WEBVTT - AMD Up To Today 0:00:04.240 --> 0:00:07.240 Welcome to tech Stuff, a production of I Heart Radios, 0:00:07.320 --> 0:00:13.800 How Stuff Works. Hey there, and welcome to tech Stuff. 0:00:13.840 --> 0:00:16.800 I'm your host, Jonathan Strickland. I'm an executive producer with 0:00:16.840 --> 0:00:19.000 How Stuff Works and I heart Radio and I love 0:00:19.160 --> 0:00:22.680 all things tech. And in our last episode, which was 0:00:22.880 --> 0:00:26.079 requested by listener Stephen, I left off with a m 0:00:26.200 --> 0:00:30.280 d's history in ninety six, when the company found itself 0:00:30.440 --> 0:00:35.200 reeling when Intel decided to cut ties. Up to that point, 0:00:35.520 --> 0:00:38.120 a m D had been in an agreement to act 0:00:38.200 --> 0:00:42.480 as a second source for Intel designed chips. Intel would 0:00:42.520 --> 0:00:44.720 get a licensing fee and a m D would be 0:00:44.800 --> 0:00:49.120 able to manufacture chips based on Intel's designs. But when 0:00:49.159 --> 0:00:52.040 a m D chips started to do better in performance 0:00:52.120 --> 0:00:57.720 tests than the Intel originals, things changed. Until ended a 0:00:57.760 --> 0:01:01.760 ten year agreement several years early, and the two companies 0:01:01.800 --> 0:01:05.200 would enter into a lengthy court battle that would ultimately 0:01:05.280 --> 0:01:07.600 go all the way to the U. S. Supreme Court. 0:01:07.959 --> 0:01:10.319 But we've got a few years to get through before 0:01:10.360 --> 0:01:14.080 we get to that. So Intel had introduced the eighty 0:01:14.240 --> 0:01:18.640 three eighty six microprocessor in nineteen eighty five, a year 0:01:18.680 --> 0:01:21.399 before it severed the agreement with a m D. The 0:01:21.480 --> 0:01:24.920 three eight six, as it was known, was a thirty 0:01:24.959 --> 0:01:28.920 two bit micro processor. It could run most older code 0:01:29.040 --> 0:01:32.759 designed for its predecessors, and was capable of faster clock rates, 0:01:33.040 --> 0:01:36.440 meaning it could run more operations per second, and it 0:01:36.520 --> 0:01:39.560 had a greater data bandwidth, which means it could run 0:01:39.600 --> 0:01:44.120 operations on larger amounts of information than the earlier chips could. 0:01:44.600 --> 0:01:46.600 Intel an a m D had set the stage for 0:01:46.680 --> 0:01:50.080 creating the standard in computer processing, and now Intel was 0:01:50.160 --> 0:01:53.480 determined to stand alone. A m D, for its part, 0:01:53.720 --> 0:01:56.680 have been designing its own version of the three eight six, 0:01:57.040 --> 0:02:01.040 called the get Ready for It a M three eight six, 0:02:01.360 --> 0:02:05.080 but Intel's decision to end the agreement through things into disarray. 0:02:05.480 --> 0:02:08.520 Intel argued that their agreement with a m D only 0:02:08.639 --> 0:02:11.920 covered the eight eighty six through the e D two 0:02:11.960 --> 0:02:15.720 eighty six family of micro processors, and that the three 0:02:15.760 --> 0:02:19.720 eight six and later iterations were excluded. A and D 0:02:19.880 --> 0:02:24.760 obviously disagreed with this interpretation of their agreement, claiming that 0:02:24.840 --> 0:02:28.480 all x eighty six derivatives were covered under this ten 0:02:28.560 --> 0:02:31.160 year plan that they had struck with Intel back in 0:02:31.240 --> 0:02:35.360 nineteen eighty two. At this same time, things were shifting 0:02:35.600 --> 0:02:38.600 in the PC market. You might remember from my episode 0:02:38.800 --> 0:02:41.639 about early computer systems that there used to be a 0:02:41.680 --> 0:02:44.920 ton of different types of PCs on the market in 0:02:44.960 --> 0:02:48.080 the late seventies and early eighties, each with its own 0:02:48.120 --> 0:02:51.400 hardware and operating system. The ones we think about today 0:02:51.720 --> 0:02:55.560 are Windows based machines and Mac computers as far as 0:02:55.600 --> 0:02:59.239 personal machines are concerned. But up through the early nineteen eighties, 0:02:59.280 --> 0:03:03.480 the field was much more crowded. You had companies like Tandy, Commodore, 0:03:03.520 --> 0:03:07.880 and Amiga and others competing with Apple and IBM. However, 0:03:08.000 --> 0:03:12.440 by the mid nineteen eighties the field had thinned out significantly. 