1
00:00:00,280 --> 00:00:02,840
Speaker 1: Brought to you by the reinvented two thousand twelve camera.

2
00:00:03,160 --> 00:00:06,280
Speaker 1: It's ready, are you hey there? Tech stuff listeners, This

3
00:00:06,440 --> 00:00:09,360
Speaker 1: is Jonathan Strickland, and I wanted to talk to you

4
00:00:09,440 --> 00:00:11,639
Speaker 1: a little bit about something cool going on at how

5
00:00:11,680 --> 00:00:14,040
Speaker 1: stuff works right now. I know all of you guys

6
00:00:14,040 --> 00:00:17,119
Speaker 1: are really creative and you love technology. Well, now you

7
00:00:17,160 --> 00:00:19,720
Speaker 1: can show us what you're made of, because Toyota is

8
00:00:19,720 --> 00:00:22,360
Speaker 1: sponsoring a new photo upload widget over at how stuff

9
00:00:22,360 --> 00:00:27,000
Speaker 1: works dot com. You can share your gadget ideas, modifications, hacks,

10
00:00:27,360 --> 00:00:30,080
Speaker 1: some great tech ideas. Show us what you're made of.

11
00:00:30,200 --> 00:00:32,640
Speaker 1: Let us know how creative you are. You can go

12
00:00:32,720 --> 00:00:37,120
Speaker 1: to www dot how stuff works dot com, slash upgrade

13
00:00:37,159 --> 00:00:39,960
Speaker 1: your tech and upload those photos. Now we want to

14
00:00:40,000 --> 00:00:45,040
Speaker 1: see what you got. Get in touch with technology with

15
00:00:45,120 --> 00:00:53,760
Speaker 1: tech stuff from how stuff works dot com. Hello again, everyone,

16
00:00:53,760 --> 00:00:56,040
Speaker 1: and welcome to tech stuff. My name is Chris Poulette

17
00:00:56,040 --> 00:00:57,920
Speaker 1: and I am an editor at how stuff works dot com.

18
00:00:57,920 --> 00:01:02,720
Speaker 1: Sitting across from me, as is senior writer Jonathan he there,

19
00:01:03,400 --> 00:01:06,280
Speaker 1: and I want to warn you all today's episode maybe

20
00:01:06,959 --> 00:01:12,440
Speaker 1: graphic that's correct, it's about graphics cards. We have received

21
00:01:12,560 --> 00:01:15,319
Speaker 1: lots of questions and requests for us to do an

22
00:01:15,319 --> 00:01:20,080
Speaker 1: episode about graphics cards, some of which hounded us through

23
00:01:20,240 --> 00:01:24,679
Speaker 1: multiple forms of media. And so today you get your wish,

24
00:01:24,720 --> 00:01:27,600
Speaker 1: we're gonna talk about graphics cards. These are the things

25
00:01:27,640 --> 00:01:34,720
Speaker 1: that got me into console gaming, which is true. No, So, Chris,

26
00:01:34,760 --> 00:01:36,960
Speaker 1: you and I remember, I mean we were we were

27
00:01:37,000 --> 00:01:40,880
Speaker 1: both using computers during the nineties when there was the

28
00:01:41,800 --> 00:01:46,319
Speaker 1: kind of revolution in personal computers where and a a

29
00:01:46,319 --> 00:01:51,240
Speaker 1: an independent secondary graphics card became a necessity if you

30
00:01:51,280 --> 00:01:54,080
Speaker 1: wanted to stay up to date with the latest games

31
00:01:54,080 --> 00:01:56,880
Speaker 1: that were coming out. Yeah. I remember with my very

32
00:01:56,880 --> 00:02:00,240
Speaker 1: first Amiga how they would talk about there were their

33
00:02:00,280 --> 00:02:04,680
Speaker 1: secondary chips on the motherboard that we handle additional operations,

34
00:02:04,720 --> 00:02:08,760
Speaker 1: not just the one processor. And I think, um, it

35
00:02:08,800 --> 00:02:12,359
Speaker 1: probably seems silly to a lot of us now, like, oh, well,

36
00:02:12,400 --> 00:02:15,240
Speaker 1: of course you know that would be important. But you know,

37
00:02:15,280 --> 00:02:19,200
Speaker 1: as computing has gone on, we've realized some that although

38
00:02:19,240 --> 00:02:23,600
Speaker 1: you have you know, Moore's law, which basically just you know,

39
00:02:23,639 --> 00:02:28,600
Speaker 1: talks about computing power growing exponentially, and you say, well, okay,

40
00:02:29,120 --> 00:02:31,120
Speaker 1: you know you've got more and more powerful computers, but

41
00:02:31,240 --> 00:02:34,520
Speaker 1: we've realized over that time too that although you may

42
00:02:34,560 --> 00:02:40,720
Speaker 1: have more powerful processors, Dividing tasks up between multiple processors

43
00:02:40,840 --> 00:02:44,239
Speaker 1: is a better way to compute because you can accomplish

44
00:02:44,280 --> 00:02:48,960
Speaker 1: more when you divide these jobs, and uh, you know,

45
00:02:49,200 --> 00:02:52,240
Speaker 1: it's it's more efficient that way. It's it's very simple. Yeah,

46
00:02:52,240 --> 00:02:56,000
Speaker 1: So a graphics card job essentially is to really manipulate

47
00:02:56,040 --> 00:03:00,440
Speaker 1: computer memory at a very rapid rate so that information

48
00:03:00,480 --> 00:03:04,320
Speaker 1: coming in can be converted into graphic information that's going

49
00:03:04,360 --> 00:03:08,120
Speaker 1: to be displayed upon a monitor of some sort. And

50
00:03:08,320 --> 00:03:11,919
Speaker 1: it's almost its own computer. Yeah, it's it's it's its

51
00:03:11,919 --> 00:03:15,919
Speaker 1: own processor, certainly. And the idea is that this takes

52
00:03:16,080 --> 00:03:20,160
Speaker 1: some of that burden from the CPU and it shoulders

53
00:03:20,200 --> 00:03:23,400
Speaker 1: it itself so that the CPU can um sort of

54
00:03:23,440 --> 00:03:27,280
Speaker 1: become like a manager. The CPU sends the instructions to

55
00:03:27,360 --> 00:03:32,200
Speaker 1: the GPU, which then takes care of calculations and and

56
00:03:32,240 --> 00:03:35,960
Speaker 1: it's it's designed to do that in a very streamlined,

57
00:03:36,120 --> 00:03:39,240
Speaker 1: fast way. And part of that design is this idea

58
00:03:39,280 --> 00:03:42,040
Speaker 1: of parallelism, which is what Chris was talking about, where

59
00:03:42,840 --> 00:03:46,640
Speaker 1: you have your your chip created in such a way

60
00:03:46,680 --> 00:03:50,320
Speaker 1: to process sets of instructions in parallel with each other

61
00:03:51,360 --> 00:03:55,480
Speaker 1: because those different sets of instructions don't depend directly on

62
00:03:55,520 --> 00:04:00,000
Speaker 1: each other. You know, if you think about computer problem.

63
00:04:00,240 --> 00:04:04,120
Speaker 1: One type of computer problem could be an ongoing uh

64
00:04:04,480 --> 00:04:09,840
Speaker 1: calculation where the outcome of the previous calculation determines what

65
00:04:09,960 --> 00:04:13,880
Speaker 1: the next calculation will be. Right, So we'll just say

66
00:04:13,920 --> 00:04:17,000
Speaker 1: that for the first calculation, we're gonna do something really simple.

67
00:04:17,480 --> 00:04:19,720
Speaker 1: Just for example, say that the first calculation is two

68
00:04:19,720 --> 00:04:23,960
Speaker 1: plus two and the the output is four, and then

69
00:04:24,040 --> 00:04:26,760
Speaker 1: the four from that first calculation is used in the

70
00:04:26,800 --> 00:04:31,159
Speaker 1: second calculation. Where you have another variable, Well, we'll say

71
00:04:31,160 --> 00:04:33,640
Speaker 1: that for the purposes of this problem, that variable is

72
00:04:33,680 --> 00:04:37,360
Speaker 1: going to be eight, and whatever the product the whatever

73
00:04:37,400 --> 00:04:40,719
Speaker 1: the sum is from that first calculation gets added to eight.

