1
00:00:00,160 --> 00:00:05,520
have you ever wondered about the placement of expansion cards in your rig

2
00:00:04,200 --> 00:00:09,559
and how they might affect your graphics card temperatures and vice versa well

3
00:00:07,839 --> 00:00:13,360
today we're going to explore that from a few different angles how much does it

4
00:00:11,400 --> 00:00:18,400
matter in what scenarios does it matter and how much do you need to worry about

5
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

6
00:00:20,320 --> 00:00:28,530
hosts on scrapyard Wars and if you'd want to see that

7
00:00:28,960 --> 00:00:31,960
again

8
00:00:32,000 --> 00:00:39,719
Intel brings ddr4 to the mainstream with their new core i76700 K and core i56600k

9
00:00:37,719 --> 00:00:44,559
processors check the link in the video description to learn more so for this

10
00:00:41,879 --> 00:00:48,440
video we can't actually use our standard test bench because an open air rig would

11
00:00:46,960 --> 00:00:52,239
really screw with the test results compared to a close-off standard case

12
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

13
00:00:54,440 --> 00:01:00,320
fairly standard scenario computer case in order to properly explore the

14
00:00:58,199 --> 00:01:03,800
temperature changes in these VAR ious scenarios from here I chose two

15
00:01:02,120 --> 00:01:09,600
different graphics cards for testing a referenc 980 TI and an open air internal

16
00:01:06,520 --> 00:01:11,240
exhaust cooled styled uh G1 gaming 980

17
00:01:09,600 --> 00:01:16,000
from Gigabyte to control the temperatures in the rest of the case we

18
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

19
00:01:18,200 --> 00:01:24,000
pretty common use case uh these days and

20
00:01:21,560 --> 00:01:29,000
we used a control graphics card to interfere with the airflow of the two

21
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

22
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

23
00:01:35,320 --> 00:01:40,720
amount of heat into the rest of our bench and messing with the card in more

24
00:01:38,840 --> 00:01:45,640
substantial ways than just blocking off some of the airflow I did our standard

25
00:01:43,159 --> 00:01:51,000
Crisis 3 Skybox test for about 10 minutes recording temperatures and clock

26
00:01:47,680 --> 00:01:53,719
speeds as I went through um using the

27
00:01:51,000 --> 00:01:57,840
current test cards numbers one thing to note is that this is not a direct

28
00:01:55,680 --> 00:02:03,479
comparison of whether a blower or internal exhaust style card is uh better

29
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

30
00:02:05,439 --> 00:02:10,640
just check it out up here in the top right hand corner there were many

31
00:02:08,440 --> 00:02:15,440
opportunities for feature creep in this video the amount of potential scenarios

32
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

33
00:02:18,040 --> 00:02:25,040
possible scenarios of which we decided on a full-size average length test

34
00:02:22,400 --> 00:02:29,400
graphics cards and a full-size average length Control Card which will be placed

35
00:02:27,040 --> 00:02:34,440
either directly against the test card as close as possible or as far away as

36
00:02:31,680 --> 00:02:38,280
possible um on a seven expansion slot case which is the bottom two slots so

37
00:02:36,800 --> 00:02:41,640
basically the question is how much of an advantage are you gaining by pulling

38
00:02:39,920 --> 00:02:45,800
your second expansion card as far away as possible from the first I think it's

39
00:02:43,400 --> 00:02:49,400
pretty common knowledge that largely blocking the airf flow to one of your

40
00:02:47,400 --> 00:02:54,599
expansion cards if you have another option isn't exactly ideal but how much

41
00:02:52,239 --> 00:02:57,720
of an impact is it really well as these are ultimately different cards and

42
00:02:55,920 --> 00:03:01,760
comparing them together isn't really the point let's look at them individually

43
00:02:59,800 --> 00:03:06,200
first up we have the blower style card this is our

44
00:03:03,360 --> 00:03:09,280
gtx980ti this card already being on the edge of its cooling capabilities with an

45
00:03:07,920 --> 00:03:14,080
ambient room temperature of only actually 24° didn't actually change what

46
00:03:12,159 --> 00:03:19,200
temperature it ran depending on where the expansion card was placed which

47
00:03:16,360 --> 00:03:23,519
sounds great at first but what it did do is pretty substantially change its

48
00:03:21,000 --> 00:03:26,760
effective clock speed due to how GPU boost Works check out the video in the

49
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

50
00:03:29,000 --> 00:03:35,239
difference here between no control card installed and the control card in the

