1
00:00:04,120 --> 00:00:07,160
Speaker 1: Get in touch with technology with tech Stuff from how

2
00:00:07,200 --> 00:00:13,640
Speaker 1: stuff works dot com. Hey there, and welcome to tech Stuff.

3
00:00:13,680 --> 00:00:16,160
Speaker 1: I'm your host job in Strickland. I'm an executive producer

4
00:00:16,200 --> 00:00:18,560
Speaker 1: with how Stuff Works, and I love all things tech.

5
00:00:18,960 --> 00:00:21,439
Speaker 1: And we left off in our last episode at a

6
00:00:21,520 --> 00:00:25,200
Speaker 1: point when in Video, a brand new company at that point,

7
00:00:25,320 --> 00:00:28,320
Speaker 1: was in bad shape at a shaky start. So if

8
00:00:28,360 --> 00:00:32,159
Speaker 1: you have not listened to the last episode where I

9
00:00:32,200 --> 00:00:36,120
Speaker 1: introduced the concept of graphics chips and cards and in video,

10
00:00:36,520 --> 00:00:40,080
Speaker 1: go listen to that one first. Anyway, In Video's first product,

11
00:00:40,120 --> 00:00:44,760
Speaker 1: the n V one, had not really sold super well

12
00:00:44,800 --> 00:00:48,519
Speaker 1: in the market. It hadn't really captured a whole lot

13
00:00:48,520 --> 00:00:51,680
Speaker 1: of market share, and the company had to cancel the

14
00:00:51,720 --> 00:00:55,760
Speaker 1: in V two chip after it's big partners, Sega, announced

15
00:00:55,760 --> 00:00:58,280
Speaker 1: it was going with a different chip designer for the

16
00:00:58,320 --> 00:01:03,400
Speaker 1: graphics for the upcoming dream Cast console. But the engineers

17
00:01:03,400 --> 00:01:06,320
Speaker 1: at Nvidia rolled with the punches and they went back

18
00:01:06,319 --> 00:01:09,400
Speaker 1: to the drawing board and designed a new graphics chip

19
00:01:09,480 --> 00:01:13,800
Speaker 1: that would rely upon the humble triangular polygon rather than

20
00:01:13,840 --> 00:01:18,440
Speaker 1: the quadrilateralals that the Envy one employed for texture mapping.

21
00:01:19,040 --> 00:01:22,400
Speaker 1: This would make their new card compatible with Windows based

22
00:01:22,560 --> 00:01:26,480
Speaker 1: machines that were running the direct three D a p I,

23
00:01:27,000 --> 00:01:31,559
Speaker 1: and the new card would include polygone texture mapping, which

24
00:01:31,600 --> 00:01:34,440
Speaker 1: reduced the amount of time it took for the graphics

25
00:01:34,440 --> 00:01:37,360
Speaker 1: card to render frames, but it had a negative impact

26
00:01:37,440 --> 00:01:40,720
Speaker 1: on image quality. So what the heck is texture mapping?

27
00:01:41,080 --> 00:01:45,000
Speaker 1: Does that even mean? It might sound like gobbledygook, Well

28
00:01:45,000 --> 00:01:47,800
Speaker 1: it pretty much is what it sounds like. It involves

29
00:01:47,800 --> 00:01:51,960
Speaker 1: taking a two dimensional surface called a texture map. Think

30
00:01:51,960 --> 00:01:55,000
Speaker 1: of it like wallpaper or wrapping paper. Think of a

31
00:01:55,040 --> 00:01:59,720
Speaker 1: pattern on that, and then virtually wrapping a three dimensional

32
00:02:00,000 --> 00:02:04,280
Speaker 1: object with that texture. So it's it's a virtual three

33
00:02:04,280 --> 00:02:08,040
Speaker 1: dimensional object. So it's kind of like taking wrapping paper

34
00:02:08,360 --> 00:02:11,120
Speaker 1: and wrapping it around a box. Let's say you've got

35
00:02:11,200 --> 00:02:13,360
Speaker 1: a box, You've got a gift you're gonna give somebody,

36
00:02:13,680 --> 00:02:16,160
Speaker 1: You have some wrapping paper. You have to take this

37
00:02:16,240 --> 00:02:19,120
Speaker 1: wrapping paper that's in a two dimensional format, that being

38
00:02:19,160 --> 00:02:21,480
Speaker 1: a flat sheet of paper, and wrap it around a

39
00:02:21,480 --> 00:02:27,480
Speaker 1: three dimensional object, that being a rectangular cubicle type box.

40
00:02:28,280 --> 00:02:30,880
Speaker 1: And you have to fold the pattern around the contours

41
00:02:30,880 --> 00:02:32,400
Speaker 1: of the box in order to wrap it, which is

42
00:02:32,760 --> 00:02:36,120
Speaker 1: easier to do with simple geometric shapes, but as you

43
00:02:36,160 --> 00:02:39,800
Speaker 1: introduce more complicated shapes, it gets harder to do very well.

44
00:02:40,200 --> 00:02:42,760
Speaker 1: And it doesn't take very much to do to make

45
00:02:42,760 --> 00:02:45,000
Speaker 1: this really hard. I mean you can if you wanted

46
00:02:45,040 --> 00:02:47,680
Speaker 1: to map a simple two D shape like a pattern

47
00:02:47,680 --> 00:02:51,120
Speaker 1: of squares on a three D object like a sphere,

48
00:02:51,639 --> 00:02:54,720
Speaker 1: you're gonna run into problems because if you took an

49
00:02:54,720 --> 00:02:58,160
Speaker 1: actual sphere, like a physical sphere, and you had a

50
00:02:58,160 --> 00:03:01,600
Speaker 1: wrapping paper that had boxes on it, like let's say

51
00:03:01,639 --> 00:03:05,000
Speaker 1: it's just black and white boxes on it, a checked

52
00:03:05,000 --> 00:03:09,360
Speaker 1: pattern almost, and you wrap that sphere with that wrapping paper,

53
00:03:09,400 --> 00:03:11,320
Speaker 1: you you're gonna find that it's going to require you

54
00:03:11,360 --> 00:03:13,520
Speaker 1: to cut the paper in such a way that some

55
00:03:13,600 --> 00:03:17,280
Speaker 1: point on the surface of the sphere the pattern would

56
00:03:17,400 --> 00:03:20,400
Speaker 1: no longer be continuous. There'd be a break in the pattern.

57
00:03:20,800 --> 00:03:23,560
Speaker 1: The same thing can happen with texture mapping, and it

58
00:03:23,600 --> 00:03:26,680
Speaker 1: means that your graphic quality can suffer as a result.

59
00:03:26,760 --> 00:03:31,280
Speaker 1: It can become noticeable. The political texture mapping strategy did

60
00:03:31,320 --> 00:03:35,400
Speaker 1: require less processing power to cover three D objects with textures, however,

61
00:03:35,840 --> 00:03:39,280
Speaker 1: so you have a trade off. The quality of the

62
00:03:39,320 --> 00:03:43,000
Speaker 1: graphics might not be quite as high, but you can

63
00:03:43,080 --> 00:03:46,080
Speaker 1: render them much more quickly, which could improve things like

64
00:03:46,200 --> 00:03:49,560
Speaker 1: the refresh rate. While in Video was working on this,

65
00:03:49,880 --> 00:03:52,080
Speaker 1: another company came out with a product that had a

66
00:03:52,160 --> 00:03:55,600
Speaker 1: big impact on the graphics card industry, and it also

67
00:03:55,720 --> 00:03:58,640
Speaker 1: led to conditions that convinced me to get out of

68
00:03:58,720 --> 00:04:03,360
Speaker 1: PC gaming for quite some time. Not putting blame here,

69
00:04:03,920 --> 00:04:08,240
Speaker 1: just that it created an environment that I got increasingly

70
00:04:10,080 --> 00:04:14,160
Speaker 1: disenchanted with the company, and their product was not a

71
00:04:14,160 --> 00:04:16,000
Speaker 1: bad one, by the way. The company was three d

72
00:04:16,320 --> 00:04:21,040
Speaker 1: f X Interactive. It launched a year after Video had incorporated,

73
00:04:21,040 --> 00:04:25,760
Speaker 1: so he launched in and they introduced a three D

74
00:04:25,880 --> 00:04:30,279
Speaker 1: accelerated graphics card called the Voodoo. And this card was

75
00:04:30,360 --> 00:04:33,600
Speaker 1: different from other graphics cards that had come before it

76
00:04:34,040 --> 00:04:37,520
Speaker 1: because up to that point, graphics cards would typically integrate

77
00:04:37,760 --> 00:04:41,640
Speaker 1: two D and three D graphics processing on a single motherboard.

