1
00:00:07,680 --> 00:00:15,639
well today I have what I believe is going to be a fascinating episode for

2
00:00:11,920 --> 00:00:17,279
you guys so let's start with video okay

3
00:00:15,639 --> 00:00:21,359
so videos everybody likes to watch videos there's a few people these days

4
00:00:19,039 --> 00:00:25,480
who own tablets there's a Motorola Zoom or smartphones there's an iPhone 4 and

5
00:00:23,760 --> 00:00:28,720
want to watch video on them but not all these devices and not all the media

6
00:00:26,960 --> 00:00:32,680
players are compatible with every kind of media that you that you download or

7
00:00:30,920 --> 00:00:38,160
rip or whatever however else you get your media so you often have to

8
00:00:35,200 --> 00:00:42,200
transcode or convert the media from one format to another so I'm going to be

9
00:00:39,559 --> 00:00:46,000
using media converter 7 for my testing today this is actually a really cool

10
00:00:43,680 --> 00:00:50,280
little piece of software it was $29.99 with a coupon code that I found very

11
00:00:47,960 --> 00:00:56,800
easily on their website okay so I'm going to be using a 200 Meg it is a 3

12
00:00:54,160 --> 00:01:02,199
minute 1080p file that was actually taken with oh here you guys can actually

13
00:00:58,879 --> 00:01:03,879
see me this camera the sx1 is 1080p I'm

14
00:01:02,199 --> 00:01:07,960
going to be outputting it to an iPad compatible format and I'm going to be

15
00:01:05,880 --> 00:01:11,840
doing it on a number of different platforms a number of different ways so

16
00:01:10,200 --> 00:01:17,320
I have three platforms I'm going to use and I'm going to be using kind of a best

17
00:01:13,439 --> 00:01:19,720
of Intel 1155 so that's the z68 chipset

18
00:01:17,320 --> 00:01:23,720
with a 2600k hyperthreaded processor that is the most expensive platform out

19
00:01:21,600 --> 00:01:30,520
of what I'll be using today I have an AMD am3 platform I have a Crosshair 4

20
00:01:26,840 --> 00:01:31,960
formula with a 6 core 1100t CPU bear in

21
00:01:30,520 --> 00:01:37,880
mind of course that the motherboard doesn't really have any impact on the performance in this case it's a CPU

22
00:01:35,399 --> 00:01:44,520
limited thing and finally I have AMD socket fm1 so this is their new Apu

23
00:01:40,479 --> 00:01:47,960
socket so I've got a 3850 quad core Apu

24
00:01:44,520 --> 00:01:50,119
here on an a75 m55 motherboard I'm going

25
00:01:47,960 --> 00:01:54,520
to be using all the same Hardware across these different boards and platforms so

26
00:01:52,280 --> 00:01:59,680
I'm going to be using 8 gigs of DDR3 memory I've got same power supply and

27
00:01:56,360 --> 00:02:02,640
Intel 510 series SSD on my test platform

28
00:01:59,680 --> 00:02:07,360
then on each platform I'm going to en uh transcode the video a different way so

29
00:02:05,000 --> 00:02:11,840
I'm going to use a CPU then I'm going to use a GTX 580 then I'm going to use a

30
00:02:10,759 --> 00:02:16,400
radon 6970 and finally I'm going to use the

31
00:02:14,480 --> 00:02:20,319
integrated GPU and we're going to compare how all these Solutions perform

32
00:02:18,879 --> 00:02:28,879
against each other whether we're using CPU Cuda direct compute or uh in Intel

33
00:02:24,280 --> 00:02:30,400
1155 case virtue or the Apu on AMD fm1

34
00:02:28,879 --> 00:02:34,440
let's see how it goes okay so my first two runs are done with

35
00:02:32,120 --> 00:02:39,280
the CPU as well as with the GTX 580 using Cuda on the 1155 platform and the

36
00:02:37,560 --> 00:02:44,239
one that I'm about to run and the reason I'm stopping to tell you guys about it

37
00:02:41,120 --> 00:02:47,239
is with Lucid virtue so all I have to do

38
00:02:44,239 --> 00:02:49,040
is turn virtue on there we go the

39
00:02:47,239 --> 00:02:55,480
processor Graphics are now enabled although I am plugged into the dedicated

40
00:02:52,080 --> 00:02:57,040
graphics card so the igp is working now

41
00:02:55,480 --> 00:03:01,280
and I'm going to run it with this one and see how it goes so I'll share

42
00:02:59,000 --> 00:03:07,120
results with you as I go here and as I discover what works better and what

43
00:03:04,680 --> 00:03:12,239
doesn't so here's a bit of an update I have the AMD fm1 system running right

