Air Cooler vs AIO Liquid Cooler – 240mm AIO vs Dark Rock Pro 4
It seems like all-in-one watercoolers like this Asus Ryuo 3 240m ARGB have become the defacto standard for gaming PC builds these days. I know I use them constantly, but with the rats nest worth of wires, the gaudy RGB lighting, the built in displays and the noise, is it actually worth it? Are air coolers actually the better choice? I’m not sure, so let’s run some tests and find out. The air cooler I’ve gone with here is the bequiet Dark Rock Pro 4 – an all black dual tower heatsink with two included fans (and mounting hardware for a third) – which should be a great fit for the Ryzen 7900X I’ll be using to test against.
It’s worth quickly looking at the AIO – this is an asetek design with a pretty massive and hefty pump unit. Built into the top is a matrix of RGB LEDs that Asus calls their “AnimeMatrix”. In a stunning move, the contents of that “display” can’t be rotated, which means if you install the pump “wrong” like I did, which is the correct orientation for every other AIO on the planet and what I’d argue is the best for it’s longevity, you will never be able to read things like your CPU temperature on it without turning your head. Amazing. What’s also frustrating is that they clearly expect you to buy two more fans from them – the RGB splitter cable has four heads on it for up to four fans, but they don’t include any spare mounting hardware for those fans.
The Dark Rock Pro 4, by comparison, is a beautiful example of simplicity and restraint. It’s all black with not an LED in sight. That’s weirdly refreshing to me. Now it isn’t perfect either – if you are installing this in an already built PC it’s a bit of a pain. You’ll need to lay the system on its side and remove your graphics card to be able to fit the central fan back in after you’ve mounted it, but once it is mounted you are good to go. If you are building your system from scratch this isn’t a problem.
Ok let’s do some testing. First off is the most obvious thing, how does each cooler handle CPU loads like Cinebench R23 on loop for 30 minutes? The nice thing about this is it’ll give me the final score too so we can see not only the temperature and power usage, but the outright performance too. Looking at the temperature graph it isn’t looking good for the air cooler. It isn’t a massive difference, only about 4.5°c over the 30 minutes on average, but it is pretty undeniable the air cooler is running hotter. The AIO runs at about 87°c, whereas the air cooler is more like 91°c. Looking at the power graph we can see the air cooler runs a little lower power too, averaging 176W over the 30 minutes, versus 180W from the AIO. This makes sense though – what doesn’t is the cinebench scores. The air cooler ended its run with almost 1000 points more performance than the AIO. Now that is only around 3% more performance, but both of the other results would suggest the air cooler should have had LESS performance, not more. If anyone has any ideas here please do leave them in the comments below.
Of course, it’s pretty rare for people to exclusively do CPU based work on their gaming PC, so what about when you’re gaming? I’ve got an RX 6900 XT in here and the case is fully closed up so lets see if that makes any difference to your temperatures while gaming, or your gaming performance. I went the Hitman 3 here as the built in benchmark can run in an endless loop, and is a little more CPU heavy which makes it more of a worst case scenario for the coolers. So, does it make a difference? No. Not in the slightest. Both coolers sat comfortably at around 65°c basically throughout the run, with very little variation between them. Performance wise, after 30 minutes of heat soaking it was within margin of error. The air cooler was slightly faster, but the room had cooled a by a few degrees by the time I ran the air cooler test so that’s within margin of error.
There are a few other factors you might want to consider though, like noise. The air cooler was practically silent – it was no louder than the rest of the case fans, and while gaming it was easily drowned out by the GPUs fans spinning up. By contrast, the AIO was considerably noisier. I could hear the AIO while gaming, and certainly while rendering. That is something that seems to be a known quantity with these Asus coolers, but any AIO is likely to be at least a little louder as you have at least the same number of fans, but more restriction on them blowing through a tightly packed radiator, and the addition of the water pump too.
The other thing to consider is longevity and reliability. The AIO will fail at some point. The fluid will slowly evaporate and the pump will fail. It might even catastrophically fail and leak all over your other components. The air cooler won’t leave you high and dry – sure the fans might fail, but that’s a lot less likely than the pump failing on the AIO, and it will still act as a decent passive heatsink until you replace them, whereas if the pump dies your CPU will just get cooked. The biggest downside to the air cooler is the weight. If you need to transport your system that can be a big problem, and it’s also just a lot of weight to be hanging from your motherboard.
In short then, for CPU specific tasks an air cooler might struggle to keep up with a closed loop water cooler, but for gaming it doesn’t make much of a difference. The air cooler is likely to be quieter and more reliable, but considerably heavier. If the gaudy RGB is your thing I’m sure you can find some RGB air coolers too, although you’d struggle to find one with a screen on top of it. It’s worth noting too that this Asus cooler is ludicrously expensive. It’s £250. The Dark Rock Pro 4 is a much more reasonable £80, although you certainly can get more reasonably priced (and potentially better) AIOs. The ARCTIC Liquid Freezer II 240mm comes to mind on that one. I’ll link to all of these in the description if you’re interested.
