WEBVTT

00:00:00.000 --> 00:00:08.330
this video on air versus water cooling

00:00:03.720 --> 00:00:13.469
has two million views it looks like crap

00:00:08.330 --> 00:00:15.870
time for an HD remake and it's not just

00:00:13.469 --> 00:00:21.660
a pure sellout move to capitalize on an idea that we already know works because

00:00:18.090 --> 00:00:23.850
a lot more has changed since that video

00:00:21.660 --> 00:00:31.800
then our production values around here in particular AMD's high performance not

00:00:27.689 --> 00:00:34.530
to mention huge and hot 180 watt thread

00:00:31.800 --> 00:00:39.469
Ripper CPUs are pushing the limits of all types of cooling systems in ways

00:00:37.170 --> 00:00:48.570
that we haven't never seen before so what's the ultimate cooler for the

00:00:42.300 --> 00:00:53.380
ultimate CPU we have the answer

00:00:55.790 --> 00:01:01.140
the white dark bass pro 900 from be

00:00:59.160 --> 00:01:05.040
quiet features a modular design to support a variety of layouts and

00:01:02.880 --> 00:01:11.210
configurations check it out at the link below we'll be testing six coolers that

00:01:09.150 --> 00:01:18.600
on previous platforms have delivered comparable results three air and three

00:01:15.090 --> 00:01:20.820
liquid four water coolers MD helpfully

00:01:18.600 --> 00:01:25.619
provided an ASA tech compatible TR for mount in the CPU box so we're able to

00:01:23.160 --> 00:01:31.979
use our existing stable of compatible course our H series a iOS which is cool

00:01:28.470 --> 00:01:35.640
because that means so can you we've got

00:01:31.979 --> 00:01:38.340
120 millimeter 240 millimeter and 280

00:01:35.640 --> 00:01:44.160
millimeter radiator versions with all of them featuring the same ASA tech design

00:01:40.170 --> 00:01:47.399
CPU block and in the H 115 eyes case a

00:01:44.160 --> 00:01:50.160
slightly more powerful pump in a stroke

00:01:47.399 --> 00:01:54.780
of either luck or AMD foresight the blocks are just big enough that even

00:01:52.890 --> 00:02:01.590
though the heat spreader sticks out the heat producing die area for thread

00:01:57.570 --> 00:02:04.979
Ripper is covered so while these are

00:02:01.590 --> 00:02:06.420
mainly meant for more square cpus we

00:02:04.979 --> 00:02:12.569
have no reason to believe that you should worry about it as for our air

00:02:08.940 --> 00:02:15.569
coolers well due to their wildly varying

00:02:12.569 --> 00:02:17.160
shapes and mounting mechanisms not many

00:02:15.569 --> 00:02:21.900
of them have been retrofitted with support for socket tr4 yet which is

00:02:20.160 --> 00:02:27.660
where naktu as new line-up comes in they're visually and for the most part

00:02:24.739 --> 00:02:33.690
functionally identical to their non tr4 cousins but when you turn them over you

00:02:31.230 --> 00:02:40.349
are immediately greeted with plates big and shiny enough to host a fancy dinner

00:02:36.480 --> 00:02:42.690
party this huge mount is common on each

00:02:40.349 --> 00:02:50.040
of them right down to the 92 millimeter Nhu 9 because Noctua decided to bake in

00:02:46.739 --> 00:02:53.879
support for AMD's on socket mounting

00:02:50.040 --> 00:02:57.090
points a good idea because well let's

00:02:53.879 --> 00:02:57.950
just say that taking this apart isn't a

00:02:57.090 --> 00:03:04.680
good idea now to testing while our old standard

00:03:02.489 --> 00:03:09.719
death version of this video discussed performance with different

00:03:06.930 --> 00:03:14.010
and configurations inside a case our video today will focus on the

00:03:11.579 --> 00:03:19.230
temperature differences between the coolers themselves right out of the box

00:03:16.409 --> 00:03:24.060
so it should then be noted that you could get slightly better performance

00:03:21.060 --> 00:03:26.700
out of these AI OS by adding more fans

00:03:24.060 --> 00:03:31.620
in a push-pull configuration and the same can be said of the two tower

00:03:28.859 --> 00:03:35.400
coolers another trick up our sleeve this time around is that we're going to be

00:03:33.450 --> 00:03:41.549
running our Threadripper nineteen fifty X in both stock and overclocked

00:03:37.739 --> 00:03:43.319
configurations furthermore this time we

00:03:41.549 --> 00:03:47.609
are in a climate controlled twenty four degrees Celsius environment not my

00:03:45.480 --> 00:03:51.299
garage during the winter so our results should be more comparable to what you

00:03:49.049 --> 00:03:58.319
can achieve at home but enough talk let's do the results starting off then

00:03:54.840 --> 00:04:01.680
with our H Adi GT and Noctua n hu nine

00:03:58.319 --> 00:04:04.139
we're seeing some pretty ok numbers for

00:04:01.680 --> 00:04:09.359
stock performance with similar low temperatures and package power

00:04:06.859 --> 00:04:17.669
overclocked on the other hand shows again similar numbers but markedly lower

