WEBVTT

00:00:00.160 --> 00:00:07.839
now wait a minute this may very well be the quietest air-cooled GPU in its class

00:00:05.839 --> 00:00:14.160
but enthusiasts have been bolting full-size case fans to their gpus for

00:00:10.639 --> 00:00:16.880
over a decade could it possibly be worth

00:00:14.160 --> 00:00:22.560
buying one of these the ASUS x Noctua rtx 3070 when you could just redneck

00:00:20.240 --> 00:00:26.560
engineer your own guess we'll find out won't we and we'll find out who our

00:00:24.560 --> 00:00:30.640
sponsor is thanks to g-skill for sponsoring this video g-skill just

00:00:28.400 --> 00:00:36.000
announced their latest flagship ddr5 kits the trident z5 series these extreme

00:00:34.000 --> 00:00:40.559
performance ddr5 kits are designed for 12th gen Intel core processors and the

00:00:38.079 --> 00:00:44.640
z690 chipset platform find out more at the link in the video description i mean

00:00:42.719 --> 00:00:49.039
it's kind of cheating isn't it ASUS and not to us performance claims may very

00:00:46.640 --> 00:00:53.199
well pan out in the end but the thing is

00:00:50.399 --> 00:00:56.600
comically large look at this

00:00:59.440 --> 00:01:07.040
i mean it's not like we haven't seen cards with taller pcbs before it's not

00:01:04.559 --> 00:01:11.760
like we haven't seen cards with long coolers on them before but what i don't

00:01:09.680 --> 00:01:18.799
know if i've seen before is this appears to be a quadruple slot card

00:01:16.720 --> 00:01:22.720
and not only that but the thing about a quadruple slot card is you're gonna need

00:01:20.720 --> 00:01:27.360
at least one next to it for these fans to get any kind of air intake i mean the

00:01:25.200 --> 00:01:34.720
truly mind-blowing to me is not only did they put full-sized 120 millimeter by 25

00:01:31.600 --> 00:01:35.920
millimeter thick nf a12 desktop fans on

00:01:34.720 --> 00:01:41.360
the thing but they actually made the shroud like

00:01:39.280 --> 00:01:44.640
taller than them you know it's the kind of thing that i

00:01:42.880 --> 00:01:48.960
have to look at and go that probably wasn't ASUS's idea that

00:01:46.960 --> 00:01:52.240
was probably a knock to his idea or something to have this part sticking out

00:01:50.960 --> 00:01:57.280
here to make sure that under no circumstances would a knock to a cooled

00:01:54.960 --> 00:02:01.520
graphics card not perform properly because you can see

00:01:58.640 --> 00:02:07.040
that right there gives these fans quite a lot of buffer to take in air because

00:02:03.439 --> 00:02:07.040
you can't have anything right next to it

00:02:07.680 --> 00:02:14.879
you know i actually kind of like the look

00:02:14.080 --> 00:02:22.000
well yeah i just i think that's going to be a bit of a controversial take it's kind of

00:02:18.640 --> 00:02:24.000
sexy and like uh an ugly way you know

00:02:22.000 --> 00:02:29.520
does that make sense you know man it's like that like having a really

00:02:26.480 --> 00:02:32.959
like scary stuffed animal you know it's

00:02:29.520 --> 00:02:34.800
like scary cute it's ugly sexy

00:02:32.959 --> 00:02:38.720
man this is such an obvious collab i can't believe it took this long good on

00:02:36.239 --> 00:02:44.400
ASUS for getting it together i'd sort of expected october to do more than just

00:02:41.440 --> 00:02:47.760
strap their fans to a graphics card though

00:02:46.080 --> 00:02:52.800
now they claim they did but der bauer did a video about this

00:02:49.920 --> 00:02:58.239
card already and as far as he could tell the claims that ASUS knocked to a you

00:02:56.000 --> 00:03:02.000
know painstakingly redesigned this heatsink component

