome said:

In your chart it says the max temp was 93.3c though….

yes, i had to take the mesh at the top intake off, i will put that inn the Q/A

for the next round im going to do these top with and without mesh against each other . that thing is very very restrictive

RevGAM · April 14

Very interesting! When I'm done crunching all my results, I'll be very interested to see the differences, especially since our PCs are different in many ways, such as 240 vs 360, 4500D vs 5000X, OC vs no OC, and, well, just about everything else.

Have you checked yet to see which other testers results match/ differ from your results? That would be interesting to know, especially if you can analyze what might have caused different results.

RevGAM · April 14

What are the specs for the fans?

NorKris · April 14

5 minutes ago, RevGAM said:

What are the specs for the fans?

sadly i cant find very much

found this tho

https://www.aliexpress.com/i/1005005182450518.html

RevGAM · April 14

5 hours ago, NorKris said:

sadly i cant find very much

found this tho

https://www.aliexpress.com/i/1005005182450518.html

Maybe this?

120mm fan (A12025-25RB-2IN-F1). This Chinese fan is rated 12V 0.16A and we know the 15RB in the same series is actually a tuned Delta rebrand. This new 25RB fan is rated to produce 52CFM airflow at a maximum speed of 2,500 rpm. Noise levels are officially rated to 29.7dBa. It uses an LDB bearing.

Or this?

Silencio FP A12025-25RB-2IN-F1. It comes from the same product line as the Cooler Master Silencio FP 120 PWM Performance Edition . Further research indicates the A12025-25RB-2IN-F1 is a fluid dynamic bearing fan specified at 0.16A for a maximum of speed of 2250 rpm. Fans with fluid dynamic bearings generally have much longer lifespans compared to sleeve bearing fans.

Or?

Silencio A12025-25RB-2IN-F1
(120mm, 12 V, 0.16 A, 2250 RPM, Loop Dynamic Bearing)

Output %	Fan speed (RPM)	Temperature intake/ outtake	Noise (dBA)
4.2	552	18 °C/ 21 °C	39.2
20	555	19 °C/ 22 °C	39.2
40	561	19°C/ 25 °C	39.2
60	567	20 °C/ 29 °C	39.2
80	662	22 °C/ 34 °C	39.4
100	943	22 °C/ 39 °C	39.4
CL 16	603	21 °C/ 35 °C	39.3
CL 21	572	21 °C/ 35 °C	39.2
CL 100	913	21 °C/ 38 °C	39.4
OL 155	2292	22 °C/ 47 °C	44.4

NorKris · April 15

8 hours ago, RevGAM said:

Maybe this?

120mm fan (A12025-25RB-2IN-F1). This Chinese fan is rated 12V 0.16A and we know the 15RB in the same series is actually a tuned Delta rebrand. This new 25RB fan is rated to produce 52CFM airflow at a maximum speed of 2,500 rpm. Noise levels are officially rated to 29.7dBa. It uses an LDB bearing.

Or this?

Silencio FP A12025-25RB-2IN-F1. It comes from the same product line as the Cooler Master Silencio FP 120 PWM Performance Edition . Further research indicates the A12025-25RB-2IN-F1 is a fluid dynamic bearing fan specified at 0.16A for a maximum of speed of 2250 rpm. Fans with fluid dynamic bearings generally have much longer lifespans compared to sleeve bearing fans.

Or?

Silencio A12025-25RB-2IN-F1
(120mm, 12 V, 0.16 A, 2250 RPM, Loop Dynamic Bearing)

i think it is the same fan as in this aio https://www.coolermaster.com/catalog/coolers/cpu-liquid-coolers/masterliquid-lite-120/

RevGAM · April 15

2 hours ago, NorKris said:

i think it is the same fan as in this aio https://www.coolermaster.com/catalog/coolers/cpu-liquid-coolers/masterliquid-lite-120/

That same model number is printed on them as on the AIO's fans?

