Top-Mounted AIO Isn't Always Best... Here's Proof

[00:00:00]Hey everyone. If you have an all-in-one cooler,  you might have wondered where the radiator should go. You know, should it be at the top of your  case, on the side, or at the front? And then, should the fans be set to intake or exhaust? Well,  I've spent a bunch of time testing this, and many people would say you should put the radiator on  the top of your case because warm air rises, but that's not actually correct. Sometimes mounting  the radiator on the side can actually be best, but only if you're then really careful about  the direction of all your other fans. So, the general way I've tested in this video is  that I've got test computer in my garage with a Corsair 360 mil all-in-one cooler. And then  I'm using the Corsair Frame 4000D case because that allows you to put the radiator on the top or  the front or the side. And for all of my tests, I've run Furmark and OCCT just to max out  the CPU and the GPU and really drive up those temperatures. And I know that the Ryzen  3600 that I've used in my test computer doesn't actually need an all-in-one cooler, but  physics is physics, and everything I cover in this video will apply to you, even if you've got  a really high-end CPU. But with all that said, let's finally dive into the results so we can  see the best mounting position for our radiators.

[00:01:03]And for this first test, I'm going to start  off with the all-in-one radiator mounted to the top. And this obviously has three 120 mil  fans since this is a 360 mil cooler. In terms of the air flow, I've also installed  two front intakes, three side intakes, and then one exhaust. And the radiator fans  are set to exhaust for this very first test, which is probably fairly normal for a gaming  computer. And the air flow is all fairly standard, too. We can see the air coming in through the two  front intakes. And this air is then being bought across the motherboard, you know, to cool the RAM  and the VRMOS. And then the GPU will also benefit from having this cool intake air coming in as  well. The side intake fans are also going to be bringing cool air in. And then this is again going  to cool like the CPU and the you know VRMOS and the motherboard in general. And then that air is  going to be exhausted out the back as we've seen so far. And you might have seen that there's  a little bit of air turbulence happening here.

[00:01:50]And this can happen when you have two sources of  intake fans really close together. In other words, the air from the front and the side intake fans  are clashing together a little bit. But I don't think this is a major issue to be honest. You  know, air flow can get a bit complicated and I'll come back to this point throughout the video.  But the reason I say it's not a bad thing is that if we look at the temperature results, we can see  that things are fairly healthy. So when adjusting for the ambient temps, both the CPU and the  GPU are below 50° C. The VRMOS is 35°C and our NVMe drive is at 25.52. And while, as I say,  these are ambient adjusted temperatures. If you actually look at the absolute temperatures, the  CPU and the GPU are actually doing fairly well, and they below 70° C, too, despite being maxed  out. So, that's a good starting point and a good baseline, but I've actually had some comments  in the past from people wondering whether it's good to have as many sort of sources of intake  air as possible near a radiator. In other words, if I flip the rear exhaust around and make it an  intake instead, surely this will benefit the CPU temps, right? Well, the actual air flow looks  fairly normal with intake air coming in and then this is exhausted out the top of the case  as expected. But it does also look like there's a bit more air turbulence happening around  the case with much of the intake air moving around a little half-hazardly in my opinion. The  rear intake fan though is cooling the radiator straight away. So that looks good. But how does it  affect temperatures? Well, it's not great. While the extra intake air is benefiting the motherboard  components with 5° less in those, both the CPU and the GPU are performing worse. And granted, this  is within the margin of error, but I do rerun these tests multiple times just to make sure that,  you know, a one-degree difference is a legitimate one-degree difference. And I think what's  happening here is that taking away exhaust air flow has just been a bit of a negative overall.  And I think that probably makes sense because we already had enough intake air in the first test.  So removing a source of exhaust air isn't actually going to help the main components much at all.  Especially if the remaining air is then like forced through the dense radiator fins and out of  the case instead of just coming out of a standard open case fan slot. You know, in terms of static  pressure, it's always better for air just to go out of an open slot than to go through an actual  radiator with fin stacks and everything else.

[00:04:01]Next up, what happens if we flip the order of the  fans on the radiator? So instead of exhausting air out of the case, they intake air through the  radiator instead. Well, you could argue that since there's more intake air coming into the  case, this will then clearly benefit the GPU and the motherboard temps. But of course, we do still  need some exhaust air. So I did flip the rear fan back to an exhaust. And in terms of the air flow,  the front and the side intake fans do naturally continue to bring some fresh air into the system,  and this is then exhausted out the back of the case as normal. We can also see that the radiators  are now bringing fresh air into the case, but there are also signs of air turbulence,  which again is not surprising considering we have so much intake air. Some of this air is  also going straight out the back of the case, which is a bit pointless to be honest. And looking  at the temps, this is the worst config so far.