78
00:04:42,040 --> 00:04:46,359
Speaker 1: But three d FX created a card solely dedicated to

79
00:04:46,520 --> 00:04:50,400
Speaker 1: three D graphics. So you could not just have a

80
00:04:50,480 --> 00:04:53,039
Speaker 1: Voodoo graphics card and a computer and have that be

81
00:04:53,120 --> 00:04:56,159
Speaker 1: your only graphics card and have that work. You actually

82
00:04:56,200 --> 00:04:59,279
Speaker 1: had to make it work with in parallel with a

83
00:04:59,360 --> 00:05:02,640
Speaker 1: two D graphics card or really in series, because what

84
00:05:02,680 --> 00:05:04,560
Speaker 1: you would do is you would take a v g

85
00:05:04,760 --> 00:05:08,800
Speaker 1: A cable from your video controller card, your your basic

86
00:05:08,920 --> 00:05:12,160
Speaker 1: graphics card in your computer. You would feed that cable

87
00:05:12,600 --> 00:05:15,680
Speaker 1: into your Voodoo card, you know, plug it in on

88
00:05:15,760 --> 00:05:18,560
Speaker 1: the back of where the expansion card is, and then

89
00:05:19,000 --> 00:05:21,880
Speaker 1: you'd take another cable and you'd hook the Voodoo card

90
00:05:22,000 --> 00:05:26,400
Speaker 1: to your monitor, so you effectively had two different graphics

91
00:05:26,400 --> 00:05:29,720
Speaker 1: cards doing work inside your computer. The Voodoo handling all

92
00:05:29,880 --> 00:05:32,960
Speaker 1: the three D stuff, and it gave gamers a new

93
00:05:33,000 --> 00:05:36,360
Speaker 1: option when building out computers to play games like Doom

94
00:05:36,440 --> 00:05:41,200
Speaker 1: and Quake, which were progressively putting more emphasis on three

95
00:05:41,320 --> 00:05:44,479
Speaker 1: D models, and they were pushing the boundaries of what

96
00:05:44,600 --> 00:05:47,680
Speaker 1: PCs could do. At that time. Before the Voodoo, the

97
00:05:47,680 --> 00:05:51,039
Speaker 1: most common way to keep up with gaming developments would

98
00:05:51,040 --> 00:05:53,880
Speaker 1: be you'd go and get an upgrade to your CPU,

99
00:05:54,120 --> 00:05:57,000
Speaker 1: which get it really expensive. So I remember when my

100
00:05:57,080 --> 00:05:59,839
Speaker 1: family went from having a two eight six IBM compatible

101
00:06:00,120 --> 00:06:02,560
Speaker 1: to a four a D six, and then later on

102
00:06:02,600 --> 00:06:08,080
Speaker 1: we got a Pentium. But these graphics accelerator cards meant

103
00:06:08,200 --> 00:06:11,200
Speaker 1: you could just buy a graphics accelerated card for for

104
00:06:11,360 --> 00:06:13,479
Speaker 1: much less money than it would cost to upgrade the

105
00:06:13,480 --> 00:06:16,840
Speaker 1: whole CPU, and you could insert it into an expansion

106
00:06:16,880 --> 00:06:20,640
Speaker 1: slot inside your computer, plug up a few cables, and

107
00:06:20,680 --> 00:06:22,960
Speaker 1: provide the processing of you would need to have really

108
00:06:22,960 --> 00:06:27,040
Speaker 1: great graphics. But it also helped usher in an era

109
00:06:27,160 --> 00:06:31,480
Speaker 1: in which new graphics cards would debut at an alarmingly

110
00:06:31,960 --> 00:06:36,840
Speaker 1: frequent rate and make the previous cards obsolete very very

111
00:06:36,880 --> 00:06:42,000
Speaker 1: shortly after they had debuted. Some were two D cards

112
00:06:42,040 --> 00:06:44,919
Speaker 1: that also had three D accelerators on them, some of

113
00:06:44,960 --> 00:06:48,520
Speaker 1: them were dedicated three D accelerators, like the Voodoo, and

114
00:06:48,600 --> 00:06:50,880
Speaker 1: more games were coming out that would push the limits

115
00:06:50,960 --> 00:06:53,400
Speaker 1: on the fastest processors on the market. So it became

116
00:06:53,440 --> 00:06:57,320
Speaker 1: this vicious cycle where you would have brand new cards

117
00:06:57,360 --> 00:07:00,360
Speaker 1: coming out that were much more capable than the ones,

118
00:07:00,480 --> 00:07:05,520
Speaker 1: so everyone wanted one because you thought, who faster means better,

119
00:07:05,560 --> 00:07:09,600
Speaker 1: more features means better. Developers would start developing games that

120
00:07:09,600 --> 00:07:13,320
Speaker 1: would take advantage of those greater capabilities. So while it

121
00:07:13,320 --> 00:07:16,240
Speaker 1: would seem in the very short term that you were

122
00:07:16,240 --> 00:07:19,080
Speaker 1: getting a head start by getting one of these powerful cards,

123
00:07:19,440 --> 00:07:23,960
Speaker 1: pretty soon developers would end up creating applications that would

124
00:07:24,000 --> 00:07:26,640
Speaker 1: push those cards to their limits, and it just became

125
00:07:26,680 --> 00:07:30,920
Speaker 1: this kind of seesaw approach. And that's when I decided

126
00:07:31,600 --> 00:07:34,280
Speaker 1: I don't want to deal with this, and I stopped

127
00:07:34,320 --> 00:07:36,760
Speaker 1: being a PC gamer for a while, and I switched

128
00:07:36,760 --> 00:07:40,080
Speaker 1: to consoles for for quite some time. I did eventually

129
00:07:40,080 --> 00:07:43,600
Speaker 1: go back to PC gamer, but it was in this

130
00:07:43,680 --> 00:07:47,679
Speaker 1: era of numerous graphics cards, and it weren't they weren't

131
00:07:47,680 --> 00:07:50,720
Speaker 1: just graphics cards too, that you also had sound cards

132
00:07:50,720 --> 00:07:52,800
Speaker 1: that were doing a very similar thing at the same time,

133
00:07:53,480 --> 00:07:55,600
Speaker 1: and it was causing all sorts of issues, things like

134
00:07:55,680 --> 00:08:00,920
Speaker 1: questions about compatibility and affordability. I just too frustrating for me.

135
00:08:01,240 --> 00:08:04,760
Speaker 1: But it was a boom time for PC gamers who were,

136
00:08:05,160 --> 00:08:08,640
Speaker 1: you know, eager to pursue that hobby, and they weren't

137
00:08:08,760 --> 00:08:11,600
Speaker 1: weren't put off by the idea that there was suddenly

138
00:08:11,640 --> 00:08:14,680
Speaker 1: a ton of different products out there now. It did

139
00:08:14,680 --> 00:08:21,120
Speaker 1: not help that by were looking at around seventy companies

140
00:08:21,160 --> 00:08:24,560
Speaker 1: that were marketing graphics cards of some sort. Some of

141
00:08:24,560 --> 00:08:27,480
Speaker 1: them were working with technology that was compatible with the

142
00:08:27,520 --> 00:08:32,040
Speaker 1: open g L standard, many more were developing cards that

143
00:08:32,080 --> 00:08:36,479
Speaker 1: were specifically geared to support Microsoft's Direct three D Application

144
00:08:36,559 --> 00:08:40,000
Speaker 1: Programming Interface, and more than a few had their own

145
00:08:40,040 --> 00:08:44,200
Speaker 1: proprietary application programmer interface that that developers could use to

146
00:08:44,320 --> 00:08:47,760
Speaker 1: create software that would look fantastic if it ran on

147
00:08:47,840 --> 00:08:51,440
Speaker 1: those cards. There was a ton of market confusion, not

148
00:08:51,520 --> 00:08:54,720
Speaker 1: just for end consumers, but also for programmers who were

149
00:08:54,760 --> 00:08:57,840
Speaker 1: making software in the first place. Well a Voodoo from

150
00:08:57,920 --> 00:09:01,160
Speaker 1: three D f X. Actually it would officially debut in

151
00:09:01,240 --> 00:09:04,960
Speaker 1: nine and the following year in Video would come out

152
00:09:05,040 --> 00:09:08,400
Speaker 1: with their second commercial product. This was their follow up

153
00:09:08,480 --> 00:09:10,480
Speaker 1: to the n V one. This was the n V

154
00:09:10,720 --> 00:09:13,679
Speaker 1: three because the n V two was canceled. This one

155
00:09:13,760 --> 00:09:17,959
Speaker 1: was better known as the RIVA on. RIVA or r

156
00:09:18,040 --> 00:09:21,840
Speaker 1: i v A stood for real Time Interactive Video and

157
00:09:21,920 --> 00:09:25,680
Speaker 1: Animation Accelerator and this was a hundred twenty eight bit

158
00:09:25,960 --> 00:09:28,960
Speaker 1: three D processor. That's why the one was in the name,

159
00:09:29,360 --> 00:09:32,520
Speaker 1: and that means the processor and busses could operate on

160
00:09:32,520 --> 00:09:36,559
Speaker 1: one eight bit integers or data units that were sixteen

161
00:09:36,640 --> 00:09:41,840
Speaker 1: octets wide. This gave the RIVA a bandwidth of one

162
00:09:41,880 --> 00:09:45,199
Speaker 1: point six gigabytes per second, which is the same thing

163
00:09:45,200 --> 00:09:50,680
Speaker 1: as twelve point eight gigabits per second. Incredible bandwidth. The

164
00:09:50,800 --> 00:09:54,000
Speaker 1: processor was a big hit with O E M or

165
00:09:54,160 --> 00:09:58,160
Speaker 1: original equipment manufacturer customers. So these were companies that either

166
00:09:58,240 --> 00:10:01,840
Speaker 1: made computers or they may graphics cards and they would

167
00:10:01,840 --> 00:10:06,640
Speaker 1: incorporate the REVA in them. It had four megabytes of

168
00:10:06,720 --> 00:10:10,960
Speaker 1: RAM and a two hundred six mega Hurts processor. Um.

169
00:10:11,000 --> 00:10:13,720
Speaker 1: There was a faster model that came out shortly thereafter,

170
00:10:13,760 --> 00:10:17,679
Speaker 1: called the Revo z X. That one had an additional

171
00:10:17,840 --> 00:10:20,280
Speaker 1: four megabytes of RAM, bringing up to eight and had

172
00:10:20,320 --> 00:10:23,320
Speaker 1: a clock speed of two hundred fifty mega hurts. This

173
00:10:23,400 --> 00:10:28,319
Speaker 1: chip was sold to those original equipment manufacturers, and one

174
00:10:28,360 --> 00:10:32,800
Speaker 1: of them was Diamond, which had also been one of

175
00:10:32,840 --> 00:10:34,880
Speaker 1: the O E M s for the original n V

176
00:10:35,040 --> 00:10:38,760
Speaker 1: one chip. They introduced the Diamond of Viper three thirty

177
00:10:38,840 --> 00:10:41,080
Speaker 1: which had a Reva one chip as its three D

178
00:10:41,160 --> 00:10:45,360
Speaker 1: graphics accelerator, and I read some reviews of the revoight

179
00:10:45,520 --> 00:10:48,079
Speaker 1: from that era. I actually found you can find reviews

180
00:10:48,120 --> 00:10:50,880
Speaker 1: online the date back to when this chip came out,

181
00:10:51,440 --> 00:10:55,760
Speaker 1: because the web was a thing in and the reviews

182
00:10:55,760 --> 00:10:59,079
Speaker 1: were pretty positive. Critics were impressed with the three D performance.

