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 --> 00:00:13,800 Speaker 1: stuff works dot com. Hey there, and welcome to tech Stuff. 3 00:00:13,840 --> 00:00:16,360 Speaker 1: I'm your host, Jonathan Strickling. I'm an executive producer with 4 00:00:16,360 --> 00:00:19,440 Speaker 1: how Stuff Works in love all Things Tech, and I 5 00:00:19,480 --> 00:00:23,000 Speaker 1: have received lots of requests over the years to cover 6 00:00:23,160 --> 00:00:26,880 Speaker 1: the company in Video and the products they make, and 7 00:00:26,920 --> 00:00:30,600 Speaker 1: in Video is chiefly known for their GPUs, or graphics 8 00:00:30,680 --> 00:00:34,680 Speaker 1: processing units. So before we jump into the history of 9 00:00:34,800 --> 00:00:38,160 Speaker 1: Nvidia and how it helped define the evolution of computers 10 00:00:38,320 --> 00:00:42,440 Speaker 1: and how it created the term GPU, how about we 11 00:00:42,479 --> 00:00:45,879 Speaker 1: explore the concept of graphics cards and GPUs in general. 12 00:00:45,920 --> 00:00:50,440 Speaker 1: A and just a note for you guys, this gets 13 00:00:50,440 --> 00:00:53,640 Speaker 1: super technical, super fast. So I am going to take 14 00:00:53,720 --> 00:00:57,000 Speaker 1: a very high level on all of this because if 15 00:00:57,040 --> 00:01:00,760 Speaker 1: you really want to know and have a but understanding 16 00:01:00,840 --> 00:01:04,479 Speaker 1: of the history of GPUs and graphics cards, you also 17 00:01:04,600 --> 00:01:08,959 Speaker 1: have to have an understanding of the evolution of how 18 00:01:09,760 --> 00:01:14,560 Speaker 1: we program graphics, things like application programmer interfaces or a 19 00:01:14,640 --> 00:01:17,840 Speaker 1: p I s um that kind of stuff. You need 20 00:01:17,880 --> 00:01:21,759 Speaker 1: to know about that evolution in addition to the evolution 21 00:01:21,959 --> 00:01:27,399 Speaker 1: of the technology itself. And it gets super technical. And 22 00:01:28,000 --> 00:01:30,640 Speaker 1: while it might be interesting to a subset of tech 23 00:01:30,640 --> 00:01:33,360 Speaker 1: stuff listeners. I feel like for a lot of people, 24 00:01:33,520 --> 00:01:35,959 Speaker 1: I would lose them very quickly. I have a feeling 25 00:01:35,959 --> 00:01:38,839 Speaker 1: this These episodes are going to lose some folks anyway, 26 00:01:39,080 --> 00:01:41,960 Speaker 1: and that's okay. But for those who are interested in 27 00:01:42,360 --> 00:01:45,720 Speaker 1: in video and graphics cards, stick around. We're gonna learn 28 00:01:45,800 --> 00:01:50,640 Speaker 1: some interesting stuff. So to understand in video's role, we 29 00:01:50,720 --> 00:01:55,240 Speaker 1: actually have to dial way back before the founding of Nvidio, 30 00:01:55,320 --> 00:01:59,640 Speaker 1: which was We're gonna go all the way back to 31 00:01:59,760 --> 00:02:03,800 Speaker 1: nine teen fifty one, just briefly, because that's when the U. S. 32 00:02:03,920 --> 00:02:06,920 Speaker 1: Navy awarded a contract to M I T to design 33 00:02:06,960 --> 00:02:10,360 Speaker 1: a flight simulator, and M I T created a simulator 34 00:02:10,440 --> 00:02:15,320 Speaker 1: called Whirlwind. That system had an early three D graphics system, 35 00:02:15,360 --> 00:02:18,239 Speaker 1: but the architecture of Whirlwind wouldn't become the basis for 36 00:02:18,320 --> 00:02:21,040 Speaker 1: the graphics card of the future. However, it's important to 37 00:02:21,120 --> 00:02:26,360 Speaker 1: note that people were thinking about trying to create graphics 38 00:02:26,360 --> 00:02:30,760 Speaker 1: that would represent three dimensional objects using computer systems as 39 00:02:30,800 --> 00:02:34,760 Speaker 1: early as the nineteen fifties. But to look at the 40 00:02:34,840 --> 00:02:38,720 Speaker 1: kind of technology that eventually evolved into the realm where 41 00:02:38,760 --> 00:02:41,519 Speaker 1: in video got involved, we have to look really more 42 00:02:41,560 --> 00:02:44,760 Speaker 1: at the birth of Arcades home video game consoles and 43 00:02:44,840 --> 00:02:48,440 Speaker 1: the personal computer, so this would be the mid nineteen seventies. 44 00:02:49,000 --> 00:02:52,960 Speaker 1: Arcade machines and consoles had specialized chips in them, and 45 00:02:53,000 --> 00:02:55,760 Speaker 1: we typically would refer to them as things like video 46 00:02:55,840 --> 00:03:01,040 Speaker 1: address generators or video shifters, and the chips or circuits 47 00:03:01,040 --> 00:03:05,840 Speaker 1: in some cases would take information from a processor and 48 00:03:05,919 --> 00:03:08,799 Speaker 1: relay that in a meaningful way to a display in 49 00:03:08,960 --> 00:03:10,760 Speaker 1: order to create the images you would see on a 50 00:03:10,760 --> 00:03:15,480 Speaker 1: computer monitor or a television. So remember that a processor 51 00:03:15,520 --> 00:03:19,239 Speaker 1: like a CPU central processing unit, will take in data 52 00:03:19,639 --> 00:03:24,040 Speaker 1: from a source. You get input that goes into the CPU. 53 00:03:24,120 --> 00:03:28,600 Speaker 1: The CPU will perform various operations on that data according 54 00:03:28,639 --> 00:03:32,440 Speaker 1: to a program running on the machine, and then produce 55 00:03:32,560 --> 00:03:36,560 Speaker 1: output that is meaningful based upon the combination of the 56 00:03:36,600 --> 00:03:41,520 Speaker 1: program and the input. And then that output will go somewhere. 57 00:03:41,920 --> 00:03:44,720 Speaker 1: It might feed into a new process, it might go 58 00:03:44,840 --> 00:03:48,600 Speaker 1: to an output device, and this is in good old 59 00:03:48,680 --> 00:03:54,200 Speaker 1: machine code. So the graphics chips or circuits their purpose 60 00:03:54,320 --> 00:03:58,640 Speaker 1: was to receive output data from the CPU that was 61 00:03:58,920 --> 00:04:02,960 Speaker 1: specially scifically intended to go toward a display and then 62 00:04:03,280 --> 00:04:05,680 Speaker 1: to take it from there. So graphics card act as 63 00:04:05,720 --> 00:04:08,839 Speaker 1: sort of a translator. They take the data from the CPU, 64 00:04:08,920 --> 00:04:11,360 Speaker 1: and they turn it into something that's meaningful for a 65 00:04:11,400 --> 00:04:14,800 Speaker 1: display to show a bunch of points of light on 66 00:04:14,840 --> 00:04:18,240 Speaker 1: a screen. Remember, the images we see on screens are 67 00:04:18,240 --> 00:04:21,520 Speaker 1: made up of pixels, and pixels are points of light. 