[00:04:46]Both the CPU and the GPU are worse off with this  configuration, particularly with the graphics card. And this probably makes sense because the  graphics card has loads of warm air that's being pumped out the back of it and then that warm  air is sort of clashing with the cool air that's coming in through the radiator. So in other words,  you know where the motherboard is, you've just got all this air clashing together. And yes, I know we  got that rear exhaust at least, but there is still a lot of turbulence happening here. But before  moving on, it's quickly worth comparing test one with test three, the one we just done, since  these had the same fan configs, but just with the radiator set to intake versus exhaust. And  clearly for the top-mounted fan radiator position, it's better to set your radiator fans to exhaust.  This is clearly helping both the CPU and the GPU temps. At this point now, I wanted to test  other mounting positions for the radiator. So, the front and the side and not just at the top.  I could have also moved the fans to the other side and then explored push-pull configurations,  but to be honest, at this point, there'd then be so many different like permutations of tests that  this video would turn out to be like 2 hours long.

[00:05:46]But I will explore push-pull configs in a future  video. So stay tuned for that. So for this next test, I've installed the radiator in the front and  I'm temporarily skipping the top fans altogether just to keep things simple, but I will add them  back later. I've also set the radiator fans to be intake. So we have six intakes and one exhaust  and the air flow is fairly good. As you can see, you know, the air at the front of the case is  being bought through the radiator fins. And this cooler air is then helping to cool the GPU  and the motherboard components before the air is exhausted out the rear. The side intake fans are  also acting in the same way with the fresh air coming in and then being exhausted out the case as  we'd expect. And looking at the temperature data, victory. We have the lowest CPU temperatures so  far. A massive 1.2° C different. Okay, that's not massive, but at least there is a difference there.  Unfortunately, the GPU and the VRMOS is slightly higher, but at least we can see that the whole hot  air rises, you know, argument kind of breaks down here because we've moved the radiator away from  the top and had a slight improvement in the CPU temps. And again, I run all these tests multiple  times. So, this isn't just sample variation.

[00:06:49]But for the next test, obviously, we need to check  for the front mounting positions. Is it best to have those fans set to intake or exhaust? Well,  to test that, I removed the radiator and I flipped the fans around. So, now putting that back, we  have three intake fans on the side and a total of four exhaust fans. One at the back and then three  in the radiator. And as we can see, the radiator is definitely acting as an exhaust with air being  blown away here. The side fans are now the only source of active intake air. But unfortunately,  most of this air is being sucked straight out of the front of the case, which probably doesn't bode  well for the actual temperature data. Even though admittedly, you know, some of this air is coming  out the rear, it's still not a really good sign, and the data proves that. So, everything is higher  than in our previous test. And while it is only 1 or 2° C, clearly in this configuration, our system  is being starved of fresh air flow by having much of that side intake air just being sucked straight  out of the case. again. But at this point, surely if we just add more intake fans, that's  going to solve the problem we just saw, and that's going to benefit temperatures throughout, right?  So, I got to work installing two new intake fans at the top of the case, meaning that we now have  five intake fans in total and then four exhausts.

[00:07:56]And it does look like these new intake fans are  helping with more air moving around the case and then being exhausted out the back of the case  than before. So, that looks good. The side intake fans are continuing to mainly bring cool air in  that then immediately cools the radiator and that gets exhausted out the case. But maybe, you know,  actually having those top intake fans is going to help some of the other temperatures, right? Well,  unfortunately, the temperature data doesn't really back that up. Everything is worse, sometimes  decently so with the GPU and the VRMOS being 3 to 4° higher and the CPU being around 2 1/2° higher.  And while this might not seem to make immediate sense, it's worth going back to what I said  earlier about air turbulence. Simply adding more fans doesn't necessarily mean lower temperatures.  Sometimes it just means that more air is clashing around inside the case, which can then end up  negatively affecting the natural or more optimal flow of air. In other words, sometimes just having  air come in the front of the case, flowing over the components, and then being exhausted out  the back can be best. It's simple and it just works. But what about a side-mounted radiator  position? You know, surely that's going to be easier and more consistent and get just better  results overall, right? So, for this next test, I set the radiator fans to exhaust out the case,  leaving us with five intake fans and four exhaust in total. And some of the air flow is moving  across the case and then cooling the graphics card before being exhausted out the back of the  case, which looks promising. Some of the air is also being exhausted out the radiator. And you  can see the air floating around behind the case, but it's actually a bit confusing from that camera  angle. Sorry, but I could actually see, you know, the air was just aimlessly floating around,  which is actually what we'd expect. The fact that there's then two intake fans nearby means  that some of this warmer exhaust air is then being sucked back into the case, which is not ideal. And  the reason I set those to intake and not exhaust was firstly so I could directly compare to the  previous test. But secondly, if they were exhaust, the front intake air would then immediately be  sucked out the case. So basically, you know, fan configurations can have lots of weird  nuances and that's why I keep saying that less can be more. Anywhoo, looking at the temperature  results, it's actually not that bad. Thankfully, it's only far ahead of our last test with most  of the figures here being a few degrees or more lower and the VR MOS being 7° lower, which is good  to see. And while that clearly is a good thing, this still isn't our best result overall.  You know, we had some front and top results which were better. So, can we actually salvage  side-mounted radiators to get better results?