74
00:04:41,279 --> 00:04:45,640
Speaker 1: So the the second calculation depends upon the outcome of

75
00:04:45,680 --> 00:04:47,880
Speaker 1: the first one. The two plus two is four. Well,

76
00:04:47,880 --> 00:04:50,839
Speaker 1: now I've got the the sum of that first calculation four.

77
00:04:50,960 --> 00:04:53,239
Speaker 1: I can now add that to eight to get twelve,

78
00:04:53,640 --> 00:04:56,000
Speaker 1: and so on and so forth. Well, with all those

79
00:04:56,040 --> 00:04:59,799
Speaker 1: calculations depending upon one another, having a single processor makes sense.

80
00:05:00,000 --> 00:05:02,040
Speaker 1: You can't really divide that up because you don't know

81
00:05:02,120 --> 00:05:04,839
Speaker 1: what the outcome is going to be of a previous calculation,

82
00:05:05,360 --> 00:05:08,719
Speaker 1: so you can't move forward until you get that. But

83
00:05:08,920 --> 00:05:11,960
Speaker 1: there are other computer problems which we have talked about

84
00:05:12,000 --> 00:05:15,160
Speaker 1: several times on this podcast that you can divide up

85
00:05:15,200 --> 00:05:19,800
Speaker 1: because they are not interdependent that way, and have a

86
00:05:19,880 --> 00:05:24,480
Speaker 1: processor or several processors work on that problem, on those problems,

87
00:05:25,240 --> 00:05:28,120
Speaker 1: and the outcome ends up creating whatever it is you

88
00:05:28,160 --> 00:05:32,400
Speaker 1: need and UM. In that case, the GPU design was

89
00:05:32,440 --> 00:05:36,440
Speaker 1: meant to do that. So we had parallel processors with

90
00:05:36,560 --> 00:05:43,680
Speaker 1: GPUs well before you started seeing multi cores in consumer computers. Yeah, yeah,

91
00:05:43,720 --> 00:05:47,000
Speaker 1: that's true. By the way, if you run across computers

92
00:05:47,000 --> 00:05:51,240
Speaker 1: that offer perpendicular processing, I don't recommend it they I

93
00:05:52,760 --> 00:05:58,760
Speaker 1: wasted a lot of money once anyway, being so obtuse UM. Actually,

94
00:05:59,200 --> 00:06:02,239
Speaker 1: when you look at the parts of a graphics card,

95
00:06:02,960 --> 00:06:07,440
Speaker 1: UM again, they seem rather you know common sense. You've

96
00:06:07,440 --> 00:06:10,719
Speaker 1: got a graphics card is its own motherboard, It has

97
00:06:10,760 --> 00:06:15,039
Speaker 1: a its own processor on it, UM, and its own memory. UM.

98
00:06:15,080 --> 00:06:18,640
Speaker 1: You may actually see, you know, specifications, I remember seeing

99
00:06:18,640 --> 00:06:22,440
Speaker 1: specifications for how much v RAM, video RAM or or

100
00:06:22,680 --> 00:06:26,279
Speaker 1: graphics memory. UM. That's all part of what you see

101
00:06:26,320 --> 00:06:30,080
Speaker 1: on the graphics card. And some graphics cards are you know, customizable.

102
00:06:30,120 --> 00:06:33,600
Speaker 1: You can you can add memory. UM depends on your

103
00:06:33,600 --> 00:06:36,720
Speaker 1: computer course, and and all of that also depends on

104
00:06:36,720 --> 00:06:39,440
Speaker 1: your your the type of graphics card you use depends

105
00:06:39,480 --> 00:06:42,960
Speaker 1: heavily upon the kind of motherboard your computer has. That's true.

106
00:06:43,120 --> 00:06:47,760
Speaker 1: That's true. Not all motherboards can support all graphics cards.

107
00:06:47,800 --> 00:06:50,400
Speaker 1: It depends on especially if you have an older computer.

108
00:06:50,480 --> 00:06:54,920
Speaker 1: There comes a point with older computers where the newest

109
00:06:56,160 --> 00:07:00,679
Speaker 1: video cards just the old motherboards can't support them. So

110
00:07:01,040 --> 00:07:03,520
Speaker 1: that's one thing to keep in mind that the motherboard

111
00:07:03,920 --> 00:07:07,400
Speaker 1: your your computer's motherboard has a you know, a shelf

112
00:07:07,440 --> 00:07:10,400
Speaker 1: life really and at some point it does become obsolete,

113
00:07:10,440 --> 00:07:13,400
Speaker 1: at least in comparison to the latest hardware that comes

114
00:07:13,560 --> 00:07:15,480
Speaker 1: or the latest components that are out on the market.

115
00:07:16,400 --> 00:07:19,800
Speaker 1: And basically, uh, well, I can't think of a computer

116
00:07:19,880 --> 00:07:22,920
Speaker 1: that doesn't have a video output of some kind on it,

117
00:07:22,960 --> 00:07:27,680
Speaker 1: even even laptops or ultra books, you know, things that

118
00:07:27,680 --> 00:07:30,160
Speaker 1: that are designed to have fewer ports than things because

119
00:07:30,160 --> 00:07:35,720
Speaker 1: it's streamlined. Even those machines generally have UM some kind

120
00:07:35,720 --> 00:07:38,160
Speaker 1: of video output. Of course, the main one on those

121
00:07:38,160 --> 00:07:40,800
Speaker 1: would be at their own screen, but um coming off

122
00:07:40,800 --> 00:07:43,160
Speaker 1: of the graphics card. You need to uh to be

123
00:07:43,200 --> 00:07:46,480
Speaker 1: able to display the results, so you know, they'll they'll

124
00:07:46,520 --> 00:07:49,760
Speaker 1: generally have some kind of connector UM and of course,

125
00:07:50,600 --> 00:07:54,080
Speaker 1: key to this all is your operating system, because your

126
00:07:54,120 --> 00:07:59,960
Speaker 1: operating system is uh, what is doing basically the traffic management.

127
00:08:00,480 --> 00:08:03,680
Speaker 1: So let's say, uh, probably the main reason we talk

128
00:08:03,720 --> 00:08:08,200
Speaker 1: about graphics cards these days are for you know, really

129
00:08:08,240 --> 00:08:14,960
Speaker 1: three things games, um, high resolution video editing or or

130
00:08:15,200 --> 00:08:17,840
Speaker 1: you know, graphics photo editing. You know, those are the

131
00:08:17,840 --> 00:08:20,560
Speaker 1: reasons we really talk about the graphics card. I mean,

132
00:08:20,560 --> 00:08:22,400
Speaker 1: if you say, yeah, I've got to surf the web

133
00:08:22,440 --> 00:08:24,200
Speaker 1: on my computer, you don't go, yeah, but what kind

134
00:08:24,200 --> 00:08:26,360
Speaker 1: of graphics card you got, man? I will add I

135
00:08:26,400 --> 00:08:28,760
Speaker 1: will add a fourth a fourth yeah, which has nothing

136
00:08:28,800 --> 00:08:31,640
Speaker 1: to do with graphics. Okay, it's what people have started

137
00:08:31,640 --> 00:08:33,880
Speaker 1: to use GPUs to do. I figured we'd talked about

138
00:08:33,880 --> 00:08:36,240
Speaker 1: that way, which goes back to the parallel processing. So

139
00:08:36,280 --> 00:08:38,640
Speaker 1: the fourth one, I would say, is parallel processing that

140
00:08:38,720 --> 00:08:43,920
Speaker 1: does not necessarily go back to video processing. So and

141
00:08:43,960 --> 00:08:45,800
Speaker 1: that can be for lots of different stuff. I know

142
00:08:45,920 --> 00:08:48,800
Speaker 1: specifically which one you're thinking about. But because we did

143
00:08:48,840 --> 00:08:52,439
Speaker 1: our our episode on on password security and talked about

144
00:08:52,440 --> 00:08:57,680
Speaker 1: how GPUs have been used to help crack passwords through

145
00:08:57,720 --> 00:09:00,199
Speaker 1: brute force attacks, but they can also be is to

146
00:09:00,240 --> 00:09:03,280
Speaker 1: do other parallel processing, not just you know, cryptography, No,

147
00:09:03,440 --> 00:09:06,640
Speaker 1: that that's true because UM, well, actually uh we had

148
00:09:06,679 --> 00:09:10,800
Speaker 1: talked about it on our our podcast about about the

149
00:09:10,840 --> 00:09:14,240
Speaker 1: mac os ten snow Leopard. One of the changes was

150
00:09:14,280 --> 00:09:18,120
Speaker 1: that they added the ability to uh um send instructions.