68 00:04:22,320 --> 00:04:25,440 Speaker 1: And the smaller the pixels, the more you can fit 69 00:04:25,520 --> 00:04:29,039 Speaker 1: on a screen, the greater the resolution, just generally speaking. 70 00:04:29,320 --> 00:04:31,760 Speaker 1: So computers need a graphics card or some sort of 71 00:04:31,760 --> 00:04:36,920 Speaker 1: graphics processing unit or interface on the computer's motherboard itself 72 00:04:37,720 --> 00:04:41,000 Speaker 1: or in an expansion card in order to do this. Now, 73 00:04:41,040 --> 00:04:44,040 Speaker 1: those early chips in the nineteen seventies had names like 74 00:04:44,360 --> 00:04:47,320 Speaker 1: the c d P eighteen sixty one from our c 75 00:04:47,520 --> 00:04:50,200 Speaker 1: A or the T I A one A which was 76 00:04:50,480 --> 00:04:56,120 Speaker 1: used in the ATAR in Motorola debut. The MC six 77 00:04:56,160 --> 00:05:00,200 Speaker 1: eight four or five video address generator. That chip give 78 00:05:00,240 --> 00:05:03,120 Speaker 1: instructions to a display that would include information such as 79 00:05:03,200 --> 00:05:06,840 Speaker 1: luminants or how bright the image should be, that pixel 80 00:05:06,880 --> 00:05:11,000 Speaker 1: should be, color, and position. And so the pixels new 81 00:05:11,320 --> 00:05:13,640 Speaker 1: quote unquote knew how bright they need to be, what 82 00:05:13,680 --> 00:05:15,799 Speaker 1: color they needed to be, and the display new which 83 00:05:15,800 --> 00:05:19,000 Speaker 1: pixels needed to be shown it or be activated at 84 00:05:19,040 --> 00:05:23,560 Speaker 1: any given time. More importantly for our story, the MC 85 00:05:23,839 --> 00:05:26,520 Speaker 1: six eight four or five chip for Motorola served as 86 00:05:26,560 --> 00:05:29,880 Speaker 1: the foundation for the first video card for a personal computer, 87 00:05:30,279 --> 00:05:34,240 Speaker 1: which was IBMS Monochrome Display Adapter, which was introduced in 88 00:05:36,120 --> 00:05:38,880 Speaker 1: Now that could not really generate graphics the same way 89 00:05:39,200 --> 00:05:41,359 Speaker 1: we would talk about today, but it could display text 90 00:05:41,400 --> 00:05:46,080 Speaker 1: and characters in monochrome initially. Eventually you also had a 91 00:05:46,120 --> 00:05:49,800 Speaker 1: color one. Intel built a graphics chip called the eight 92 00:05:50,040 --> 00:05:54,320 Speaker 1: two seven twenty which could display up to eight colors. Wow, 93 00:05:54,839 --> 00:05:57,760 Speaker 1: can also support a resolution of two hundred fifty six 94 00:05:57,760 --> 00:06:01,160 Speaker 1: by two fifty six pixels. Now, if you swopped over 95 00:06:01,200 --> 00:06:04,279 Speaker 1: to monochrome, you could actually have a better resolution of 96 00:06:04,279 --> 00:06:08,200 Speaker 1: five twelve by five twelve pixels. And for several years, 97 00:06:08,320 --> 00:06:12,479 Speaker 1: video game consoles and arcade machines were miles ahead of 98 00:06:12,480 --> 00:06:17,280 Speaker 1: personal computers when it came to graphic quality. The circuits 99 00:06:17,279 --> 00:06:20,920 Speaker 1: in the gaming machines were dedicated to that task, and 100 00:06:21,040 --> 00:06:25,400 Speaker 1: early arcade games had systems that were essentially hard coded 101 00:06:25,480 --> 00:06:28,680 Speaker 1: onto circuits, which meant if you wanted to change a 102 00:06:28,760 --> 00:06:31,240 Speaker 1: game out in an arcade cabinet. Let's say that you've 103 00:06:31,440 --> 00:06:34,479 Speaker 1: got an arcade cabinet, and inside that cabinet you have 104 00:06:35,160 --> 00:06:37,720 Speaker 1: I don't know, let's say it's Galica and you want 105 00:06:37,720 --> 00:06:41,279 Speaker 1: to swap out Galaga for pac Man, Well, you can't 106 00:06:41,320 --> 00:06:44,000 Speaker 1: just take a chip out and replace it with another chip. 107 00:06:44,040 --> 00:06:46,839 Speaker 1: You would actually have to essentially gut the whole cabinet 108 00:06:47,200 --> 00:06:51,919 Speaker 1: and replace the INNERDS with the new ones from pac Man. 109 00:06:52,200 --> 00:06:54,920 Speaker 1: There was no swapping a flash drive or anything like that. 110 00:06:55,480 --> 00:07:01,120 Speaker 1: The advantage of the dedicated systems was that a could 111 00:07:01,160 --> 00:07:05,440 Speaker 1: really create pretty good graphics, especially compared to the stuff 112 00:07:05,480 --> 00:07:08,360 Speaker 1: you would see on computers. You get really good results. 113 00:07:08,640 --> 00:07:11,480 Speaker 1: But the downside was it was not a viable approach 114 00:07:11,480 --> 00:07:14,000 Speaker 1: if you wanted to create a general purpose machine like 115 00:07:14,040 --> 00:07:16,800 Speaker 1: a personal computer, and so PCs for the most part, 116 00:07:17,000 --> 00:07:21,400 Speaker 1: lagged behind these specialty machines, at least in the graphics department. 117 00:07:21,720 --> 00:07:24,320 Speaker 1: In the mid nineteen eighties, a company called A t 118 00:07:24,600 --> 00:07:27,560 Speaker 1: I began to manufacture graphics cards. They would end up 119 00:07:27,600 --> 00:07:32,040 Speaker 1: becoming incredibly important in that realm. Their early products were 120 00:07:32,080 --> 00:07:36,760 Speaker 1: o e M, or original equipment manufacturer cards, which meant 121 00:07:37,160 --> 00:07:40,120 Speaker 1: they were producing chips that were meant to be included 122 00:07:40,280 --> 00:07:43,520 Speaker 1: in finished products that other companies were making. So they 123 00:07:43,520 --> 00:07:46,560 Speaker 1: were they were making a component that would be included 124 00:07:46,640 --> 00:07:50,520 Speaker 1: in a larger product that some other manufacturer was producing. 