[00:10:16]And that brings us to our next test, which might  seem a bit barmy, but I wanted to see what would happen if we just had less intake fans overall  since I could tell that, you know, air turbulence was a bit of an issue. So, I was wondering if it  was an issue as well in that previous test. So, I unplugged the front intake fans, meaning  that we ended up with just two intakes and four exhausts. And in terms of the air flow, we can  see that the front intakes are now doing nothing, of course. So, instead, air is only really being  drawn in through the top of the case. And this is then helping to cool the GPU and the motherboard.  And then this air is exhausted out the, you know, front and the rear of the case. And while  this is clearly an unconventional setup, at least there's no obvious signs of, you know,  air turbulence and air clashing and things like that. But unfortunately, the temperature results  were bad. The worst so far, with most of the results being 5 to 6° C worse when adjusting  ambient temperatures. So, that was a failed test.

[00:11:03]But let's actually go back to a more standard  test config so we can actually look at the last per one test and directly compare those. And as  a reminder, that particular test had five intakes and four exhausts. So for this new test, I then  installed three intakes in the front. And then I switched the top fans to exhaust and then move  them backwards a bit so that they'd actually be exhausting properly. This means that we then have  six intakes and three exhaust, which isn't the exact same fan config as the last but one test,  but it's the same overall number of fans, and it's the best for head-to-head comparison so that we  can then better test whether it's best to have the radiator fan set to intake or exhaust. And looking  at the air flow, the front intake fans do appear to be helping with much of the air blowing  across the case, helping to cool the VRMOS, the NVME drives, the RAM, and and all of that  along with the GPU. And then this air is being exhausted out the top and the back of the case.  There also thankfully doesn't seem to be much air turbulence caused by the fact that the side  all-in-one cooler radiator fans are set to intake and not exhaust. Oh yeah, before I forget, I know  the Grogu isn't here right now, but don't worry, I did test with and without Grogu and it didn't  affect thermals whatsoever. I just wanted to say that just to pre-empt any possible comments. Okay,  thanks. Bye. Anyway, looking at the thermals of this test, it's actually one of the best results.  The ambient adjusted CPU temps are below 45° C, while the GPU temps are below 48°. These two  are actually some of the best results we've had, even if the NVMe and the VRMOS ones are slightly  higher. Overall though, having a more standard fan config does seem to have benefited things here.  And when the radiator is mounted on the side, it seems clear that the fan should be set to intake  because this delivered lower CPU and GPU temps.

[00:12:43]And with all these videos, I do get lots of  comments saying, you know, why didn't you test this specific thing? But yeah, in this particular  video, I just wanted to actually test with, you know, mounting the radiator on the front  and the side and the top because that's going to cover like 95% of people's actual setups. But  please do leave suggestions in the comments for what you'd want to see in a future video. And  as I mentioned, I will look at push pull in a future video, too. But before wrapping up, I  had two main things to talk about. So, firstly, if you want to put your all-in-one radiator in the  front or the side of the case, then in general, setting the fans to be intake is probably going  to be best. And yes, this can vary depending on your case and your hardware. But this seems to  be the best overall because the radiator gets cooled quite quickly. In other words, it's  going to deliver lower CPU temps. Whereas, if you were thinking of mounting your radiator  in the top, then setting the fans to exhaust will probably work well because all that warm air  will just be pushed out the case. However, my second point is it's never that simple and there's  always going to be trade-offs. With a top-mounted all-in-one cooler, the warm GPU air will be  pushed through the radiator. So, in other words, in a top-mounted config, the GPU is probably  going to benefit more than the CPU will. While with a side or a front- mounted all-in-one cooler,  the CPU will get cooled immediately, but then the air inside the case is going to be warmer, and  that's then going to be worse for the GPU. So, if your priority is gaming and having a cooler  graphics card, then mounting your radiator to the top might actually work out best, even if this  leads to slightly higher CPU temps. But really, a lot of these results are quite close to  each other. So, as long as you, you know, do fairly normal things, then just try not to  worry about it too much. You know, you're not going to see a massive difference. And sometimes  the mounting position can come down to your own preferences and your case aesthetics and things  like that. But I'd love to hear from you down in the comments. So, let me know what you think of  all of this. And if you enjoyed this video, please click the thumbs up button and please subscribe  to see more videos like this. Please watch him.