0:03:13.000 --> 0:03:17.520 IBM had secured valuable deals with corporations, becoming known as 0:03:17.560 --> 0:03:21.840 the computer of choice for office workstations. Apple maintained a 0:03:21.880 --> 0:03:25.320 more niche market of users interested in the creative power 0:03:25.400 --> 0:03:29.840 of the Macintosh. Everyone else sort of began to fade away, 0:03:29.880 --> 0:03:33.560 and this left the IBM PC and it's compatible clones 0:03:33.680 --> 0:03:36.960 with the lion's share of the market. In fact, by 0:03:37.000 --> 0:03:39.600 the time Intel was trying to block a m D, 0:03:40.200 --> 0:03:43.400 the IBM PC market share had grown to about eighty 0:03:43.600 --> 0:03:47.880 four percent of all personal computers, so this was a 0:03:48.000 --> 0:03:51.640 really big deal. A m D had helped cement the 0:03:51.840 --> 0:03:55.839 X eight six chip as the go to microprocessor for computers, 0:03:56.280 --> 0:03:58.480 and now it looks like Intel was going to run 0:03:58.480 --> 0:04:02.200 away with the whole thing. Things were starting to smell 0:04:02.520 --> 0:04:06.360 a little anti competitive. Am D did still have an 0:04:06.360 --> 0:04:09.960 agreement to the underlying instruction set for the X eight 0:04:10.080 --> 0:04:14.000 six family of processors, so in some respects a m 0:04:14.080 --> 0:04:16.840 D was still in the game, but Intel wasn't going 0:04:16.880 --> 0:04:20.320 to share the actual physical design of the three six 0:04:20.360 --> 0:04:24.040 microprocessor with a m D. So the engineers at a 0:04:24.160 --> 0:04:27.520 m D set out to reverse engineer the three six 0:04:27.720 --> 0:04:30.320 and build their own version of it while the legal 0:04:30.360 --> 0:04:34.760 battle continued in the courts. Reverse engineering alone is a 0:04:34.800 --> 0:04:39.880 pretty fascinating subject. The basic concept is fairly intuitive. You 0:04:39.960 --> 0:04:43.520 take a technology and you examine it closely and you 0:04:43.560 --> 0:04:46.200 figure out what makes it tick, how does the tech 0:04:46.560 --> 0:04:49.960 actually do whatever it does. Then you go back and 0:04:50.000 --> 0:04:53.440 you build your own version of that technology based upon 0:04:53.520 --> 0:04:58.080 your understanding of how the starting example worked. So you're 0:04:58.080 --> 0:05:00.480 not starting off with some sort of blueprint or set 0:05:00.520 --> 0:05:03.560 of plans or instructions. You're sussing it out on your 0:05:03.560 --> 0:05:07.280 own based on existing instances of the technology. So it's 0:05:07.279 --> 0:05:11.080 a bit like detective work. Between reverse engineering and the 0:05:11.200 --> 0:05:13.680 legal battles. It would be years before a m D 0:05:13.800 --> 0:05:16.480 could bring its own three eight six chips to market. 0:05:16.920 --> 0:05:21.839 The company began releasing its version starting in nineteen, and 0:05:22.080 --> 0:05:25.359 once again a m D S version of Intel's chips 0:05:25.480 --> 0:05:29.400 were clocking in at a faster clock rate than the competition. 0:05:29.920 --> 0:05:32.919 Intel's three eight six chips maxed out at thirty three 0:05:32.920 --> 0:05:36.159 mega hurts, whereas a m D S could hit forty 0:05:36.440 --> 0:05:40.000 mega hurts. The legal battles continued, and a m D 0:05:40.160 --> 0:05:43.800 began to invest in designing its own microcode for chips. 0:05:44.279 --> 0:05:49.599 Intel's next microprocessor was predictably the eight four eight six, 0:05:49.920 --> 0:05:52.400 and a m D created its own version the A 0:05:52.640 --> 0:05:56.200 M four eight six. Some A M for eight six 0:05:56.279 --> 0:05:59.400 chips had Intel microcode from the x A D six 0:05:59.440 --> 0:06:03.279 agreement and others had a m D microcode, making it 0:06:03.320 --> 0:06:07.880 a little confusing, and all of them were outpacing Intel's 0:06:08.080 --> 0:06:11.080 version of the same chip. Even the top of the 0:06:11.200 --> 0:06:16.000 line microprocessor in Intel's forty six line was left behind. 