RevGAM · April 15

18 hours ago, NorKris said:

Hi guys, back with more test results

The goal: where is it best to place the AIO-radiator, in this testing i was assuming the person did not buy any additional case fans. more on this later.

About the tests:
- 10m runs of cine R23 multi + 720p Furmark

- Radiator Fans: 1500 RPM (fixed)

- case fans: 1250 RPM (fixed)

- Room Temp was measured about 30cm from the PC case with a probe

- System power was taken with a "in line power meter" ( power plug with read outs)

- the case is a "Corsair 4500D" (read the Q/A for more info)

- every config is run'd 3 times to eliminate strange spikes or other strange behaviors.

- the numbers in the result is a average of 3 runs for every config.

- Setup 7 was run without the top mesh due to how restrictive this was, more testing on this later ( but this is a case specific thing ofc)

the results:

Here is a fast Q/A of things you guys might wonder about

Reveal hidden contents

Why only “push” on the fans?

To make the only variables the rad location

Why the same fan speeds?
- to make the only variables the rad’s location (or in vs out)

Why only run push on the fans?

- to make the only variables the rad’s location (or in vs out)

Why buy and mod a 4000D instead of buying a 5000D?

The 4000D was avaiable in a local small pc store, so fast delivery + i wanted to support my local store,

What is the Mod?

- Side mount for radiator and mesh on the sidepanel, i named it a 4500D (dont sue me corsair )

Why run furmark (not on full resolution) on the graphics card?

To simulate a gaming load

Why only run 2 case fans and 4 in total?

- using the fans from the case and the fans from the AIO was a combo i found to be a realistic scenario for many if not most users.

What CPU was used?

- 7940x OC’d to 4,1 Ghz

What cooler was used?

- Cooler master ML240L

What GPU was used?

- A GTX 680, i didn't have any newer, sry

Why 10m Cine r23 and furmark test?

- i found that with a 240 AIO at this heat load that I'm reaching equilibrium after one cine run.

How is the intakes?

Every intake has a mesh of some kind

Why is the CPU watt just “approximately”?

- my mother board has never been available to report CPU watt/power

What's up with the arrows in the pictures?

- it's to indicate if are has touched/moved over a heat producing component. (blue for none, purple for one and red for two)

Why is the heatload on the CPU so high?

- To eliminate margin of error and to highlight the "extremes abit", and cpus are getting hotter and hotter...

here is some pics of the "4500D"

Reveal hidden contents

my toughts:

1. some of these are inline with what i have seen from other testers

2. the temps for top exhaust(8) is higher than i suspected. and this can be down to 2 things:

my test setup being limited in these runs to case fans + what u get with an AIO. i want ur guys inputs on how many fans u want me to test with next time is it another number u find to be more realistic?

or it could be how restrictive the top mesh is on the 4000D, cuz when i ran config (7) with the mesh i got 100c after 2m... so i had to run it without.

i hope you enjoy'd the reading and apricate the testing. plz stay on topic this time guys

I think that the main reason for the difference in results between your tests and mine (which show that side-mounted is the best), and this is important for people to keep in mind, is airflow. I cannot compare the AIOs very effectively since the fans are similar but the page for the AIO does not give any useful pump specs. I tracked clocks, but not watts. @TatamiMatthas agreed to help me by making the graphs, so I hope he'll be able to dig into this more deeply. Thanks, MatMatt!

Some things stand out in regards to airflow:

glass panels on the front and top of the 5000x but not on the 4500D
tubes disrupting the airflow (speculative)
I used a large 3-fan GC (XFX RX 6800 XT) on an EZDIY-FAB adjustable vertical mount that sits on the PSU shroud
- Kris, which type of GTX 680 do you have: blower, 2 fan or 3 fan?

One of the main differences in my 20 tests using Prime 95, CB24, and Time Spy is I tested in both push and pull configs, tubes up and down (but only tubes forward on top), and intake and exhaust on each location, and the only thing I did with the case was to remove the dust filter when exhausting AIO air from that position. Keep in mind that the 3 filters are not identical for whatever reason.

