1 00:00:04,400 --> 00:00:07,800 Speaker 1: Welcome to Tech Stuff, a production from I Heart Radio. 2 00:00:12,160 --> 00:00:14,880 Speaker 1: Hey there, and welcome to tech Stuff. I'm your host, 3 00:00:15,040 --> 00:00:18,200 Speaker 1: Jonathan Strickland. I'm an executive producer with I Heart Radio, 4 00:00:18,239 --> 00:00:21,840 Speaker 1: and I love all things tech. And when it comes 5 00:00:22,000 --> 00:00:27,360 Speaker 1: to micro processors, there are a few names that tend 6 00:00:27,520 --> 00:00:31,639 Speaker 1: to pop up. Intel is obviously a big one. A 7 00:00:31,840 --> 00:00:34,760 Speaker 1: M D is another, and those are the two they 8 00:00:34,800 --> 00:00:38,280 Speaker 1: get talked about when you're discussing stuff like dusktop computers, 9 00:00:38,280 --> 00:00:42,880 Speaker 1: you know, PCs. But when it comes to more lightweight devices, 10 00:00:43,240 --> 00:00:47,640 Speaker 1: you know, like mobile devices, there's another name, ARM a 11 00:00:48,040 --> 00:00:52,519 Speaker 1: r M. Recently, news broke that the graphics card company 12 00:00:52,600 --> 00:00:57,200 Speaker 1: in Vidio would be acquiring ARM for a staggering forty 13 00:00:57,240 --> 00:01:01,320 Speaker 1: billion dollars a princely some so I thought it would 14 00:01:01,360 --> 00:01:04,040 Speaker 1: be a good idea to kind of go on a 15 00:01:04,120 --> 00:01:08,160 Speaker 1: full rundown on what ARM is, its history, and what 16 00:01:08,280 --> 00:01:12,759 Speaker 1: this acquisition means for the industry and for people like 17 00:01:12,840 --> 00:01:15,800 Speaker 1: you and me. And this is gonna be a two 18 00:01:15,880 --> 00:01:18,440 Speaker 1: parter because ARM has been around for a while and 19 00:01:18,480 --> 00:01:21,480 Speaker 1: its story is actually really interesting. Plus it gives me 20 00:01:21,520 --> 00:01:25,080 Speaker 1: opportunities to go off on crazy tangents and tell you guys, 21 00:01:25,080 --> 00:01:29,039 Speaker 1: how various stuff works, which you know is kind of 22 00:01:29,080 --> 00:01:33,720 Speaker 1: my jam. As the kids say, you know, fifteen years ago. 23 00:01:34,560 --> 00:01:38,960 Speaker 1: We'll start with some history lessons. Now, typically when I 24 00:01:39,000 --> 00:01:42,360 Speaker 1: cover the history of a company or a technology, I 25 00:01:42,440 --> 00:01:46,560 Speaker 1: run into a few cases where dates maybe a little confusing. 26 00:01:46,959 --> 00:01:50,120 Speaker 1: Sometimes one source will have a specific date for an 27 00:01:50,120 --> 00:01:54,360 Speaker 1: event that conflicts with a date that's found in another source, 28 00:01:54,400 --> 00:01:57,120 Speaker 1: and so at that point I will usually say that 29 00:01:58,480 --> 00:02:02,320 Speaker 1: I'm sorry, I apologize. I can't get too specific. And 30 00:02:02,360 --> 00:02:06,320 Speaker 1: you would think that ARM wouldn't have this issue. The 31 00:02:06,360 --> 00:02:09,120 Speaker 1: company isn't that old. We can measure its age in 32 00:02:09,160 --> 00:02:13,000 Speaker 1: a few decades, but we only go back to the 33 00:02:13,080 --> 00:02:16,640 Speaker 1: nineteen eighties or so to to look at origins really 34 00:02:16,680 --> 00:02:19,600 Speaker 1: the late nineteen seventies, And yet when it comes to 35 00:02:19,760 --> 00:02:24,200 Speaker 1: particulars such as which events really got things started for ARM, 36 00:02:24,240 --> 00:02:26,920 Speaker 1: there's actually a lot of disagreements. So I'm going to 37 00:02:26,960 --> 00:02:31,200 Speaker 1: give you a a version of ARMS history. But you know, 38 00:02:31,280 --> 00:02:34,960 Speaker 1: don't think of this as the definitive version, because some 39 00:02:35,000 --> 00:02:37,679 Speaker 1: people say, no, you shouldn't trace its history to that point. 40 00:02:37,760 --> 00:02:41,560 Speaker 1: That silly go to this other point instead. Here in 41 00:02:41,600 --> 00:02:45,320 Speaker 1: the United States, when we talk about the early days 42 00:02:45,639 --> 00:02:49,760 Speaker 1: of personal computers, the names that typically pop up in 43 00:02:49,800 --> 00:02:56,480 Speaker 1: those discussions are Apple, Texas Instruments, Commodore, maybe Tandy, and 44 00:02:56,520 --> 00:03:00,560 Speaker 1: then IBM would follow not too long behind does. But 45 00:03:00,960 --> 00:03:04,480 Speaker 1: across the pond in the UK there was another computer 46 00:03:04,600 --> 00:03:08,160 Speaker 1: company that was trying to get an early part of 47 00:03:08,200 --> 00:03:11,799 Speaker 1: the personal computer era, and this company was called Acorn 48 00:03:12,080 --> 00:03:16,200 Speaker 1: Computers Limited. The three co founders of the company where 49 00:03:16,280 --> 00:03:21,000 Speaker 1: Chris Curry, Hermann Hauser and Andy Hopper, whom I suppose 50 00:03:21,120 --> 00:03:23,360 Speaker 1: was just not dedicated enough to go all in with 51 00:03:23,440 --> 00:03:26,920 Speaker 1: the illiterate names of the other two founders, way to 52 00:03:26,960 --> 00:03:31,840 Speaker 1: go and y. Chris Curry was born in nineteen forty 53 00:03:31,919 --> 00:03:35,720 Speaker 1: six in Cambridge, England. He studied math and physics in 54 00:03:35,840 --> 00:03:39,240 Speaker 1: school and went on to work for various technology companies, 55 00:03:39,280 --> 