151
00:09:18,120 --> 00:09:20,040
Speaker 1: And that's really what I was sort of getting at

152
00:09:20,080 --> 00:09:22,760
Speaker 1: with the the operating system is, you know, it's let's

153
00:09:22,760 --> 00:09:25,400
Speaker 1: say you're playing a game and uh you know, you

154
00:09:25,440 --> 00:09:28,600
Speaker 1: have the game itself, the instructions that are going out

155
00:09:28,679 --> 00:09:32,200
Speaker 1: about where your character is and where it's moving across

156
00:09:32,240 --> 00:09:35,280
Speaker 1: the screen, and it's sending other instructions about how to

157
00:09:35,320 --> 00:09:38,480
Speaker 1: render that on the screen to the graphics processor card,

158
00:09:38,520 --> 00:09:41,920
Speaker 1: which takes care of computation for things like the number

159
00:09:41,960 --> 00:09:46,000
Speaker 1: of polygons that you see on the screen, how fast

160
00:09:46,080 --> 00:09:49,240
Speaker 1: those that information can be displayed. So they work in

161
00:09:49,320 --> 00:09:53,600
Speaker 1: concert through the help of the operating system to uh

162
00:09:53,800 --> 00:09:57,280
Speaker 1: manage the rest of the stuff and the graphics specific

163
00:09:57,360 --> 00:10:00,640
Speaker 1: stuff with the two different the two different cars, and

164
00:10:01,679 --> 00:10:04,480
Speaker 1: mac Os ten is able in the newer incarnations of

165
00:10:04,520 --> 00:10:07,440
Speaker 1: the operating system is able to uh um, and I

166
00:10:07,480 --> 00:10:10,520
Speaker 1: assume that's true for newer versions of Windows and Lenox

167
00:10:10,559 --> 00:10:14,520
Speaker 1: as well as you know, able to send instructions to

168
00:10:14,640 --> 00:10:16,840
Speaker 1: the graphics cards. So Let's say you are surfing the

169
00:10:16,880 --> 00:10:20,440
Speaker 1: web and maybe you have a music playing on you know,

170
00:10:20,559 --> 00:10:23,760
Speaker 1: music player service, and you've got email going on, and uh,

171
00:10:24,120 --> 00:10:26,200
Speaker 1: you know, five or ten other little things going on

172
00:10:26,280 --> 00:10:30,920
Speaker 1: in the background, your calendar software. UM, your graphics card

173
00:10:31,000 --> 00:10:34,640
Speaker 1: may not be working over time. So the processing uh,

174
00:10:34,800 --> 00:10:38,120
Speaker 1: you know, the the operating system goes, hey, since you're

175
00:10:38,160 --> 00:10:41,040
Speaker 1: not busy and you handle some of these other instructions,

176
00:10:41,080 --> 00:10:44,319
Speaker 1: kind of take a load off this couch. Yeah, exactly.

177
00:10:44,679 --> 00:10:49,400
Speaker 1: So it goes in and helps move the couch, um,

178
00:10:49,640 --> 00:10:51,719
Speaker 1: so to speak, and take some of the load off

179
00:10:51,760 --> 00:10:55,640
Speaker 1: the main processor, which which is ingenious that any operating

180
00:10:55,640 --> 00:10:58,120
Speaker 1: system would be able to do that. Um, And it

181
00:10:58,160 --> 00:11:00,520
Speaker 1: just so happens. I know about that one because we

182
00:11:00,520 --> 00:11:03,680
Speaker 1: were talking about it in that But you know, the

183
00:11:03,720 --> 00:11:06,360
Speaker 1: operating systems have become so much more sophisticated and to

184
00:11:06,400 --> 00:11:09,360
Speaker 1: be able to take advantage of that specifically, UM, to

185
00:11:09,480 --> 00:11:12,280
Speaker 1: take a load off of the main processors is excellent.

186
00:11:13,080 --> 00:11:17,720
Speaker 1: The things that determine how fast a graphics card is, uh,

187
00:11:17,800 --> 00:11:21,840
Speaker 1: it's more than just the GPU clock speed of the processors.

188
00:11:21,920 --> 00:11:25,720
Speaker 1: So keep in mind the heart of this graphics card

189
00:11:25,880 --> 00:11:29,600
Speaker 1: is a processor. We call the GPU. The graphics processing unit.

190
00:11:30,120 --> 00:11:33,440
Speaker 1: And and it's it's very similar to a CPU. I

191
00:11:33,440 --> 00:11:36,480
Speaker 1: mean that it's job is to crunch data. It's just

192
00:11:36,520 --> 00:11:38,679
Speaker 1: in this case most of the time that data has

193
00:11:38,720 --> 00:11:43,240
Speaker 1: to do with rendering graphics. UM, so the clock speed

194
00:11:43,400 --> 00:11:46,960
Speaker 1: is definitely important. And clock speed, of course, that's generally speaking,

195
00:11:47,000 --> 00:11:51,400
Speaker 1: the number of instructions a CPU can carry out within

196
00:11:51,440 --> 00:11:56,240
Speaker 1: a second. Uh, that's the number of cycles. Because keeping

197
00:11:56,240 --> 00:11:58,720
Speaker 1: in mind that when I say instruction, really instruct it's

198
00:11:58,880 --> 00:12:02,640
Speaker 1: really cycles, and some instructions require multiple cycles to complete.

199
00:12:02,840 --> 00:12:05,080
Speaker 1: I want to make sure I get that before people

200
00:12:05,080 --> 00:12:07,360
Speaker 1: are like, well, you know actually, and people are going

201
00:12:07,400 --> 00:12:12,520
Speaker 1: to mention the fact that processors can be overclocked or underclocked. Yeah,

202
00:12:12,559 --> 00:12:16,040
Speaker 1: that that speed is a rating that the manufacturer gives

203
00:12:16,040 --> 00:12:18,800
Speaker 1: it that basically says, this will operate the way we've

204
00:12:18,800 --> 00:12:20,920
Speaker 1: set it up on this board. It will operate at

205
00:12:20,960 --> 00:12:23,160
Speaker 1: this speed it's gonna and it's a safe speed. It's

206
00:12:23,160 --> 00:12:26,920
Speaker 1: not going to cause the processor, processor to overheat falls.

207
00:12:27,000 --> 00:12:29,920
Speaker 1: It falls within certain operating parameters that are considered to

208
00:12:30,000 --> 00:12:33,760
Speaker 1: be optimal for that system. So yeah, it's the rated speed. Yeah,

209
00:12:33,800 --> 00:12:38,560
Speaker 1: and just like CPUs, you can overclock GPUs aboutely want to.

210
00:12:38,679 --> 00:12:40,439
Speaker 1: In fact, there are a lot of video gamers out there,

211
00:12:40,559 --> 00:12:43,960
Speaker 1: that's what they do. Uh. So that's one thing that

212
00:12:44,000 --> 00:12:46,920
Speaker 1: determines how fast a video graphics card is able to

213
00:12:47,000 --> 00:12:50,120
Speaker 1: handle graphics. But that's not the only thing, because you

214
00:12:50,160 --> 00:12:53,000
Speaker 1: also have to worry about getting information to and from

215
00:12:53,040 --> 00:12:57,720
Speaker 1: that uh that that processor. So the size of the

216
00:12:57,760 --> 00:13:01,160
Speaker 1: memory bus is really important. The memory bus this is

217
00:13:01,200 --> 00:13:02,480
Speaker 1: kind of you can think of that is sort of

218
00:13:02,520 --> 00:13:06,920
Speaker 1: like the highway between the memory and the processor. And

219
00:13:07,080 --> 00:13:10,000
Speaker 1: you want a nice wide highway so that you can

220
00:13:10,080 --> 00:13:13,200
Speaker 1: shove as much data towards that processor as possible. Because

221
00:13:13,200 --> 00:13:15,280
Speaker 1: the processor, if it's really fast and is able to

222
00:13:15,320 --> 00:13:18,160
Speaker 1: handle a lot of instructions, that doesn't matter so much.