183
00:10:59,520 --> 00:11:03,960
Speaker 1: The review is also said that, you know, obviously the

184
00:11:04,040 --> 00:11:07,680
Speaker 1: quality of the performance didn't come just from in video,

185
00:11:07,760 --> 00:11:12,480
Speaker 1: but also from the manufacturers that produced the other equipment,

186
00:11:12,520 --> 00:11:16,280
Speaker 1: the the graphics cards themselves, like the motherboard for the

187
00:11:16,280 --> 00:11:18,640
Speaker 1: the that the chip was sitting on, so it was

188
00:11:18,679 --> 00:11:22,520
Speaker 1: more than justin video, but in video overall came out

189
00:11:22,559 --> 00:11:26,520
Speaker 1: looking pretty good. From the Reva however, it did create

190
00:11:26,559 --> 00:11:28,720
Speaker 1: a brand new problem for in Video but this one

191
00:11:28,840 --> 00:11:32,320
Speaker 1: was a welcome problem, not a bad one. It turned

192
00:11:32,320 --> 00:11:33,920
Speaker 1: out the company was not able to keep up with

193
00:11:34,000 --> 00:11:38,400
Speaker 1: demand because the Reva was a huge economic and critical

194
00:11:38,440 --> 00:11:41,480
Speaker 1: success for the company, and so in Video would sign

195
00:11:41,520 --> 00:11:46,920
Speaker 1: a manufacturing deal with T s MC to supplement manufacturing

196
00:11:46,920 --> 00:11:50,840
Speaker 1: that was already going on with their established partner. They

197
00:11:50,920 --> 00:11:55,319
Speaker 1: had been manufacturing chips through a company called st Micro

198
00:11:56,160 --> 00:12:01,040
Speaker 1: or Microsystems, and they decided to supplement that and use

199
00:12:01,080 --> 00:12:05,120
Speaker 1: a second second partner to provide even more of these.

200
00:12:05,559 --> 00:12:09,439
Speaker 1: By the end, in Video had gone from a struggling

201
00:12:09,440 --> 00:12:13,920
Speaker 1: company on the virginal collapse to owning twenty fo of

202
00:12:13,960 --> 00:12:17,480
Speaker 1: the market share in three D graphics processors. Not a

203
00:12:17,520 --> 00:12:21,880
Speaker 1: bad turnaround, and as it turns out, they had higher

204
00:12:21,880 --> 00:12:24,800
Speaker 1: heights to climb. I'll say more in just a moment,

205
00:12:24,840 --> 00:12:27,280
Speaker 1: but first let's take a quick break to thank our sponsor.

206
00:12:35,679 --> 00:12:38,720
Speaker 1: At this stage, the market was starting to clear out

207
00:12:39,200 --> 00:12:42,080
Speaker 1: a little bit. Some of the companies that had entered

208
00:12:42,120 --> 00:12:44,480
Speaker 1: the graphics cards space we're starting to go out of business.

209
00:12:45,040 --> 00:12:48,920
Speaker 1: But there were three companies that were essentially dominating the

210
00:12:49,080 --> 00:12:53,480
Speaker 1: high performance graphics cards space. Those three companies were a

211
00:12:53,600 --> 00:12:58,160
Speaker 1: T I, three D f X and Nvidia. Some companies,

212
00:12:58,200 --> 00:13:01,000
Speaker 1: in an effort to survive in this increase singly narrow market,

213
00:13:01,120 --> 00:13:04,680
Speaker 1: began to focus on creating budget graphics cards that had

214
00:13:05,040 --> 00:13:09,120
Speaker 1: fewer features, They had a slightly lower performance level, and

215
00:13:09,200 --> 00:13:12,160
Speaker 1: they were generally of a lower cost, so that would

216
00:13:12,679 --> 00:13:15,520
Speaker 1: put them in the range of a different market for gamers.

217
00:13:16,080 --> 00:13:19,640
Speaker 1: But in video was on pretty solid ground now. They

218
00:13:19,679 --> 00:13:21,880
Speaker 1: wanted to follow up on the success of the Reva

219
00:13:23,120 --> 00:13:26,720
Speaker 1: with their next card, which was the Riva TNT also

220
00:13:26,760 --> 00:13:30,000
Speaker 1: known as the n V four T n T stood

221
00:13:30,000 --> 00:13:34,119
Speaker 1: for twin Texel because this card actually added a parallel

222
00:13:34,160 --> 00:13:38,800
Speaker 1: pixel pipeline. It's a lot of ps there to get

223
00:13:38,840 --> 00:13:42,199
Speaker 1: the pop filter going, but this would double the rendering

224
00:13:42,280 --> 00:13:45,160
Speaker 1: speed for the card, and it was a two D

225
00:13:45,520 --> 00:13:48,040
Speaker 1: and three D graphics card. It had support for thirty

226
00:13:48,040 --> 00:13:51,600
Speaker 1: two bit true color graphics. They had expanded the RAM

227
00:13:51,679 --> 00:13:54,360
Speaker 1: up to sixteen megabytes, and it was available both in

228
00:13:54,440 --> 00:13:58,000
Speaker 1: PC I and A g P formats. Those would be

229
00:13:58,040 --> 00:14:02,880
Speaker 1: the buses that you could use with personal computers, and

230
00:14:02,920 --> 00:14:05,920
Speaker 1: there was two different formats for buses. The A g

231
00:14:06,080 --> 00:14:08,880
Speaker 1: P was the newer one of the two and the

232
00:14:08,960 --> 00:14:11,680
Speaker 1: superior one. As far as performance goes of the two,

233
00:14:12,160 --> 00:14:15,440
Speaker 1: but they in video. This time we're mostly focused on

234
00:14:15,480 --> 00:14:20,080
Speaker 1: PC I simply because a g P had not rolled

235
00:14:20,120 --> 00:14:22,840
Speaker 1: out to a very wide user base yet it was

236
00:14:22,880 --> 00:14:28,240
Speaker 1: still new enough where it was not widely adopted as

237
00:14:28,280 --> 00:14:32,280
Speaker 1: the news standard yet. The one downside to in Video's

238
00:14:32,320 --> 00:14:35,560
Speaker 1: design was that the complexity of the chip they had

239
00:14:35,600 --> 00:14:38,360
Speaker 1: made with the tn T meant they weren't able to

240
00:14:38,440 --> 00:14:41,920
Speaker 1: run it at what they had originally planned with their

241
00:14:41,960 --> 00:14:45,360
Speaker 1: memory core speed. They originally wanted to have a memory

242
00:14:45,360 --> 00:14:48,000
Speaker 1: core speed of a hundred twenty five Mega hurts, but

243
00:14:48,280 --> 00:14:51,720
Speaker 1: they found out that because they had made the chips

244
00:14:51,720 --> 00:14:54,040
Speaker 1: soaked or the card so complex, they had packed it

245
00:14:54,080 --> 00:14:56,840
Speaker 1: with so much stuff that if they tried to run

246
00:14:56,840 --> 00:15:00,280
Speaker 1: it at a Mega hurts, the card would over heat,

247
00:15:01,480 --> 00:15:04,040
Speaker 1: so they were forced to throttle the speed down to

248
00:15:04,240 --> 00:15:07,840
Speaker 1: nine d Mega hurts to avoid those overheating issues. The

249
00:15:07,840 --> 00:15:10,400
Speaker 1: big competitor to the t n T was three D

250
00:15:10,640 --> 00:15:14,400
Speaker 1: f x IS Voodoo Too card, which in benchmark tests

251
00:15:14,960 --> 00:15:18,160
Speaker 1: pretty much dominated the three D acceleration world. But the

252
00:15:18,240 --> 00:15:21,960
Speaker 1: Voodoo Too still required a second two D graphics card

253
00:15:22,000 --> 00:15:24,840
Speaker 1: to work. It was not a standalone all in one card,

254
00:15:25,400 --> 00:15:28,000
Speaker 1: so if you were to add up costs. If you

255
00:15:28,000 --> 00:15:31,520
Speaker 1: were building a brand new PC from the ground up,

256
00:15:31,920 --> 00:15:34,600
Speaker 1: you're buying all the different components. If you went with

257
00:15:34,640 --> 00:15:37,440
Speaker 1: a Voodoo too, the costs actually would go up a

258
00:15:37,480 --> 00:15:40,840
Speaker 1: bit because you had to have a secondary graphics card

259
00:15:40,880 --> 00:15:45,160
Speaker 1: in order to run the Voodoo two, whereas in videos

260
00:15:45,200 --> 00:15:47,520
Speaker 1: T and T wasn't all in one you could just

261
00:15:47,840 --> 00:15:50,920
Speaker 1: buy one card and get all of that, So it

262
00:15:50,960 --> 00:15:53,400
Speaker 1: all depended on what your needs were and what you

263
00:15:53,480 --> 00:15:57,720
Speaker 1: already had at your disposal. Also, behind the scenes, three

264
00:15:57,800 --> 00:16:00,760
Speaker 1: d FX was trying to head off competitors like in

265
00:16:00,920 --> 00:16:03,880
Speaker 1: Vidio and A T I by marketing and selling their

266
00:16:03,880 --> 00:16:08,120
Speaker 1: graphics card boards themselves directly. So three d f X,

267
00:16:08,160 --> 00:16:10,760
Speaker 1: in order to do this, purchased a company called STB