00:03:43,640 Speaker 1: including Pie Limited that's a p y e. Not you 56 00:03:43,680 --> 00:03:46,240 Speaker 1: know the kind of pie that I love, The Royal 57 00:03:46,360 --> 00:03:51,320 Speaker 1: Radar Establishment and Sinclair Radionics. While his stints at Pie 58 00:03:51,360 --> 00:03:54,880 Speaker 1: and Royal Radar were fairly short, he stuck around at 59 00:03:54,880 --> 00:03:59,480 Speaker 1: Sinclair for several years. By the late nineteen seventies, Curry 60 00:03:59,800 --> 00:04:03,840 Speaker 1: was interested in developing computers, but he was finding no 61 00:04:04,080 --> 00:04:08,119 Speaker 1: real support at Sinclair. He had been working on some 62 00:04:08,160 --> 00:04:11,280 Speaker 1: stuff he was trying to pitch the idea to Sinclair, 63 00:04:11,320 --> 00:04:13,920 Speaker 1: but he wasn't finding them very receptive. So we're gonna 64 00:04:14,000 --> 00:04:17,680 Speaker 1: leave off for now with with Curry and move on 65 00:04:17,880 --> 00:04:22,080 Speaker 1: and we'll regroup in a second. Swapping over to Hermann Hauser, 66 00:04:22,200 --> 00:04:26,400 Speaker 1: who was born in Austria in nineteen forty eight. He 67 00:04:26,520 --> 00:04:30,320 Speaker 1: came to Cambridge as a teenager to attend school and 68 00:04:30,440 --> 00:04:34,360 Speaker 1: learn English, but he really enjoyed it. He went on 69 00:04:34,480 --> 00:04:38,479 Speaker 1: to pursue advanced studies in places like Vienna University, but 70 00:04:38,680 --> 00:04:43,040 Speaker 1: he came back to England attending King's College in Cambridge 71 00:04:43,360 --> 00:04:46,320 Speaker 1: and getting an advanced degree. They're smart, dude. His work 72 00:04:46,360 --> 00:04:49,279 Speaker 1: in physics led him to become friends with Curry, and 73 00:04:49,320 --> 00:04:52,760 Speaker 1: when Curry was ready to start a company to manufacture 74 00:04:52,800 --> 00:04:57,560 Speaker 1: and market personal computers, Houser was on board. Andy Hopper 75 00:04:58,080 --> 00:05:01,479 Speaker 1: was the youngest of the three co founders, having been 76 00:05:01,480 --> 00:05:06,080 Speaker 1: born in nineteen fifty three in Warsaw, Poland. Hopper studied 77 00:05:06,200 --> 00:05:10,960 Speaker 1: in London and later Swansea University before pursuing postgraduate studies 78 00:05:11,040 --> 00:05:14,760 Speaker 1: at the University of Cambridge. He focused on computer science 79 00:05:15,160 --> 00:05:19,040 Speaker 1: and he was researching networking technologies. He co founded a 80 00:05:19,080 --> 00:05:23,719 Speaker 1: company called Orbis Limited, which focused mostly on Networking Tech, 81 00:05:23,960 --> 00:05:27,839 Speaker 1: and this entity would end up merging with Curry and 82 00:05:27,920 --> 00:05:32,599 Speaker 1: Houser's efforts to bring Acorn Computers Limited to life. Although 83 00:05:32,640 --> 00:05:36,840 Speaker 1: the original name for the actual company was Cambridge Processing 84 00:05:37,040 --> 00:05:42,320 Speaker 1: Unit Limited CPU cute right, but the co founders decided 85 00:05:42,360 --> 00:05:46,320 Speaker 1: to use the name Acorn Computers as the trading name 86 00:05:46,560 --> 00:05:50,719 Speaker 1: for the company, allegedly choosing the name Acorn so that 87 00:05:50,800 --> 00:05:55,960 Speaker 1: their computers would appear ahead of rival Apple Computers whenever 88 00:05:56,000 --> 00:05:59,960 Speaker 1: it was in an alphabetical listing. One of the earliest 89 00:06:00,160 --> 00:06:03,880 Speaker 1: jobs that the group had was to develop micro controllers 90 00:06:04,160 --> 00:06:09,400 Speaker 1: for fruit machines, and folks, I consider myself an Anglo 91 00:06:09,480 --> 00:06:12,680 Speaker 1: file but when I first read that, I have to 92 00:06:12,760 --> 00:06:16,200 Speaker 1: admit I had no idea what the heck it meant. 93 00:06:16,480 --> 00:06:19,880 Speaker 1: In my imagination, it was some sort of harvesting device 94 00:06:19,960 --> 00:06:24,680 Speaker 1: that depended on micro controllers to do something like pick apples. 95 00:06:24,760 --> 00:06:27,880 Speaker 1: But no, that, of course, is not what a fruit 96 00:06:27,960 --> 00:06:30,640 Speaker 1: machine is. And my guess is there's more than a 97 00:06:30,640 --> 00:06:33,400 Speaker 1: few of you out there giggling at my ignorance right now. 98 00:06:34,120 --> 00:06:37,960 Speaker 1: It's well warranted. So a fruit machine in this context 99 00:06:38,040 --> 00:06:41,680 Speaker 1: is what Brits call a slot machine. I guess they 100 00:06:41,720 --> 00:06:45,600 Speaker 1: do it because of the symbols of fruit that appear 101 00:06:46,200 --> 00:06:50,120 Speaker 1: on the various parts of the slot machine. So the 102 00:06:50,120 --> 00:06:53,680 Speaker 1: first big job that this new company had was designing 103 00:06:53,720 --> 00:06:57,520 Speaker 1: micro controllers for these gambling machines the slot machines to 104 00:06:57,560 --> 00:06:59,960 Speaker 1: make the more difficult to tamper with, as there were 105 00:07:00,040 --> 00:07:02,839 Speaker 1: some clever hackers who are finding ways to rig big 106 00:07:02,880 --> 00:07:06,760 Speaker 1: payouts from the machines. And while the machines are designed 107 00:07:06,760 --> 00:07:09,640 Speaker 1: to take money from you, uh, they don't like the 108 00:07:09,680 --> 00:07:12,840 Speaker 1: design to go the opposite direction. The house is not 109 00:07:12,960 --> 00:07:16,040 Speaker 1: a big fan of that. A few years later, the 110 00:07:16,200 --> 00:07:20,720 Speaker 1: UK launched an initiative to put a computer in every classroom. 