223
00:13:18,160 --> 00:13:21,520
Speaker 1: If it can't get the data it needs fast enough,

224
00:13:22,080 --> 00:13:24,760
Speaker 1: it may just be that, yeah, I'm super smart, but

225
00:13:24,800 --> 00:13:27,840
Speaker 1: you're giving me math problems at a rate that is,

226
00:13:28,400 --> 00:13:30,959
Speaker 1: you know, nowhere near my capacity, and so a lot

227
00:13:31,000 --> 00:13:33,840
Speaker 1: of this power is going to waste. So the memory

228
00:13:33,840 --> 00:13:38,319
Speaker 1: bus is a very important part of that equation, you know,

229
00:13:38,520 --> 00:13:42,600
Speaker 1: because you you you said the memory buses. It's like well,

230
00:13:42,640 --> 00:13:45,360
Speaker 1: and and that's the funny part about that is if

231
00:13:45,360 --> 00:13:47,360
Speaker 1: you're thinking about it in terms if that if you

232
00:13:47,480 --> 00:13:50,760
Speaker 1: made that visual connection to it's not um it really is.

233
00:13:51,160 --> 00:13:55,520
Speaker 1: The bus is the connection itself, not because I was

234
00:13:55,559 --> 00:13:59,040
Speaker 1: actually thinking of bus and what in the term given

235
00:13:59,120 --> 00:14:01,520
Speaker 1: in my in the right context, not in the vehicular

236
00:14:01,520 --> 00:14:06,200
Speaker 1: one I want. But also other things that that determine

237
00:14:06,200 --> 00:14:09,800
Speaker 1: the speed include the amount of available memory, so that

238
00:14:09,920 --> 00:14:13,040
Speaker 1: onboard memory that's on that graphics card plays a big part.

239
00:14:13,080 --> 00:14:16,000
Speaker 1: That's just that's just as true in your computer for

240
00:14:16,080 --> 00:14:19,080
Speaker 1: general computing purposes. You know, your computer needs a good

241
00:14:19,080 --> 00:14:21,040
Speaker 1: amount of memory, so the CPU does not have to

242
00:14:21,200 --> 00:14:25,320
Speaker 1: constantly look for the information within your systems hard drive.

243
00:14:25,440 --> 00:14:29,040
Speaker 1: That slows things down. So keeping more and more data

244
00:14:29,040 --> 00:14:32,560
Speaker 1: in memory means the CPU doesn't have to consult the

245
00:14:32,560 --> 00:14:35,120
Speaker 1: hard drive so much and things just move at a

246
00:14:35,160 --> 00:14:38,800
Speaker 1: faster clip. Very important when you're rendering graphics, because you know,

247
00:14:39,080 --> 00:14:40,600
Speaker 1: you want to be able to render graphics at a

248
00:14:40,760 --> 00:14:45,040
Speaker 1: really good speed. Humans like in general twenty five frames

249
00:14:45,040 --> 00:14:48,000
Speaker 1: per second. That's a good speed for us. It's hard

250
00:14:48,040 --> 00:14:52,520
Speaker 1: for us to detect uh problems at anything that's faster

251
00:14:52,640 --> 00:14:57,680
Speaker 1: than that. But for high you know, really action oriented

252
00:14:57,720 --> 00:15:00,240
Speaker 1: video games with lots of high speed stuff. Stick Steve

253
00:15:00,240 --> 00:15:04,360
Speaker 1: frames per second is generally considered, that's your goal. You

254
00:15:04,400 --> 00:15:06,360
Speaker 1: want to hit sixty frames per second because you're not

255
00:15:06,360 --> 00:15:09,760
Speaker 1: going to have too much blurring or artifacts or other

256
00:15:09,840 --> 00:15:13,800
Speaker 1: problems if you're able to display information at that speed

257
00:15:14,240 --> 00:15:18,640
Speaker 1: and uh and so that's really important as well. The

258
00:15:19,000 --> 00:15:24,280
Speaker 1: memory clock rate is also important. Memory bandwidth um and uh.

259
00:15:24,360 --> 00:15:30,760
Speaker 1: If you're using an ancient system, ramdack speed. So ramdack

260
00:15:31,040 --> 00:15:34,320
Speaker 1: that's uh. When when the process is handling this information,

261
00:15:34,400 --> 00:15:37,960
Speaker 1: it's all digital, right, it's all ones and zeros. Some

262
00:15:38,040 --> 00:15:42,600
Speaker 1: of us who have older computer systems may have display

263
00:15:42,640 --> 00:15:46,440
Speaker 1: that's not digital. We might have an analog display, which

264
00:15:46,480 --> 00:15:48,960
Speaker 1: means you have to convert those signals from digital to analog.

265
00:15:49,040 --> 00:15:51,640
Speaker 1: That's what the ram dex job is. But these days

266
00:15:52,320 --> 00:15:55,240
Speaker 1: most displays are digital. If you were to go out

267
00:15:55,240 --> 00:15:57,400
Speaker 1: and buy a system now, it would be it would

268
00:15:57,400 --> 00:16:00,320
Speaker 1: have a digital display. Analog just doesn't really that's not

269
00:16:00,360 --> 00:16:02,440
Speaker 1: a thing anymore. So it's really for people who have

270
00:16:02,480 --> 00:16:04,240
Speaker 1: older systems they would have to worry about the ram

271
00:16:04,360 --> 00:16:07,840
Speaker 1: dag issue. In fact, I'm sure that it's being phased

272
00:16:07,840 --> 00:16:11,160
Speaker 1: out of almost everything at this point because I mean,

273
00:16:11,360 --> 00:16:14,520
Speaker 1: you know, why would you be using this older monitor

274
00:16:14,520 --> 00:16:16,160
Speaker 1: when you could go out and get a new display.

275
00:16:17,960 --> 00:16:21,120
Speaker 1: Tune in this week is Captain Atomic takes on the

276
00:16:21,160 --> 00:16:27,160
Speaker 1: evil of ram Dak. Yeah, yeah, different different Ramdackum you

277
00:16:27,200 --> 00:16:29,960
Speaker 1: will phase him out. So and like you were saying that,

278
00:16:30,160 --> 00:16:35,920
Speaker 1: the job of this is to to to the the

279
00:16:36,040 --> 00:16:39,560
Speaker 1: job that we were intending graphics cards to do, I

280
00:16:39,560 --> 00:16:43,720
Speaker 1: should say, was to render graphics. And uh so like

281
00:16:43,800 --> 00:16:46,640
Speaker 1: the polygon issue, that's generally speaking, it tends to be

282
00:16:47,120 --> 00:16:51,320
Speaker 1: triangles a second, and uh and and these these little

283
00:16:51,360 --> 00:16:53,680
Speaker 1: bit tiny triangles are what make up the various surfaces

284
00:16:53,680 --> 00:16:57,160
Speaker 1: you see in represented in computer graphics. Yeah, we're talking

285
00:16:57,200 --> 00:17:02,200
Speaker 1: primarily three dimensional type stuff, three D graphics, so that

286
00:17:02,240 --> 00:17:04,560
Speaker 1: there's at least an illusion of depth. In fact, that's

287
00:17:04,600 --> 00:17:07,960
Speaker 1: really where the graphics cards came out of. You know. Um,

288
00:17:07,960 --> 00:17:11,840
Speaker 1: before that, you might have a GPU in a computer,

289
00:17:12,000 --> 00:17:15,560
Speaker 1: but it would be uh incorporated directly onto the motherboard.