125 00:07:51,200 --> 00:07:56,240 Speaker 1: Um they were not selling directly to the end customers, 126 00:07:56,320 --> 00:08:00,320 Speaker 1: So you wouldn't go into a store and buy an 127 00:08:00,320 --> 00:08:04,760 Speaker 1: a TI chip back in the mid nineteen eighties. Instead 128 00:08:04,760 --> 00:08:07,880 Speaker 1: you would go and buy a computer that had an 129 00:08:07,880 --> 00:08:11,440 Speaker 1: a TI chip in it, or maybe a card that 130 00:08:11,560 --> 00:08:14,560 Speaker 1: had an a TI chip incorporated in it. That would 131 00:08:14,640 --> 00:08:17,480 Speaker 1: change in the late nineteen eighties. You would start seeing 132 00:08:17,520 --> 00:08:21,200 Speaker 1: a t I producing its own cards supporting stuff like 133 00:08:21,280 --> 00:08:23,440 Speaker 1: e G A and v G A graphics and I'm 134 00:08:23,480 --> 00:08:27,119 Speaker 1: not going to go into those. It would be ridiculously 135 00:08:27,560 --> 00:08:30,600 Speaker 1: long winded to go into all those different graphics and 136 00:08:30,760 --> 00:08:33,280 Speaker 1: they largely are unimportant by the time we get around 137 00:08:33,280 --> 00:08:37,120 Speaker 1: to talking about in video anyway. So this, however, was 138 00:08:37,160 --> 00:08:38,520 Speaker 1: one of the big moves that led to the birth 139 00:08:38,559 --> 00:08:41,920 Speaker 1: of the graphics card industry as a consumer facing business. 140 00:08:42,640 --> 00:08:46,800 Speaker 1: By the early nineteen nineties, when Nvidio would debut, more 141 00:08:46,880 --> 00:08:50,840 Speaker 1: companies started making two D graphics cards for computers and 142 00:08:50,960 --> 00:08:55,240 Speaker 1: three D accelerators typically paired with those two D graphics cards, 143 00:08:55,960 --> 00:08:58,040 Speaker 1: and then a t I came out with the Mack 144 00:08:58,160 --> 00:09:02,480 Speaker 1: sixty four that was the the graphics card that also 145 00:09:02,520 --> 00:09:06,800 Speaker 1: included support for a TV tuner and full motion video playback. Acceleration, 146 00:09:07,320 --> 00:09:09,520 Speaker 1: and things were really starting to pick up in those 147 00:09:09,520 --> 00:09:12,480 Speaker 1: early nineties. Now let's take a step back and consider 148 00:09:12,520 --> 00:09:14,880 Speaker 1: why this is a big deal. When you're looking at 149 00:09:14,920 --> 00:09:18,560 Speaker 1: a three D image on a computer screen, meaning one 150 00:09:18,559 --> 00:09:22,079 Speaker 1: that uses various techniques to simulate depths, a lot has 151 00:09:22,120 --> 00:09:24,839 Speaker 1: to happen to produce those results. Because what you're looking 152 00:09:24,840 --> 00:09:29,200 Speaker 1: at is a two dimensional representation of a three dimensional object. 153 00:09:30,040 --> 00:09:34,200 Speaker 1: The graphics processing unit has to construct a wire frame 154 00:09:34,559 --> 00:09:37,360 Speaker 1: for the shape you're looking at, to build out sort 155 00:09:37,400 --> 00:09:42,720 Speaker 1: of a an exoskeleton, if you will, for the entire scene, 156 00:09:43,080 --> 00:09:46,839 Speaker 1: everything from characters in it to backgrounds, et cetera. If 157 00:09:46,840 --> 00:09:51,200 Speaker 1: it's doing a full three dimensional h field of view, 158 00:09:51,880 --> 00:09:55,840 Speaker 1: that frame ultimately consists of a bunch of straight lines typically, 159 00:09:56,160 --> 00:09:58,959 Speaker 1: but the lines can be really short and at angles, 160 00:09:59,320 --> 00:10:03,880 Speaker 1: and if you have the lines small enough and at 161 00:10:04,400 --> 00:10:07,640 Speaker 1: slight enough angles from each other, you can simulate curved shapes. 162 00:10:08,320 --> 00:10:10,880 Speaker 1: And once the wire frame is there, the graphics card 163 00:10:11,120 --> 00:10:14,200 Speaker 1: raster rises the image, which means it fills in the 164 00:10:14,200 --> 00:10:17,520 Speaker 1: pixels kind of like a coloring book. You're filling in 165 00:10:18,040 --> 00:10:22,200 Speaker 1: the spaces for that wire frame. Next, it adds elements 166 00:10:22,240 --> 00:10:26,080 Speaker 1: like color and texture and lighting effects. And that's just 167 00:10:26,240 --> 00:10:30,640 Speaker 1: for one moment, one instance of that image. So if 168 00:10:30,640 --> 00:10:34,000 Speaker 1: you're playing a game, typically you're having stuff change on 169 00:10:34,120 --> 00:10:38,160 Speaker 1: screen pretty regularly and sometimes quickly, so the image needs 170 00:10:38,160 --> 00:10:42,400 Speaker 1: to be updated frequently, and a fast paced shooter, for example, 171 00:10:42,679 --> 00:10:45,040 Speaker 1: will not be a fun experience. If your computer can 172 00:10:45,080 --> 00:10:48,320 Speaker 1: only display a couple of new images every second, you'd 173 00:10:48,320 --> 00:10:51,560 Speaker 1: be toast. If it were a competitive game, thirty frames 174 00:10:51,559 --> 00:10:55,280 Speaker 1: per second is considered slow or at least subpar, But 175 00:10:55,360 --> 00:10:58,360 Speaker 1: if you can get thirty frames per second regularly, tend 176 00:10:58,520 --> 00:11:01,000 Speaker 1: things tend to be playable. You really want something that's 177 00:11:01,000 --> 00:11:04,520 Speaker 1: sixty frames per second at least for a smooth experience. 178 00:11:04,880 --> 00:11:08,040 Speaker 1: And the more graphically rich the image, the more work 179 00:11:08,120 --> 00:11:11,280 Speaker 1: that entails. Right, the more detail you have in your 180 00:11:11,360 --> 00:11:15,480 Speaker 1: in your individual images, the more work your your graphics 181 00:11:15,480 --> 00:11:18,240 Speaker 1: processor is going to have to do in order to 182 00:11:18,360 --> 00:11:23,000 Speaker 1: refresh that sixty times a second. So the graphics processing 183 00:11:23,080 --> 00:11:25,440 Speaker 1: unit has to be able to relay the appropriate information 184 00:11:25,440 --> 00:11:28,280 Speaker 1: at a bandwidth large enough and a speed fast enough 185 00:11:28,320 --> 00:11:31,440 Speaker 1: to produce those results, and it's a big job. Graphics 186 00:11:31,480 --> 00:11:36,240 Speaker 1: cards typically have four main components. They have a motherboard, 187 00:11:36,360 --> 00:11:38,960 Speaker 1: which is a circuit board upon which all the other 188 00:11:39,040 --> 00:11:44,760 Speaker 1: components are on top of UH. The graphics processor and 189 00:11:44,840 --> 00:11:48,880 Speaker 1: other connections are all sitting on this circuit board, and 190 00:11:48,960 --> 00:11:51,920 Speaker 1: the motherboard also has a connector to plug into a computer, 191 00:11:52,320 --> 00:11:56,280 Speaker 1: and that connector allows the motherboard to accept data and 192 00:11:56,400 --> 00:11:59,559 Speaker 1: power from the computer and send data back to the computer. 