268
00:16:11,040 --> 00:16:13,600
Speaker 1: Systems for a hundred forty one million dollars in the

269
00:16:13,720 --> 00:16:17,720
Speaker 1: stock deal. But it turned out the cost of manufacturing

270
00:16:17,880 --> 00:16:21,880
Speaker 1: ended up being higher with STB Systems foundry than the

271
00:16:21,920 --> 00:16:24,480
Speaker 1: foundries that A T and nine and Video were using,

272
00:16:24,560 --> 00:16:31,000
Speaker 1: So the the actual expenses of producing the boards was

273
00:16:31,120 --> 00:16:35,840
Speaker 1: higher than competitors costs, and so that would eat into

274
00:16:35,880 --> 00:16:38,840
Speaker 1: your profit margin. If you're still trying to price your

275
00:16:38,960 --> 00:16:44,200
Speaker 1: your cards competitively against A T I and Nvidia. So

276
00:16:44,600 --> 00:16:47,520
Speaker 1: three d f x is misstep became in Video's opportunity

277
00:16:47,600 --> 00:16:50,560
Speaker 1: because some of three d f x's former partners began

278
00:16:50,600 --> 00:16:54,160
Speaker 1: to drift over to work within Video instead. But while

279
00:16:54,160 --> 00:16:57,720
Speaker 1: in Video was doing well against his competitors, specifically through

280
00:16:57,760 --> 00:17:00,600
Speaker 1: the t n T card, it was dealing with other

281
00:17:00,680 --> 00:17:04,200
Speaker 1: issues that were not so great. Silicon Graphics Incorporated now

282
00:17:04,440 --> 00:17:08,800
Speaker 1: s g I filed a lawsuit against Nvidia. The company

283
00:17:08,840 --> 00:17:12,080
Speaker 1: alleged that in Vidia had infringed on a patent related

284
00:17:12,119 --> 00:17:16,000
Speaker 1: to texture mapping. That dispute would stretch from the spring

285
00:17:16,119 --> 00:17:20,439
Speaker 1: of n all the way into nine but eventually the

286
00:17:20,480 --> 00:17:23,080
Speaker 1: two companies would reach a settlement, and as part of

287
00:17:23,119 --> 00:17:25,280
Speaker 1: that settlement, in Video would actually get s g i

288
00:17:25,359 --> 00:17:29,439
Speaker 1: s professional graphics portfolio, and they hired on several of

289
00:17:29,600 --> 00:17:32,800
Speaker 1: s g i s low level graphics team members, and

290
00:17:33,000 --> 00:17:36,520
Speaker 1: SGI sort of transitioned out of the graphics business at

291
00:17:36,560 --> 00:17:40,280
Speaker 1: that point and started to focus on other areas of business.

292
00:17:40,320 --> 00:17:43,480
Speaker 1: It never really worked out for that company. The company

293
00:17:43,600 --> 00:17:46,159
Speaker 1: s g I was already reeling from other problems, and

294
00:17:46,160 --> 00:17:49,240
Speaker 1: it would ultimately go into bankruptcy in two thousand nine.

295
00:17:49,800 --> 00:17:52,720
Speaker 1: And Video released the t n T two and the

296
00:17:52,760 --> 00:17:55,679
Speaker 1: t n T to Ultra the following year, which effectively

297
00:17:56,200 --> 00:17:59,680
Speaker 1: rested the title of most powerful three D accelerator away

298
00:17:59,760 --> 00:18:02,040
Speaker 1: from three D f X and the Voodoo three card.

299
00:18:02,400 --> 00:18:05,800
Speaker 1: At least in most bitchmark tests. Graphics cards are weird

300
00:18:05,880 --> 00:18:10,359
Speaker 1: by the way you typically test these by running various

301
00:18:10,880 --> 00:18:15,240
Speaker 1: tests that that check the different settings and different features

302
00:18:15,320 --> 00:18:17,920
Speaker 1: of these graphics cards, and it's not always a one

303
00:18:17,920 --> 00:18:20,959
Speaker 1: to one match. You don't always have features that measure

304
00:18:21,080 --> 00:18:25,639
Speaker 1: up where there's an equivalent on a competing card. So

305
00:18:25,960 --> 00:18:28,720
Speaker 1: depending on the test, you might have one card come

306
00:18:28,720 --> 00:18:31,159
Speaker 1: out on top, and then you change tests and you

307
00:18:31,280 --> 00:18:33,280
Speaker 1: run those same two cards through it, and the other

308
00:18:33,320 --> 00:18:36,600
Speaker 1: card comes out on top. So it's a little complicated

309
00:18:36,600 --> 00:18:38,920
Speaker 1: to talk about, but this was a point where in

310
00:18:39,040 --> 00:18:43,600
Speaker 1: Videous performance was look looked at as best in class,

311
00:18:43,640 --> 00:18:46,600
Speaker 1: at least at that moment. Uh. And also keep in mind,

312
00:18:46,640 --> 00:18:48,800
Speaker 1: we're still talking about a world that had different A

313
00:18:48,880 --> 00:18:51,000
Speaker 1: p I s out there that developers were using to

314
00:18:51,000 --> 00:18:54,639
Speaker 1: build software that had three D graphics in it, and

315
00:18:54,720 --> 00:18:59,080
Speaker 1: so sometimes you would find a program that would run

316
00:18:59,240 --> 00:19:02,639
Speaker 1: really well on a computer that had a three D

317
00:19:02,760 --> 00:19:05,679
Speaker 1: f X card, but not as well on a computer

318
00:19:05,760 --> 00:19:08,320
Speaker 1: that had in video card, and then meanwhile you might

319
00:19:08,320 --> 00:19:13,400
Speaker 1: get a different program made by a different developer that

320
00:19:13,520 --> 00:19:18,160
Speaker 1: the opposite is true. So it's very complicated. On October one,

321
00:19:18,480 --> 00:19:22,080
Speaker 1: and Video would define a new era because it announced

322
00:19:22,080 --> 00:19:26,399
Speaker 1: a new graphics card called the g Force two. It

323
00:19:26,520 --> 00:19:29,840
Speaker 1: previously had the code name in V ten. This was

324
00:19:29,880 --> 00:19:33,480
Speaker 1: the first chip to get the designation of GPU, or

325
00:19:33,560 --> 00:19:37,880
Speaker 1: graphics processing unit. The very first GPU was an Nvidia

326
00:19:38,640 --> 00:19:42,439
Speaker 1: product called the g Force two. So what was different

327
00:19:42,440 --> 00:19:45,520
Speaker 1: about this card? What was it that merited the creation

328
00:19:45,640 --> 00:19:48,959
Speaker 1: of a new term. Well, it had a transformation and

329
00:19:49,080 --> 00:19:53,680
Speaker 1: lighting engine which tackled floating point calculations to figure out

330
00:19:53,720 --> 00:19:57,880
Speaker 1: how to render transforming virtual three dimensional objects and display

331
00:19:57,920 --> 00:20:01,480
Speaker 1: them in a two dimensional representation of the image. Wait

332
00:20:01,480 --> 00:20:04,720
Speaker 1: what all right? This is actually not as complicated as

333
00:20:04,960 --> 00:20:06,679
Speaker 1: I just made it sound. In fact, a lot of

334
00:20:06,720 --> 00:20:08,760
Speaker 1: you probably are way ahead of me. But just in case,

335
00:20:09,080 --> 00:20:11,760
Speaker 1: here's what I actually mean by that. Let's imagine for

336
00:20:11,800 --> 00:20:15,440
Speaker 1: a second that you can pop into the virtual world

337
00:20:15,680 --> 00:20:18,600
Speaker 1: of a computer, kind of like Entron. So you are

338
00:20:18,640 --> 00:20:21,679
Speaker 1: in a virtual world, and in this virtual world you

339
00:20:21,720 --> 00:20:25,240
Speaker 1: can see three dimensional objects. Right, there are three dimensional

340
00:20:25,240 --> 00:20:29,560
Speaker 1: objects all around you. They truly have those three dimensions now,

341
00:20:29,560 --> 00:20:32,560
Speaker 1: and in the computer world, that's all represented by math.

342
00:20:32,760 --> 00:20:36,960
Speaker 1: They are just virtual objects. But let's pretend that we

343
00:20:37,040 --> 00:20:39,800
Speaker 1: are in that virtual world. We can see them, and

344
00:20:39,840 --> 00:20:43,320
Speaker 1: we can see how those objects transform as they move

345
00:20:43,400 --> 00:20:47,160
Speaker 1: through a space and environment. Maybe they change shape. Maybe

346
00:20:47,200 --> 00:20:50,479
Speaker 1: it represents a character in a video game, and as

347
00:20:50,520 --> 00:20:53,679
Speaker 1: the character moves, we see how this shape changes. We

348
00:20:53,720 --> 00:20:57,520
Speaker 1: can see how light from fixed sources can play over

349
00:20:57,560 --> 00:21:01,360
Speaker 1: the surfaces of these three dimensional images. We can see

350
00:21:01,400 --> 00:21:05,679
Speaker 1: where shadows form, where highlights are. Now, let's say that

351
00:21:05,760 --> 00:21:08,359
Speaker 1: we we have these three dimensional images that exist in

352
00:21:08,400 --> 00:21:11,560
Speaker 1: this virtual world, and now it's time for us to

353
00:21:11,680 --> 00:21:16,040
Speaker 1: display those images to the real world. And our our

354
00:21:16,119 --> 00:21:19,600
Speaker 1: method of doing that. Our medium is a computer display.

355
00:21:19,640 --> 00:21:23,439
Speaker 1: But computer displays are not three dimensional. They are two dimensional.

356
00:21:23,720 --> 00:21:26,040
Speaker 1: They have width and length, but they don't have depth.