290
00:17:15,640 --> 00:17:18,560
Speaker 1: It wasn't an additional card like you said. The Amiga

291
00:17:18,600 --> 00:17:21,320
Speaker 1: had them. The Amiga is one of the earliest systems

292
00:17:21,359 --> 00:17:25,240
Speaker 1: to have its own GPU. Before that, you could have

293
00:17:25,400 --> 00:17:29,720
Speaker 1: graphics that were vector based graphics, like you know, the

294
00:17:29,920 --> 00:17:34,360
Speaker 1: the video game Asteroids was vector based graphics. Yeah, these

295
00:17:34,359 --> 00:17:38,960
Speaker 1: are games vector vector versus raster we're talking about here, Yes,

296
00:17:39,440 --> 00:17:43,280
Speaker 1: I love those. We can stop giving examples, but I

297
00:17:43,280 --> 00:17:45,680
Speaker 1: guess it helps to explain what the differences between vector

298
00:17:45,720 --> 00:17:50,280
Speaker 1: graphics and raster graphics. Vector graphics graphics are mathematical there,

299
00:17:50,400 --> 00:17:55,040
Speaker 1: they're like line art exactly. It's it's geometrical shapes that

300
00:17:55,080 --> 00:18:01,040
Speaker 1: are based on mathematic uh, calculations. So these are lines

301
00:18:01,080 --> 00:18:05,960
Speaker 1: and curves points. They could be shapes like polygons and uh.

302
00:18:06,040 --> 00:18:09,680
Speaker 1: The nice thing about vector graphics is they are really scalable.

303
00:18:10,280 --> 00:18:14,399
Speaker 1: You can it's dependent upon what your machine can do,

304
00:18:14,640 --> 00:18:18,359
Speaker 1: not dependent upon the file. Uh. So in other words,

305
00:18:18,840 --> 00:18:21,920
Speaker 1: you can a vector graphic at a certain size is

306
00:18:21,960 --> 00:18:24,840
Speaker 1: going to look just as good or not as good

307
00:18:25,320 --> 00:18:28,600
Speaker 1: as it will be at a size many many manion

308
00:18:28,640 --> 00:18:31,840
Speaker 1: times larger than that. It all depends on the equipment,

309
00:18:31,880 --> 00:18:35,399
Speaker 1: not on the graphics file. Yeah. Now, now, raster graphic

310
00:18:35,480 --> 00:18:39,840
Speaker 1: or bitmapped graphics UM is made up of basically made

311
00:18:39,880 --> 00:18:42,680
Speaker 1: up of little dots and um. That's the way are

312
00:18:42,880 --> 00:18:47,000
Speaker 1: are our printers work, That's the way our displays works.

313
00:18:47,080 --> 00:18:51,920
Speaker 1: So this is raster graphics that's much more common. Yeah.

314
00:18:52,040 --> 00:18:55,760
Speaker 1: So well, if you talk about a good example, I

315
00:18:55,800 --> 00:18:59,600
Speaker 1: think for what you were mentioning was photos. Yeah, if

316
00:18:59,640 --> 00:19:02,240
Speaker 1: you've ever taking a photo, put on your computer and

317
00:19:02,280 --> 00:19:05,440
Speaker 1: decided you wanted to blow it up, you may notice

318
00:19:05,600 --> 00:19:10,160
Speaker 1: that it starts to look rather grainy, block and yeah

319
00:19:10,320 --> 00:19:12,800
Speaker 1: or blocky. You can see a little sharp angles where

320
00:19:12,800 --> 00:19:16,520
Speaker 1: the which or the outlines of the pixels really yeah,

321
00:19:16,560 --> 00:19:18,560
Speaker 1: and and then and what it's doing is the computer

322
00:19:18,720 --> 00:19:23,840
Speaker 1: is basically having to make up information in between the

323
00:19:23,920 --> 00:19:27,560
Speaker 1: little dots of information that came with the picture. So

324
00:19:28,000 --> 00:19:30,440
Speaker 1: you know, it's it's going, well, that's kind of green

325
00:19:30,520 --> 00:19:34,440
Speaker 1: and this kind of orange. All I'll split the difference

326
00:19:35,200 --> 00:19:38,000
Speaker 1: and uh and uh, you know, when you make something smaller,

327
00:19:38,040 --> 00:19:41,840
Speaker 1: it's easy because you're compressing the information. It can go,

328
00:19:41,880 --> 00:19:43,880
Speaker 1: oh well, I don't need these pixels, so I'll toss

329
00:19:43,960 --> 00:19:46,520
Speaker 1: them out and uh, you know things that you can

330
00:19:46,560 --> 00:19:50,520
Speaker 1: only shrink a photo without distorting it, but expanding it

331
00:19:50,520 --> 00:19:54,720
Speaker 1: as a different question and a graphics vector graphics aren't

332
00:19:54,720 --> 00:19:57,080
Speaker 1: like that because when you blow something up, it goes,

333
00:19:57,080 --> 00:19:59,400
Speaker 1: oh well, basically it's this point in that point, I'm

334
00:19:59,800 --> 00:20:04,119
Speaker 1: you do the math. It looks fine. So the games

335
00:20:04,160 --> 00:20:06,800
Speaker 1: and stuff that we play, those are raster graphics. That's

336
00:20:06,800 --> 00:20:09,280
Speaker 1: where we've got those those images that are made up

337
00:20:09,280 --> 00:20:14,360
Speaker 1: of pixels. So, um, you know, that's trying to create

338
00:20:14,560 --> 00:20:22,280
Speaker 1: a uh, a detailed graphic representation of a character, especially

339
00:20:22,359 --> 00:20:24,320
Speaker 1: character that's in motion and doing lots of stuff at

340
00:20:24,320 --> 00:20:27,200
Speaker 1: a at a good frame rate. That's that's a pretty

341
00:20:27,200 --> 00:20:31,040
Speaker 1: intensive process as far as processing power is concerned. Um.

342
00:20:31,119 --> 00:20:36,240
Speaker 1: And that's why we had companies start to develop graphics

343
00:20:36,240 --> 00:20:39,760
Speaker 1: cards to help take this load off the onboard graphics

344
00:20:39,840 --> 00:20:43,040
Speaker 1: chips that computers had because they just weren't up to

345
00:20:43,080 --> 00:20:45,800
Speaker 1: the task of processing that amount of information. Which meant

346
00:20:45,800 --> 00:20:48,760
Speaker 1: that if you went out and you had a regular PC,

347
00:20:48,920 --> 00:20:52,600
Speaker 1: will off the shelf PC, no additional graphics card other

348
00:20:52,640 --> 00:20:56,199
Speaker 1: than what the onboard graphics card was, uh, and you

349
00:20:56,280 --> 00:20:59,679
Speaker 1: come home, plug it in and you load up that

350
00:21:00,560 --> 00:21:04,120
Speaker 1: first person shooter game that's brand new, it would not

351
00:21:04,359 --> 00:21:07,359
Speaker 1: run very well. You would have to run it at

352
00:21:07,280 --> 00:21:11,040
Speaker 1: a very low detail quality, so that it's having to

353
00:21:11,160 --> 00:21:14,840
Speaker 1: do fewer polygons on the screen at a time because

354
00:21:14,840 --> 00:21:16,639
Speaker 1: it just didn't have the processing power to keep up.