111 00:07:21,200 --> 00:07:25,800 Speaker 1: Acorn Computers secured the contract to provide these computers to 112 00:07:25,880 --> 00:07:30,520 Speaker 1: produce them, and it was called the BBC Micro. The 113 00:07:30,640 --> 00:07:34,320 Speaker 1: Micro used a processor called the six five O two. 114 00:07:34,640 --> 00:07:38,240 Speaker 1: This was an eight bit processor from Rockwell h Though 115 00:07:38,280 --> 00:07:42,560 Speaker 1: engineers from Most Technology originally developed the six five O 116 00:07:42,720 --> 00:07:45,600 Speaker 1: two processor, the six five O two was a low 117 00:07:45,680 --> 00:07:49,640 Speaker 1: cost processor that totally changed the processor market. It was 118 00:07:49,800 --> 00:07:53,760 Speaker 1: much less expensive than Intel's d D processor at the time, 119 00:07:54,160 --> 00:07:56,800 Speaker 1: and as such, the six five O two had found 120 00:07:56,840 --> 00:08:02,080 Speaker 1: its way into numerous technologies in the including the Atrey 121 00:08:02,840 --> 00:08:06,360 Speaker 1: video game console and the Apple two Computer Systems, among others. 122 00:08:06,880 --> 00:08:12,880 Speaker 1: The micro contract gave Acorn Computers some momentum. In three 123 00:08:13,200 --> 00:08:17,160 Speaker 1: the company wanted to free itself from dependence upon processors 124 00:08:17,280 --> 00:08:22,360 Speaker 1: from other companies. To computer scientists from Cambridge University, Sophie 125 00:08:22,360 --> 00:08:26,400 Speaker 1: Wilson and Steve Ferber became the head designers for the 126 00:08:26,600 --> 00:08:31,120 Speaker 1: new thirty two bit processor. Ferber focused on the actual 127 00:08:31,200 --> 00:08:35,040 Speaker 1: physical design of the chips architecture, while Wilson was focusing 128 00:08:35,200 --> 00:08:40,000 Speaker 1: on the instruction set. The limited resources forced the pair 129 00:08:40,120 --> 00:08:43,280 Speaker 1: to come up with a simplified approach to processors, and 130 00:08:43,320 --> 00:08:45,920 Speaker 1: they chose to go with a specific approach to processor 131 00:08:46,000 --> 00:08:51,720 Speaker 1: design in a category called reduced instruction set computing or 132 00:08:52,080 --> 00:08:56,400 Speaker 1: risk r I s C. This is in contrast with 133 00:08:56,559 --> 00:09:01,360 Speaker 1: complex instruction set computing or see I s C CISC. 134 00:09:01,960 --> 00:09:06,000 Speaker 1: But what does that actually mean? Well, let's take a 135 00:09:06,080 --> 00:09:10,240 Speaker 1: step back to understand this. A processor's job is to 136 00:09:10,280 --> 00:09:15,280 Speaker 1: perform arithmetic and logic operations on data, and this includes 137 00:09:15,360 --> 00:09:18,560 Speaker 1: basic stuff like you know, adding and subtracting, kind of 138 00:09:18,600 --> 00:09:22,599 Speaker 1: like your basic calculator, and it can also involve transferring 139 00:09:22,679 --> 00:09:26,960 Speaker 1: numbers and comparing different numbers to one another. The data 140 00:09:27,240 --> 00:09:31,200 Speaker 1: comes in as binary or let's be zeros and ones. 141 00:09:32,000 --> 00:09:35,560 Speaker 1: The processor follows instructions given to it by a program. 142 00:09:35,920 --> 00:09:39,760 Speaker 1: So the processor gets its instructions like, you know, add 143 00:09:39,840 --> 00:09:42,839 Speaker 1: the next two numbers together and then send it on, 144 00:09:43,520 --> 00:09:47,679 Speaker 1: and then the processor executes that instruction on the supplied 145 00:09:47,920 --> 00:09:51,000 Speaker 1: zeros and ones that come in. And that's basically what's 146 00:09:51,000 --> 00:09:54,400 Speaker 1: going on with a processor at a very high level. 147 00:09:55,000 --> 00:09:59,119 Speaker 1: In addition, we describe the number of operations a processor 148 00:09:59,200 --> 00:10:02,720 Speaker 1: can complete in a second as its clock speed, which 149 00:10:02,760 --> 00:10:06,040 Speaker 1: we talk about in hurts. A single pulse of the 150 00:10:06,080 --> 00:10:10,760 Speaker 1: processor is a cycle. A one Hurts processor would only 151 00:10:10,800 --> 00:10:14,240 Speaker 1: be able to complete a single operation every second. It 152 00:10:14,240 --> 00:10:19,720 Speaker 1: would be unfathomably slow to us. A decent processor speed 153 00:10:19,720 --> 00:10:23,520 Speaker 1: today is somewhere between two point five and three point 154 00:10:23,559 --> 00:10:26,920 Speaker 1: five giga hurts. That's decent. I'm not talking about top 155 00:10:26,960 --> 00:10:29,319 Speaker 1: of the line, but that would mean two point five 156 00:10:29,520 --> 00:10:33,760 Speaker 1: to three point five billion cycles per second. So the 157 00:10:33,840 --> 00:10:39,560 Speaker 1: processors in modern computers are pulsing billions of times every second, 158 00:10:39,559 --> 00:10:44,120 Speaker 1: and each pulse can power and operation. Now, some instructions 159 00:10:44,160 --> 00:10:47,840 Speaker 1: are pretty simple and they might only require one or 160 00:10:48,000 --> 00:10:52,720 Speaker 1: two cycles to complete that instruction. Other instructions are more 161 00:10:52,760 --> 00:10:56,400 Speaker 1: complicated and might have lots more steps involved, and this 162 00:10:56,480 --> 00:11:00,559 Speaker 1: is where we get to the risk versus CISK approach. 