Liquid Cooling vs. Air Cooling Benchmark In-Depth (NH-D15, NZXT X62, & More)

We're answering the question of whether "liquid cooling or air cooling is better," ft. the Noctua NH-D15, Deepcool Assassin III, NZXT Kraken X62 / X72 (280mm, 360mm), and more.

In this video, testing focuses on big air coolers versus big liquid coolers. We're testing for time-to-max temperature, noise-normalized thermals at 35dBA (and 40dBA), max fan speed thermals, and talking about compatibility, ease-of-installation, reliability and uptime, and liquid cooler failure rate and lifespan.

Thermal benchmarking for this type of content requires a lot of testing and retesting. Each cooler underwent a minimum of 4 validation passes for its result (averaged, outliers retested if present), with each cooler requiring two tests per CPU, so that's a minimum of 16 test passes per device. This kind of work costs a lot for us to do -- totaling in the days, and thousands of dollars of time -- and watching this content is the best way to support it. Other ways are linked below. For this testing, we wanted to definitively provide the real answer for "should I buy air cooling or liquid cooling?" questions that we so commonly get. The answer, ultimately, is that it depends, but we're going to give you the most detailed "it depends" possible. Our goal is that you can make a properly informed decision now and in the future, as this content won't age much. A lot of people are also curious about if liquid cooler failures really are that common, or what the risks are with a closed-loop liquid cooler. For the most part, they're fine, but the ones that are bad really make the rounds in media. We talk about that in the video. For PC building with AMD Ryzen CPUs, like the AMD R9 3950X or R7 3800X, we tested air coolers and liquid coolers on Ryzen 3000 products to help builders pick the right parts. We also have Intel HEDT system testing for a larger IHS analysis.

To support us even more than just watching the video, consider these options:

TIMESTAMPS

00:00 - Liquid vs Air
03:23 - Time-to-Max
06:00 - AMD R9 3950X Noise-Normalized
10:48 - Intel HEDT
13:18 - Coldplate Design
14:59 - AMD R9 3950X 100% Fan
15:58 - AMD R7 3800X Noise-Normalized
17:46 - AMD R7 3800X 100% Fan
18:55 - Conclusions

** Please like, comment, and subscribe for more! **

Links to Amazon and Newegg are typically monetized on our channel (affiliate links) and may return a commission of sales to us from the retailer. This is unrelated to the product manufacturer. Any advertisements or sponsorships are disclosed within the video ("this video is brought to you by") and above the fold in the description. We do not ever produce paid content or "sponsored content" (meaning that the content is our idea and is not funded externally aside from whatever ad placement is in the beginning) and we do not ever charge manufacturers for coverage.

Follow us in these locations for more gaming and hardware updates:

Editorial, Testing: Steve Burke
Video: Josh Svoboda, Andrew Coleman, Keegan Gallick

[meejmuas]

I love Linus for the entertainment but I tend to make my buying decisions by watching this channel.

[SilveradoNL]

Steve is right, the fight between air/liquid cooling is stupid.
Passive cooling is obviously best!

[BubbaFett]

Just say it: "liquid is pretty and cool, but good air coolers get the job done for the life of the PC with zero maintenance."

[radicalxedward]

Wait a minute... you’re saying just setting the RGB to blue _doesn’t_ make my system run super cool? What if I set it to ice white?

[FritzVanZyl]

I just love the no nonsense approach, the techno talk rap in all monotone, the sensory overload of information and factual instinct, the effort, dedication and time to create these master pieces, .. I just love GN ;)

[TheGamerUnknown]

0:00 - Introduction
0:51 - Conductonaut sponsor
1:18 - Test methodology and more intro
3:22 - First test ("Time to max", aka time to steady state temperature)
6:00 - Ryzen R9 3950X CPU (16 core) fixed frequency/voltage test 35DBA noise normalized temperatures
10:58 - Intel HEDT Noise Normalized Thermals test (40DBA)
13:19 - Coldplate discussion
14:59 - 100% fan speed test
15:58 - R7 3800X 35 DBA normalized 35DBA noise normalized temperatures
17:45 - R7 3800X 100% fan speed test
18:55 - Liquid Vs. Air discussion
21:52 - Liquid cooler failure discussion
27:25 - Conclusion on Air Vs. Liquid

[shootinbruin]

"The level of emotion people seem to express in their angry internet comments about why someone else is an idiot for using a cooler that this particular person didn't use . . ." How I wish this kind of insanity was only with CPU coolers and not seemingly EVERY SINGLE THING on the internet

[fireday]

"Not included in these tests but it is pink" That's pretty great.

[dirkstrider]

It just dawned on me, GN feels more like a news source than anything I see on TV. every topic they cover is so thorough and whenever disclaimers are needed they cover them completely. Not to mention this man is basically a news anchor, but he is nowhere near as insufferable to listen to as people from "real" media sources. in summary, keep up the good work yo, keep setting a outstanding example for what journalism in general should be.

[uss_]

“Some people just want a blood bath between Air and Liquid”
I guess you can say, some people just want to see the world Thermal Throttle”

[brianjackson]

Attention to detail here is impressive. I like the blue bars on the edges of the charts that let us know how long we have to view it.

[shadowpuppet]

I've had my Noctua NH-D14 since 2011& Noctua sends me free mounting hardware if i need it for a new socket....I have NO complaints!

[bardofhighrenown]

My biggest take away from this is that a premium tower cooler is roughly equivalent to a decent 240mm AIO in terms of thermal performance, which is great because I was trying to figure out if I should get an air cooler or a 240, turns out it's mostly irrelevant and I can get whatever I want.

[stephenwakeman]

"The fan should actively not suck" - I'm pretty sure the fan should actively suck AND blow 😂

[JohnSmith-hfkh]

This is the video we needed. You and your team are our heros, Steve!

[patoenojado]

"Air coolers have one thing that needs to actively not suck"

Or in the case of a pull-push config, it needs to actively suck and blow.

[nathanmontgomery]

Steve over here just quietly making 2 of Ltt's videos totally pointless lol

[riot]

Those timing bars on the side for the charts are the best! Wish more would do this.

[Seedyrom]

These videos trigger my PTSD from university. Steve is like that lecturer who talks at a million miles an hour, and starts rubbing the blackboard off just after he’s finished talking.

[ygny]

CPU consumes less than 100 watts, must liquid cool it, GPU consumes 250 watts, two fans is fine.