193 00:12:00,080 --> 00:12:03,480 Speaker 1: Then you've got the processor. It's the processor's job to 194 00:12:03,520 --> 00:12:06,400 Speaker 1: interpret information from the computer and send data to the 195 00:12:06,480 --> 00:12:09,280 Speaker 1: display so that the right pixels show up at the 196 00:12:09,360 --> 00:12:11,280 Speaker 1: right time at the right part of the screen. And 197 00:12:11,320 --> 00:12:15,600 Speaker 1: the processor specializes in performing geometric calculations and can include 198 00:12:15,640 --> 00:12:19,679 Speaker 1: features like anti aliasing that's a smoothing technology to take 199 00:12:19,720 --> 00:12:23,240 Speaker 1: some of those pointy edges off the graphics, since ultimately 200 00:12:23,679 --> 00:12:25,679 Speaker 1: all the shapes you're looking at are made up of 201 00:12:27,160 --> 00:12:31,720 Speaker 1: straight lines. Or there's also an isotropic filtering, which creates 202 00:12:31,920 --> 00:12:36,520 Speaker 1: a more crisp image. Then there's another component on these 203 00:12:36,520 --> 00:12:39,560 Speaker 1: cards that's memory. Graphics cards have memory to store information 204 00:12:39,559 --> 00:12:44,199 Speaker 1: about each of the pixels and temporarily store completed images. UH. 205 00:12:44,280 --> 00:12:47,240 Speaker 1: Then there's some sort of connector that allows you to 206 00:12:47,280 --> 00:12:50,760 Speaker 1: plug the display into the graphics card so that way 207 00:12:50,920 --> 00:12:55,720 Speaker 1: you can actually see the wonderful graphics this card is creating. 208 00:12:55,880 --> 00:12:59,480 Speaker 1: If you don't have a connector, then there's the display 209 00:12:59,480 --> 00:13:03,400 Speaker 1: has no way of receiving that information, right. So also 210 00:13:03,480 --> 00:13:08,319 Speaker 1: these days you might end up using a connector on 211 00:13:08,360 --> 00:13:12,360 Speaker 1: the back to plug one graphics card into a second 212 00:13:12,400 --> 00:13:16,680 Speaker 1: graphics card and run them in series so that you 213 00:13:16,720 --> 00:13:20,240 Speaker 1: can boost the performance. That is becoming more and more 214 00:13:20,280 --> 00:13:24,080 Speaker 1: common where you start seeing big gaming rigs that have 215 00:13:24,200 --> 00:13:29,360 Speaker 1: multiple high performing graphics cards inside of them. Now, these 216 00:13:29,440 --> 00:13:32,080 Speaker 1: cards tend to produce a lot of heat because they're 217 00:13:32,080 --> 00:13:33,880 Speaker 1: working very, very hard, and they've got a lot of 218 00:13:33,880 --> 00:13:36,640 Speaker 1: components on them, and so they tend to also have 219 00:13:36,760 --> 00:13:40,200 Speaker 1: multiple cores. That means the processors can work on parallel problems. 220 00:13:40,840 --> 00:13:44,800 Speaker 1: Most graphics rendering falls into that category, but as it 221 00:13:44,800 --> 00:13:48,040 Speaker 1: turns out, so do other computational problems, which is why 222 00:13:48,120 --> 00:13:51,680 Speaker 1: you will hear about large groups using GPUs to tackle 223 00:13:51,760 --> 00:13:55,280 Speaker 1: certain types of challenges like mining bitcoins. I'll talk more 224 00:13:55,280 --> 00:13:58,120 Speaker 1: about that in our second episode, but that's a good 225 00:13:58,120 --> 00:14:00,880 Speaker 1: basic grounding of graphics cards. When we come back, I'm 226 00:14:00,880 --> 00:14:04,400 Speaker 1: going to talk about the founding of Nvidia and the 227 00:14:04,520 --> 00:14:16,200 Speaker 1: environment it found itself in those early days. In nine three, 228 00:14:16,600 --> 00:14:22,160 Speaker 1: Jensen Huang, Chris Malachowski, and Curtis priam founded in Vidia. 229 00:14:22,480 --> 00:14:25,480 Speaker 1: But where did they come from? Well, Huang, who is 230 00:14:25,520 --> 00:14:28,760 Speaker 1: president and CEO of Nvidia to this day, was born 231 00:14:28,800 --> 00:14:32,440 Speaker 1: in Taiwan, and then he lived in Thailand as a 232 00:14:32,480 --> 00:14:34,760 Speaker 1: young boy until his family sent him and his brother 233 00:14:34,800 --> 00:14:38,640 Speaker 1: to the United States because Thailand was becoming politically unstable, 234 00:14:39,000 --> 00:14:41,680 Speaker 1: and they moved in with some family members over in 235 00:14:41,680 --> 00:14:44,640 Speaker 1: the States. His aunt and uncle enrolled Huang in a 236 00:14:44,720 --> 00:14:48,120 Speaker 1: boarding school in Kentucky. They thought it was a prep school. 237 00:14:48,200 --> 00:14:50,040 Speaker 1: It turned out that it was actually a school for 238 00:14:50,080 --> 00:14:53,320 Speaker 1: troubled youth. He and his brother were reunited with their 239 00:14:53,360 --> 00:14:58,040 Speaker 1: family and moved to Oregon. Huang would attend Oregon State University. 240 00:14:58,080 --> 00:15:02,640 Speaker 1: He studied electrical engineering there and graduated with an undergraduate degree. 241 00:15:02,840 --> 00:15:06,160 Speaker 1: Then he went on to enroll in Stanford University to 242 00:15:06,240 --> 00:15:10,080 Speaker 1: earn his master's degree in electrical engineering. After that, he 243 00:15:10,160 --> 00:15:13,200 Speaker 1: went to work for Advanced micro Devices or a m D. 244 00:15:13,680 --> 00:15:16,440 Speaker 1: Interesting because we'll hear more about a m D in 245 00:15:16,480 --> 00:15:19,200 Speaker 1: the next episode, and he also then worked for l 246 00:15:19,360 --> 00:15:22,440 Speaker 1: s I Logic before he would leave to co found 247 00:15:22,480 --> 00:15:26,160 Speaker 1: in Video Chris Malachowski, who today is part of in 248 00:15:26,240 --> 00:15:30,720 Speaker 1: Vidio's executive staff and serves as a senior technology executive. 