00:03:00.000 --> 00:03:06.959
don't appear to hold any water it seems to be exactly the same heat sink from

00:03:04.000 --> 00:03:13.120
ASUS rtx 3070 tough gaming let's play count the slots shall we so

00:03:09.280 --> 00:03:13.120
you got your one two three

00:03:14.080 --> 00:03:24.640
hold on one two three four four slots but

00:03:21.280 --> 00:03:26.400
can i actually install a card

00:03:24.640 --> 00:03:29.360
right next to it in that fifth one

00:03:29.440 --> 00:03:35.680
holy shnikeys it's actually going to push it a little bit

00:03:33.519 --> 00:03:40.159
and it's not even this isn't even a long enough card to interfere with the the

00:03:38.000 --> 00:03:43.840
peak of the bulge it's like it's like 4.1

00:03:42.640 --> 00:03:49.040
slots i guess it kind of makes sense i mean you look at high-end motherboards these

00:03:46.799 --> 00:03:53.680
days and lots of them don't have a full complement of pci express slots like

00:03:51.040 --> 00:03:57.519
they used to they'll have just one 16x slot up at the top and then maybe a

00:03:55.599 --> 00:04:02.159
couple of more because so many of the platforms PCIe lanes are taken up by

00:03:59.599 --> 00:04:06.159
storage now so if you're not going to be installing anything in it anyway i guess

00:04:04.159 --> 00:04:11.519
you might as well have a thick chunky ass GPU like this one

00:04:10.000 --> 00:04:15.840
of course you might not as well have it unless

00:04:13.599 --> 00:04:20.239
there's actually a performance benefit let's get it on the bench although first

00:04:17.919 --> 00:04:26.000
this is here for reference so we can see what uh i'm gonna call this a regular

00:04:22.960 --> 00:04:27.759
rtx 3070 with a triple fan cooler on

00:04:26.000 --> 00:04:31.919
here is gonna get us in terms of acoustic results

00:04:29.440 --> 00:04:36.560
to be clear this is a thick card and around 44 decibels while turboing to

00:04:34.160 --> 00:04:42.160
1950 megahertz and do maternal with GPU temps sitting at 63 degrees on the die

00:04:39.360 --> 00:04:45.759
and 73 degrees on the hot spot very respectable the Noctua one's just gonna

00:04:44.240 --> 00:04:50.160
have to do better if it's gonna take up a fourth slot as we shut down the system here we'll

00:04:48.479 --> 00:04:53.840
get a good opportunity to show you guys our noise floor here in the workshop

00:04:51.759 --> 00:04:57.400
with the hvac off so i'm gonna shut up for a second here

00:04:58.080 --> 00:05:04.160
at 28 29 decibels it's not an anechoic chamber in here by any stretch of the

00:05:02.560 --> 00:05:09.759
imagination but it's good enough for us to pick up

00:05:06.800 --> 00:05:17.199
a more normal rtx 3070 so it should be good enough for us to tell if this is

00:05:12.400 --> 00:05:17.199
better this thing looks so ridiculous

00:05:18.000 --> 00:05:24.320
i know right like the hilarious thing is the CPU

00:05:22.479 --> 00:05:29.520
cooler back here is what used to pass for you know a big CPU cooler now it

00:05:26.479 --> 00:05:32.400
looks like a toy getting CPU cooler envy

00:05:29.520 --> 00:05:36.639
over here ah something to note is that just like the tough gaming card that

00:05:34.080 --> 00:05:40.880
this is based on there's a performance and quiet mode switch

00:05:38.400 --> 00:05:44.320
on the back of the card here it ships in quiet mode but my assumption is that

00:05:42.880 --> 00:05:48.639
even performance mode on this thing would probably end up being pretty quiet

00:05:46.560 --> 00:05:53.360
i mean this is not my first kick at the knock to a can here i know that they

00:05:50.800 --> 00:05:57.039
make really quiet fans but damn it's a little awkward we went to

00:05:55.440 --> 00:05:59.680
get a reading off of it just to show you that it's under the noise floor in the