0:06:16.400 --> 0:06:19.040 The fastest four D six from Intel had the top 0:06:19.080 --> 0:06:22.000 clock speed of one hundred mega hurts. A m D 0:06:22.200 --> 0:06:24.880 S version was able to reach speeds of one D 0:06:25.000 --> 0:06:30.200 twenty mega hurts. Now this was in When that legal 0:06:30.240 --> 0:06:33.960 battle I talked about finally concluded, the courts found in 0:06:34.080 --> 0:06:37.160 favor of A m D, granting the company some royalty 0:06:37.200 --> 0:06:40.480 free use of some of Intel's patents and awarding A 0:06:40.600 --> 0:06:43.760 m D millions of dollars in the process. But the 0:06:43.800 --> 0:06:46.400 whole endeavor had taught the engineers at A m D 0:06:46.600 --> 0:06:49.960 a valuable lesson. While they had won this battle, there 0:06:50.040 --> 0:06:53.120 was no guarantee that things would remain stable between Intel 0:06:53.160 --> 0:06:56.440 and a m D, so the company did release another 0:06:56.600 --> 0:06:59.000 x E D six derived chip. This one was called 0:06:59.040 --> 0:07:02.440 the A m D five x eight six, or five 0:07:02.560 --> 0:07:04.960 by eighty six if you wanted to think of it 0:07:05.000 --> 0:07:07.640 that way. And you may be thinking, ha, I never 0:07:07.680 --> 0:07:10.520 heard of a five eight six computer. I'm pretty sure 0:07:10.560 --> 0:07:14.120 that Intel switched over from four A D six to pentium, 0:07:14.160 --> 0:07:16.440 and you would be right. The A m D five 0:07:16.800 --> 0:07:20.160 x eight six chip was based off the same architecture 0:07:20.440 --> 0:07:23.920 as the A M four eight six microprocessor, but it 0:07:24.000 --> 0:07:27.400 did manage an even faster clock speed out of the box. 0:07:27.440 --> 0:07:30.040 It was a hundred thirty Mega hurts, but original equipment 0:07:30.080 --> 0:07:33.640 manufacturers or o e m s in the biz could 0:07:33.680 --> 0:07:35.800 get an even faster version than maxed out at a 0:07:35.880 --> 0:07:40.600 hundred fifty Mega hurts now according to Tom's Hardware, a website, 0:07:40.640 --> 0:07:43.160 which is, by the way, a great resource if you 0:07:43.160 --> 0:07:45.280 ever want to learn everything there is to know about 0:07:45.360 --> 0:07:48.640 just about any computer component you can think of. The 0:07:48.680 --> 0:07:50.960 A M four, A D six, and the five X 0:07:51.000 --> 0:07:55.480 eighty six processors also moved the floating point unit or 0:07:55.600 --> 0:07:59.760 FPU over to the central processing unit or CPU itself. 0:08:00.240 --> 0:08:02.720 Up until then, it had been customary to have separate 0:08:02.760 --> 0:08:06.240 CPUs and FPUs that would connect to each other through 0:08:06.240 --> 0:08:09.160 the motherboard. So I guess it's time to give a 0:08:09.240 --> 0:08:13.480 quick explanation about what these things actually mean. The CPU, 0:08:13.680 --> 0:08:15.920 I'm sure you've all heard of, right, It's sort of 0:08:15.960 --> 0:08:20.040 the head manager of your computer. It executes basic instructions. 0:08:20.520 --> 0:08:22.720 In the event that the instructions require the use of 0:08:22.760 --> 0:08:26.920 a specialized chip like a graphics processing unit also known 0:08:26.920 --> 0:08:30.120 as a GPU, the CPU can delegate those tasks to 0:08:30.200 --> 0:08:34.720 the appropriate hardware. It's a high functioning component of a computer. 0:08:34.960 --> 0:08:37.280 We often refer to it as the brains of the computer, 0:08:37.320 --> 0:08:40.600 but really it's just calling the shots at the highest level. 0:08:41.040 --> 0:08:44.160 The floating point unit carries out instructions on what are 0:08:44.160 --> 0:08:47.880 called floating point numbers. A floating point number is a 0:08:47.920 --> 0:08:51.040 workaround for a particular problem, which is how a computer 0:08:51.120 --> 0:08:55.360 represents real numbers. The range of real numbers is infinite, 0:08:55.800 --> 0:08:58.320 but a computer can't handle that. A computer has a 0:08:58.360 --> 0:09:02.640 limited capacity, so programmers use floating point numbers, so called 0:09:02.640 --> 0:09:06.000 because the decimal point has no fixed number of digits 0:09:06.040 --> 0:09:09.720 that have to appear before or after it. This allows 0:09:09.760 --> 0:09:13.600 programmers to represent numbers separated by many orders of magnitude. 