44
00:03:09,640 --> 00:03:16,519
now so I've got most of my results done for that and I'm actually running it

45
00:03:13,560 --> 00:03:23,000
with the 6970 right now so you can see that the GTX 580 due to its very high

46
00:03:20,080 --> 00:03:27,560
performance actually outperforms the CPU by itself which we expect I mean the CPU

47
00:03:25,400 --> 00:03:33,000
was the slowest out of all of our results even on the 1155 platform where

48
00:03:30,319 --> 00:03:39,879
it's a significantly faster CPU with hyperthreading um the igp did beat the

49
00:03:36,680 --> 00:03:42,560
CPU but uh not maybe not by as much as I

50
00:03:39,879 --> 00:03:47,640
would like to see and uh actually here's something interesting I also added on

51
00:03:44,200 --> 00:03:50,000
the 1155 platform I tested a GTX 550 to

52
00:03:47,640 --> 00:03:55,040
see how it would compare to the 580 so it's almost as slow as a CPU so and

53
00:03:52,840 --> 00:03:58,959
definitely slower than the igp so if you're using 1155 and a lower end

54
00:03:57,280 --> 00:04:04,120
graphics card well anything but a super high-end graphics card you might

55
00:04:00,760 --> 00:04:07,360
actually be better off with uh with

56
00:04:04,120 --> 00:04:09,840
virtue or using the CPU versus using

57
00:04:07,360 --> 00:04:14,140
Cuda so uh here we go we've got our result for the

58
00:04:11,439 --> 00:04:15,680
6970 so here it is it is

59
00:04:15,680 --> 00:04:20,959
133 so you can see the 6970 once again

60
00:04:18,720 --> 00:04:26,400
doesn't perform quite as well as the 580 but it's close and then the igp trails

61
00:04:24,160 --> 00:04:30,880
significantly behind those two due to it well being substantially slower so I'm

62
00:04:29,400 --> 00:04:36,600
pretty much much finished my data at this point only to discover that it was

63
00:04:34,080 --> 00:04:42,680
all pretty much for nothing so what I accidentally did was I used a video file

64
00:04:41,080 --> 00:04:49,919
this one which had already been converted from its uh large High bit

65
00:04:46,440 --> 00:04:52,360
rate self into a Windows media audio

66
00:04:49,919 --> 00:04:56,759
file so what I'm going to do Second Time Around is I'll be using this one this is

67
00:04:54,000 --> 00:05:02,840
a 1.64 gig file whereas this is a 200 Meg file by

68
00:04:58,160 --> 00:05:05,919
aent okay so this is a OV file that is

69
00:05:02,840 --> 00:05:08,479
what is natively recorded by my camera

70
00:05:05,919 --> 00:05:14,600
then I'm going to convert this into an iPad format so you can see the data I've

71
00:05:10,759 --> 00:05:17,360
got here with the other file so the uh

72
00:05:14,600 --> 00:05:23,240
Intel Quicks sync is a good performer uh Cuda also performs quite well open CL is

73
00:05:20,199 --> 00:05:26,880
a little bit behind Cuda and then CPU is

74
00:05:23,240 --> 00:05:30,600
uh inv variably uh the slowest solution

75
00:05:26,880 --> 00:05:33,919
however I just ran quicksync on the new

76
00:05:30,600 --> 00:05:36,639
file which is the Intel Core i7 extreme

77
00:05:33,919 --> 00:05:41,880
Edition 990x or maybe it's a 980x unboxing and I it was done in 56 seconds

78
00:05:40,240 --> 00:05:46,479
I am now running it on opencl and you can see the difference in performance

79
00:05:44,000 --> 00:05:51,960
between quicksync and opencl on that already transcoded once file this ran in

80
00:05:49,400 --> 00:05:57,520
yeah 56 seconds and it's already taken well over 2 minutes with open CL when

81
00:05:54,560 --> 00:06:00,520
we're doing a far more complicated file so I'm going to just rerun everything

82
00:05:59,240 --> 00:06:04,600
which is good going to take me look at the time it is now already which is

83
00:06:02,240 --> 00:06:08,360
going to take me another you know hour or hour and a half or whatever and then

84
00:06:06,240 --> 00:06:13,000
I will be back with proper results for you guys okay well I'm finally done took

85
00:06:10,479 --> 00:06:19,479
me another couple hours to uh finish up here but I have the results so I've just

86
00:06:16,280 --> 00:06:22,479
uh used my old chart but all of the

87
00:06:19,479 --> 00:06:25,280
bottom results are the ones using the

88
00:06:22,479 --> 00:06:29,960
larger file and converting it to an iPad compatible format with media converter 7