357
00:21:26,840 --> 00:21:29,080
Speaker 1: And so you have to figure out how do I

358
00:21:29,200 --> 00:21:33,399
Speaker 1: portray this virtual three dimensional image that has stuff like

359
00:21:33,560 --> 00:21:36,520
Speaker 1: lighting effects going on with it in a way that

360
00:21:36,640 --> 00:21:40,600
Speaker 1: makes sense within a two dimensional medium, so that you

361
00:21:40,680 --> 00:21:43,720
Speaker 1: get the idea of what that object, how that object

362
00:21:43,800 --> 00:21:48,200
Speaker 1: is shaped. Before the g fource two six, the CPU

363
00:21:48,400 --> 00:21:51,920
Speaker 1: handled that sort of work, this transformation and lighting work.

364
00:21:52,280 --> 00:21:55,320
Speaker 1: But the CPU was also doing all the other work too,

365
00:21:55,440 --> 00:21:59,120
Speaker 1: besides the other graphics processing, and that meant that if

366
00:21:59,160 --> 00:22:02,159
Speaker 1: you added in too much detail in your game, everything

367
00:22:02,200 --> 00:22:05,320
Speaker 1: would bog down and run super slowly or maybe even

368
00:22:05,400 --> 00:22:09,240
Speaker 1: crash because the CPU would be so busy trying to

369
00:22:09,640 --> 00:22:13,639
Speaker 1: calculate all the information required to display that kind of

370
00:22:13,680 --> 00:22:17,200
Speaker 1: stuff that it couldn't do anything else. But by moving

371
00:22:17,200 --> 00:22:21,520
Speaker 1: that requirement over to the graphics card, the GPU could

372
00:22:21,560 --> 00:22:24,359
Speaker 1: free the CPU up to do other calculations, and the

373
00:22:24,400 --> 00:22:28,080
Speaker 1: bandwidth the GPU was super high, high enough to handle

374
00:22:28,359 --> 00:22:31,399
Speaker 1: a massive amount of data, so as long as the

375
00:22:31,440 --> 00:22:36,159
Speaker 1: processor was fast, it could go through this information very quickly.

376
00:22:36,359 --> 00:22:38,720
Speaker 1: In fact, that g Force two six had a throughput

377
00:22:38,960 --> 00:22:42,359
Speaker 1: that was about five times greater than that of a

378
00:22:42,400 --> 00:22:47,520
Speaker 1: five fifty Mega Hurts Pentium three processor, So for certain

379
00:22:47,560 --> 00:22:51,520
Speaker 1: types of applications, a GPU is going to be much

380
00:22:51,560 --> 00:22:55,440
Speaker 1: better than a CPU. Plus, the GPU had on board memory,

381
00:22:56,040 --> 00:22:59,560
Speaker 1: it included a second set of parallel pixel pipelines that

382
00:22:59,720 --> 00:23:02,480
Speaker 1: meant you had four total that was double the TNT

383
00:23:02,680 --> 00:23:06,199
Speaker 1: design and game design developers now could create games with

384
00:23:06,280 --> 00:23:08,760
Speaker 1: much more detail without fear of causing a total slow

385
00:23:08,840 --> 00:23:12,200
Speaker 1: down on a system. It was a literal game changer.

386
00:23:12,840 --> 00:23:15,040
Speaker 1: The clock speed for the g Force two fifty six

387
00:23:15,080 --> 00:23:18,480
Speaker 1: was technically slower than that of the Riva T and

388
00:23:18,560 --> 00:23:22,040
Speaker 1: T two, so they actually took a step down in

389
00:23:22,080 --> 00:23:24,879
Speaker 1: their clock speed. They didn't make it faster than the

390
00:23:24,920 --> 00:23:29,520
Speaker 1: previous graphics card. However, the improvements they made in the

391
00:23:29,680 --> 00:23:33,160
Speaker 1: architecture meant that the g Force two fifty six still

392
00:23:33,200 --> 00:23:37,600
Speaker 1: performed at at faster speed than the t n T two,

393
00:23:37,840 --> 00:23:41,119
Speaker 1: even though the clock speed was slower. Now, up to

394
00:23:41,160 --> 00:23:43,840
Speaker 1: this point, I've gone through the Nvidia cards more or

395
00:23:43,880 --> 00:23:46,479
Speaker 1: less one by one because they were really important for

396
00:23:46,520 --> 00:23:49,280
Speaker 1: the foundation and success of the company and its rise

397
00:23:49,359 --> 00:23:52,760
Speaker 1: from startup to industry giant. But from this point forward,

398
00:23:52,800 --> 00:23:55,440
Speaker 1: I'm going to be jumping around a bit. Otherwise, all

399
00:23:55,440 --> 00:23:57,560
Speaker 1: I'm gonna be doing is giving you a long list

400
00:23:57,560 --> 00:24:00,280
Speaker 1: of graphics cards and their specifications, and that is not

401
00:24:00,640 --> 00:24:05,200
Speaker 1: terribly interesting. It's like reading a catalog. However, around this time,

402
00:24:05,720 --> 00:24:08,919
Speaker 1: things were still in that pretty chaotic stage in the

403
00:24:08,960 --> 00:24:12,400
Speaker 1: gaming rig world. You had a small number of companies

404
00:24:12,440 --> 00:24:15,640
Speaker 1: that were dominating the graphics cards scene at the very top,

405
00:24:15,960 --> 00:24:18,280
Speaker 1: but there were lots of different budget companies out there.

406
00:24:18,359 --> 00:24:21,000
Speaker 1: There were other companies that would occasionally create a high

407
00:24:21,040 --> 00:24:24,320
Speaker 1: performing graphics card to try and compete with a t

408
00:24:24,480 --> 00:24:27,920
Speaker 1: I three D f X and Nvidia, And there were

409
00:24:27,960 --> 00:24:31,760
Speaker 1: also the companies that were specializing in sound cards. There

410
00:24:31,760 --> 00:24:35,320
Speaker 1: were the different application programming interfaces, there were the different

411
00:24:35,400 --> 00:24:39,359
Speaker 1: drivers that were meant to try and make more compatibility

412
00:24:39,400 --> 00:24:43,840
Speaker 1: between different hardware and software. It was still a huge,

413
00:24:44,480 --> 00:24:47,800
Speaker 1: confusing mess for a lot of people, including myself. I

414
00:24:47,800 --> 00:24:49,639
Speaker 1: remember the days of looking at a game and wondering

415
00:24:49,640 --> 00:24:51,560
Speaker 1: if I could even play it because it might not

416
00:24:51,600 --> 00:24:54,879
Speaker 1: be compatible with one or more of the components inside

417
00:24:54,920 --> 00:24:57,960
Speaker 1: my PC. It was a real mess. And not all

418
00:24:57,960 --> 00:25:00,200
Speaker 1: the companies that were around at that point would serve five,

419
00:25:01,240 --> 00:25:05,400
Speaker 1: but in Video would. Now, when we come back, i'll

420
00:25:05,400 --> 00:25:08,080
Speaker 1: give you a little last look at what in Video

421
00:25:08,119 --> 00:25:10,480
Speaker 1: has been up to, and we'll kind of rush through

422
00:25:11,119 --> 00:25:13,919
Speaker 1: the rest of its of its history to kind of

423
00:25:13,960 --> 00:25:16,240
Speaker 1: give you the highlights. But first let's take a quick

424
00:25:16,240 --> 00:25:27,480
Speaker 1: break to thank our sponsor. In late nine and Video

425
00:25:27,560 --> 00:25:31,200
Speaker 1: announced a new line of products called Quadripe cards. These

426
00:25:31,240 --> 00:25:35,240
Speaker 1: were meant not for gaming rigs, but for professional workstations,

427
00:25:35,280 --> 00:25:38,280
Speaker 1: so in videos branching out. At this point, the GPU

428
00:25:38,440 --> 00:25:41,679
Speaker 1: was showing that it had some interesting uses outside of

429
00:25:41,760 --> 00:25:45,119
Speaker 1: rendering graphics for gamers, although at this point it was

430
00:25:45,160 --> 00:25:50,679
Speaker 1: really still about high end graphically intensive workstations. But speaking

431
00:25:50,680 --> 00:25:53,920
Speaker 1: of games, and video also landed a pretty sweet contract

432
00:25:54,000 --> 00:25:59,320
Speaker 1: with Microsoft. Microsoft was developing its own game console, which

433
00:25:59,320 --> 00:26:02,080
Speaker 1: at that point was still a super secret project and

434
00:26:02,119 --> 00:26:05,880
Speaker 1: it wouldn't come out until November two thousand one. Obviously,

435
00:26:05,960 --> 00:26:09,520
Speaker 1: that was the original Xbox, and Microsoft went in Video

436
00:26:09,600 --> 00:26:11,920
Speaker 1: to build the graphics chip that would be inside the system.

437
00:26:12,040 --> 00:26:15,480
Speaker 1: So the initial payment for that chip, which was called

438
00:26:15,520 --> 00:26:18,960
Speaker 1: the n V two A internally at Nvidia, was two

439
00:26:19,040 --> 00:26:23,800
Speaker 1: hundred million dollars, healthy chunk of change, one might even say,

440
00:26:24,040 --> 00:26:27,520
Speaker 1: a princely some. By the end of the fiscal year,

441
00:26:27,600 --> 00:26:30,400
Speaker 1: in Video had made a profit of forty one million dollars,

442
00:26:30,760 --> 00:26:34,600
Speaker 1: which was ten times more than the previous year's profit.

443
00:26:35,119 --> 00:26:37,880
Speaker 1: That's a heck of a lot of growth. It's really

444
00:26:37,960 --> 00:26:39,879
Speaker 1: nice to be able to point to a number and

445
00:26:39,920 --> 00:26:42,600
Speaker 1: say we made ten times as much money this year,

446
00:26:43,240 --> 00:26:46,080
Speaker 1: not not ten times the revenue, ten times the profit.