0:09:13.600 --> 0:09:16.960 You can have incredibly large numbers paired with incredibly small 0:09:17.040 --> 0:09:21.600 numbers using this approach. UH typically you would use a 0:09:21.679 --> 0:09:25.319 variant of scientific notation for those really big or really 0:09:25.360 --> 0:09:28.440 small numbers. However, this does mean that much of the 0:09:28.440 --> 0:09:34.080 work computers do happens as approximations rather than as precise calculations, 0:09:34.720 --> 0:09:38.559 and this introduces the possibility of error. The more you 0:09:38.760 --> 0:09:42.880 are approximating something, the less accurate or precise it's going 0:09:42.920 --> 0:09:46.440 to be, particularly as you perform more calculations based on 0:09:46.559 --> 0:09:51.360 previous approximations. As these approximations start to add up, you 0:09:51.440 --> 0:09:55.160 can potentially get further and further away from a correct 0:09:55.280 --> 0:09:58.520 or true answer. But never mind that that's a discussion 0:09:58.520 --> 0:10:01.439 for a different episode. Now, after the four a D six, 0:10:01.800 --> 0:10:05.600 Intel came out with the first Pentium processor. So why 0:10:05.600 --> 0:10:08.439 did Intel change things up? Why did Intel go from 0:10:08.440 --> 0:10:11.679 four D six to pentium? Because the Pentium still followed 0:10:11.720 --> 0:10:14.720 the x A D six architecture and instruction set and 0:10:14.760 --> 0:10:19.280 spoiler alert, so do today's computers. So why would Intel 0:10:19.400 --> 0:10:22.760 choose pentium instead of sticking with the naming convention it 0:10:22.800 --> 0:10:25.880 had created. Why wasn't it the five eight six? Well, 0:10:25.920 --> 0:10:28.520 the main reason was, as I'm sure many of you 0:10:28.720 --> 0:10:33.040 have guessed that companies like A m D were the 0:10:33.080 --> 0:10:35.760 cause of this. Intel decided because of a m D, 0:10:36.240 --> 0:10:40.640 Intel couldn't trademark a number. Intel couldn't have five eighty 0:10:40.720 --> 0:10:44.080 six trademarked. You can't just trademarket basic number like that. 0:10:44.200 --> 0:10:47.760 So if it had stuck with the numbering system, a 0:10:47.920 --> 0:10:50.000 m D could then come out with its A m 0:10:50.120 --> 0:10:53.040 D five a D six and with its reputation for 0:10:53.080 --> 0:10:59.520 outpacing Intel's comparable chips, that could hurt Intel's sales. But Pentium, 0:11:00.080 --> 0:11:02.439 that was different because Pentium was a name. You can 0:11:02.480 --> 0:11:06.240 trademark a name, and that's what Intel did. It trademarked 0:11:06.280 --> 0:11:10.000 the term pentium, which prevented a m D and other 0:11:10.040 --> 0:11:14.280 competitors from using that name on their own chips. So 0:11:14.480 --> 0:11:19.120 now it added a marketing concern for these competitors. How 0:11:19.160 --> 0:11:22.600 would they be able to market their own chips and 0:11:22.720 --> 0:11:28.120 compare them against Intel's chips without using a trademarked name 0:11:28.200 --> 0:11:30.199 that they did not have the rights to. It was 0:11:30.280 --> 0:11:33.160 kind of throwing a monkey wrench into things. Now. The 0:11:33.200 --> 0:11:36.480 way companies got around this was to include a number 0:11:36.640 --> 0:11:40.640 that they referred to as PR, which essentially stood for 0:11:40.800 --> 0:11:45.080 pentium rating. The number next to the PR designation would 0:11:45.120 --> 0:11:48.800 indicate the comparable pentium clock speed that the chip in 0:11:48.920 --> 0:11:53.800 question would be most like. So if you came out 0:11:53.840 --> 0:11:56.600 with a microchip and you gave it a PR rating 0:11:56.679 --> 0:12:00.520 of one hundred, what that tells the end consumer is 0:12:00.559 --> 0:12:03.440 that the chip you have put out is equivalent to 0:12:03.520 --> 0:12:06.360 an Intel pentium processor that has a clock speed of 0:12:06.440 --> 0:12:08.960 one mega