00:04:13.859 --> 00:04:20.070
package power for the H Adi GT meaning

00:04:17.669 --> 00:04:25.349
that it's throttling more to maintain that same 87 degrees so in the

00:04:23.159 --> 00:04:32.760
lightweight category air cooling comes away with a clear victory the next step

00:04:28.320 --> 00:04:36.180
up is the NH u 12 s and H 100 IV two and

00:04:32.760 --> 00:04:38.900
again we see almost identical numbers

00:04:36.180 --> 00:04:44.039
between these guys at stock speeds looking now at the overclocking results

00:04:41.370 --> 00:04:49.440
we've got a much narrower lead for the Noctua fan over the Corsair AIO as far

00:04:46.949 --> 00:04:53.780
as package power goes meaning that they've throttled just about evenly so

00:04:51.960 --> 00:05:01.199
we're gonna call this one a tie bringing us to the main event the Nhu 14

00:04:58.229 --> 00:05:06.509
s weather versus the h 115 i've rhaegar both of them lasted all 12 rounds

00:05:04.139 --> 00:05:12.030
maintaining maximum performance in both the stock and overclocked tests and this

00:05:10.080 --> 00:05:19.010
is incredible considering we've got all 16 cores pinned to 3.9 gigahertz but

00:05:16.790 --> 00:05:23.900
this kind of a match isn't about pinning and the knock - I ran a few degrees

00:05:21.290 --> 00:05:30.440
cooler so we have to award it the win by technical knockout of course performance

00:05:27.740 --> 00:05:37.190
is only part of the equation aesthetics and physical compatibility are key

00:05:33.680 --> 00:05:39.140
considerations as well water coolers do

00:05:37.190 --> 00:05:44.870
come with an additional mechanical point of failure the pump but they look sexy

00:05:42.170 --> 00:05:49.730
and it's becoming more and more common for small form-factor cases to have

00:05:47.150 --> 00:05:55.640
mounts for 120 or 240 millimeter radiators but not enough clearance for

00:05:52.340 --> 00:05:58.070
comparable air coolers and it's also

00:05:55.640 --> 00:06:03.890
worth considering that large AIO coolers like the h 115 i are generally far

00:06:01.280 --> 00:06:09.530
easier to work around inside a chassis than a larger air cooler especially

00:06:06.320 --> 00:06:13.520
where RAM slots are concerned with that

00:06:09.530 --> 00:06:16.190
said price is kind of a big deal - and

00:06:13.520 --> 00:06:21.020
not too as solutions even though they're not the cheapest air coolers out there

00:06:18.040 --> 00:06:26.750
generally come out ahead of the water cooling equivalents by a significant

00:06:23.000 --> 00:06:28.670
margin here so TL DR then if you're not

00:06:26.750 --> 00:06:34.730
planning on overclocking Threadripper it's actually a relatively easy beast to

00:06:31.910 --> 00:06:40.340
tame but overclock it and you'll need to set aside some budget for a beefy cooler

00:06:37.360 --> 00:06:46.250
also if your priority is performance you'll reach for an air cooler or custom

00:06:43.310 --> 00:06:51.620
water cooling but water cooling is a great choice if a more compact easy to

00:06:49.400 --> 00:06:58.600
work in and subjectively better-looking machine is what you're after so then as

00:06:54.770 --> 00:07:00.880
always the answer is

00:06:58.600 --> 00:07:06.260
it depends if you haven't already now's the time to

00:07:03.620 --> 00:07:11.450
try out Dollar Shave Club it's more than just razors Dollar Shave Club makes it

00:07:09.380 --> 00:07:14.630
easy for you to upgrade your shave and your bathroom you don't even have to

00:07:13.520 --> 00:07:19.310
leave the house to get high-quality shaving and grooming products they

00:07:16.700 --> 00:07:24.080
deliver them right to your door and for a limited time Dollar Shave Club is

00:07:21.140 --> 00:07:30.710
basically giving away their shower shave starter step to new members for only 5

00:07:28.190 --> 00:07:34.490
bucks this starter set features their executive razor and three trial size

00:07:32.630 --> 00:07:38.300
versions of their most popular products to help you stay fresh and clean

00:07:35.960 --> 00:07:43.220
everywhere you get their shave butter body wash and one wipe Charlies butt

00:07:41.419 --> 00:07:46.639
wipes you will also get the executive razor which includes their premium

00:07:44.780 --> 00:07:51.169
weighty handle and a full cassette of cartridges then after the first box

00:07:49.430 --> 00:07:55.610
replacement cartridges are just a few bucks a month so head over to

00:07:52.610 --> 00:07:58.160
dollarshaveclub.com slash Lynas to check

00:07:55.610 --> 00:08:01.280
it out so thanks for watching guys if you dislike this video you can hit that

00:07:59.450 --> 00:08:04.669
button but if you liked it hit the like button get subscribed maybe consider

00:08:03.380 --> 00:08:10.130
checking out where to buy the stuff we featured at the link in the video description also down there is a link to

00:08:08.750 --> 00:08:15.160
our community forum which you should totally join and our merch store which

00:08:11.750 --> 00:08:15.160
has cool shirts like this one