The reason for the glass panels not being removed was that many people have pets and some pets are too inquisitive or destructive, so the glass panels are a great barrier to allow airflow without giving pets access. When I had a cat, these panels were a boon because he would've scratched the dust filter to ribbons. Thus, I successfully simulated some types of users - those who don't remove the glass panels (for whatever reason), and those who would remove a dust filter slowing down exhausting air. Of course, I will not be doing another 40 tests (I plan to do another 20 tests

In my testing, without exception, the side mount was the best option. I am convinced by Kris's results that, if I were to do the testing without one or both of the glass panels, that the results would likely have been different, and possibly in favor of front-mounting.

One difference that I will note that I think is significant, and that I expanded upon at home (where I post results first), is the fact that when I looked at the CB24 and Time Spy results, they told a DIFFERENT story. Front and top, in that order, DOMINATED in both benchmarks.

The Best Configurations for CPU-Oriented Software (by Scores & FPS)

TIME SPY CINEBENCH 2024 Multi-Core CPU Test

1. Front mount, tubes down, intake (#8 for the CPU temps) Front mount, tubes up, intake (#4 for the CPU temps)

Time Spy CPU Score: 15,607, 52.44 FPS CB24 Score: 1,213

2. Top mount, tubes forward, intake (#7 for the CPU temps) Top mount, tubes forward, exhaust (#6 for the CPU temps)

Time Spy CPU Score: 15,577, 52.33 FPS CB24 Score: 1,194

3. Side mount, tubes up, intake (#3 for the CPU temps) Front mount, tubes down, intake (#8 for the CPU temps)

Time Spy CPU Score: 15,463, 51.95 FPS CB24 Score: 1,191

4. Side mount, tubes up, exhaust (#1 for the CPU temps) Front mount, tubes down, exhaust (#10 for the CPU temps)

Time Spy CPU Score: 15,413, 51.78 FPS CB24 Score: 1,190

5. Front mount, tubes down, exhaust (#10 for the CPU temps) Side mount, tubes up, exhaust (#1 for the CPU temps)

Time Spy CPU Score: 15,340, 51.54 FPS CB24 Score: 1,186

The Best Configurations for GPU-Oriented Software (by Scores & FPS)

TIME SPY Graphics Score with FPS

1. Front mount, tubes up, intake (#6 for the GPU temps)

Score: 20,136: 131.99 (#2) & 114.87 (#2) FPS

2. Top mount, tubes forward, exhaust (#4 for the GPU temps)

Score: 20,130: 132.04 (#1) & 114.77 (#3)

3. Side mount, tubes up, exhaust (#2 for the GPU temps)

Score: 20,122: 131.78 (#4) & 114.89 (#1)

4. Front mount, tubes up, exhaust (#7 for the GPU temps)

Score: 20,106: 131.63 (#3) & 114.59 (#5)

5. Top mount, tubes forward, intake (#10 for the GPU temps)

Score: 20,100: 131.68 (#5) & 114.72 (#4)

The strangest thing about these results is that they seem largely divorced from temperatures in the CPU, GC, mobo and VRMs. This tells me that, as long as parts aren't overheating, temperatures are not the most important factor!

I am not allowed to link to my results so anyone who is interested will need to PM me to request the link. Sorry.

I think that this covers the points I wanted to share.

@NorKrisWhich GC model? Were the tubes up or down?

NorKris · April 15

5 hours ago, RevGAM said:

I think that the main reason for the difference in results between your tests and mine (which show that side-mounted is the best), and this is important for people to keep in mind, is airflow. I cannot compare the AIOs very effectively since the fans are similar but the page for the AIO does not give any useful pump specs. I tracked clocks, but not watts. @TatamiMatthas agreed to help me by making the graphs, so I hope he'll be able to dig into this more deeply. Thanks, MatMatt!