51
00:03:32,519 --> 00:03:40,599
worst possible position right next to our test card is a whopping 114 MHz

52
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

53
00:03:43,840 --> 00:03:50,599
story ends as the difference between the control card being right below the test

54
00:03:48,680 --> 00:03:56,879
card and the control card being as far away as possible is still 88 MHz which

55
00:03:54,680 --> 00:04:00,200
is still a very noticeable performance Improvement just for optimizing your

56
00:03:58,760 --> 00:04:06,720
expansion card layout but what about the internal exhaust card

57
00:04:03,400 --> 00:04:08,480
well the G1 gaming 980 actually did

58
00:04:06,720 --> 00:04:13,799
manage to get away from the good old classic NVIDIA standard of

59
00:04:10,640 --> 00:04:15,840
83° while by itself in the case meaning

60
00:04:13,799 --> 00:04:20,280
that when we added the control card in the furthest away position the clock

61
00:04:18,359 --> 00:04:27,120
speed didn't go down like it did with the blower card but the temperature did

62
00:04:23,400 --> 00:04:28,960
go up slightly but it still did go up

63
00:04:27,120 --> 00:04:34,400
that being said when we put the control card right up against it the temperature

64
00:04:31,199 --> 00:04:37,960
shot up to the previously mentioned 83°

65
00:04:34,400 --> 00:04:42,120
and the clock speed shot down by a huge

66
00:04:37,960 --> 00:04:44,759
139 mahz whoa I had already assumed this

67
00:04:42,120 --> 00:04:49,039
card would be impacted more heavily but that's

68
00:04:45,880 --> 00:04:50,960
huge ah so in conclusion this whole

69
00:04:49,039 --> 00:04:55,400
thing probably wasn't as surprising as it was interesting to me at least sure

70
00:04:53,600 --> 00:04:59,840
bad expansion card placement might not always even be up to you you might have

71
00:04:57,120 --> 00:05:03,320
to do certain things and sure like I just said some scenarios you can't help

72
00:05:01,360 --> 00:05:07,720
it and most people won't even run into this problem because they'll probably

73
00:05:05,240 --> 00:05:12,880
only have one card but there are some scenarios even if you do only have one

74
00:05:09,720 --> 00:05:15,919
card the ever more popular as time goes

75
00:05:12,880 --> 00:05:17,720
on small ITX style systems where you may

76
00:05:15,919 --> 00:05:21,520
have to put your card directly against a side panel for instance could benefit

77
00:05:20,039 --> 00:05:26,400
from this knowledge and you may want to stay wary of that when choosing a case

78
00:05:24,400 --> 00:05:30,280
tunnel bear the easy to ous privacy app for mobile and desktop internet browsing

79
00:05:28,479 --> 00:05:34,840
you can use tunnel bear on on iOS Android Mac PC and they even have a

80
00:05:32,560 --> 00:05:38,840
Chrome extension that lets you tunnel or browse the internet as if you're in one

81
00:05:36,680 --> 00:05:42,520
of up to 14 different countries with ease one of my favorite uses of tunnel

82
00:05:40,680 --> 00:05:47,039
bear is the ability to access websites that are Geo blocked within your country

83
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

84
00:05:49,080 --> 00:05:55,639
even have support bears standing by 24 hours a day to help they currently serve

85
00:05:53,520 --> 00:05:59,440
over 5 million users and you can become a user absolutely free by heading over

86
00:05:57,600 --> 00:06:04,280
to tunnel bear.com L TT they give you 500 megabytes of data

87
00:06:02,319 --> 00:06:08,240
for free every month but if that's not enough for you they have unlimited plan

88
00:06:06,120 --> 00:06:12,000
starting at a very reasonable about $6.99 per

89
00:06:09,919 --> 00:06:15,319
month thanks for watching guys if this video sucked you know what to do but if

90
00:06:13,560 --> 00:06:19,280
it was awesome get subscribed hit the like button or even consider supporting

91
00:06:17,080 --> 00:06:24,000
us directly by using our affiliate code to shop at Amazon buying a cool t-shirt

92
00:06:21,800 --> 00:06:26,880
that probably isn't from Star Wars or with a direct monthly contribution

93
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

94
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

95
00:06:34,240 --> 00:06:41,039
the impact of GPU boost hopefully someone will kind of step up soon so I

96
00:06:38,560 --> 00:06:45,080
can look more into how it affects internal exhaust cards that aren't

97
00:06:42,680 --> 00:06:47,720
always pinned against their temperature limits like some blower cards that I

98
00:06:46,759 --> 00:06:50,720
tested earlier