WEBVTT

00:00:00.160 --> 00:00:05.520
have you ever wondered about the placement of expansion cards in your rig

00:00:04.200 --> 00:00:09.559
and how they might affect your graphics card temperatures and vice versa well

00:00:07.839 --> 00:00:13.360
today we're going to explore that from a few different angles how much does it

00:00:11.400 --> 00:00:18.400
matter in what scenarios does it matter and how much do you need to worry about

00:00:16.000 --> 00:00:22.000
this kind of problem hit the like button if you like the idea of three or more

00:00:20.320 --> 00:00:28.530
hosts on scrapyard Wars and if you'd want to see that

00:00:28.960 --> 00:00:31.960
again

00:00:32.000 --> 00:00:39.719
Intel brings ddr4 to the mainstream with their new core i76700 K and core i56600k

00:00:37.719 --> 00:00:44.559
processors check the link in the video description to learn more so for this

00:00:41.879 --> 00:00:48.440
video we can't actually use our standard test bench because an open air rig would

00:00:46.960 --> 00:00:52.239
really screw with the test results compared to a close-off standard case

00:00:50.039 --> 00:00:56.440
which is the majority of what users will have these days so we need to go with a

00:00:54.440 --> 00:01:00.320
fairly standard scenario computer case in order to properly explore the

00:00:58.199 --> 00:01:03.800
temperature changes in these VAR ious scenarios from here I chose two

00:01:02.120 --> 00:01:09.600
different graphics cards for testing a referenc 980 TI and an open air internal

00:01:06.520 --> 00:01:11.240
exhaust cooled styled uh G1 gaming 980

00:01:09.600 --> 00:01:16.000
from Gigabyte to control the temperatures in the rest of the case we

00:01:13.119 --> 00:01:21.560
used a 240 mm aiio cooler to pull the heat away from the CPU and what is a

00:01:18.200 --> 00:01:24.000
pretty common use case uh these days and

00:01:21.560 --> 00:01:29.000
we used a control graphics card to interfere with the airflow of the two

00:01:26.520 --> 00:01:33.759
other test cards I use a Titan X as our control card as it isn't too long and it

00:01:31.119 --> 00:01:36.680
isn't too short but being a blower style card it shouldn't be dumping a huge

00:01:35.320 --> 00:01:40.720
amount of heat into the rest of our bench and messing with the card in more

00:01:38.840 --> 00:01:45.640
substantial ways than just blocking off some of the airflow I did our standard

00:01:43.159 --> 00:01:51.000
Crisis 3 Skybox test for about 10 minutes recording temperatures and clock

00:01:47.680 --> 00:01:53.719
speeds as I went through um using the

00:01:51.000 --> 00:01:57.840
current test cards numbers one thing to note is that this is not a direct

00:01:55.680 --> 00:02:03.479
comparison of whether a blower or internal exhaust style card is uh better

00:02:01.119 --> 00:02:06.719
at cooling itself but if you want to see this Linus already has a video on it so

00:02:05.439 --> 00:02:10.640
just check it out up here in the top right hand corner there were many

00:02:08.440 --> 00:02:15.440
opportunities for feature creep in this video the amount of potential scenarios

00:02:12.800 --> 00:02:20.280
was a little insane so we had to tone it down to what we saw as the most common

00:02:18.040 --> 00:02:25.040
possible scenarios of which we decided on a full-size average length test

00:02:22.400 --> 00:02:29.400
graphics cards and a full-size average length Control Card which will be placed

00:02:27.040 --> 00:02:34.440
either directly against the test card as close as possible or as far away as

00:02:31.680 --> 00:02:38.280
possible um on a seven expansion slot case which is the bottom two slots so

00:02:36.800 --> 00:02:41.640
basically the question is how much of an advantage are you gaining by pulling

00:02:39.920 --> 00:02:45.800
your second expansion card as far away as possible from the first I think it's

00:02:43.400 --> 00:02:49.400
pretty common knowledge that largely blocking the airf flow to one of your

00:02:47.400 --> 00:02:54.599
expansion cards if you have another option isn't exactly ideal but how much

00:02:52.239 --> 00:02:57.720
of an impact is it really well as these are ultimately different cards and

00:02:55.920 --> 00:03:01.760
comparing them together isn't really the point let's look at them individually

00:02:59.800 --> 00:03:06.200
first up we have the blower style card this is our

00:03:03.360 --> 00:03:09.280
gtx980ti this card already being on the edge of its cooling capabilities with an

00:03:07.920 --> 00:03:14.080
ambient room temperature of only actually 24° didn't actually change what

00:03:12.159 --> 00:03:19.200
temperature it ran depending on where the expansion card was placed which

00:03:16.360 --> 00:03:23.519
sounds great at first but what it did do is pretty substantially change its

00:03:21.000 --> 00:03:26.760
effective clock speed due to how GPU boost Works check out the video in the