249 00:15:31,000 --> 00:15:35,119 Speaker 1: He attended the University of Florida and received his bachelor's 250 00:15:35,120 --> 00:15:38,080 Speaker 1: degree there and then went to Santa Clara University for 251 00:15:38,120 --> 00:15:41,120 Speaker 1: his masters. He went on to work as a manufacturing 252 00:15:41,160 --> 00:15:44,080 Speaker 1: design engineer at Hewlett Packard in nineteen eighty and in 253 00:15:44,160 --> 00:15:48,640 Speaker 1: nineteen seven he started working for Sun Microsystems and became 254 00:15:48,680 --> 00:15:51,360 Speaker 1: a senior staff engineer, and while he was there, he 255 00:15:51,520 --> 00:15:55,480 Speaker 1: co invented the g X graphics architecture and led design 256 00:15:55,520 --> 00:16:01,040 Speaker 1: work on graphics interfaces. Curtis Priam attended rin Alert Polytechnic 257 00:16:01,520 --> 00:16:04,240 Speaker 1: Institute and he earned a degree in electrical engineering in 258 00:16:04,280 --> 00:16:07,880 Speaker 1: nineteen eighty two. He worked on the IBM Professional Graphics 259 00:16:07,880 --> 00:16:11,720 Speaker 1: Adapter and like Malachowski, Prium would end up working at 260 00:16:11,760 --> 00:16:14,480 Speaker 1: Sun Microsystems and was also one of the inventors of 261 00:16:14,480 --> 00:16:18,040 Speaker 1: the g X graphics architecture and now the three founders 262 00:16:18,080 --> 00:16:20,120 Speaker 1: for in Nvidia. He is the one who is no 263 00:16:20,160 --> 00:16:23,080 Speaker 1: longer working with the company. He retired from in Vidia 264 00:16:23,160 --> 00:16:27,280 Speaker 1: in two thousand three. So back in pri Um and 265 00:16:27,320 --> 00:16:33,680 Speaker 1: Malachowski are steaming. They're incredibly frustrated because they had been 266 00:16:33,760 --> 00:16:37,480 Speaker 1: encountering roadblocks at Sun micro Systems. They both felt the 267 00:16:37,520 --> 00:16:40,240 Speaker 1: company was pushing R and D in the wrong direction 268 00:16:40,400 --> 00:16:44,160 Speaker 1: for graphics, and Wong, who was working at l S 269 00:16:44,160 --> 00:16:48,000 Speaker 1: I he wasn't really disenchanted by his work, but he 270 00:16:48,120 --> 00:16:50,520 Speaker 1: was thinking about what the future of processing was going 271 00:16:50,560 --> 00:16:52,960 Speaker 1: to be all about. And the three sat down to 272 00:16:53,000 --> 00:16:57,640 Speaker 1: talk things over at dinner at a Denny's restaurant, which 273 00:16:57,680 --> 00:17:01,240 Speaker 1: I think is great. The idea that a hugely successful 274 00:17:01,280 --> 00:17:05,720 Speaker 1: company was born over a conversation at a Denny's is 275 00:17:05,760 --> 00:17:09,320 Speaker 1: fantastic to me. I've had a lot of conversations at Denny's. 276 00:17:10,080 --> 00:17:14,240 Speaker 1: None of them have led to a monumentally successful company, 277 00:17:14,280 --> 00:17:16,640 Speaker 1: so I'm not doing it right. But anyway, the three 278 00:17:16,640 --> 00:17:20,160 Speaker 1: of them, I started talking about the future of computing, 279 00:17:20,160 --> 00:17:24,760 Speaker 1: and they agreed that graphics based accelerated computing was the future. 280 00:17:25,400 --> 00:17:30,440 Speaker 1: And gosh, they were incredibly prescient about that. We'll we'll 281 00:17:30,440 --> 00:17:32,720 Speaker 1: talk more about that, I mean, but I kind of 282 00:17:32,760 --> 00:17:35,880 Speaker 1: alluded to it earlier, about how graphics processing units are 283 00:17:35,920 --> 00:17:38,680 Speaker 1: now used in lots of different applications that have nothing 284 00:17:38,720 --> 00:17:43,160 Speaker 1: to do with making pretty pictures on a display. Uh. 285 00:17:43,320 --> 00:17:47,200 Speaker 1: They kind of saw that coming down the pipe. Maybe 286 00:17:47,280 --> 00:17:49,760 Speaker 1: not quite in those terms, but they did think that 287 00:17:49,920 --> 00:17:55,280 Speaker 1: was the future of computing. However, to pursue this belief. 288 00:17:55,520 --> 00:17:58,719 Speaker 1: It would mean forming a new company and and leaving 289 00:17:58,760 --> 00:18:02,280 Speaker 1: their current jobs, and then designing the technology from the 290 00:18:02,320 --> 00:18:06,680 Speaker 1: ground up, designing graphics chips from the from the beginning, 291 00:18:07,000 --> 00:18:10,120 Speaker 1: and so it would mean entering into a crowded marketplace 292 00:18:10,200 --> 00:18:13,040 Speaker 1: because the early nineties already had a lot of different 293 00:18:13,080 --> 00:18:17,240 Speaker 1: companies that were creating graphics cards, and there were probably 294 00:18:17,280 --> 00:18:20,760 Speaker 1: around two dozen competitors already on the market. So it's 295 00:18:20,800 --> 00:18:24,320 Speaker 1: becoming a huge headache already because software developers had to 296 00:18:24,320 --> 00:18:27,160 Speaker 1: figure out which formats they were going to support, which 297 00:18:27,200 --> 00:18:30,919 Speaker 1: a p I s should they use to develop their 298 00:18:30,960 --> 00:18:35,119 Speaker 1: software that because the APIs would be compatible with specific 299 00:18:35,200 --> 00:18:40,240 Speaker 1: types of graphics cards. Graphics cards would support programs that 300 00:18:40,280 --> 00:18:43,920 Speaker 1: were developed using certain APIs, and you could get things 301 00:18:43,920 --> 00:18:48,280 Speaker 1: like drivers to help support uh all other API s. 302 00:18:48,359 --> 00:18:52,800 Speaker 1: But if a company is making a graphics card with 303 00:18:52,960 --> 00:18:57,719 Speaker 1: the intent to support a specific API, that kind of 304 00:18:57,720 --> 00:19:00,240 Speaker 1: software is going to perform better on that hardware air 305 00:19:00,440 --> 00:19:04,080 Speaker 1: than other software. So this was becoming a confusing marketplace 306 00:19:04,160 --> 00:19:06,760 Speaker 1: for developers as well as for customers. Let's say that 307 00:19:06,800 --> 00:19:08,720 Speaker 1: you're out there on the market. You you want to 308 00:19:09,040 --> 00:19:13,520 Speaker 1: buy a machine that is really graphically intensive so that 309 00:19:13,600 --> 00:19:16,480 Speaker 1: you can do stuff like maybe maybe you're doing architecture 310 00:19:16,480 --> 00:19:21,560 Speaker 1: design or or computer assistant design CAD programming. You want 311 00:19:21,600 --> 00:19:23,720 Speaker 1: to make sure that the systems you buy are going 312 00:19:23,800 --> 00:19:26,880 Speaker 1: to be compatible with the software that you depend upon. 313 00:19:27,320 --> 00:19:29,720 Speaker 1: And that was starting to get kind of a little 314 00:19:29,720 --> 00:19:31,880 Speaker 1: complicated in the early nineties, and it would only get 315 00:19:31,920 --> 00:19:35,080 Speaker 1: worse before it would get better. But this environment did 316 00:19:35,160 --> 00:19:39,480 Speaker 1: lead to opportunities like the founding of companies like alien Ware. 317 00:19:40,080 --> 00:19:43,960 Speaker 1: You know, alien Ware being the boutique gaming rig company 318 00:19:44,520 --> 00:19:47,000 