447
00:26:47,160 --> 00:26:50,840
Speaker 1: The following year saw in Video sweep up a former competitor,

448
00:26:51,080 --> 00:26:54,520
Speaker 1: or begin to anyway, three d f X, which had

449
00:26:54,600 --> 00:26:57,680
Speaker 1: transformed the PC gaming industry in the mid nineties when

450
00:26:57,680 --> 00:27:01,480
Speaker 1: they launched the Voodoo three D accelerator car had essentially

451
00:27:01,480 --> 00:27:05,719
Speaker 1: started to burn itself out. It had made some bad acquisitions, uh,

452
00:27:05,840 --> 00:27:08,440
Speaker 1: some of the cards that had created just weren't performing

453
00:27:08,480 --> 00:27:11,440
Speaker 1: well in the marketplace, and in Video made a move

454
00:27:11,480 --> 00:27:14,920
Speaker 1: to acquire three D effects, mostly for the intellectual property

455
00:27:14,960 --> 00:27:17,960
Speaker 1: and for some of their engineers, and it began the

456
00:27:18,000 --> 00:27:22,240
Speaker 1: acquisition process starting at the tail end of two thousand.

457
00:27:22,760 --> 00:27:26,640
Speaker 1: But this acquisition process was incredibly painful and troublesome. There

458
00:27:26,640 --> 00:27:29,080
Speaker 1: were a ton of different court cases that came out

459
00:27:29,520 --> 00:27:33,200
Speaker 1: from creditors who said that, you know, three D f

460
00:27:33,359 --> 00:27:35,360
Speaker 1: X had owed them a lot of money and therefore

461
00:27:35,440 --> 00:27:39,400
Speaker 1: they deserved a larger slice of cash from the acquisition.

462
00:27:39,880 --> 00:27:44,639
Speaker 1: So that process did not officially conclude until two thousand eight,

463
00:27:45,240 --> 00:27:49,200
Speaker 1: So it's like an eight year acquisition process that's pretty painful.

464
00:27:49,760 --> 00:27:52,440
Speaker 1: At this point, A t I was still the dominant

465
00:27:52,520 --> 00:27:55,720
Speaker 1: name in graphics cards in Video was in second place,

466
00:27:56,000 --> 00:27:58,560
Speaker 1: but that would quickly switch where in Video was finally

467
00:27:58,600 --> 00:28:01,919
Speaker 1: able to kind of overtake A t I. The follow

468
00:28:02,000 --> 00:28:04,040
Speaker 1: up of the G Force two fifty six was the

469
00:28:04,040 --> 00:28:07,080
Speaker 1: g Force two just seems like a prequel to me,

470
00:28:07,480 --> 00:28:10,480
Speaker 1: but anyway, g Force two line of cards were the

471
00:28:10,600 --> 00:28:15,119
Speaker 1: first from Nvidia to support multiple monitor configurations, which is

472
00:28:15,119 --> 00:28:17,760
Speaker 1: something we take for granted these days. Then after that

473
00:28:17,880 --> 00:28:20,520
Speaker 1: came the g Force three that debuted in February two

474
00:28:20,800 --> 00:28:24,600
Speaker 1: one and in video introduced some new technologies with these cards,

475
00:28:24,640 --> 00:28:29,200
Speaker 1: implementing processors saving features, like one called light Speed Memory

476
00:28:29,320 --> 00:28:32,840
Speaker 1: Architecture or l m A. Essentially, what lm A did

477
00:28:33,080 --> 00:28:35,640
Speaker 1: two pixels is the same thing that MP three files

478
00:28:35,720 --> 00:28:39,000
Speaker 1: due to sound. It's a it's sort of a compression strategy.

479
00:28:39,320 --> 00:28:42,640
Speaker 1: So what I mean by that is MP three compression.

480
00:28:42,960 --> 00:28:45,160
Speaker 1: One of the strategy is MP three uses is it

481
00:28:45,240 --> 00:28:49,959
Speaker 1: analyzes the sound within a file and it says, anything

482
00:28:50,320 --> 00:28:53,320
Speaker 1: that appears to be outside the range of human hearing,

483
00:28:53,560 --> 00:28:55,520
Speaker 1: we're just gonna get rid of that data. We don't

484
00:28:55,520 --> 00:28:57,560
Speaker 1: need it because no one could hear it anyway. If

485
00:28:57,560 --> 00:28:59,320
Speaker 1: you can't hear it, we don't need to save it.

486
00:28:59,640 --> 00:29:02,840
Speaker 1: And that save space, right, We aren't saving all that

487
00:29:02,920 --> 00:29:06,200
Speaker 1: extraneous information. L m A does the same thing, except

488
00:29:06,280 --> 00:29:10,800
Speaker 1: with pixels. It would indicate like, well, these pixels are

489
00:29:10,880 --> 00:29:15,320
Speaker 1: being blocked on the screen by some other component, so

490
00:29:15,800 --> 00:29:19,040
Speaker 1: there's no need to save this information because those pixels

491
00:29:19,040 --> 00:29:21,320
Speaker 1: are not going to be displayed. Let's say it's pixels

492
00:29:21,320 --> 00:29:25,400
Speaker 1: that represent a character and the character passes behind something

493
00:29:26,080 --> 00:29:30,000
Speaker 1: that is uh blocking your view of the character. Well,

494
00:29:30,040 --> 00:29:33,160
Speaker 1: there's no reason to keep the information of the pixels

495
00:29:33,160 --> 00:29:35,160
Speaker 1: that relate to the character at that point because you

496
00:29:35,200 --> 00:29:40,080
Speaker 1: can't see it anyway. So uh. It was called a

497
00:29:40,120 --> 00:29:44,160
Speaker 1: process called z occlusion culling, and it would help save

498
00:29:46,080 --> 00:29:49,120
Speaker 1: save bandwidth because you don't need as much You're not

499
00:29:49,160 --> 00:29:52,360
Speaker 1: sending as much data through, so you can, uh you

500
00:29:52,360 --> 00:29:54,720
Speaker 1: don't have to worry about taking up so much bandwidth space.

501
00:29:55,280 --> 00:29:57,240
Speaker 1: This marked a turning point for in video. This was

502
00:29:57,280 --> 00:30:01,320
Speaker 1: the point wherein Video overtook a TI as the industry leader. Actually,

503
00:30:01,360 --> 00:30:03,640
Speaker 1: there was a couple of shifts here. At this point,

504
00:30:03,720 --> 00:30:08,720
Speaker 1: in Vidia commanded of the graphics market. Intel was in

505
00:30:08,800 --> 00:30:11,640
Speaker 1: second place with and a t I had fallen to

506
00:30:11,680 --> 00:30:16,080
Speaker 1: third place. Now that would not remain steady over the

507
00:30:16,080 --> 00:30:20,280
Speaker 1: next few years. Companies would swap places a couple of times,

508
00:30:20,760 --> 00:30:23,760
Speaker 1: but it did show that in Vidia had now become

509
00:30:24,440 --> 00:30:28,200
Speaker 1: a truly dominant player in this space. Then we get

510
00:30:28,240 --> 00:30:30,719
Speaker 1: the g Force four that came out in February two.

511
00:30:31,880 --> 00:30:34,280
Speaker 1: There were several budget models that came out along with

512
00:30:34,360 --> 00:30:37,440
Speaker 1: that series, and they were in a line called m X.

513
00:30:37,600 --> 00:30:41,760
Speaker 1: They were actually focused or built largely upon the architecture

514
00:30:41,800 --> 00:30:44,800
Speaker 1: from g Force two, not g Force four. So even

515
00:30:44,840 --> 00:30:46,960
Speaker 1: though they were part of the g Force four series,

516
00:30:47,600 --> 00:30:50,200
Speaker 1: they were based on an on an earlier and some

517
00:30:50,280 --> 00:30:53,800
Speaker 1: would say outdated architecture, and they had limited features, but

518
00:30:53,800 --> 00:30:55,520
Speaker 1: they also had a lower price point. So here in

519
00:30:55,640 --> 00:30:59,080
Speaker 1: Video was saying, I'm seeing all these other companies coming

520
00:30:59,120 --> 00:31:06,200
Speaker 1: out with budget graphics cards aimed at a different target audience.

521
00:31:06,240 --> 00:31:08,280
Speaker 1: What if we win after those people and we just

522
00:31:08,320 --> 00:31:13,880
Speaker 1: start creating budget, lower feature graphics cards. And that's sort

523
00:31:13,920 --> 00:31:17,720
Speaker 1: of what this this was. Fortune would name in Video

524
00:31:17,880 --> 00:31:20,560
Speaker 1: the fastest growing company in the United States in two

525
00:31:20,560 --> 00:31:24,120
Speaker 1: thousand two, and in two thousand three, in Video acquired

526
00:31:24,120 --> 00:31:26,760
Speaker 1: a company called media Q that specialized in graphics and

527
00:31:26,800 --> 00:31:30,080
Speaker 1: MultiMate technology for wireless devices. Is where we start seeing

528
00:31:30,120 --> 00:31:33,760
Speaker 1: in Video getting into mobile technology as well. This is

529
00:31:33,800 --> 00:31:37,400
Speaker 1: also the year when in Video co founder Curtis Prium retires,

530
00:31:38,120 --> 00:31:40,160
Speaker 1: and it's also the year and Video would release the

531
00:31:40,240 --> 00:31:44,640
Speaker 1: first cards in its g Force f X series, including

532
00:31:44,680 --> 00:31:49,320
Speaker 1: one that became infamous. The g Force hundred, had a

533
00:31:49,440 --> 00:31:54,320
Speaker 1: nickname in the industry, it was the dust Buster. Why well,

534
00:31:54,360 --> 00:31:57,080
Speaker 1: the dust Buster or g Force fifty hundred had a

535
00:31:57,200 --> 00:32:02,200
Speaker 1: dual slot cooling mechanism double and cooling takes up two slots,

536
00:32:02,920 --> 00:32:07,400
Speaker 1: high powered, very loud. So people said, oh, it sounds

537
00:32:07,440 --> 00:32:10,000
Speaker 1: like that you've got a vacuum cleaner inside your computer,

538
00:32:10,320 --> 00:32:15,760
Speaker 1: because it's it's so distracting, so loud. This was a

539
00:32:15,760 --> 00:32:18,760
Speaker 1: a piece of technology that was built in part by

540
00:32:18,800 --> 00:32:21,400
Speaker 1: some of the engineers that in video had pulled over

541
00:32:21,480 --> 00:32:26,440
Speaker 1: from three Dfx, and that it got a lot of

542
00:32:26,520 --> 00:32:29,080
Speaker 1: ribbing in the industry for the fact that it was

543
00:32:29,200 --> 00:32:32,360
Speaker 1: it was so loud, but others in the f X

544
00:32:32,440 --> 00:32:36,320
Speaker 1: line would end up getting some acclaim. The company created

545
00:32:36,360 --> 00:32:39,880
Speaker 1: a demo as well in this in two thousand three.