Some things stand out in regards to airflow:

glass panels on the front and top of the 5000x but not on the 4500D

tubes disrupting the airflow (speculative)

I used a large 3-fan GC (XFX RX 6800 XT) on an EZDIY-FAB adjustable vertical mount that sits on the PSU shroud

Kris, which type of GTX 680 do you have: blower, 2 fan or 3 fan?

One of the main differences in my 20 tests using Prime 95, CB24, and Time Spy is I tested in both push and pull configs, tubes up and down (but only tubes forward on top), and intake and exhaust on each location, and the only thing I did with the case was to remove the dust filter when exhausting AIO air from that position. Keep in mind that the 3 filters are not identical for whatever reason.

The reason for the glass panels not being removed was that many people have pets and some pets are too inquisitive or destructive, so the glass panels are a great barrier to allow airflow without giving pets access. When I had a cat, these panels were a boon because he would've scratched the dust filter to ribbons. Thus, I successfully simulated some types of users - those who don't remove the glass panels (for whatever reason), and those who would remove a dust filter slowing down exhausting air. Of course, I will not be doing another 40 tests (I plan to do another 20 tests

In my testing, without exception, the side mount was the best option. I am convinced by Kris's results that, if I were to do the testing without one or both of the glass panels, that the results would likely have been different, and possibly in favor of front-mounting.

One difference that I will note that I think is significant, and that I expanded upon at home (where I post results first), is the fact that when I looked at the CB24 and Time Spy results, they told a DIFFERENT story. Front and top, in that order, DOMINATED in both benchmarks.

I think that this covers the points I wanted to share.

@NorKrisWhich GC model? Were the tubes up or down?

old 680 a blower style with the shroud off to simulate the more modern form of Graphics cards. kicking heat into the system is the norm now

Tubes up always, 2 reasons:

i dont belive it to make any difference

im after 2 variables only: where i the rad and what way i the air flowing

thrasher_565 · April 15

41 minutes ago, NorKris said:

old 680 a blower style with the shroud off to simulate the more modern form of Graphics cards. kicking heat into the system is the norm now

Tubes up always, 2 reasons:

i dont belive it to make any difference

im after 2 variables only: where i the rad and what way i the air flowing

tubes up vs down was mostly for custom loops because you didn't want to trap the bubble in the rad and try and push it out to the res.

in a good filled aio wont make a difference how ever some that are not filled right it would be best to have it tubes down and the rad hier then the cpu block so you could trap the air bubble in the rad and not thow the cpu block witch cause problems over time.

NorKris · April 15

2 minutes ago, thrasher_565 said:

tubes up vs down was mostly for custom loops because you didn't want to trap the bubble in the rad and try and push it out to the res.

in a good filled aio wont make a difference how ever some that are not filled right it would be best to have it tubes down and the rad hier then the cpu block so you could trap the air bubble in the rad and not thow the cpu block witch cause problems over time.

agree, its just that in MSI's testing ( the same folks that sells 4-6 months lasting AIO's) tubs down is much better than up

the only reason for this was probably that their aio was so clogged up that i didn't even know if it was mounted or not

thrasher_565 · April 15

2 minutes ago, NorKris said:

agree, its just that in MSI's testing ( the same folks that sells 4-6 months lasting AIO's) tubs down is much better than up

the only reason for this was probably that their aio was so clogged up that i didn't even know if it was mounted or not

ya i dont like msi or coolermaster. even thow msi alot of times has "cool" stuff at the lowest cost...

as an aio loses water then tubes down would still be the best.

the clogs was do to improper liquid additive and its a know problem in aios...

NorKris · April 15

1 minute ago, thrasher_565 said:

ya i dont like msi or coolermaster. even thow msi alot of times has "cool" stuff at the lowest cost...

as an aio loses water then tubes down would still be the best.

the clogs was do to improper liquid additive and its a know problem in aios...

CM of the win!! best120 ARGB fans best 200mm ARGB fans

best case!!

thrasher_565 · April 15

Just now, NorKris said:

CM of the win!! best120 ARGB fans best 200mm ARGB fans

best case!!

i have a biso with them when i buy my first rgb 2 fans out of 6 had bad barings...