Speaker 1: that that all started because the founder of alien Ware 319 00:19:47,920 --> 00:19:52,680 Speaker 1: would end up building gaming rigs for friends because he 320 00:19:52,800 --> 00:19:57,400 Speaker 1: was not confused by all the different types of cards 321 00:19:57,440 --> 00:19:59,240 Speaker 1: and motherboards and things like that out there, and so 322 00:19:59,280 --> 00:20:03,240 Speaker 1: he would build gaming rigs for them. That was possible 323 00:20:03,280 --> 00:20:06,360 Speaker 1: because of this massive confusion in the marketplace. Now, one 324 00:20:06,440 --> 00:20:12,159 Speaker 1: year before in Video launched, so back in there was 325 00:20:12,200 --> 00:20:16,199 Speaker 1: another company called Silicon Graphics Incorporated, later known as s 326 00:20:16,240 --> 00:20:20,240 Speaker 1: g I. They will release a multi platform application programming 327 00:20:20,240 --> 00:20:23,080 Speaker 1: interface for two D and three D graphics called open 328 00:20:23,160 --> 00:20:27,560 Speaker 1: g L one. Oh. Now, it was not initially intended 329 00:20:27,920 --> 00:20:31,240 Speaker 1: to be an a p I for developers to develop 330 00:20:31,400 --> 00:20:34,600 Speaker 1: three D games or games that would use three D graphics, 331 00:20:34,760 --> 00:20:37,800 Speaker 1: but it was quickly adapted for that purpose. And meanwhile 332 00:20:38,080 --> 00:20:42,560 Speaker 1: Microsoft was preparing a rival API release called Direct three D. 333 00:20:43,119 --> 00:20:47,960 Speaker 1: And Microsoft didn't concern itself too much to make sure 334 00:20:48,000 --> 00:20:51,320 Speaker 1: that open g L would run smoothly on Windows machines, 335 00:20:51,920 --> 00:20:53,960 Speaker 1: which is the kind of thing that would get Microsoft 336 00:20:53,960 --> 00:20:56,760 Speaker 1: in trouble down the line. They would get accused of 337 00:20:56,800 --> 00:21:00,520 Speaker 1: engaging in anti competitive practices, in other words, making sure 338 00:21:00,560 --> 00:21:05,160 Speaker 1: that their operating systems would support their own software and 339 00:21:05,240 --> 00:21:10,199 Speaker 1: maybe not do so much to support other people's software. 340 00:21:11,640 --> 00:21:14,080 Speaker 1: But it also helps illustrate how confusing this world would 341 00:21:14,080 --> 00:21:18,119 Speaker 1: have become already before the video had even hit the market. 342 00:21:18,520 --> 00:21:21,440 Speaker 1: But while in Video was forming, other graphics card companies 343 00:21:21,480 --> 00:21:25,320 Speaker 1: were already folding or getting absorbed into larger companies. Uh So, 344 00:21:25,520 --> 00:21:27,879 Speaker 1: in Vidia is coming into a market where companies with 345 00:21:28,080 --> 00:21:34,440 Speaker 1: names like Gemini Technology, Headland Technology, Tamarack, Acumos, those were 346 00:21:34,440 --> 00:21:38,000 Speaker 1: starting to go away. Some companies were still around, but 347 00:21:38,040 --> 00:21:41,560 Speaker 1: they shut down divisions that had been working in graphics processing. 348 00:21:41,880 --> 00:21:45,320 Speaker 1: Motorola did that with their division, Acer did that. So 349 00:21:45,440 --> 00:21:48,000 Speaker 1: in Video was getting into a very uncertain market and 350 00:21:48,040 --> 00:21:50,800 Speaker 1: the three co founders didn't have any idea how to 351 00:21:50,840 --> 00:21:53,199 Speaker 1: create a business plan, and they didn't even have a 352 00:21:53,240 --> 00:21:57,840 Speaker 1: product yet, but they did hold rounds of venture capital financing. 353 00:21:57,920 --> 00:22:00,919 Speaker 1: To get the startup capital. They needed to design a 354 00:22:01,000 --> 00:22:04,000 Speaker 1: graphics card and partner with a manufacturer to make it. 355 00:22:04,359 --> 00:22:06,639 Speaker 1: Because they did not have the manufacturing facility, they were 356 00:22:06,680 --> 00:22:08,440 Speaker 1: not actually gonna build the cards. They were going to 357 00:22:08,560 --> 00:22:11,320 Speaker 1: design them and send those designs to a manufacturer that 358 00:22:11,359 --> 00:22:13,320 Speaker 1: would be in charge of actually putting the stuff together. 359 00:22:13,840 --> 00:22:17,200 Speaker 1: Wong has said in interviews that he tried to make 360 00:22:17,320 --> 00:22:20,760 Speaker 1: a business plan to present to potential investors. He really tried, 361 00:22:21,680 --> 00:22:25,040 Speaker 1: but he just wasn't able to do it. And essentially 362 00:22:25,040 --> 00:22:27,320 Speaker 1: he told investors, I tried to make one, but I 363 00:22:27,400 --> 00:22:29,080 Speaker 1: ran out of time before I could finish it. And 364 00:22:29,119 --> 00:22:32,320 Speaker 1: the response was essentially, that's okay. We wouldn't have believed 365 00:22:32,359 --> 00:22:35,840 Speaker 1: your business plan anyway. But the three engineers knew their 366 00:22:35,880 --> 00:22:39,360 Speaker 1: technology backward and forward, and they were able to get 367 00:22:39,520 --> 00:22:42,320 Speaker 1: enough confidence and investors to secure funding to the tune 368 00:22:42,359 --> 00:22:45,800 Speaker 1: of around twenty million dollars. And I would like to 369 00:22:45,800 --> 00:22:48,199 Speaker 1: take this opportunity to say that I also do not 370 00:22:48,320 --> 00:22:53,520 Speaker 1: have a business plan. Money please, I guess it doesn't 371 00:22:53,960 --> 00:22:58,639 Speaker 1: work for me. Well, anyway, this whole process took about 372 00:22:58,640 --> 00:23:02,119 Speaker 1: a year, So from ninety three to ninety four, the 373 00:23:02,160 --> 00:23:04,439 Speaker 1: company is just kind of getting started. They're starting to 374 00:23:04,480 --> 00:23:09,240 Speaker 1: design their work, they're raising finances. They form a strategic 375 00:23:09,280 --> 00:23:13,560 Speaker 1: partnership with a company called s G S Thompson micro Electrics. 376 00:23:14,640 --> 00:23:17,639 Speaker 1: This company would actually be the one to manufacture the 377 00:23:17,680 --> 00:23:22,560 Speaker 1: first graphical user interface accelerator chip that in Vidio designed. 378 00:23:22,920 --> 00:23:27,240 Speaker 1: They also partnered with another company called Diamond Multimedia Systems 379 00:23:27,600 --> 00:23:33,119 Speaker 1: to take these chips and install them in multimedia accelerator boards. 