163 00:11:00,640 --> 00:11:05,480 Speaker 1: A risk based processor handles very simple instructions, so it 164 00:11:05,559 --> 00:11:10,239 Speaker 1: handles each individual instruction very quickly, like within a cycle. 165 00:11:10,880 --> 00:11:15,880 Speaker 1: CISC systems can handle much more complicated instructions. The flip 166 00:11:16,040 --> 00:11:19,320 Speaker 1: side of that is that while a risk based processor 167 00:11:19,440 --> 00:11:23,439 Speaker 1: can execute individual instructions very very quickly, you might need 168 00:11:23,480 --> 00:11:27,440 Speaker 1: a lot more instructions to complete your overall task. The 169 00:11:27,559 --> 00:11:31,400 Speaker 1: CISK approach might take longer to execute a single instruction, 170 00:11:31,679 --> 00:11:35,800 Speaker 1: but you need fewer instructions overall to complete your task. 171 00:11:36,120 --> 00:11:39,880 Speaker 1: Now that is a little confusing, So I'll use an analogy. 172 00:11:40,200 --> 00:11:42,160 Speaker 1: If I told the typical person, I need you to 173 00:11:42,200 --> 00:11:46,000 Speaker 1: go outside and check the weather, that's a deceptively complicated 174 00:11:46,040 --> 00:11:49,400 Speaker 1: instruction because there's a lot of other stuff that's nested 175 00:11:49,559 --> 00:11:52,640 Speaker 1: in that request. For example, if I were to try 176 00:11:52,679 --> 00:11:54,840 Speaker 1: and tell this to a robot, I might have to 177 00:11:54,880 --> 00:11:58,079 Speaker 1: include what direction the robot needs to go in, how 178 00:11:58,160 --> 00:12:01,720 Speaker 1: fast it should move, where of the door is, whether 179 00:12:01,800 --> 00:12:05,840 Speaker 1: that door opens inward or outward, the actual mechanism the 180 00:12:05,920 --> 00:12:08,920 Speaker 1: robot would have to manipulate to open the door, and 181 00:12:08,960 --> 00:12:12,000 Speaker 1: so on. So what appears to be a simple task 182 00:12:12,400 --> 00:12:16,760 Speaker 1: is actually when you break it down into its individual components, 183 00:12:17,160 --> 00:12:20,800 Speaker 1: much more complicated. So a program running on a risk 184 00:12:20,880 --> 00:12:24,160 Speaker 1: based processor has to break down instructions kind of in 185 00:12:24,200 --> 00:12:28,200 Speaker 1: that way in a longer series of simple tasks that 186 00:12:28,320 --> 00:12:31,960 Speaker 1: add up to whatever you're end goal is. But risk 187 00:12:32,160 --> 00:12:36,400 Speaker 1: chips are highly optimized, so for certain applications a risk 188 00:12:36,480 --> 00:12:40,160 Speaker 1: based chip can be ideal. These days, we have a 189 00:12:40,200 --> 00:12:43,440 Speaker 1: lot of risk chips and stuff like mobile devices, for example, 190 00:12:43,480 --> 00:12:48,040 Speaker 1: because these devices are rarely running super complicated software and 191 00:12:48,080 --> 00:12:53,199 Speaker 1: they need that low power, high efficiency output. One related 192 00:12:53,280 --> 00:12:56,520 Speaker 1: thing I'd like to mention, though it doesn't tie directly 193 00:12:56,600 --> 00:13:00,080 Speaker 1: into arms history or anything, is what is called a 194 00:13:00,160 --> 00:13:04,640 Speaker 1: semantic gap. Now, remember when I said that processors taken 195 00:13:04,760 --> 00:13:08,320 Speaker 1: data in the form of zeros and ones. This binary 196 00:13:08,400 --> 00:13:11,440 Speaker 1: code is a type of machine language, or the kind 197 00:13:11,440 --> 00:13:15,920 Speaker 1: of information a machine can actually process. Machines are not 198 00:13:16,080 --> 00:13:20,720 Speaker 1: able to process information in other forms directly. The information 199 00:13:20,800 --> 00:13:24,720 Speaker 1: must ultimately convert into machine language, in this case zeros 200 00:13:24,720 --> 00:13:29,800 Speaker 1: and ones, and information that we can express pretty succinctly 201 00:13:29,960 --> 00:13:33,400 Speaker 1: with language and numerals. Beyond just the ones and zeros, 202 00:13:33,960 --> 00:13:35,880 Speaker 1: that ends up taking up a lot of space. You 203 00:13:35,920 --> 00:13:37,760 Speaker 1: have to use a lot of ones and zeros to 204 00:13:37,880 --> 00:13:41,360 Speaker 1: represent that kind of information. But computers are really good 205 00:13:41,360 --> 00:13:46,359 Speaker 1: at processing machine code. It happens lightning fast. However, programming 206 00:13:46,400 --> 00:13:51,400 Speaker 1: computers in machine code is really really hard. I mean, 207 00:13:51,440 --> 00:13:54,640 Speaker 1: imagine having to type in a string of tens of 208 00:13:54,760 --> 00:13:58,520 Speaker 1: thousands of zeros and ones while you're trying to program 209 00:13:58,520 --> 00:14:01,120 Speaker 1: a machine, and you know that if you make just 210 00:14:01,280 --> 00:14:04,600 Speaker 1: one mistake, you mess up the whole program because the 211 00:14:04,600 --> 00:14:07,360 Speaker 1: whole chain is screwed up after that. Heck, if you 212 00:14:07,400 --> 00:14:10,080 Speaker 1: did make a mistake, it would be really hard for 213 00:14:10,120 --> 00:14:13,040 Speaker 1: you to track down where you made the mistake in 214 00:14:13,080 --> 00:14:16,240 Speaker 1: the programming. You would have to compare two different, very 215 00:14:16,320 --> 00:14:19,480 Speaker 1: long sheets of zeros and ones, and you'd probably lose 216 00:14:19,520 --> 00:14:22,960 Speaker 1: your mind. That's one of the big reasons computer scientists 217 00:14:23,000 --> 00:14:26,880 Speaker 1: have developed various programming languages. The idea is that the 218 00:14:26,960 --> 00:14:30,640 Speaker 1: programming language is something that's easier for human beings to 219 00:14:30,720 --> 00:14:35,160 Speaker 1: work with, but machines can't understand programming languages without the 220 00:14:35,280 --> 00:14:39,520 Speaker 1: use of something called a compiler. The compiler's job is 221 00:14:39,640 --> 00:14:42,920 Speaker 1: essentially that of a translator. It takes the program that's 222 00:14:42,920 --> 00:14:46,920 Speaker 1: been written in whatever programming language and converts the instructions 223 00:14:46,920 --> 00:14:50,200 Speaker 1: into machine code so that the computer can process it. 224 00:14:50,960 --> 00:14:55,120 Speaker 1: The compiler is essentially a middleman between the program and 225 00:14:55,160 --> 00:14:59,560 Speaker 1: the processor. We describe programming languages by calling them stuff 226 00:14:59,600 --> 00:15:03,480 Speaker 1: like level or high level. This refers to how closely 227 00:15:03,560 --> 00:15:07,400 Speaker 1: the language resembles the machine code. So a low level 228 00:15:07,480 --> 00:15:11,000 Speaker 1: programming language is really only a couple of steps away 229 00:15:11,040 --> 00:15:14,840 Speaker 1: from machine code itself. It's much easier for a compiler 230 00:15:14,880 --> 00:15:17,760 Speaker 1: to handle that kind of language, but it's much harder 231 00:15:17,760 --> 00:15:20,840 Speaker 1: to program in. You have to frame your programming closer 232 00:15:20,880 --> 00:15:25,000 Speaker 1: to machine code, but it's still easier than programming instructions 233 00:15:25,000 --> 00:15:29,160 Speaker 1: than just ones and zeros. A high level programming language 234 00:15:29,840 --> 00:15:33,320 Speaker 1: is modeled closer to how we would think in terms 235 00:15:33,400 --> 00:15:36,640 Speaker 1: of a typical language. There are still rules you have 236 00:15:36,680 --> 00:15:39,680 Speaker 1: to follow, and if you're not familiar with that particular 237 00:15:39,720 --> 00:15:42,360 Speaker 1: language and you're looking at an example of it, it's 238 00:15:42,360 --> 00:15:44,240 Speaker 1: not likely going to make a whole lot of sense 239 00:15:44,280 --> 00:15:46,960 Speaker 1: to you. But it's much easier for humans to work 240 00:15:47,000 --> 00:15:50,440 Speaker 1: with these kind of languages. However, it's less efficient for 241 00:15:50,560 --> 00:15:54,640 Speaker 1: compilers to handle that and compile that into machine language. 242 00:15:55,280 --> 00:15:59,960 Speaker 1: We call this adding layers of abstraction. The programming lay. 243 00:16:00,120 --> 00:16:04,480 Speaker 1: WHIGE provides an abstract platform that represents the various tasks 244 00:16:04,520 --> 00:16:08,000 Speaker 1: that the processor will ultimately carry out, and the gap 245 00:16:08,240 --> 00:16:11,680 Speaker 1: between what the programming language says and what the processor 246 00:16:11,840 --> 00:16:17,200 Speaker 1: does is the semantic gap. CISK and risk designs deal 247 00:16:17,280 --> 00:16:21,280 Speaker 1: with this gap in different ways. A CISK design includes 248 00:16:21,320 --> 00:16:25,040 Speaker 1: a lot of addressing modes and lots of different instructions. 249 00:16:25,080 --> 00:16:28,800 Speaker 1: A RISK design has a much more simplified instruction set 250 00:16:29,200 --> 00:16:32,400 Speaker 1: that can meet the requirements of user programs. It's really 251 00:16:32,440 --> 00:16:36,640 Speaker 1: just two different methods to achieve a similar result. Depending 252 00:16:36,720 --> 00:16:41,320 Speaker 1: upon the history you read, Acorn Computer slash Cambridge Processing 253 00:16:41,400 --> 00:16:47,640 Speaker 1: Unit called their risk based design Acorn Risk Machines, or 254 00:16:47,680 --> 00:16:51,800 Speaker 1: they called it Advanced Risk Machines. Most histories say that 255 00:16:51,880 --> 00:16:55,120 Speaker 1: originally it was Acorn Risk Machines and only later changed 256 00:16:55,160 --> 00:16:59,240 Speaker 1: to Advanced Risk Machines. But either way, the initialism for 257 00:16:59,320 --> 00:17:04,080 Speaker 1: this technolog g became a r M or ARM. When 258 00:17:04,080 --> 00:17:07,040 Speaker 1: we come back, i'll talk about how this technology would 259 00:17:07,080 --> 00:17:11,600 Speaker 1: ultimately transcend the company that spawned it, but first let's 260 00:17:11,640 --> 00:17:23,240 Speaker 1: take a quick break. Steve Ferber, Sophie Wilson, and Robert 261 00:17:23,240 --> 00:17:27,600 Speaker 1: Heaton program the initial instruction set for the ARM processor 262 00:17:27,720 --> 00:17:31,960 Speaker 1: in Basic. That's a programming language that originated back in 263 00:17:32,119 --> 00:17:37,439 Speaker 1: nineteen sixty four. Basic stands for Beginners all Purpose Symbolic 264 00:17:37,560 --> 00:17:41,680 Speaker 1: instruction Code. It's a high level programming language that, as 265 00:17:41,720 --> 00:17:47,080 Speaker 1: the name implies, simplifies things for beginner programmers. The Acorn 266 00:17:47,119 --> 00:17:50,600 Speaker 1: team weren't beginners, but they wanted to keep instructions as 267 00:17:50,680 --> 00:17:54,760 Speaker 1: simple as possible to optimize the processors. Their first effort 268 00:17:54,840 --> 00:18:00,520 Speaker 1: yielded the ARM one processor That endeavor took two years 269 00:18:00,680 --> 00:18:05,119 Speaker 1: of development, with the ARM one debuting in nine only 270 00:18:05,240 --> 00:18:13,119 Speaker 1: debuting internally v l s I Technology, another company fabricating company, 271 00:18:13,160 --> 00:18:17,840 Speaker 1: they actually produced the chip the working chips. The chip 272 00:18:17,920 --> 00:18:21,480 Speaker 1: had fewer than twenty five thousand transistors on it and 273 00:18:21,560 --> 00:18:26,480 Speaker 1: used a process with a resolution of three microns or micrometers. 274 00:18:26,520 --> 00:18:31,639 Speaker 1: That's one millionth of a meter for comparisons sake, Today's 275 00:18:31,760 --> 00:18:35,959 Speaker 1: Intel processors have more than a billion transistors and they 276 00:18:36,040 --> 00:18:38,520 Speaker 1: use a fabrication process with a resolution of just a 277 00:18:38,560 --> 00:18:42,680 Speaker 1: few nanometers, and a nanometer is one billionth of a meter, 278 00:18:42,840 --> 00:18:46,880 Speaker 1: so we've definitely come a long way since the early eighties. 279 00:18:47,440 --> 00:18:50,080 Speaker 1: The team learned a great deal through their experience of 280 00:18:50,119 --> 00:18:53,920 Speaker 1: developing the ARM one, and rather than immediately go into production, 281 00:18:54,080 --> 00:18:57,120 Speaker 1: the design team began to work on refining their product, 282 00:18:57,440 --> 00:19:01,120 Speaker 1: creating the next generation of processors based on the architecture. 283 00:19:01,400 --> 00:19:05,040 Speaker 1: They wanted to improve certain processes, and they added instructions 284 00:19:05,040 --> 00:19:10,200 Speaker 1: for stuff like multiply and multiply and accumulate. They built 285 00:19:10,200 --> 00:19:13,560 Speaker 1: in capabilities that would allow the processor to perform real 286 00:19:13,720 --> 00:19:18,240 Speaker 1: time digital signal processing, a necessity if they wanted the 287 00:19:18,240 --> 00:19:21,720 Speaker 1: processor to be able to handle processes meant to you know, 288 00:19:21,800 --> 00:19:25,280 Speaker 1: generate sounds for example, which the company considered an important 289 00:19:25,359 --> 00:19:29,439 Speaker 1: part of a computer's capabilities. They increased the number of 290 00:19:29,440 --> 00:19:33,040 Speaker 1: transistors on the microprocessor from twenty five thousand from ARM 291 00:19:33,080 --> 00:19:37,440 Speaker 1: one to thirty thousand for ARM two. The team also 292 00:19:37,520 --> 00:19:42,400 Speaker 1: developed a coprocessor, which, as that name implies, is a 293 00:19:42,440 --> 00:19:46,800 Speaker 1: processor that can work in concert with the primary processor. 294 00:19:47,200 --> 00:19:52,640 Speaker 1: Coprocessors typically handle specific tasks. They're meant to kick in 295 00:19:52,680 --> 00:19:57,720 Speaker 1: when something specific happens, and it offloads those tasks from 296 00:19:57,840 --> 00:20:01,120 Speaker 1: the responsibility of the primary process sessor. So this would 297 00:20:01,119 --> 00:20:04,080 Speaker 1: be kind of like having two people dividing up work 298 00:20:04,119 --> 00:20:08,160 Speaker 1: among them, and one person handles a subset of chores 299 00:20:08,560 --> 00:20:10,359 Speaker 1: and the other person has to do all the rest 300 00:20:10,359 --> 00:20:13,119 Speaker 1: of the chores. Now, in this case, the coprocessor was 301 00:20:13,160 --> 00:20:18,280 Speaker 1: powering a floating point accelerator. Ah, but that leads us 302 00:20:18,280 --> 00:20:23,399 Speaker 1: to ask what is a floating point? Well, I'm sad 303 00:20:23,480 --> 00:20:25,600 Speaker 1: to be the one to have to tell you this. 304 00:20:25,840 --> 00:20:31,280 Speaker 1: Please set yourself down and and prepare yourself. Computers have 305 00:20:31,440 --> 00:20:36,400 Speaker 1: a limited capacity. Computing memory is not infinite, and so 306 00:20:36,520 --> 00:20:39,760 Speaker 1: we have to start making some concessions when we're working 307 00:20:40,040 --> 00:20:43,680 Speaker 1: with numbers. Now, as you may be aware, some numbers 308 00:20:44,040 --> 00:20:49,239 Speaker 1: can be really really big or really really small, and 309 00:20:49,280 --> 00:20:53,120 Speaker 1: they might have a super long, perhaps even infinite number 310 00:20:53,359 --> 00:20:58,160 Speaker 1: of numbers behind a decimal point. Computers can't cope with that. 311 00:20:58,680 --> 00:21:02,000 Speaker 1: They have limitation on what they can handle. So we 312 00:21:02,080 --> 00:21:05,640 Speaker 1: have to make some concessions, and floating points are one 313 00:21:05,640 --> 00:21:08,480 Speaker 1: of the ways we make concessions. Now, at some point 