355
00:21:17,480 --> 00:21:20,800
Speaker 1: Or you'd have to really and or you'd have to

356
00:21:20,840 --> 00:21:24,520
Speaker 1: really reduce the frame rate so that the actual experience

357
00:21:24,600 --> 00:21:28,280
Speaker 1: is not as smooth or pleasant. And you might think, wow,

358
00:21:28,400 --> 00:21:31,000
Speaker 1: this is a terrible game, but really it's because the

359
00:21:31,000 --> 00:21:33,639
Speaker 1: hardware you're running on isn't up to the task of

360
00:21:33,680 --> 00:21:37,520
Speaker 1: running the game at the the experience that the game

361
00:21:37,560 --> 00:21:42,159
Speaker 1: developers intended. So the graphics cards were what allowed you

362
00:21:42,240 --> 00:21:45,399
Speaker 1: to plug it in, usually in a well. Back in

363
00:21:45,440 --> 00:21:48,720
Speaker 1: the day, it was PC I or UM white a

364
00:21:48,800 --> 00:21:52,080
Speaker 1: g P. Yeah, those were the two card slots to

365
00:21:52,080 --> 00:21:55,760
Speaker 1: would plug them into. Today it's PC I Express almost

366
00:21:55,800 --> 00:22:00,200
Speaker 1: exclusively um where which has got a It's it's a

367
00:22:00,240 --> 00:22:04,879
Speaker 1: more efficient and faster way of transferring information from a

368
00:22:04,920 --> 00:22:09,040
Speaker 1: card to the motherboard and uh and out to a display.

369
00:22:09,160 --> 00:22:12,399
Speaker 1: So you would plug you would plug this in, and

370
00:22:12,440 --> 00:22:15,280
Speaker 1: it was designed to be pretty easy, I mean relatively speaking.

371
00:22:15,320 --> 00:22:18,879
Speaker 1: You're not soldering or anything, opening up a computer case,

372
00:22:19,200 --> 00:22:23,200
Speaker 1: sliding a card in an expansion slot, screwing it tight

373
00:22:23,280 --> 00:22:26,919
Speaker 1: so that it's not gonna wiggle in the case, close

374
00:22:26,960 --> 00:22:31,440
Speaker 1: it back up, plug in your your peripherals, and you're

375
00:22:31,440 --> 00:22:35,600
Speaker 1: good to go. It's a little more complicated now because

376
00:22:35,640 --> 00:22:40,520
Speaker 1: if you're getting a state of the art, fastest possible

377
00:22:40,680 --> 00:22:43,720
Speaker 1: processor you can get for a GP, especially you're getting

378
00:22:43,760 --> 00:22:45,600
Speaker 1: two of them, because there are a lot of motherboards

379
00:22:45,600 --> 00:22:50,600
Speaker 1: out there now that support dual graphics cards. Um, some

380
00:22:50,680 --> 00:22:53,600
Speaker 1: of these are so powerful that they require more power

381
00:22:53,600 --> 00:22:57,119
Speaker 1: than what the motherboard can supply. More power. Yeah, so

382
00:22:57,160 --> 00:23:00,280
Speaker 1: they have to plug into your computer's actual powers apply,

383
00:23:00,720 --> 00:23:05,480
Speaker 1: not the motherboard. And uh, particularly if they are so

384
00:23:05,560 --> 00:23:09,800
Speaker 1: powerful that they have their own onboard fans. The fan

385
00:23:09,960 --> 00:23:13,400
Speaker 1: used to help distribute the hot air so that the

386
00:23:13,440 --> 00:23:16,800
Speaker 1: processor does not overheat. Yeah, well that's that's one of

387
00:23:16,880 --> 00:23:22,400
Speaker 1: the problems that that prevent high end graphics cards from

388
00:23:22,440 --> 00:23:27,760
Speaker 1: being included with some laptops, which is simply because it's

389
00:23:27,840 --> 00:23:30,800
Speaker 1: easier to put in a high end graphics card that

390
00:23:30,840 --> 00:23:35,040
Speaker 1: requires its own fan in a tower case, really a

391
00:23:35,080 --> 00:23:39,919
Speaker 1: towered desktop computer case, because there's simply more room for it. Um.

392
00:23:40,200 --> 00:23:44,600
Speaker 1: Less room on a laptop in general, less ability to

393
00:23:45,320 --> 00:23:47,840
Speaker 1: find a way to disperse that extra heat. So not

394
00:23:47,920 --> 00:23:51,040
Speaker 1: only are you're you dealing with the main central processing

395
00:23:51,080 --> 00:23:54,680
Speaker 1: unit CPU, you're also dealing with the GPU. And you're going, wow,

396
00:23:54,880 --> 00:23:57,000
Speaker 1: this is going to burn someone's legs off when they

397
00:23:57,080 --> 00:23:59,080
Speaker 1: put it on their lap because I don't actually have

398
00:23:59,080 --> 00:24:00,600
Speaker 1: a good way to disperse a seat, or it's going

399
00:24:00,680 --> 00:24:02,879
Speaker 1: to cause the machine to shut down after twenty minutes

400
00:24:02,920 --> 00:24:06,600
Speaker 1: of working, which is probably not quite as bad as

401
00:24:06,600 --> 00:24:11,080
Speaker 1: burning somebody, but it's irritating, irritating, but not disfiguring. So

402
00:24:11,560 --> 00:24:14,639
Speaker 1: there's that you take that home. But but yeah, I

403
00:24:14,680 --> 00:24:18,240
Speaker 1: mean and and uh also easier to uh to get

404
00:24:18,400 --> 00:24:22,000
Speaker 1: into a desktop machine to swap a card or add

405
00:24:22,040 --> 00:24:24,720
Speaker 1: a card um, as you were pointing out, some you

406
00:24:24,760 --> 00:24:28,600
Speaker 1: know support multiple Um. Then it would be necessarily with

407
00:24:28,680 --> 00:24:32,119
Speaker 1: the laptop, especially something like an ultra book that is

408
00:24:33,000 --> 00:24:35,960
Speaker 1: put together just so so that it makes that nice

409
00:24:36,000 --> 00:24:39,359
Speaker 1: slim form factor that those are pretty much you're you're

410
00:24:39,400 --> 00:24:44,240
Speaker 1: out of luck there. Um. The yeah, it's uh, it's

411
00:24:44,400 --> 00:24:48,240
Speaker 1: the the graphics cards tend to be more of a

412
00:24:48,320 --> 00:24:53,640
Speaker 1: desktop type of component than laptop. And I mean, for

413
00:24:53,640 --> 00:24:57,239
Speaker 1: for the reasons you you pointed out, um, which is

414
00:24:57,280 --> 00:25:02,359
Speaker 1: also one of the reasons why uh, serious hardcore video

415
00:25:02,400 --> 00:25:06,919
Speaker 1: gamers on the PC tend to prefer desktop setups rather

416
00:25:06,960 --> 00:25:09,439
Speaker 1: than a laptop. Because even if you go out and

417
00:25:09,480 --> 00:25:11,800
Speaker 1: you buy a state of the art gaming laptop right

418
00:25:11,840 --> 00:25:14,760
Speaker 1: now for thousands of dollars, and it might it might

419
00:25:14,880 --> 00:25:19,919
Speaker 1: right now rival the best desktop computers out there. The

420
00:25:19,920 --> 00:25:22,680
Speaker 1: advantage of the desktop computers have is that, in general,

421
00:25:22,880 --> 00:25:29,600
Speaker 1: their machines are much more customizable and easily modified than laptops. There.

422
00:25:29,600 --> 00:25:31,960
Speaker 1: I'm sure there are exceptions out there, but that's the

423
00:25:32,000 --> 00:25:34,520
Speaker 1: general rule, and it's a pretty safe one because it's

424
00:25:34,520 --> 00:25:37,920
Speaker 1: just tends to be a lot easier to fiddle with

425
00:25:38,040 --> 00:25:40,359
Speaker 1: stuff that's inside a desktop case because you've got a

426
00:25:40,400 --> 00:25:43,120
Speaker 1: lot more space there than you do with a laptop.

427
00:25:43,440 --> 00:25:47,240
Speaker 1: There may be cards that a laptop just physically cannot

428
00:25:47,320 --> 00:25:50,400
Speaker 1: accommodate because there's just not the room. And if you've

429
00:25:50,440 --> 00:25:52,879
Speaker 1: seen some of these video cards, it's not hard to

430
00:25:52,920 --> 00:25:56,719
Speaker 1: imagine because some of them look like they are you know,

431
00:25:56,920 --> 00:25:58,639
Speaker 1: some of them look like they came straight out of

432
00:25:58,640 --> 00:26:02,399
Speaker 1: a video game, like this looks like it's a magazine

433
00:26:02,440 --> 00:26:06,600
Speaker 1: clip for a futuristic machine gun or something. Uh. Some

434
00:26:06,680 --> 00:26:08,240
Speaker 1: of them am like wow, I don't know that I

435
00:26:08,280 --> 00:26:11,400
Speaker 1: have a computer pretty enough to install this thing in.