hurts. So it's kind of a way of 0:12:09.000 --> 0:12:12.400 getting around the fact that they could not call their 0:12:12.400 --> 0:12:16.640 own chips their variance of the Pendium processor. Now it 0:12:16.720 --> 0:12:18.840 was clear that Intel was going to put up a 0:12:18.880 --> 0:12:21.880 fight and resist as much as it could. It would 0:12:21.920 --> 0:12:24.400 make little sense for a m D to depend solely 0:12:24.480 --> 0:12:28.560 upon being a second source for Intel. Chips, particularly when 0:12:28.600 --> 0:12:32.640 Intel wasn't really interested in cooperating fully, and so A 0:12:32.800 --> 0:12:35.800 m D began work on designing its own x eight 0:12:35.960 --> 0:12:40.920 six based microprocessor, which would be released in n and 0:12:40.960 --> 0:12:43.840 it became known as the A m D K five. 0:12:44.559 --> 0:12:47.760 Well why was it called the K five? Well, by 0:12:47.840 --> 0:12:50.920 A m D s reckoning, it represented the fifth generation 0:12:51.000 --> 0:12:54.080 microprocessor family that A m D had built, the other 0:12:54.120 --> 0:12:57.640 four being second source Intel chips, but the K five 0:12:58.080 --> 0:13:01.040 was a totally new architecture that was based on the 0:13:01.200 --> 0:13:07.240 x A D six instruction set. So why the K Well, 0:13:07.280 --> 0:13:12.199 because K is also the letter that starts the word kryptonite, 0:13:12.600 --> 0:13:16.280 the substance that could bring down Superman. And I think 0:13:16.280 --> 0:13:21.400 we can all guess who was Superman in this particular scenario. 0:13:22.360 --> 0:13:25.320 A m D designed the K five entirely in house, 0:13:25.640 --> 0:13:27.720 and it was the first x A D six processor 0:13:27.800 --> 0:13:30.000 from A m D to have architecture designed by the 0:13:30.040 --> 0:13:33.320 A m D team itself, as opposed to either following 0:13:33.320 --> 0:13:38.040 Intel's detailed instructions to make a clone of their chips 0:13:38.160 --> 0:13:42.320 or through reverse engineering and existing Intel microchip. The K 0:13:42.480 --> 0:13:46.000 five copied some elements from the earlier A M twenty 0:13:46.080 --> 0:13:50.280 nine thousand microprocessor that was a risk or R I 0:13:50.640 --> 0:13:53.720 s C microchip the company made a few years earlier. 0:13:53.760 --> 0:13:57.360 I talked about that in the previous episode, and I 0:13:57.400 --> 0:14:00.640 think that was a pretty good choice. It gave them 0:14:00.679 --> 0:14:04.080 a starting point to work from, and they were able 0:14:04.080 --> 0:14:06.439 to really build on that and make a success out 0:14:06.480 --> 0:14:11.080 of it. The K five's design was a little bit complicated, 0:14:11.120 --> 0:14:13.640 and that placed limits on how much clock speed a 0:14:13.800 --> 0:14:15.280 m D could get out of it. But at the 0:14:15.320 --> 0:14:18.240 same time, the A m D engineers had made the 0:14:18.240 --> 0:14:22.240 operations really efficient, so while it might have a technically 0:14:22.360 --> 0:14:28.040 lower clock speed than a competing microprocessor, this increased efficiency 0:14:28.080 --> 0:14:31.200 helped balance things out so that at the end result 0:14:32.040 --> 0:14:34.520 it seemed like the K five was actually faster than 0:14:34.560 --> 0:14:38.280 its counterparts that technically had higher clock speeds. Yes, the 0:14:38.320 --> 0:14:42.720 other microprocessors could run more operations per second, but K 0:14:42.880 --> 0:14:45.280 five's efficiency was such that was able to make up 0:14:45.280 --> 0:14:48.080 for that lost ground. Now, when we come back, i'll 0:14:48.120 --> 0:14:50.720 talk more about a m d s experiences in the 0:14:50.800 --> 0:14:55.040 nineteen nineties and beyond, but first let's take a quick break. 