314 00:21:08,800 --> 00:21:12,280 Speaker 1: we have to cut off numbers, and when and where 315 00:21:12,320 --> 00:21:15,119 Speaker 1: we cut off numbers depends upon what we're doing. So, 316 00:21:15,200 --> 00:21:18,760 Speaker 1: for example, if we are making a tool like a rake, 317 00:21:19,760 --> 00:21:22,840 Speaker 1: you know, just an old lawn rake, and you want 318 00:21:22,920 --> 00:21:25,120 Speaker 1: the handle for this rake to be five ft long, 319 00:21:25,640 --> 00:21:28,920 Speaker 1: you probably don't actually care if the handle comes out 320 00:21:28,960 --> 00:21:32,919 Speaker 1: to be four ft eleven inches and some change, or 321 00:21:33,280 --> 00:21:36,200 Speaker 1: five foot and a fraction of an inch that level 322 00:21:36,200 --> 00:21:39,639 Speaker 1: of precision isn't really important to you. It needs to 323 00:21:39,640 --> 00:21:43,280 Speaker 1: be five ft ish, but if it's not exactly at 324 00:21:43,280 --> 00:21:46,400 Speaker 1: five ft it's not a deal breaker. But let's say 325 00:21:46,400 --> 00:21:50,200 Speaker 1: you're building a transistor for a processor. Well, in that case, 326 00:21:50,240 --> 00:21:55,600 Speaker 1: you're working in a very very very small frame of reference, 327 00:21:55,920 --> 00:21:58,120 Speaker 1: and so the difference of a fraction of a meter 328 00:21:58,280 --> 00:22:02,560 Speaker 1: represents a gargantu one difference. On the flip side, you're 329 00:22:02,600 --> 00:22:05,679 Speaker 1: not likely to ever have to worry about distances of 330 00:22:05,880 --> 00:22:08,879 Speaker 1: a centimeter. That would be way too big. So you 331 00:22:09,040 --> 00:22:12,280 Speaker 1: just have to have a way to maintain accuracy relative 332 00:22:12,320 --> 00:22:15,280 Speaker 1: to what you're doing. You have a different, you know, 333 00:22:15,600 --> 00:22:19,800 Speaker 1: context for your work. This gets a bit more complicated 334 00:22:19,800 --> 00:22:22,239 Speaker 1: when you need to work with both really big and 335 00:22:22,440 --> 00:22:25,960 Speaker 1: really small numbers at the same time. For example, let's 336 00:22:25,960 --> 00:22:29,480 Speaker 1: say you're a scientist and you're working with Newton's gravitational constant. 337 00:22:29,920 --> 00:22:32,600 Speaker 1: That is a very small number that starts with a decimal. 338 00:22:33,040 --> 00:22:36,159 Speaker 1: Then you have ten zeros before you get to the 339 00:22:36,160 --> 00:22:39,280 Speaker 1: first non zero number, which is a six. By the way, 340 00:22:39,480 --> 00:22:41,760 Speaker 1: you might also be working with the speed of light. 341 00:22:42,680 --> 00:22:46,720 Speaker 1: That's a very big number, but the computer memory can't 342 00:22:46,760 --> 00:22:50,320 Speaker 1: really handle number sizes that include that wide a spectrum 343 00:22:50,320 --> 00:22:54,000 Speaker 1: of numbers, and that's why floating points are used, and 344 00:22:54,040 --> 00:22:58,080 Speaker 1: they're sort of like using numbers and scientific notation. You've 345 00:22:58,080 --> 00:23:02,679 Speaker 1: got a significant with contains the digit of the number 346 00:23:02,760 --> 00:23:05,479 Speaker 1: or the digits of whatever number you're talking about, and 347 00:23:05,480 --> 00:23:08,560 Speaker 1: you've got an exponent which tells you where the decimal 348 00:23:08,600 --> 00:23:11,760 Speaker 1: point needs to be in relation to the first digit 349 00:23:11,880 --> 00:23:15,280 Speaker 1: in the significant. So if I have a significant of 350 00:23:15,520 --> 00:23:19,320 Speaker 1: one point seven and I have an exponent of six, 351 00:23:19,920 --> 00:23:22,040 Speaker 1: it would be the same as if I wrote that 352 00:23:22,160 --> 00:23:25,280 Speaker 1: number in the scientific notation as one point seven times 353 00:23:25,280 --> 00:23:27,920 Speaker 1: ten to the sixth power, which is the same thing 354 00:23:27,960 --> 00:23:31,720 Speaker 1: as one million, seven hundred thousand. These are all just 355 00:23:31,800 --> 00:23:34,800 Speaker 1: different ways to represent the same value. So one point 356 00:23:34,840 --> 00:23:38,160 Speaker 1: seven significant with an exponent of six is one million, 357 00:23:38,240 --> 00:23:41,960 Speaker 1: seven hundred thousand. Likewise, if I had a significant of 358 00:23:42,000 --> 00:23:46,280 Speaker 1: one point seven and an exponent of negative six, this 359 00:23:46,280 --> 00:23:48,520 Speaker 1: would be the same as one point seven times ten 360 00:23:48,640 --> 00:23:54,160 Speaker 1: to the negative sixth power or point zero zero zero 361 00:23:54,359 --> 00:23:59,440 Speaker 1: zero zero one seven. By using floating points, we can 362 00:23:59,480 --> 00:24:04,159 Speaker 1: simplify how we represent numbers without damaging the value of 363 00:24:04,200 --> 00:24:07,240 Speaker 1: those numbers. And let's get around the limitations of computer 364 00:24:07,320 --> 00:24:10,600 Speaker 1: memory and how many bits a processor can handle at 365 00:24:10,600 --> 00:24:15,120 Speaker 1: a single time. We call the operations that processors perform 366 00:24:15,240 --> 00:24:19,199 Speaker 1: on these types of numbers floating point operations, and we 367 00:24:19,280 --> 00:24:23,040 Speaker 1: measure it in flops, which stands for floating