436
00:26:11,960 --> 00:26:13,600
Speaker 1: And in fact, some of them, I would argue the

437
00:26:14,040 --> 00:26:18,080
Speaker 1: design here. They've created aesthetic for the video card, which

438
00:26:18,680 --> 00:26:22,600
Speaker 1: never used to be a concern. Right. These were cards

439
00:26:22,600 --> 00:26:25,040
Speaker 1: that just fit inside a computer case. You never saw

440
00:26:25,040 --> 00:26:27,160
Speaker 1: them unless you had to open the case. But now

441
00:26:27,200 --> 00:26:30,560
Speaker 1: we've got these cases that have exposed sides. So suddenly

442
00:26:30,640 --> 00:26:33,080
Speaker 1: the aesthetic design of the graphics card has become a

443
00:26:33,080 --> 00:26:37,480
Speaker 1: consideration well also, um, people who play a lot of

444
00:26:37,920 --> 00:26:42,160
Speaker 1: I would argue this is probably the case for uh,

445
00:26:42,359 --> 00:26:46,600
Speaker 1: hardcore serious gamers. Two, is it's nice that it looks

446
00:26:46,600 --> 00:26:49,480
Speaker 1: cool because you're buying this, you know, updated graphics card

447
00:26:49,480 --> 00:26:51,000
Speaker 1: that's going to be able to do all this extra

448
00:26:51,400 --> 00:26:54,400
Speaker 1: pixel rendering, and you know you want to look cool.

449
00:26:54,640 --> 00:26:57,399
Speaker 1: And some of these graphics cards cost as much or

450
00:26:57,480 --> 00:27:00,960
Speaker 1: more than video game consoles, which brings me back to

451
00:27:01,240 --> 00:27:04,800
Speaker 1: the whole thing about saying, you know, video cards are

452
00:27:04,840 --> 00:27:08,399
Speaker 1: what made me a console gamer. It's not quite as

453
00:27:08,440 --> 00:27:10,560
Speaker 1: simple as that, but it is very close in the

454
00:27:10,600 --> 00:27:15,560
Speaker 1: sense that I saw that, you know, because I was

455
00:27:15,600 --> 00:27:19,320
Speaker 1: there when the video game cards, the standalone graphics cards

456
00:27:19,359 --> 00:27:23,080
Speaker 1: started to become a thing, um, and I could see

457
00:27:23,080 --> 00:27:26,399
Speaker 1: that within six months to a year, the cards that

458
00:27:26,480 --> 00:27:29,480
Speaker 1: had been out on the market were no longer able

459
00:27:29,560 --> 00:27:34,440
Speaker 1: to run the most current games at their ideal settings.

460
00:27:35,280 --> 00:27:37,199
Speaker 1: And that was a pretty rough wake up call to

461
00:27:37,280 --> 00:27:40,119
Speaker 1: me saying that even if I had saved up the

462
00:27:40,160 --> 00:27:43,040
Speaker 1: money to buy a video card for my computer and

463
00:27:43,080 --> 00:27:45,520
Speaker 1: I had swapped it out, and I thought, wow, this

464
00:27:45,560 --> 00:27:48,359
Speaker 1: is an amazing experience. Six months to a year down

465
00:27:48,400 --> 00:27:51,679
Speaker 1: the road, I would feel like I'd have to do

466
00:27:51,720 --> 00:27:54,480
Speaker 1: that all over again just to keep up with the

467
00:27:54,560 --> 00:27:58,200
Speaker 1: video games that were coming out at that time. It's

468
00:27:58,200 --> 00:28:01,040
Speaker 1: a constant cycle, or can be a concert cycle of

469
00:28:01,080 --> 00:28:04,680
Speaker 1: needing more memory and a new processor and a new

470
00:28:04,680 --> 00:28:08,320
Speaker 1: graphics processor with more memory, and having to upgrade one

471
00:28:08,440 --> 00:28:10,520
Speaker 1: or the other or all of the above. Right you

472
00:28:10,520 --> 00:28:12,359
Speaker 1: you eventually get to a point, like we said, where

473
00:28:12,359 --> 00:28:15,520
Speaker 1: the motherboard can't support the latest graphics card. Then you're like,

474
00:28:15,520 --> 00:28:19,280
Speaker 1: all right, now I gotta get a new, entirely new system. Uh.

475
00:28:19,320 --> 00:28:21,440
Speaker 1: And so now, if I were the kind of person

476
00:28:21,520 --> 00:28:25,200
Speaker 1: who I was constantly building computers, that might not be

477
00:28:25,240 --> 00:28:28,560
Speaker 1: as big a deal because I could replace components as

478
00:28:28,600 --> 00:28:31,560
Speaker 1: I needed to. And one can argue that maybe through

479
00:28:31,560 --> 00:28:35,280
Speaker 1: the lifetime of a video game PC you spend less

480
00:28:35,280 --> 00:28:38,000
Speaker 1: money than you would with a console game, depending upon

481
00:28:38,000 --> 00:28:40,400
Speaker 1: how much gaming you do, because console games do tend

482
00:28:40,480 --> 00:28:43,200
Speaker 1: to be more expensive, and they do tend to be

483
00:28:43,280 --> 00:28:47,760
Speaker 1: more expensive for longer than PC games. PC games get

484
00:28:47,800 --> 00:28:53,000
Speaker 1: marked down, generally speaking, faster than console games do, so

485
00:28:53,680 --> 00:28:55,640
Speaker 1: there there is an argument to be made that you

486
00:28:55,680 --> 00:28:58,360
Speaker 1: could save money over the long term. If you are

487
00:28:58,480 --> 00:29:01,240
Speaker 1: an avid gamer, you buy a lot of games. Um,

488
00:29:01,280 --> 00:29:03,760
Speaker 1: if you go with a PC gaming rig, especially if

489
00:29:03,760 --> 00:29:06,200
Speaker 1: you decide to build one yourself, as opposed to buying

490
00:29:06,280 --> 00:29:09,200
Speaker 1: something that was customized off the shelf, like if you

491
00:29:09,240 --> 00:29:12,400
Speaker 1: actually go out and get the components. Um. It also

492
00:29:12,440 --> 00:29:15,360
Speaker 1: gives the developer a little bit of an added advantage too,

493
00:29:15,400 --> 00:29:18,760
Speaker 1: because if they're developing a title for say the PlayStation three,

494
00:29:19,560 --> 00:29:21,680
Speaker 1: they know exactly what kind of graphics card is in there,

495
00:29:22,160 --> 00:29:25,600
Speaker 1: so they can play to its strengths. Yeah. And actually

496
00:29:25,680 --> 00:29:27,800
Speaker 1: that was a real problem early on too, is that

497
00:29:27,840 --> 00:29:30,920
Speaker 1: you would get games that would be compatible with certain

498
00:29:30,960 --> 00:29:34,840
Speaker 1: graphics cards but not others. And then you know, you

499
00:29:34,920 --> 00:29:38,800
Speaker 1: might have a state of the art gaming computer back

500
00:29:38,840 --> 00:29:40,840
Speaker 1: in the mid nineties, and then find out that even

501
00:29:40,840 --> 00:29:42,640
Speaker 1: though you have a state of art gaming computer, you

502
00:29:42,760 --> 00:29:45,400
Speaker 1: still can't play certain games because they are not compatible

503
00:29:45,440 --> 00:29:48,720
Speaker 1: with the the hardware you have. Um. That's when you

504
00:29:48,720 --> 00:29:50,840
Speaker 1: shake your fist in the air and screen con Yeah.

505
00:29:50,840 --> 00:29:58,280
Speaker 1: We should also mention that effectively, the video card industry

506
00:29:58,600 --> 00:30:03,920
Speaker 1: is a duopoli. There are two companies that that's essentially

507
00:30:04,000 --> 00:30:08,200
Speaker 1: that's who you go to. Yeah, there's Nvidia and there's

508
00:30:08,320 --> 00:30:11,120
Speaker 1: a m D Yeah, also known as a t I.