00:05:58.560 --> 00:06:04.000
room and the fans went turned off because it has a zero RPM mode so let's

00:06:01.680 --> 00:06:08.080
go ahead and fire up a light game theoretically this is one of those cards

00:06:05.759 --> 00:06:11.919
that has such a large heatsink on it that you should be able to run lighter

00:06:10.000 --> 00:06:15.440
titles without the fans even turning on but we'll see if that actually holds up

00:06:13.919 --> 00:06:20.479
and immediately there are a couple of observations we can make one is that the

00:06:17.919 --> 00:06:26.240
fans did in fact spin up so sayonara zero RPM gaming and number two

00:06:23.919 --> 00:06:31.199
is that roman der bauer observed that there was some coil whine

00:06:28.560 --> 00:06:34.479
on his card and i can verify that i'm getting some as

00:06:33.600 --> 00:06:39.360
well yeah i mean it's not loud in a way the

00:06:37.759 --> 00:06:43.440
product is kind of a victim of its own success because i wouldn't be able to

00:06:41.280 --> 00:06:47.759
hear that over the noise of a regular cooler it's just that i wish part of the

00:06:45.680 --> 00:06:53.520
collaboration between noctu and ASUS had been to design a board that managed

00:06:50.680 --> 00:06:58.479
extremely quiet power delivery i need a piece of metal like a side panel ah my

00:06:56.319 --> 00:07:03.759
lovely assistant james is over here helping prepare our victim for

00:07:01.759 --> 00:07:07.199
the thing you saw in the intro time for an extremely scientific experiment i'm

00:07:05.599 --> 00:07:12.039
going to get within earshot of the coil wine and put this panel in between us

00:07:16.240 --> 00:07:22.160
confirmed then as long as you've got it in a case like a normal person the coil

00:07:20.080 --> 00:07:27.759
wind's not likely to be a problem let's all just put the ball down huh huh

00:07:25.599 --> 00:07:31.360
they done did it they took my stuff get out of here boomer

00:07:29.280 --> 00:07:36.240
okay let's have a look if it's reached equilibrium yet

00:07:33.919 --> 00:07:39.680
wow has it ever

00:07:40.039 --> 00:07:48.240
55 degrees with 66 on the hot spot

00:07:45.759 --> 00:07:53.120
and that's clocked at 19.50 just like we saw before my fans are off

00:07:50.639 --> 00:07:56.080
the fans are freaking off of course rocket league

00:07:54.240 --> 00:08:00.160
not as demanding as do maternal let's fire that up next and we didn't even

00:07:57.919 --> 00:08:02.479
check the acoustics okay turn back on come on

00:08:03.520 --> 00:08:09.919
it's open oh excellent i don't know i exactly know what you wanted

00:08:07.919 --> 00:08:12.800
i found all the screws yes

00:08:11.599 --> 00:08:17.759
okay bye no you're not done

00:08:15.599 --> 00:08:22.879
you want the fans gone too yeah okay because i'm gonna put my own fans on it

00:08:20.160 --> 00:08:25.919
yeah that's the thing that happens when you pull off the female receptacle

00:08:24.800 --> 00:08:31.360
and you're trying to remove the male connector uh well it's fine because it's going to

00:08:28.240 --> 00:08:34.880
go back onto there anyway whoops

00:08:31.360 --> 00:08:34.880
i've seen dogs get stuck together

00:08:37.279 --> 00:08:42.560
wow what a weird thing to say um here you go

00:08:47.279 --> 00:08:49.680
wow

00:08:50.480 --> 00:08:57.200
that sounds like it's spinning really fast but this is like an 11 blade fan or

00:08:55.360 --> 00:09:01.200
something like that so it hits your finger a lot more

00:08:58.959 --> 00:09:04.160
frequently per revolution if that makes sense

00:09:02.240 --> 00:09:10.560
i don't think it's gonna heat up any more than it did that is outstanding