380 00:23:33,160 --> 00:23:35,840 Speaker 1: So if you think of a graphics card like an 381 00:23:35,880 --> 00:23:38,840 Speaker 1: actual expansion card that you would put in a computer, 382 00:23:39,480 --> 00:23:43,879 Speaker 1: Diamond Multimedia Systems built the card and the chip that 383 00:23:44,000 --> 00:23:48,840 Speaker 1: powers the card was made by Nvidia. So producing graphics 384 00:23:48,840 --> 00:23:51,640 Speaker 1: cards required a lot more work than just one company, 385 00:23:51,720 --> 00:23:54,840 Speaker 1: is my my point here. The first product out of 386 00:23:55,040 --> 00:23:59,159 Speaker 1: Nvidia would come to market in n and it was 387 00:23:59,240 --> 00:24:03,919 Speaker 1: the in V one nd PC. Consumers would know it 388 00:24:04,000 --> 00:24:07,600 Speaker 1: as the Diamond Edge three D graphics card, and the 389 00:24:07,760 --> 00:24:10,600 Speaker 1: chip was chosen by Sega to go into the Sega 390 00:24:10,640 --> 00:24:15,119 Speaker 1: Saturn game console. The Diamond Edge three D was a 391 00:24:15,160 --> 00:24:18,960 Speaker 1: conventional pc I card. Pc I stands for Peripheral Component 392 00:24:19,040 --> 00:24:25,000 Speaker 1: interconnect They would later end up being replaced by other formats, 393 00:24:25,040 --> 00:24:29,399 Speaker 1: but this was the standard that we used to plug 394 00:24:29,440 --> 00:24:31,959 Speaker 1: in expansion cards. It was a type of computer bus. 395 00:24:32,359 --> 00:24:35,880 Speaker 1: A computer bus is essentially a channel or a pathway 396 00:24:35,920 --> 00:24:39,520 Speaker 1: that exists between different components and a computer. So PC 397 00:24:39,680 --> 00:24:42,240 Speaker 1: I card slots would allow PC owners to swap out 398 00:24:42,280 --> 00:24:46,080 Speaker 1: expansion cards that would allow you to augment your PCs capabilities. 399 00:24:46,080 --> 00:24:48,240 Speaker 1: It might be a sound card, or a graphics card 400 00:24:48,320 --> 00:24:52,040 Speaker 1: or something along those lines. So how did Nvidio's first 401 00:24:52,080 --> 00:24:55,480 Speaker 1: product do well? According to A A Wong, not so great. 402 00:24:56,040 --> 00:24:58,240 Speaker 1: He said that when they got their finished card back 403 00:24:58,240 --> 00:25:02,080 Speaker 1: from the manufacturers, they on that of what they had 404 00:25:02,160 --> 00:25:05,679 Speaker 1: designed wasn't working in the actual build out. And the 405 00:25:05,720 --> 00:25:08,360 Speaker 1: card was supposed to process both two D and three 406 00:25:08,440 --> 00:25:11,320 Speaker 1: D graphics, plus act as a sound card, plus a 407 00:25:11,400 --> 00:25:14,840 Speaker 1: game pad port for Sega Saturn controllers, plus a PC 408 00:25:15,040 --> 00:25:18,119 Speaker 1: joystick port, and more so it was packing a ton 409 00:25:18,320 --> 00:25:22,240 Speaker 1: of features onto a single card. And it might have 410 00:25:22,280 --> 00:25:25,560 Speaker 1: been a bit too ambitious. Uh. It nearly meant the 411 00:25:25,560 --> 00:25:29,720 Speaker 1: invidios story could end before it could really begin. But 412 00:25:29,800 --> 00:25:33,560 Speaker 1: as we'll find out, and video was able to weather 413 00:25:33,640 --> 00:25:36,080 Speaker 1: that storm. We'll come back in just a moment, but 414 00:25:36,160 --> 00:25:46,240 Speaker 1: let's take a quick break to thank our sponsors. Wong 415 00:25:46,400 --> 00:25:50,040 Speaker 1: has talked about a demo his company held that would 416 00:25:50,080 --> 00:25:53,480 Speaker 1: show off this en V one ship, and they showed 417 00:25:53,480 --> 00:25:56,560 Speaker 1: it off at a trade show called Compute Text. He said, 418 00:25:56,600 --> 00:25:59,119 Speaker 1: even though many of the features were not working in 419 00:25:59,160 --> 00:26:02,639 Speaker 1: this demo, it still made a really good impression. But 420 00:26:03,160 --> 00:26:06,680 Speaker 1: then in Video came back the following year to Compute 421 00:26:06,680 --> 00:26:10,600 Speaker 1: Text and they showed off the exact same demo, and 422 00:26:10,640 --> 00:26:13,680 Speaker 1: then he saw that the reactions were not as positive. 423 00:26:13,960 --> 00:26:16,080 Speaker 1: And that's where he learned that if you show up 424 00:26:16,119 --> 00:26:19,560 Speaker 1: to the same conference twice, you know, in two different years, 425 00:26:19,600 --> 00:26:22,480 Speaker 1: and you just show the same demo, that's a big mistake. 426 00:26:22,600 --> 00:26:25,080 Speaker 1: It sends a bad message. And he said we almost 427 00:26:25,160 --> 00:26:29,080 Speaker 1: went out of business. In a piece published in I E. 428 00:26:29,080 --> 00:26:32,639 Speaker 1: E E. Spectrum that has the title in Video co 429 00:26:32,800 --> 00:26:37,159 Speaker 1: founders remember their startup Roller Coaster Ride, Huang said quote, 430 00:26:37,680 --> 00:26:40,879 Speaker 1: we had this idea that computer graphics was going to 431 00:26:40,920 --> 00:26:44,199 Speaker 1: be the driving force of technology and fuel would be 432 00:26:44,359 --> 00:26:47,800 Speaker 1: video games. A year and a half later, we nearly 433 00:26:47,840 --> 00:26:51,320 Speaker 1: went out of business. The technology was wrong, the market 434 00:26:51,359 --> 00:26:56,359 Speaker 1: strategy was wrong, architecture was wrong, and the execution wasn't 435 00:26:56,560 --> 00:26:59,560 Speaker 1: good end quote. So what do you mean the architecture 436 00:26:59,640 --> 00:27:02,720 Speaker 1: was wrong. Well, for one thing, the way the n 437 00:27:02,800 --> 00:27:06,720 Speaker 1: V one would render graphics was through a process called 438 00:27:06,840 --> 00:27:11,560 Speaker 1: quad texture mapping, But just as the cards were starting 439 00:27:11,600 --> 00:27:17,800 Speaker 1: to ship out of manufacturing facilities, Microsoft was finalizing Direct 440 00:27:18,040 --> 00:27:22,320 Speaker 1: X one point Oh, that's what incorporated Direct three D 441 00:27:23,040 --> 00:27:27,320 Speaker 1: and Direct three D did not use quad texture mapping. 