509
00:30:11,200 --> 00:30:13,280
Speaker 1: Back in the day, but A m D purchased them,

510
00:30:13,320 --> 00:30:16,920
Speaker 1: acquired a t I. There are there are others until

511
00:30:17,040 --> 00:30:22,320
Speaker 1: makes some graphics cards, um, you know, and they're they're

512
00:30:22,360 --> 00:30:25,640
Speaker 1: really others. But yeah, when when we're talking about this, uh,

513
00:30:25,760 --> 00:30:28,400
Speaker 1: probably the listeners that wanted us to talk about this

514
00:30:28,400 --> 00:30:31,440
Speaker 1: would already know that those two are the big players

515
00:30:32,040 --> 00:30:35,760
Speaker 1: and they they're in a constant game of one upsmanship. Uh.

516
00:30:36,120 --> 00:30:39,720
Speaker 1: The fastest card I could find so far, at least

517
00:30:39,720 --> 00:30:42,640
Speaker 1: as far as clock speed goes, not performance. Actually, as

518
00:30:42,640 --> 00:30:45,040
Speaker 1: far as performance goes, is not the fastest. But with

519
00:30:45,200 --> 00:30:50,880
Speaker 1: clock speed, uh is uh without over clocking is uh?

520
00:30:50,920 --> 00:30:53,560
Speaker 1: The A m D Radio on h D seven seven

521
00:30:53,640 --> 00:30:57,520
Speaker 1: seven zero Giga Hurts edition, which it's one giga Hurts

522
00:30:57,560 --> 00:31:03,360
Speaker 1: GPU and um it again that's without overclocking. But uh,

523
00:31:03,840 --> 00:31:06,200
Speaker 1: that's not that fast. But yeah, when it when all

524
00:31:06,240 --> 00:31:10,160
Speaker 1: it's doing, when it's handling primarily graphics, having that dedicated

525
00:31:10,320 --> 00:31:13,880
Speaker 1: processing power behind it, you know that don't sneeze at

526
00:31:13,880 --> 00:31:17,960
Speaker 1: the one gigga hurt. Now again, through the testing, it

527
00:31:18,320 --> 00:31:22,200
Speaker 1: did not perform at the same level as some of

528
00:31:22,240 --> 00:31:26,160
Speaker 1: the top of the line, most expensive cards that are

529
00:31:26,200 --> 00:31:28,920
Speaker 1: out there. Uh. And part of that is just because

530
00:31:28,960 --> 00:31:31,640
Speaker 1: you know, the Again, it's it's more than just the

531
00:31:31,680 --> 00:31:36,360
Speaker 1: clock speed. We talked about that. So that's that's kind

532
00:31:36,360 --> 00:31:38,880
Speaker 1: of the roundup on graphics cards in general, Like we

533
00:31:38,920 --> 00:31:42,280
Speaker 1: had said, they're being used to process information beyond just

534
00:31:42,440 --> 00:31:46,040
Speaker 1: graphics because it's got this parallel process and capability, so

535
00:31:46,560 --> 00:31:49,760
Speaker 1: you can use it for things that require lots of

536
00:31:50,080 --> 00:31:54,640
Speaker 1: parallel computational problems, things that you know, there's a lot

537
00:31:54,680 --> 00:31:59,520
Speaker 1: of scientific and medical applications for that kind of technology,

538
00:31:59,560 --> 00:32:02,720
Speaker 1: but the one that we talked about most recently is

539
00:32:03,000 --> 00:32:08,280
Speaker 1: cryptography and cracking codes and using parallel processing to run

540
00:32:09,880 --> 00:32:12,440
Speaker 1: a brute force attack where you're just running lots and

541
00:32:12,480 --> 00:32:16,120
Speaker 1: lots of variations on various passwords to try and determine

542
00:32:16,680 --> 00:32:20,360
Speaker 1: what the hashing algorithm is for a for for a

543
00:32:20,360 --> 00:32:25,920
Speaker 1: password site. So the unless the operating system is specifically

544
00:32:25,960 --> 00:32:29,560
Speaker 1: assigning it tasks, you would have to have software that

545
00:32:29,920 --> 00:32:34,320
Speaker 1: specifically directs information to the graphics card. And um, the

546
00:32:34,360 --> 00:32:38,880
Speaker 1: people who yeah, there there are programs that specifically tell

547
00:32:38,920 --> 00:32:41,840
Speaker 1: the GPU to to go to work on these these problems.

548
00:32:41,880 --> 00:32:45,920
Speaker 1: So um, and then it's not specifically specific enough to

549
00:32:46,160 --> 00:32:48,520
Speaker 1: where the hardware is. It's like, well, it's not a

550
00:32:48,520 --> 00:32:51,400
Speaker 1: piece of graphics, so I can't do this. You know,

551
00:32:51,440 --> 00:32:55,600
Speaker 1: it's very capable of handling that. So alright, So I

552
00:32:55,640 --> 00:32:58,200
Speaker 1: think that wraps up this discussion on video cards. I

553
00:32:58,240 --> 00:33:01,880
Speaker 1: hope that that was That was satisfying to our listeners

554
00:33:01,880 --> 00:33:07,640
Speaker 1: who were asking us repeatedly to cover it. Actually, you know,

555
00:33:07,720 --> 00:33:10,240
Speaker 1: I sound like I'm I'm exhausted by it. I'm not.

556
00:33:10,320 --> 00:33:13,120
Speaker 1: I really I like seeing enthusiasm from our listeners. That's

557
00:33:13,200 --> 00:33:16,680
Speaker 1: very It's a good thing. So if you are enthusiastically

558
00:33:16,760 --> 00:33:19,560
Speaker 1: interested in a particular topic and we have not covered it,

559
00:33:19,760 --> 00:33:21,680
Speaker 1: or perhaps we covered it a long time ago and

560
00:33:21,760 --> 00:33:25,520
Speaker 1: it's darn near time that we recover it, let us know.

561
00:33:26,160 --> 00:33:28,000
Speaker 1: Send us a message. You can send us an email

562
00:33:28,000 --> 00:33:31,239
Speaker 1: our addresses tech stuff at Discovery dot com, or let

563
00:33:31,320 --> 00:33:33,320
Speaker 1: us know on Facebook or Twitter. Our handle at both

564
00:33:33,320 --> 00:33:35,800
Speaker 1: of those is text stuff h sw and Chris and

565
00:33:35,800 --> 00:33:39,560
Speaker 1: I will talk to you again really soon for more

566
00:33:39,560 --> 00:33:41,880
Speaker 1: on this and thousands of other topics because it has

567
00:33:41,920 --> 00:33:48,080
Speaker 1: staff works dot com. See guys, I told you we

568
00:33:48,200 --> 00:33:51,280
Speaker 1: talked to you again really soon. That really soon is

569
00:33:51,400 --> 00:33:54,080
Speaker 1: right now. I'm just reminding you that we have our

570
00:33:54,080 --> 00:33:57,680
Speaker 1: photo upload widget live on the site at www dot

571
00:33:57,680 --> 00:34:01,000
Speaker 1: How stuff works dot com slash upgrade your tech Toyota

572
00:34:01,080 --> 00:34:04,160
Speaker 1: is giving us the chance to let you share your creativity.

573
00:34:04,200 --> 00:34:08,080
Speaker 1: So send us those pictures of your modifications, your tech ideas,

574
00:34:08,280 --> 00:34:11,040
Speaker 1: those gadgets that you've created, all those hacks. If you're

575
00:34:11,080 --> 00:34:14,680
Speaker 1: steampunking everything in sight, put on your goggles and show

576
00:34:14,760 --> 00:34:17,560
Speaker 1: that to us. We can't wait to see them. Brought

577
00:34:17,600 --> 00:34:20,800
Speaker 1: to you by the reinvented two thousand twelve camera. It's ready,

578
00:34:20,960 --> 00:34:21,399
Speaker 1: are you