00:09:07.680 --> 00:09:15.519
no crazy frequency on this GPU we're at hitting 1860 but that's to be expected

00:09:12.959 --> 00:09:20.240
this is not the oc version and then in terms of temps

00:09:16.880 --> 00:09:24.000
and we managed to get up to 57

00:09:20.240 --> 00:09:26.720
and 68 on the hotspot and the craziest

00:09:24.000 --> 00:09:32.160
part is i don't think these fans even spun up anymore at all

00:09:29.120 --> 00:09:32.160
here everyone shh

00:09:34.160 --> 00:09:38.880
yeah we're right back to the noise floor you should note though you're not gonna

00:09:37.200 --> 00:09:43.040
see those kinds of temperatures in the real world because we're in a pretty

00:09:40.560 --> 00:09:47.360
cool room about 18 and a half degrees and we're running on an open bench which

00:09:45.040 --> 00:09:52.720
means that the GPU isn't sucking up hot air that was preheated by the CPU or any

00:09:50.080 --> 00:09:57.920
other components however that's a pretty compelling use case for a card like this

00:09:55.440 --> 00:10:02.720
if you have a stuffier case maybe a silence optimized one having a card that

00:10:00.320 --> 00:10:06.959
can run 10 degrees colder than the competition means that your case temps

00:10:05.279 --> 00:10:10.880
could be 10 degrees higher and the things still wouldn't thermal throttle

00:10:09.519 --> 00:10:16.000
not to mention the fans hadn't even wrapped up yet honestly this looks fine

00:10:13.440 --> 00:10:22.000
to me because you've got all the screws for this like heat spreader plate

00:10:18.880 --> 00:10:24.079
and you've got the four main screws for

00:10:22.000 --> 00:10:28.320
the actual fins leaving only one question

00:10:26.000 --> 00:10:33.680
is there truly anything special about this ASUS knock to a collab or could i

00:10:31.120 --> 00:10:38.240
achieve the same results with a couple of off-the-shelf Noctua fans and a roll

00:10:36.959 --> 00:10:42.880
of duct tape you ready andy

00:10:40.399 --> 00:10:46.720
all right i'm uh i'm gonna position my fan i'm doing some careful simulation

00:10:44.959 --> 00:10:49.920
engineering here i'm going to position my fan right at the edge of this

00:10:48.000 --> 00:10:54.560
heatsink because i want to take advantage of as much of this

00:10:51.600 --> 00:10:58.560
pass-through cooling as i can so i'm going to put that to put that

00:10:56.560 --> 00:11:04.000
right there i'm going to make myself a little uh kind of shroud okay okay so

00:11:01.360 --> 00:11:08.399
that's gonna go right there so we don't uh wait no no i can't do that no i'm not

00:11:06.560 --> 00:11:12.640
gonna make myself any kind of shroud the only

00:11:10.000 --> 00:11:15.680
uh hmm yeah what is the best way to take advantage of this because this is where

00:11:14.079 --> 00:11:18.640
the air has got to come out wait no no yeah no no not on this side

00:11:17.519 --> 00:11:22.640
yeah i know i was right in the first place dude okay

00:11:20.880 --> 00:11:29.519
only on this side where there's pass through this is so stupid

00:11:26.000 --> 00:11:30.399
stupid like a fox

00:11:29.519 --> 00:11:32.640
um sure

00:11:34.079 --> 00:11:39.360
look all right let's see you engineer uh

00:11:38.000 --> 00:11:46.959
a cooler in 10 minutes what do the kids say podcamp is this

00:11:43.360 --> 00:11:48.640
popchamp GPU edition

00:11:46.959 --> 00:11:55.279
i mean in fairness mine really doesn't look that much uglier and and my design

00:11:52.160 --> 00:11:56.720
is quite a bit more compact okay i think

00:11:55.279 --> 00:12:02.320
it's only three slots and change

00:11:59.760 --> 00:12:05.920
it's a full slot less thanks andy i was waiting for somebody