89
00:06:28,479 --> 00:06:36,479
so let's have a look at the chart here and let's go we'll go through it together so this is the

90
00:06:33,400 --> 00:06:38,360
2600k with the CPU alone it clocks in at

91
00:06:36,479 --> 00:06:44,280
1 minute 40 seconds which is better than the 1100t better than the 3850 and

92
00:06:41,800 --> 00:06:48,720
better than the I 32105 so that makes a lot of sense cuz

93
00:06:47,400 --> 00:06:52,240
this is more expensive than this which is more expensive than this which is

94
00:06:50,840 --> 00:06:56,360
almost the same as this so this one you can see due to its four cores this is

95
00:06:54,720 --> 00:07:00,720
two cores with hyperthreading this is six cores four hyperthreaded cores so

96
00:06:58,960 --> 00:07:05,800
this guy with this four cores performs better than the equivalently priced

97
00:07:02,800 --> 00:07:07,960
Intel CPU with only two cores all right

98
00:07:05,800 --> 00:07:13,120
now let's factor in Cuda so we've got a GTX 580 now so they actually all perform

99
00:07:11,400 --> 00:07:18,000
fairly similarly this was a bit of a glitch but I ran it again and I don't

100
00:07:15,120 --> 00:07:25,400
know why but the 1100t did not perform quite as well with my Cuda GTX 580 as

101
00:07:21,840 --> 00:07:27,720
the uh the other cards did with the 6950

102
00:07:25,400 --> 00:07:33,479
we seem to see a lot more benefit from additional CPU power so you can see that

103
00:07:29,759 --> 00:07:34,240
the 6950 is still an improvement over

104
00:07:33,479 --> 00:07:41,800
the CPU but it doesn't like Cuda was very

105
00:07:38,759 --> 00:07:44,479
similar for with a GTX 580 across the

106
00:07:41,800 --> 00:07:48,039
board whereas the 6970 it seems like if you have a more powerful CPU to feed it

107
00:07:46,479 --> 00:07:54,520
it is really able to ramp up the performance so with the I

108
00:07:50,000 --> 00:07:58,000
32105 it clocked in at 246 with the 3850

109
00:07:54,520 --> 00:08:01,319
226 209 with the 1100t and a whopping 1

110
00:07:58,000 --> 00:08:04,960
minute 36 seconds seconds with the

111
00:08:01,319 --> 00:08:09,840
2600k now igp performance so the

112
00:08:04,960 --> 00:08:12,759
1155 the 2600k and the 2105 were both

113
00:08:09,840 --> 00:08:17,520
within spitting distance of each other and in spite of this being a quad core

114
00:08:15,000 --> 00:08:21,440
CPU that is clocked higher this being a dual core CPU so you can see that the

115
00:08:19,599 --> 00:08:25,199
igp is doing all the work here it's not getting a whole lot of help from the CPU

116
00:08:23,759 --> 00:08:33,519
side like we saw with the 6950 as for the fm1 solution remember

117
00:08:28,919 --> 00:08:36,240
unlike the Intel igp with using Lucid

118
00:08:33,519 --> 00:08:40,959
virtue on the fm1 solution you can only use it if you are running Onboard video

119
00:08:38,320 --> 00:08:47,160
this can be used if you're using a z68 or h67 chipset here's my z68 board this

120
00:08:45,040 --> 00:08:52,120
can be used even if you are running a discret graphics card so that's kind of

121
00:08:49,560 --> 00:08:56,720
a cool feature as well so here you can see that the igp did not perform as well

122
00:08:54,120 --> 00:09:02,279
as the 6970 but was still better than the CPU on its own so hopefully this was

123
00:08:59,920 --> 00:09:07,079
helpful thanks for checking out my video on the uh different platforms and their

124
00:09:04,240 --> 00:09:11,040
video transcoding performance Baron ma here's another quick thing to bear in

125
00:09:08,279 --> 00:09:17,040
mind you guys so you can see here that the with the h67 or z68 chipset the 1155

126
00:09:14,959 --> 00:09:22,839
chips whether it's an i7 or an I3 just smoke everything else but only a few

127
00:09:20,200 --> 00:09:26,640
software applications actually support all of these different Hardware

128
00:09:24,720 --> 00:09:30,800
accelerations so if you're using anything that doesn't in terms of share

129
00:09:28,480 --> 00:09:37,200
CPU perform performance you can see that the fm1 3850 due to its additional cores

130
00:09:34,160 --> 00:09:39,079
blows away the equivalently priced uh

131
00:09:37,200 --> 00:09:42,240
product from Intel and then you have to spend quite a bit more to get better

132
00:09:40,640 --> 00:09:45,839
performance on the Intel side don't forget to subscribe guys