Agall · April 15

Radiator positioning really just depends on your objective. That's whether the goal is prioritizing fresh air to the GPU or CPU. Some cases can get both depending on the design, like a 5000D with front and side fans and a well designed fan curve.

I generally stick to the top mount for an AIO, since then the front fans can be biased to provide a fresh air channel to the top radiator and the GPU. Most cases that support two front 140mm fans can do this by just spacing them apart.

It also helps having a well designed exhaust path for the GPU. Something the FD North (mesh) has inherently with a mesh side panel.

NorKris · April 15

1 minute ago, Agall said:

Radiator positioning really just depends on your objective. That's whether the goal is prioritizing fresh air to the GPU or CPU.

true but GPUs are so much more easier to cool

Agall · April 15

2 minutes ago, NorKris said:

true but GPUs are so much more easier to cool

GPUs are generally easier to cool, the die is usually larger providing better heat flux with less power spiking while also not having two extra heat transfer layers of the solder and IHS. They're also usually a lot higher wattage.

People also focus too heavily on operating temperature in my opinion. The extra efficiency granted by running a processor at a lower temperature that's still within operating limits is insignificant, especially if it involves running the fans at a proportionally higher wattage.

3D v-cache being the focal point of this discussion now a days. Before that, it was the era before they used solder to attach the IHS, like the 4770k vs 4790k.

NorKris · April 15

i still need input on how many fans to test for the next round

RevGAM · April 15

40 minutes ago, NorKris said:

i still need input on how many fans to test for the next round

I would suggest that you stick to the same number unless your goals dictate a change.

NorKris · April 15

6 minutes ago, RevGAM said:

I would suggest that you stick to the same number unless your goals dictate a change.

but why would i test the same again?

RevGAM · April 15

1 hour ago, NorKris said:

but why would i test the same again?

Why are you going to do further tests? Why do you want to change the number of fans? If I can understand that, that would help. If you definitely want to use another quantity of fans, by all means do so.

NorKris · April 15

Just now, RevGAM said:

Why are you going to do further tests? Why do you want to change the number of fans? If I can understand that, that would help. If you definitely want to use another quantity of fans, by all means do so.

some ppl might say only 2 case fans are too few

RevGAM · April 15

2 hours ago, NorKris said:

some ppl might say only 2 case fans are too few

True. I just gave you my thoughts, and I see nothing wrong with adding another fan if that's what you think you should do. It might be interesting to use that side mount you created to try fan placement. You could do that with some sort of wind diverter to blow the air towards the mobo, leave it open, or both.

I would personally be interested in seeing you test different fan configurations with an air cooler, too.

Sign In

Where to mount the AIO radiator? .. my findings

Link to comment

Share on other sites

Link to post

Share on other sites

Link to comment

Share on other sites

Link to post

Share on other sites

Link to comment

Share on other sites

Link to post

Share on other sites

Link to comment

Share on other sites

Link to post

Share on other sites

Link to comment

Share on other sites

Link to post

Share on other sites

Link to comment

Share on other sites

Link to post

Share on other sites

Link to comment

Share on other sites

Link to post

Share on other sites

Link to comment

Share on other sites

Link to post

Share on other sites

Link to comment

Share on other sites

Link to post

Share on other sites

Link to comment

Share on other sites

Link to post

Share on other sites

Link to comment

Share on other sites

Link to post

Share on other sites

Link to comment

Share on other sites

Link to post

Share on other sites

Link to comment

Share on other sites

Link to post

Share on other sites

Link to comment

Share on other sites

Link to post

Share on other sites

Link to comment

Share on other sites

Link to post

Share on other sites

Link to comment

Share on other sites

Link to post

Share on other sites

Link to comment

Share on other sites

Link to post

Share on other sites

Link to comment

Share on other sites

Link to post

Share on other sites

Link to comment

Share on other sites

Link to post

Share on other sites

Link to comment

Share on other sites