00:12:04.320 --> 00:12:10.079
to say it they're definitely spinning faster but i

00:12:08.480 --> 00:12:14.240
think part of the problem is that those are tied to CPU fan speeds

00:12:13.200 --> 00:12:19.760
ooh i have an idea i could go into the BIOS and reconfigure

00:12:17.279 --> 00:12:23.360
the fan curves but i got wancho in just a few minutes here so i'll just throw a

00:12:21.519 --> 00:12:26.639
couple of noise adapters on them let's go ahead and get an acoustic reading now

00:12:24.959 --> 00:12:30.959
it's not going to change because the fan speed is not tied to GPU temperatures in

00:12:28.800 --> 00:12:35.519
any way which could be dangerous unless our cooler was way overkill which i

00:12:33.600 --> 00:12:39.160
suspect this probably is so here i'll shut up let's uh

00:12:39.600 --> 00:12:45.760
just like the official knock to a GPU we are below the noise floor of this room

00:12:43.120 --> 00:12:49.920
which is around 29 decibels which is really freaking quiet

00:12:48.000 --> 00:12:53.440
but what about performance let's check out our canadian engineering

00:12:52.000 --> 00:12:58.720
here oh it's not as good not even close

00:12:56.720 --> 00:13:06.560
we suck oh no okay we're still turboing to

00:13:02.399 --> 00:13:07.920
1875 1890 megahertz but we've got a

00:13:06.560 --> 00:13:15.600
problem we hit GPU temps of 83 degrees

00:13:11.760 --> 00:13:17.440
and a hot spot temp of 96.

00:13:15.600 --> 00:13:21.680
so to answer the question

00:13:19.200 --> 00:13:25.040
can you just build one for yourself the answer is no

00:13:23.120 --> 00:13:29.519
not unless you're willing to compromise on the silence that Noctua and ASUS have

00:13:27.839 --> 00:13:34.560
achieved with this collaborative product because we have demonstrated in the past

00:13:31.600 --> 00:13:38.800
that you absolutely can just clip 120 millimeter fan to your GPU

00:13:36.959 --> 00:13:42.639
cooler and it will perform pretty darn well you just aren't going to get away

00:13:40.639 --> 00:13:46.800
with using low noise adapters if you want to have proper performance

00:13:45.120 --> 00:13:51.440
so dang good product jesus Noctua good job guys

00:13:50.240 --> 00:13:56.720
and good job Linus segwaying to our sponsor

00:13:54.880 --> 00:14:00.160
ting mobile has new rates that make it easier than ever to see how much you can

00:13:58.639 --> 00:14:03.920
save by switching they have unlimited talking text for 10

00:14:01.760 --> 00:14:08.720
data plans starting at 15 and then there's their set 12 plan which

00:14:06.000 --> 00:14:12.639
has 12 gigs of data for 35 bucks and they've even got unlimited data for 45

00:14:11.040 --> 00:14:16.160
dollars a month if you liked their previous pay for only what you use plans

00:14:14.720 --> 00:14:20.560
they're still there they're called ting mobile's flex plans now and they charge

00:14:18.000 --> 00:14:24.480
just five dollars per gig data can even be shared if you have a family plan so

00:14:22.160 --> 00:14:27.600
you can connect more phones to save more you get the same nationwide coverage in

00:14:26.079 --> 00:14:31.839
the us and award-winning customer service and pretty much any phone will

00:14:29.600 --> 00:14:34.959
work with ting mobile so don't wait they've got the perfect plan for

00:14:33.360 --> 00:14:40.399
everybody no matter what your needs are check them out at Linus.ting.com and you

00:14:37.279 --> 00:14:42.399
can get a 25 service credit if you guys

00:14:40.399 --> 00:14:46.800
enjoyed this video maybe go check out our previous GPU upgrade video where we

00:14:44.959 --> 00:14:52.639
used a real add-in GPU cooling product to achieve

00:14:50.320 --> 00:14:55.199
significantly better results good job guys