0:15:02.440 --> 0:15:04.520 A m D would follow up the K five with 0:15:04.560 --> 0:15:08.960 a microprocessor called the now wait for it, the Case six, 0:15:09.360 --> 0:15:12.680 But the K six wasn't designed by a m D engineers, 0:15:13.000 --> 0:15:16.520 nor did it follow the K five architecture. Instead, A 0:15:16.680 --> 0:15:20.880 m D acquired another microchip manufacturing company called next Gen 0:15:21.160 --> 0:15:24.400 an e x G E N. Next Gen was getting 0:15:24.440 --> 0:15:27.560 ready to release a CPU it called the n X 0:15:27.640 --> 0:15:30.680 six eight six, but then A m D swooped in, 0:15:31.000 --> 0:15:34.200 bought up next Gen, and then repurpose the as yet 0:15:34.320 --> 0:15:38.200 unreleased n X six eight six to become the K six. 0:15:38.560 --> 0:15:41.680 A m D marketed it as an alternative for Intel's 0:15:41.680 --> 0:15:45.440 Pentium two processor, claiming that for less money you could 0:15:45.440 --> 0:15:49.280 get the same level of performance, and that was mostly true, 0:15:49.480 --> 0:15:52.920 though the Pentium two had some advantages over the Case six, 0:15:53.240 --> 0:15:57.880 namely a better math coprocessor or FPU. At this point, 0:15:58.240 --> 0:16:00.840 the K six and the variance I'll talk about in 0:16:00.880 --> 0:16:05.560 a second we're still compatible with Intel designed motherboards. The 0:16:05.640 --> 0:16:09.000 Case six was also cheaper than the pent Um two chips, 0:16:09.280 --> 0:16:11.640 and so the Case six became a popular choice for 0:16:11.720 --> 0:16:14.360 both O E M s and people building their own machines. 0:16:14.920 --> 0:16:17.080 A m D would follow up the Case six with 0:16:17.160 --> 0:16:21.840 the Case six two and the Case six three in nine. 0:16:23.240 --> 0:16:26.320 The Case six three paired two hundred fifty six L 0:16:26.400 --> 0:16:29.480 two cash memory on the CPU die in an effort 0:16:29.480 --> 0:16:32.800 to speed up processing and increasing the amount of data 0:16:32.920 --> 0:16:36.000 the CPU could access at any given time. The Case 0:16:36.000 --> 0:16:40.160 six too, was phenomenally successful, so much so that some 0:16:40.240 --> 0:16:44.720 analysts estimated that seventy percent of the under one thousand 0:16:44.720 --> 0:16:48.800 dollar PC market in nine had a m D K 0:16:49.000 --> 0:16:52.520 six two chips powering them. So, if you were building 0:16:52.840 --> 0:16:54.680 a computer on a budget and you wanted to get 0:16:54.720 --> 0:16:58.360 the most umph for your dollars, chances are you were 0:16:58.360 --> 0:17:01.120 going with a m D. The company would also release 0:17:01.200 --> 0:17:04.440 the Case six two plus in the Case six three 0:17:04.480 --> 0:17:08.320 plus in two thousands. These were microprocessors meant specifically for 0:17:08.359 --> 0:17:12.119 the mobile market, and they'd be the final entries in 0:17:12.160 --> 0:17:17.200 the Case six line of CPUs. Meanwhile, Jerry Sanders, whom 0:17:17.240 --> 0:17:18.960 you might remember from the first episode, he was the 0:17:19.000 --> 0:17:21.359 first president of a m D. He was a co 0:17:21.520 --> 0:17:25.480 founder and at this point was the CEO of the company, 0:17:25.760 --> 0:17:29.760 was riding high. He predicted astronomical share prices for the 0:17:29.760 --> 0:17:33.320 company in the near future. He continued the company's practice 0:17:33.640 --> 0:17:37.560 of building fabrication plants at a breakneck pace. He was 0:17:37.600 --> 0:17:42.360 building plants to manufacture microprocessors and semiconductor chips all over 0:17:42.359 --> 0:17:46.040 the world. A m D had been incredibly aggressive in 0:17:46.119 --> 0:17:49.480 building and staffing these fabrication facilities in order to meet 0:17:49.520 --> 0:17:54.000 the demand for microprocessors and actually to anticipate the next 0:17:54.160 --> 0:17:58.760 demand for them, and Sanders had adopted a reportedly lavish lifestyle, 0:17:58.880 --> 0:18:02.360 maintaining an off us in Beverly Hills, which is pretty 0:18:02.440 --> 0:18:06.080 darn far from Silicon Valley and the headquarters of a 0:18:06.200 --> 0:18:08.359 m D. I guess he never really gave up his 0:18:08.480 --> 0:18:12.240 dream of going into the recording industry, but a lavish 0:18:12.280 --> 0:18:15.400 lifestyle might be fine if things continued to go well 0:18:15.440 --> 0:18:18.320 for the company, and sadly that would not be the case. 0:18:18.760 --> 0:18:22.199 Sanders spending also seemed to trickle its way into the 0:18:22.240 --> 0:18:25.440 