546
00:32:40,600 --> 00:32:44,440
Speaker 1: The demo was called Dawn d A w N and

547
00:32:44,480 --> 00:32:46,800
Speaker 1: it was all meant to show off the various features

548
00:32:46,960 --> 00:32:52,760
Speaker 1: of what the the g Force Effex line of graphics

549
00:32:52,760 --> 00:32:55,640
Speaker 1: cards was capable of doing. And the way it did

550
00:32:55,720 --> 00:33:02,320
Speaker 1: this was it showed a female ferry, little wings, translucent wings,

551
00:33:03,440 --> 00:33:08,760
Speaker 1: translucent outfit. Um. The the naughty bits were covered by

552
00:33:08,840 --> 00:33:12,360
Speaker 1: opaque costuming, but the rest of it was see through

553
00:33:13,080 --> 00:33:16,760
Speaker 1: and the ferry was very curvacious a lot of people

554
00:33:17,200 --> 00:33:19,440
Speaker 1: called out a video saying it was the most blatant

555
00:33:19,560 --> 00:33:22,280
Speaker 1: use of sex appeal to try and sell a graphics

556
00:33:22,280 --> 00:33:26,320
Speaker 1: card up to that date. UH, So they got some

557
00:33:26,560 --> 00:33:34,400
Speaker 1: criticism for their use of this particular visual effect in

558
00:33:34,480 --> 00:33:37,440
Speaker 1: order to try and sell cards. In two thousand four,

559
00:33:38,280 --> 00:33:40,960
Speaker 1: s l I technology would debut, and that would allow

560
00:33:41,040 --> 00:33:44,720
Speaker 1: users to link together multiple GPUs to boost performance even further.

561
00:33:44,880 --> 00:33:48,280
Speaker 1: That it's where you start seeing these crazy gaming rigs

562
00:33:48,280 --> 00:33:54,239
Speaker 1: that have multiple GPUs in link together, plus typically some

563
00:33:54,320 --> 00:33:59,200
Speaker 1: sort of crazy cooling system, because as you start adding

564
00:33:59,200 --> 00:34:02,200
Speaker 1: more CPU and GPU power to a system, it's going

565
00:34:02,240 --> 00:34:05,760
Speaker 1: to generate more heat, and heat and electronics don't go

566
00:34:05,920 --> 00:34:09,120
Speaker 1: well together. If they get hot enough, things tend to

567
00:34:09,120 --> 00:34:12,840
Speaker 1: break down. So there are a lot of innovative cooling

568
00:34:12,880 --> 00:34:14,719
Speaker 1: systems that came out around this time as well to

569
00:34:14,760 --> 00:34:19,319
Speaker 1: try and keep those temperatures at at nominal levels so

570
00:34:19,640 --> 00:34:22,960
Speaker 1: things would still operate well. In two thousand five, in

571
00:34:23,200 --> 00:34:27,480
Speaker 1: Video would UH get the contract to develop the graphics

572
00:34:27,560 --> 00:34:31,440
Speaker 1: processor for Sony's PlayStation three console. So they had already

573
00:34:31,480 --> 00:34:33,600
Speaker 1: developed the one for Xbox. Now they were developing the

574
00:34:33,600 --> 00:34:36,000
Speaker 1: one from PlayStation three, and in two thousand six and

575
00:34:36,120 --> 00:34:39,759
Speaker 1: Video introduced the seventy x two, which was the first

576
00:34:39,760 --> 00:34:43,000
Speaker 1: of in videous cards to feature two GPU single boards

577
00:34:43,360 --> 00:34:47,200
Speaker 1: packaged as a single product, so it was a dual

578
00:34:47,440 --> 00:34:52,719
Speaker 1: GPU all in one package. On October two thousand six,

579
00:34:52,800 --> 00:34:57,319
Speaker 1: a m D would acquire a t I, the competitor

580
00:34:57,440 --> 00:35:00,719
Speaker 1: to in video. That acquisition was a huge deal. Is

581
00:35:00,760 --> 00:35:05,600
Speaker 1: like five point four billion dollars, an enormous deal. And

582
00:35:05,719 --> 00:35:07,920
Speaker 1: now it was no longer in video versus a t I.

583
00:35:08,080 --> 00:35:10,440
Speaker 1: Now it was in video versus a m D. And

584
00:35:10,520 --> 00:35:15,760
Speaker 1: remember that H. Jensen Huang, the president and CEO of Nvidia,

585
00:35:15,840 --> 00:35:18,279
Speaker 1: the co founder of Nvidia, had you had worked for

586
00:35:18,320 --> 00:35:20,920
Speaker 1: a m D for a while. In June two seven,

587
00:35:21,480 --> 00:35:26,480
Speaker 1: and Video introduced a line of math coprocessors called Tesla.

588
00:35:26,680 --> 00:35:29,600
Speaker 1: Now these initially used architecture that was originally built for

589
00:35:29,640 --> 00:35:33,240
Speaker 1: the g Force and Quadrille products, and they are general

590
00:35:33,320 --> 00:35:38,160
Speaker 1: purpose GPUs because now there was this emerging opportunity to

591
00:35:38,280 --> 00:35:42,920
Speaker 1: use GPUs to tackle other problems besides rendering graphics. GPUs

592
00:35:42,960 --> 00:35:47,920
Speaker 1: can power supercomputers, and they can perform calculations that have

593
00:35:48,000 --> 00:35:51,600
Speaker 1: high floating operations per seconds or or flops. They can

594
00:35:51,920 --> 00:35:55,240
Speaker 1: handle huge amounts of data and and and perform lots

595
00:35:55,320 --> 00:35:58,560
Speaker 1: of operations in a short amount of time. The big

596
00:35:58,600 --> 00:36:01,719
Speaker 1: difference between the Tesla GPUs and the ones that we're

597
00:36:01,840 --> 00:36:06,160
Speaker 1: powering gaming rigs is that the Tesla versions didn't have

598
00:36:06,239 --> 00:36:09,120
Speaker 1: any output to displays because that's not what they were for.

599
00:36:09,280 --> 00:36:11,680
Speaker 1: They weren't meant to create graphics. They were meant to

600
00:36:11,719 --> 00:36:16,520
Speaker 1: crunch numbers. And uh, I think this really does mark

601
00:36:16,560 --> 00:36:22,680
Speaker 1: the beginning of using GPUs to do some pretty phenomenal stuff.

602
00:36:23,120 --> 00:36:25,320
Speaker 1: And we see it and a lot of parallel processing

603
00:36:25,360 --> 00:36:29,520
Speaker 1: applications where you have these problems that can be broken

604
00:36:29,600 --> 00:36:33,040
Speaker 1: up into smaller parts. A lot of GPUs are really

605
00:36:33,040 --> 00:36:35,640
Speaker 1: good at handling that kind of stuff. The g Force

606
00:36:35,800 --> 00:36:39,279
Speaker 1: eight series would come out from Video and included chip

607
00:36:39,320 --> 00:36:44,680
Speaker 1: sets that had high lead solder and inadequate cooling mechanisms.

608
00:36:45,360 --> 00:36:48,399
Speaker 1: So there were a substantial number, not all of them,

609
00:36:48,440 --> 00:36:51,960
Speaker 1: but a substantial number of cards that had g Force

610
00:36:52,040 --> 00:36:56,440
Speaker 1: eight series chips in them that ended up suffering failures

611
00:36:56,920 --> 00:37:03,440
Speaker 1: because of overheating issues. This led to in Vidia writing

612
00:37:03,480 --> 00:37:07,960
Speaker 1: off about a couple of hundred million dollars worth of product.