442 00:27:27,640 --> 00:27:33,760 Speaker 1: It used triangular polygons to map textures. And Microsoft's dominance 443 00:27:34,000 --> 00:27:37,960 Speaker 1: in the operating systems space for PCs meant the game 444 00:27:37,960 --> 00:27:42,320 Speaker 1: developers were more likely to build games tailored to Microsoft's 445 00:27:42,320 --> 00:27:47,400 Speaker 1: approach because most of their customers were Microsoft customers. They 446 00:27:47,400 --> 00:27:50,200 Speaker 1: weren't gonna go the quad texture mapping strategy that in 447 00:27:50,320 --> 00:27:53,439 Speaker 1: Video had used, so in Video tried to compensate for this. 448 00:27:53,520 --> 00:27:57,920 Speaker 1: They released a driver that would uh if you installed 449 00:27:57,920 --> 00:28:01,159 Speaker 1: the driver if you had the Envy one chip in 450 00:28:01,200 --> 00:28:05,239 Speaker 1: your computer, it would try to wrap triangles as if 451 00:28:05,280 --> 00:28:09,480 Speaker 1: they were quadratic surfaces to give the Envy one a 452 00:28:09,680 --> 00:28:13,120 Speaker 1: semi compatible platform with Direct three D. But the NV 453 00:28:13,240 --> 00:28:17,000 Speaker 1: one failed to attract many developers apart from those early 454 00:28:17,160 --> 00:28:22,600 Speaker 1: Sega ports. Sega obviously had tapped them for the Saturn, 455 00:28:22,760 --> 00:28:26,600 Speaker 1: so ports from Sega Saturn to the PC. If you 456 00:28:26,680 --> 00:28:30,000 Speaker 1: had a in Video V one, they ran pretty good, 457 00:28:30,600 --> 00:28:35,160 Speaker 1: really good actually, And I think it's interesting that Wong 458 00:28:35,320 --> 00:28:40,160 Speaker 1: was very candid about this, about how this decision was 459 00:28:40,720 --> 00:28:44,680 Speaker 1: not a good one early on in the company's history. 460 00:28:44,720 --> 00:28:48,520 Speaker 1: But Wong and Malachowski have both credited in Videos success 461 00:28:48,560 --> 00:28:51,480 Speaker 1: at least in part to an acceptance that sometimes things 462 00:28:51,560 --> 00:28:53,440 Speaker 1: just don't go the way you planned and you have 463 00:28:53,520 --> 00:28:58,040 Speaker 1: to pivot and make changes. They say their work was 464 00:28:58,080 --> 00:29:01,160 Speaker 1: adaptable and that's what allowed them to survive. They could 465 00:29:01,240 --> 00:29:06,200 Speaker 1: adapt their thinking and their strategies and their technologies to 466 00:29:07,480 --> 00:29:09,480 Speaker 1: make the best of a bad situation. They could learn 467 00:29:09,520 --> 00:29:12,520 Speaker 1: from their mistakes and work to avoid them in the future, well, 468 00:29:12,560 --> 00:29:16,640 Speaker 1: you know, making brand new mistakes. Despite the issues with 469 00:29:16,680 --> 00:29:19,960 Speaker 1: the NV one, and Video was able to land a 470 00:29:20,080 --> 00:29:23,480 Speaker 1: deal with Sega, and ports of Sega games to the 471 00:29:23,520 --> 00:29:26,400 Speaker 1: PC depended upon that in V one architecture, and n 472 00:29:26,480 --> 00:29:30,000 Speaker 1: V one had the honor of becoming the first commercial 473 00:29:30,360 --> 00:29:35,240 Speaker 1: graphics processor that could do three D rendering, video acceleration 474 00:29:35,560 --> 00:29:41,280 Speaker 1: and graphics user interface acceleration. However, the sales of the 475 00:29:41,360 --> 00:29:44,320 Speaker 1: NV one were a disappointment, but they did help keep 476 00:29:44,320 --> 00:29:47,360 Speaker 1: the company above water, so they began work on its 477 00:29:47,360 --> 00:29:52,880 Speaker 1: successor the design of the n V two. But then 478 00:29:53,000 --> 00:29:57,880 Speaker 1: the company got some bad news. Sega had decided they 479 00:29:57,880 --> 00:30:01,560 Speaker 1: were not going to go within video or uh the 480 00:30:01,760 --> 00:30:05,440 Speaker 1: graphics chip for their next video game console, which was 481 00:30:05,480 --> 00:30:08,040 Speaker 1: the Dreamcast. Instead, they were going to go with power 482 00:30:08,200 --> 00:30:11,800 Speaker 1: VR for that. So in Video had to figure out 483 00:30:12,040 --> 00:30:15,840 Speaker 1: what they were going to do next. And speaking of that, 484 00:30:16,520 --> 00:30:20,760 Speaker 1: in our next episode, we're going to look at how 485 00:30:20,840 --> 00:30:24,800 Speaker 1: in Video came from this shaky start and we're able 486 00:30:24,840 --> 00:30:30,320 Speaker 1: to cement themselves as industry leaders in the graphics card arena, 487 00:30:30,400 --> 00:30:34,800 Speaker 1: how they were able to invent the term graphics processing unit, 488 00:30:34,880 --> 00:30:37,880 Speaker 1: what that actually means, what is a graphics processing unit, 489 00:30:37,920 --> 00:30:42,960 Speaker 1: what made their first graphics processing unit a GPU, And 490 00:30:43,080 --> 00:30:46,120 Speaker 1: we'll also talk about some of the problems the company 491 00:30:46,240 --> 00:30:50,600 Speaker 1: has encountered since then. I hope you enjoy this episode 492 00:30:50,600 --> 00:30:52,440 Speaker 1: in the next one, and if you guys have any 493 00:30:52,440 --> 00:30:55,360 Speaker 1: suggestions for future episodes of tech Stuff. Maybe it's a technology, 494 00:30:55,440 --> 00:30:57,640 Speaker 1: maybe it's a company, Maybe it's a person in tech. 495 00:30:58,320 --> 00:31:01,120 Speaker 1: Maybe there's someone I should interview. Let me know. Send 496 00:31:01,160 --> 00:31:04,160 Speaker 1: me an email the addresses tech Stuff at how stuff 497 00:31:04,160 --> 00:31:07,000 Speaker 1: works dot com, or draw me a line on Facebook 498 00:31:07,080 --> 00:31:08,840 Speaker 1: or Twitter. The handle at both of those is tech 499 00:31:08,880 --> 00:31:12,600 Speaker 1: Stuff hs W. You can check out our merchandise store 500 00:31:12,640 --> 00:31:15,680 Speaker 1: over at t public dot com slash tech stuff. I'm 501 00:31:15,720 --> 00:31:18,240 Speaker 1: wearing one of those shirts as I sit here recording 502 00:31:18,840 --> 00:31:22,200 Speaker 1: right now. I'm wearing my Princely sum T shirt. It's 503 00:31:22,320 --> 00:31:25,720 Speaker 1: so comfy and I look so snazzy in it. I 504 00:31:25,760 --> 00:31:28,160 Speaker 1: don't even mind that I spent a princely sum on it. 505 00:31:28,600 --> 00:31:31,240 Speaker 1: You shouldn't either, t public dot com slash tech stuff, 506 00:31:31,680 --> 00:31:35,280 Speaker 1: and make sure you follow us over on Instagram and 507 00:31:35,280 --> 00:31:43,800 Speaker 1: I'll talk to you again about Nvidia really soon for 508 00:31:43,880 --> 00:31:46,200 Speaker 1: more on this and thousands of other topics, because it 509 00:31:46,280 --> 00:31:57,320 Speaker 1: how stuff works dot com