corporate culture of a m D overall, with executives and 0:18:25.520 --> 0:18:29.000 high ranking salesforce professionals spending greater amounts of money to 0:18:29.040 --> 0:18:34.000 curate an image of luxury and sophistication. Spending was getting 0:18:34.000 --> 0:18:36.320 out of hand, and a lot of that spending had 0:18:36.359 --> 0:18:40.400 to do with those fabrication facilities. According to Autika Raza, 0:18:40.600 --> 0:18:43.959 who had led next Gen before a m D had 0:18:43.960 --> 0:18:47.080 acquired that company and then later became the president and 0:18:47.160 --> 0:18:51.960 chief operating officer or CEO of a m D. Sanders 0:18:52.240 --> 0:18:55.960 was in a bad habit of building fabrication facilities too 0:18:55.960 --> 0:19:01.240 far in advance, at least according to Raza's analysis. His perspective, 0:19:01.400 --> 0:19:04.760 that is, Roza's perspective, was that the company should hold 0:19:04.840 --> 0:19:09.480 off building new facilities until the need was there. Sanders 0:19:09.760 --> 0:19:12.400 was building them ahead of the game, but that would 0:19:12.480 --> 0:19:15.560 mean that a m D was constantly raising money to 0:19:15.600 --> 0:19:18.720 build out the next facility in advance of any revenue 0:19:18.840 --> 0:19:21.919 it was generating, and if the industry were to ever dip, 0:19:22.160 --> 0:19:25.040 then it would leave a m D over extended. So 0:19:25.160 --> 0:19:28.000 Raza wanted to take a different route. He wanted to 0:19:28.119 --> 0:19:32.320 use revenues from current successes to fuel expansion on an 0:19:32.359 --> 0:19:34.879 as needed basis. In other words, you don't need to 0:19:34.880 --> 0:19:37.320 go out and build a new fabrication plant until the 0:19:37.400 --> 0:19:41.080 demand requires you to do it. Raza, who at one 0:19:41.080 --> 0:19:44.000 time had been viewed as a potential successor to Sanders, 0:19:44.240 --> 0:19:47.879 found himself in direct disagreement with the founder, and he 0:19:47.920 --> 0:19:51.760 would actually leave a m D in n reportedly after 0:19:51.840 --> 0:19:54.440 a massive falling out with Sanders, with whom he would 0:19:54.480 --> 0:19:58.520 never speak again. Now his successor was a guy named 0:19:58.640 --> 0:20:02.480 Hector Ruez, who had up to that time been heading 0:20:02.560 --> 0:20:06.240 up a division over at Motorola. Ruez was first wary 0:20:06.400 --> 0:20:11.040 of taking this job. It was more technically oriented industry 0:20:11.080 --> 0:20:13.800 than he had been used to, and he knew about 0:20:13.880 --> 0:20:18.520 Sanders and his reputation of alienating senior level staff. And 0:20:18.640 --> 0:20:20.440 he saw that there had been a string of chief 0:20:20.440 --> 0:20:24.960 operating officers, several of whom were rumored to be groomed 0:20:25.080 --> 0:20:28.000 as the heir apparent to a m D who had 0:20:28.000 --> 0:20:31.640 subsequently left the company. But he figured that Sanders might 0:20:31.720 --> 0:20:36.360 have issues relinquishing control to others, that this could cause issues, 0:20:36.560 --> 0:20:42.280 but Sanders was still a very impressive person. A m 0:20:42.359 --> 0:20:44.880 D was an impressive company, so Hector decided to take 0:20:44.880 --> 0:20:48.879 the job. Then Jerry Sanders would retire in the early 0:20:48.920 --> 0:20:51.879 two thousand's and Ruez would take over the company, and 0:20:51.920 --> 0:20:54.639 he began to clean house. He got rid of several 0:20:54.680 --> 0:20:57.920 top executives who had been around for quite some time 0:20:57.920 --> 0:21:01.240 in the Sanders era, and he started bring in new people, 0:21:01.320 --> 0:21:05.160 new talent. So Ruiz also saw that the market was changing, 0:21:05.760 --> 0:21:08.440 and while a m D was being innovative in CPUs 0:21:08.520 --> 0:21:12.760 and other microchips meant for personal computers, it was really 0:21:12.800 --> 0:21:17.520 making most of its profits from selling flash memory not CPUs, 0:21:18.000 --> 0:21:21.520 and so he started to refocus the company to that endeavor, 0:21:21.680 --> 0:21:24.159 but he also found that a m D was holding 0:21:24.240 --> 0:21:27.040 an odd place in the market. Ruez would write a 0:21:27.040 --> 0:21:30.879 book about his experiences, stating that Sanders had created this 0:21:31.200 --> 0:21:34.439 sort of weird paradox in a m D because Sanders 0:21:34.480 --> 0:21:37.080 had a real can do attitude, a a never say 0:21:37.280 --> 0:21:41.040 die approach to business. But at the same time, no 0:21:41.119 --> 0:21:43.600 one in the company ever seemed convinced that a m 0:21:43.680 --> 0:21:46.320 D could really go toe to toe with Intel, that 0:21:46.400 --> 0:21:49.400 a m D would always be a tiny company compared 0:21:49.440 --> 0:21:53.480 to Intel, that could never really take over as the 0:21:53.680 --> 0:21:58.040 leading microchip manufacturer in the industry. That that was just 0:21:58.080 --> 0:22:02.000 sort of this underlineing philosophy at a m D. And 0:22:02.000 --> 0:22:04.640 as possibly because a m D had built its business 0:22:04.720 --> 0:22:08.400 largely on being a second source chip company. So Rule 0:22:08.440 --> 0:22:10.840 has tried to change things, directing a m d s 0:22:10.880 --> 0:22:14.360 efforts at not just flash memory, but also developing premium 0:22:14.400 --> 0:22:18.200 processors for stuff like Internet servers, which were just starting 0:22:18.200 --> 0:22:21.240 to become a serious thing at the time. Now, around 0:22:21.280 --> 0:22:23.920 that same time, a m D and Intel faced off 0:22:24.000 --> 0:22:27.840 again in courtrooms. This time it was in the European Union. 0:22:28.280 --> 0:22:31.199 A m D complained to the European Commission that Intel 0:22:31.280 --> 0:22:35.399 was engaging in anti competitive behavior, violating the law, primarily 0:22:35.720 --> 0:22:39.960 through what am D described as abusive marketing campaigns. A 0:22:40.119 --> 0:22:42.520 m D even tried to use legal means to secure 0:22:42.560 --> 0:22:46.359 documents from a separate case against Intel. This one was 0:22:46.400 --> 0:22:50.080 brought against Intel by a company called Intergraph. But then 0:22:50.160 --> 0:22:54.080 the Intergraph case eventually settled out of court and things 0:22:54.119 --> 0:22:58.760 were obviously still very choppy between a m D and Intel, 0:22:58.920 --> 0:23:02.600 despite the fact that they still had this cross licensing agreement. 0:23:03.359 --> 0:23:05.639 The next chip from a m D, the Case seven, 0:23:05.800 --> 0:23:10.320 better known as the Athlon processor, changed things up again. 0:23:10.600 --> 0:23:13.760 Now the details get pretty technical, but an easy thing 0:23:13.800 --> 0:23:16.320 to understand is that the company was able to push 0:23:16.359 --> 0:23:19.840 clock rates up to one giga hurts. A m D 0:23:20.000 --> 0:23:23.879 also began to manufacture its own motherboards, anticipating that the 0:23:23.960 --> 0:23:27.840 day might come when compatibility with Intel's motherboards would come 0:23:27.880 --> 0:23:30.600 to an end. So, hey, what the heck is a motherboard? 0:23:30.600 --> 0:23:32.600 I mentioned it a couple of times in this episode. 0:23:33.040 --> 0:23:36.480 A motherboard is just a printed circuit board. It's sort 0:23:36.520 --> 0:23:40.560 of the highway system for information inside a computer. The 0:23:40.640 --> 0:23:44.560 motherboard typically has connectors into which you can plug other 0:23:44.680 --> 0:23:47.520 circuits like a CPU as a circuit, so you can 0:23:47.520 --> 0:23:51.520 plug a CPU into a motherboard or a GPU a 0:23:51.520 --> 0:23:56.080 graphics processing unit. The motherboard provides the physical connections between 0:23:56.119 --> 0:23:59.399 all these different components so that these circuits can send 0:23:59.400 --> 0:24:02.360 proper command ends to the right places. Now, not all 0:24:02.440 --> 0:24:07.480 CPUs or GPUs for that matter, are compatible with all motherboards. 0:24:07.800 --> 0:24:11.240 Motherboards can accept certain types of CPUs and not others, 0:24:11.640 --> 0:24:14.679 and that's one of the reasons it's really important to 0:24:14.800 --> 0:24:17.880 research first before you set out to build your first computer. 0:24:18.240 --> 0:24:22.160