00:03:25.200 --> 00:03:30.720
top right hand corner here if you want to know what the heck GPU boost does the

00:03:29.000 --> 00:03:35.239
difference here between no control card installed and the control card in the

00:03:32.519 --> 00:03:40.599
worst possible position right next to our test card is a whopping 114 MHz

00:03:39.080 --> 00:03:46.480
which would actually be quite a noticeable difference in terms of performance but that's not where the

00:03:43.840 --> 00:03:50.599
story ends as the difference between the control card being right below the test

00:03:48.680 --> 00:03:56.879
card and the control card being as far away as possible is still 88 MHz which

00:03:54.680 --> 00:04:00.200
is still a very noticeable performance Improvement just for optimizing your

00:03:58.760 --> 00:04:06.720
expansion card layout but what about the internal exhaust card

00:04:03.400 --> 00:04:08.480
well the G1 gaming 980 actually did

00:04:06.720 --> 00:04:13.799
manage to get away from the good old classic NVIDIA standard of

00:04:10.640 --> 00:04:15.840
83° while by itself in the case meaning

00:04:13.799 --> 00:04:20.280
that when we added the control card in the furthest away position the clock

00:04:18.359 --> 00:04:27.120
speed didn't go down like it did with the blower card but the temperature did

00:04:23.400 --> 00:04:28.960
go up slightly but it still did go up

00:04:27.120 --> 00:04:34.400
that being said when we put the control card right up against it the temperature

00:04:31.199 --> 00:04:37.960
shot up to the previously mentioned 83°

00:04:34.400 --> 00:04:42.120
and the clock speed shot down by a huge

00:04:37.960 --> 00:04:44.759
139 mahz whoa I had already assumed this

00:04:42.120 --> 00:04:49.039
card would be impacted more heavily but that's

00:04:45.880 --> 00:04:50.960
huge ah so in conclusion this whole

00:04:49.039 --> 00:04:55.400
thing probably wasn't as surprising as it was interesting to me at least sure

00:04:53.600 --> 00:04:59.840
bad expansion card placement might not always even be up to you you might have

00:04:57.120 --> 00:05:03.320
to do certain things and sure like I just said some scenarios you can't help

00:05:01.360 --> 00:05:07.720
it and most people won't even run into this problem because they'll probably

00:05:05.240 --> 00:05:12.880
only have one card but there are some scenarios even if you do only have one

00:05:09.720 --> 00:05:15.919
card the ever more popular as time goes

00:05:12.880 --> 00:05:17.720
on small ITX style systems where you may

00:05:15.919 --> 00:05:21.520
have to put your card directly against a side panel for instance could benefit

00:05:20.039 --> 00:05:26.400
from this knowledge and you may want to stay wary of that when choosing a case

00:05:24.400 --> 00:05:30.280
tunnel bear the easy to ous privacy app for mobile and desktop internet browsing

00:05:28.479 --> 00:05:34.840
you can use tunnel bear on on iOS Android Mac PC and they even have a

00:05:32.560 --> 00:05:38.840
Chrome extension that lets you tunnel or browse the internet as if you're in one

00:05:36.680 --> 00:05:42.520
of up to 14 different countries with ease one of my favorite uses of tunnel

00:05:40.680 --> 00:05:47.039
bear is the ability to access websites that are Geo blocked within your country

00:05:44.919 --> 00:05:51.639
it's super easy to install and use and they even have if you have issues they

00:05:49.080 --> 00:05:55.639
even have support bears standing by 24 hours a day to help they currently serve

00:05:53.520 --> 00:05:59.440
over 5 million users and you can become a user absolutely free by heading over

00:05:57.600 --> 00:06:04.280
to tunnel bear.com L TT they give you 500 megabytes of data

00:06:02.319 --> 00:06:08.240
for free every month but if that's not enough for you they have unlimited plan

00:06:06.120 --> 00:06:12.000
starting at a very reasonable about $6.99 per

00:06:09.919 --> 00:06:15.319
month thanks for watching guys if this video sucked you know what to do but if

00:06:13.560 --> 00:06:19.280
it was awesome get subscribed hit the like button or even consider supporting

00:06:17.080 --> 00:06:24.000
us directly by using our affiliate code to shop at Amazon buying a cool t-shirt

00:06:21.800 --> 00:06:26.880
that probably isn't from Star Wars or with a direct monthly contribution

00:06:25.479 --> 00:06:31.360
through the Forum now that you're done doing all that stuff you're probably wondering what to watch next so click

00:06:30.039 --> 00:06:36.319
the little button in the top right hand corner to check out the video I mentioned earlier where I investigate

00:06:34.240 --> 00:06:41.039
the impact of GPU boost hopefully someone will kind of step up soon so I

00:06:38.560 --> 00:06:45.080
can look more into how it affects internal exhaust cards that aren't

00:06:42.680 --> 00:06:47.720
always pinned against their temperature limits like some blower cards that I

00:06:46.759 --> 00:06:50.720
tested earlier