613
00:37:08,680 --> 00:37:12,000
Speaker 1: Not to mention it led to lawsuits that alleged that

614
00:37:12,200 --> 00:37:17,319
Speaker 1: the CEO Jensen Huang and the CFO Marvin Burkett were

615
00:37:17,360 --> 00:37:21,720
Speaker 1: aware of these manufacturing problems, but they didn't say anything

616
00:37:21,760 --> 00:37:25,720
Speaker 1: about it. They attempted to go forward despite these issues,

617
00:37:26,360 --> 00:37:29,360
Speaker 1: and according to one source, the full amount that the the

618
00:37:29,400 --> 00:37:33,799
Speaker 1: company incurred as a cost relating to this problem ended

619
00:37:33,880 --> 00:37:38,719
Speaker 1: up being around four hundred seventies six million dollars, and

620
00:37:38,920 --> 00:37:42,879
Speaker 1: one of the UH sources I was reading said, it's

621
00:37:42,920 --> 00:37:46,480
Speaker 1: unfortunate that this happened. It ended up being such a

622
00:37:46,520 --> 00:37:49,719
Speaker 1: black mark against in Video's reputation that a lot of

623
00:37:49,719 --> 00:37:53,839
Speaker 1: people to this day hold it against the company. Um

624
00:37:54,239 --> 00:37:59,080
Speaker 1: because it was it came across is so deceptive and

625
00:37:59,239 --> 00:38:01,680
Speaker 1: um and see like it was exploiting the customer base,

626
00:38:02,040 --> 00:38:05,279
Speaker 1: and Video would stumble again, at least according to a

627
00:38:05,280 --> 00:38:09,040
Speaker 1: lot of critics. With the G Force nine series, a

628
00:38:09,040 --> 00:38:11,160
Speaker 1: lot of people said that it didn't really have very

629
00:38:11,280 --> 00:38:16,000
Speaker 1: much very many innovative factors, but in Video would introduce

630
00:38:16,080 --> 00:38:20,880
Speaker 1: new micro architecture code named FIRMI in with a launch

631
00:38:20,920 --> 00:38:24,840
Speaker 1: of a news series, the G Force four series, which

632
00:38:25,320 --> 00:38:30,719
Speaker 1: did not It didn't equal the performance of a m

633
00:38:30,800 --> 00:38:34,400
Speaker 1: d S flagship cards, but it was more affordable and

634
00:38:34,440 --> 00:38:37,160
Speaker 1: it performed well enough. It just wasn't at the same

635
00:38:38,120 --> 00:38:41,160
Speaker 1: level of metrics that a m d S flagships were

636
00:38:41,200 --> 00:38:45,600
Speaker 1: performing at, and in Video had also throttled the clock

637
00:38:45,680 --> 00:38:51,240
Speaker 1: speed of the cards on that on that series, largely

638
00:38:51,280 --> 00:38:55,040
Speaker 1: in order to make sure that power consumption stayed at

639
00:38:55,080 --> 00:38:57,520
Speaker 1: nominal levels. But it did mean that if you were

640
00:38:57,520 --> 00:39:01,239
Speaker 1: a serious gamer and you didn't mind getting your your

641
00:39:01,280 --> 00:39:06,439
Speaker 1: hands a little dirty, uh metaphorically speaking, you could over

642
00:39:06,560 --> 00:39:11,839
Speaker 1: clock your GPU. You could make it run faster than

643
00:39:11,920 --> 00:39:14,880
Speaker 1: what it was supposed to run. Because in Video had

644
00:39:14,920 --> 00:39:18,880
Speaker 1: been very conservative and had throttled, the actual top speed

645
00:39:19,560 --> 00:39:22,400
Speaker 1: of the processing units was much higher than what in

646
00:39:22,520 --> 00:39:25,319
Speaker 1: Video put it out to be, So if you wanted to,

647
00:39:25,400 --> 00:39:27,920
Speaker 1: you could overclock it get some better performance out of it.

648
00:39:28,200 --> 00:39:31,000
Speaker 1: And Video followed up for me with other micro architecture

649
00:39:31,040 --> 00:39:37,239
Speaker 1: designs you had Kepler had, Maxwell, Pascal and Turing and

650
00:39:37,560 --> 00:39:40,719
Speaker 1: just in general, the way micro architecture tends to work.

651
00:39:41,200 --> 00:39:45,319
Speaker 1: Intel does this as well, is that you get what

652
00:39:45,480 --> 00:39:51,479
Speaker 1: Intel calls the TICK talk approach, where in TICK you

653
00:39:52,239 --> 00:39:57,920
Speaker 1: create a new layout of your of your microprocessor that

654
00:39:58,000 --> 00:40:02,080
Speaker 1: has smaller components. It's when you've shrunk things down so

655
00:40:02,120 --> 00:40:05,400
Speaker 1: that you can fit more individual components on the same

656
00:40:05,520 --> 00:40:11,160
Speaker 1: amount of chip space as the previous generation. So UH

657
00:40:11,400 --> 00:40:14,120
Speaker 1: think of it as like cramming more employees into the

658
00:40:14,160 --> 00:40:18,120
Speaker 1: same office space. And and so you can keep doing that,

659
00:40:18,520 --> 00:40:20,920
Speaker 1: especially if you get smaller employees right they take up

660
00:40:21,000 --> 00:40:26,240
Speaker 1: less space individually. Then the tox side is where you

661
00:40:26,719 --> 00:40:32,080
Speaker 1: end up fine tuning the layout of the office space

662
00:40:32,160 --> 00:40:35,799
Speaker 1: to accommodate the fact that you've crammed more employees in there.

663
00:40:36,800 --> 00:40:40,920
Speaker 1: So when I talk about these different micro architecture designs,

664
00:40:41,080 --> 00:40:45,000
Speaker 1: it's a combination of things. It's creating dye sets that

665
00:40:45,120 --> 00:40:49,520
Speaker 1: have smaller components so you can cram more of them

666
00:40:49,600 --> 00:40:54,680
Speaker 1: onto that chip, and it's rearranging the orientation of those

667
00:40:54,680 --> 00:40:58,080
Speaker 1: components so that they operate at a greater efficiency. And

668
00:40:58,120 --> 00:41:01,440
Speaker 1: so each of those changes in micro arch texture represent

669
00:41:01,600 --> 00:41:06,880
Speaker 1: an attempt to produce more power more efficiently than the

670
00:41:06,920 --> 00:41:12,600
Speaker 1: previous generation of chips. And that's generally how we see

671
00:41:12,640 --> 00:41:16,879
Speaker 1: all microprocessors, whether they're CPUs or GPUs, how they how

672
00:41:16,920 --> 00:41:21,080
Speaker 1: they continue and we're still seeing it today in video

673
00:41:21,120 --> 00:41:25,400
Speaker 1: is still one of the top names in graphics cards

674
00:41:26,160 --> 00:41:30,319
Speaker 1: and UH their CEO, Jensen Wong, is worth more than

675
00:41:30,400 --> 00:41:35,320
Speaker 1: six billion dollars. He is the sixty one highest paid CEO.

676
00:41:36,040 --> 00:41:41,080
Speaker 1: Not bad for somebody who back in said yeah, why

677
00:41:41,080 --> 00:41:44,520
Speaker 1: don't we give us a shot? So a pretty amazing story.

678
00:41:44,840 --> 00:41:47,680
Speaker 1: And like I said, I really skipped over quite a

679
00:41:47,719 --> 00:41:50,399
Speaker 1: lot in this last section. You know, I didn't go

680
00:41:50,920 --> 00:41:54,759
Speaker 1: card by card because frankly, to go through all those

681
00:41:54,800 --> 00:41:58,759
Speaker 1: specifications and explain their relevance would mean having to do

682
00:41:59,040 --> 00:42:05,240
Speaker 1: companion episode odes about the evolution of graphics in UH,

683
00:42:05,280 --> 00:42:10,720
Speaker 1: in computers in general, UH, and programming and the whole

684
00:42:11,360 --> 00:42:15,080
Speaker 1: thought process behind graphics cards as well, and it would

685
00:42:15,080 --> 00:42:20,319
Speaker 1: require a lot more, a lot several more episodes. The

686
00:42:20,360 --> 00:42:22,520
Speaker 1: important thing to remember now is that in videos not

687
00:42:22,600 --> 00:42:25,759
Speaker 1: just making these graphics cards for for gamers, although they

688
00:42:25,760 --> 00:42:29,000
Speaker 1: still do that obviously, but that they have branched out

689
00:42:29,040 --> 00:42:32,279
Speaker 1: into other areas. GPUs these days are being used in

690
00:42:32,360 --> 00:42:36,480
Speaker 1: all sorts of fields, including artificial intelligence and deep learning,

691
00:42:36,719 --> 00:42:39,960
Speaker 1: machine learning, that kind of stuff. So we're seeing a

692
00:42:40,040 --> 00:42:45,240
Speaker 1: lot more relevance of GPUs, also used in bitcoin mining,

693
00:42:45,960 --> 00:42:52,000
Speaker 1: u in in UH cryptography, we're seeing tons of applications

694
00:42:52,000 --> 00:42:57,080
Speaker 1: of GPUs far beyond making Laura Croft look amazing as

695
00:42:57,120 --> 00:43:02,360
Speaker 1: she runs through the jungle. So prey the phenomenal story.

696
00:43:02,840 --> 00:43:07,160
Speaker 1: I hope you enjoyed this episode or these episodes about Nvidia. Uh.

697
00:43:07,280 --> 00:43:10,160
Speaker 1: In our next episodes, we're gonna cover totally different topics,

698
00:43:10,160 --> 00:43:13,400
Speaker 1: so stick around. If you guys have suggestions for future topics.

699
00:43:13,840 --> 00:43:16,520
Speaker 1: Maybe there's a company I should cover, a person in tech,

700
00:43:16,640 --> 00:43:19,120
Speaker 1: a technology in particular that you would like to hear

701
00:43:19,160 --> 00:43:22,680
Speaker 1: more about. Send me an email the addresses tech stuff

702
00:43:22,800 --> 00:43:26,000
Speaker 1: at how stuff works dot com or drop me a

703
00:43:26,040 --> 00:43:28,719
Speaker 1: line on Facebook or Twitter to handle with both of those.

704
00:43:28,840 --> 00:43:32,960
Speaker 1: Is tech stuff h s W. Go to T public

705
00:43:33,120 --> 00:43:36,480
Speaker 1: dot com slash tech stuff that's T E E Public

706
00:43:36,840 --> 00:43:40,280
Speaker 1: dot com slash tech stuff for all your tech stuff

707
00:43:40,360 --> 00:43:43,880
Speaker 1: merchandise needs. We've got some cool stuff over there, and

708
00:43:43,960 --> 00:43:46,640
Speaker 1: every purchase helps out the show, so we appreciate it,

709
00:43:47,080 --> 00:43:50,000
Speaker 1: and don't forget follow us on Instagram and I'll talk

710
00:43:50,040 --> 00:43:58,719
Speaker 1: to you again really soon for more on this and

711
00:43:58,760 --> 00:44:01,279
Speaker 1: bathands of other topics. Is that how stuff works dot

712
00:44:01,360 --> 00:44:11,320
Speaker 1: com