point operations 368 00:24:23,080 --> 00:24:27,280 Speaker 1: per second. A giga flop would be a billion floating 369 00:24:27,359 --> 00:24:33,679 Speaker 1: point operations per second. The Japanese super computer Fugaku can 370 00:24:33,760 --> 00:24:38,200 Speaker 1: reach more than four hundred fifteen peda flops. A pedal 371 00:24:38,240 --> 00:24:44,119 Speaker 1: flop is a thousand million million floating point operations per second, 372 00:24:44,520 --> 00:24:47,440 Speaker 1: So a pedal flop would be a one followed by 373 00:24:47,600 --> 00:24:55,000 Speaker 1: fifteen zeros yauza. So the ARM to architecture included a 374 00:24:55,000 --> 00:24:59,920 Speaker 1: coprocessor for floating point acceleration, not a full floating point process, 375 00:25:00,240 --> 00:25:04,720 Speaker 1: but to accelerate floating point operation calculations, as well as 376 00:25:04,720 --> 00:25:09,480 Speaker 1: the possibility of adding other coprocessors with the basic ARM architecture. 377 00:25:09,520 --> 00:25:12,200 Speaker 1: It was kind of a sort of a modular design. 378 00:25:12,720 --> 00:25:16,040 Speaker 1: This generation was called, fittingly enough, ARMED two, and the 379 00:25:16,119 --> 00:25:19,880 Speaker 1: first product to market that featured the ARM two wasn't 380 00:25:19,960 --> 00:25:23,800 Speaker 1: a fully fledged computer, but rather the ARM development system, 381 00:25:23,840 --> 00:25:27,880 Speaker 1: which included the ARM processor, four megabytes of RAM, three 382 00:25:27,920 --> 00:25:31,240 Speaker 1: support chips, and some development tools. So essentially this was 383 00:25:31,280 --> 00:25:34,320 Speaker 1: a product meant for programmers. It wasn't like it was 384 00:25:34,400 --> 00:25:38,800 Speaker 1: meant for your average end consumer. Meanwhile, at the company 385 00:25:38,840 --> 00:25:43,200 Speaker 1: at large, things were not going so super well. Acorn 386 00:25:43,240 --> 00:25:46,520 Speaker 1: Computers was in a bit of a financial crisis and 387 00:25:46,680 --> 00:25:50,760 Speaker 1: an Italian company known for computer systems and office equipment 388 00:25:50,760 --> 00:25:54,719 Speaker 1: in Europe called Olivetti ing s c. And I know 389 00:25:54,840 --> 00:25:58,280 Speaker 1: I've butchered it, but Olivetti is what's best known as 390 00:25:58,680 --> 00:26:02,480 Speaker 1: it swept in and it acquired the English computer company. 391 00:26:02,840 --> 00:26:07,480 Speaker 1: At the time, Olivetti was reportedly unaware that within Acorn 392 00:26:07,560 --> 00:26:10,840 Speaker 1: Computers there were engineers who are working on new processors 393 00:26:10,880 --> 00:26:14,879 Speaker 1: because the original Acorn computers we're using processors made from 394 00:26:14,920 --> 00:26:18,480 Speaker 1: other companies, so the acquisition would slow things down a 395 00:26:18,480 --> 00:26:20,480 Speaker 1: little bit. That's one of the reasons why there was 396 00:26:20,520 --> 00:26:24,000 Speaker 1: a delay between the development of the original ARM one 397 00:26:24,119 --> 00:26:28,720 Speaker 1: processor and an actual Acorn computer system running on an 398 00:26:28,840 --> 00:26:32,080 Speaker 1: ARMED two processor. However, the day did eventually come around, 399 00:26:32,320 --> 00:26:35,639 Speaker 1: and that day arrived in n seven, and that is 400 00:26:35,720 --> 00:26:40,040 Speaker 1: when Acorn Computers launched the Archimedes. It was a home 401 00:26:40,080 --> 00:26:43,840 Speaker 1: computer running on an ARM two processor with a clock 402 00:26:43,920 --> 00:26:46,480 Speaker 1: speed of eight mega hurts, meaning it would send out 403 00:26:46,560 --> 00:26:50,560 Speaker 1: eight million pulses per second. I wish I could say 404 00:26:50,600 --> 00:26:55,040 Speaker 1: that the Archimedes revolutionized computing right away, but that just 405 00:26:55,200 --> 00:26:59,040 Speaker 1: wouldn't be true. The delays meant that Acorn Computers was 406 00:26:59,240 --> 00:27:04,040 Speaker 1: way behind the chief competitor, which in seven was IBM, 407 00:27:04,240 --> 00:27:08,359 Speaker 1: or rather computers running on IBM's design a k a 408 00:27:08,840 --> 00:27:13,360 Speaker 1: IBM compatibles. While Acorn Computers was working on developing its 409 00:27:13,440 --> 00:27:17,560 Speaker 1: ARM processor technology and then afterward as it sorted itself 410 00:27:17,600 --> 00:27:22,200 Speaker 1: out post acquisition from Olivetti, the computing world was consolidating 411 00:27:22,240 --> 00:27:26,560 Speaker 1: behind the IBM compatible approach. Apple's market share was already 412 00:27:26,840 --> 00:27:30,080 Speaker 1: heading forward decline. At this point. The company had released 413 00:27:30,080 --> 00:27:34,240 Speaker 1: the Macintosh computer. In four Steve Jobs had been ousted 414 00:27:34,359 --> 00:27:37,520 Speaker 1: or had left in a huff. Reports differ on this. 415 00:27:38,000 --> 00:27:41,439 Speaker 1: IBM had taken aim at dominating the office computer space 416 00:27:41,760 --> 00:27:45,719 Speaker 1: and then expanded beyond to home computing. But IBM had 417 00:27:45,720 --> 00:27:49,040 Speaker 1: also made some decisions that allowed some other manufacturers to 418 00:27:49,080 --> 00:27:53,359 Speaker 1: build machines with essentially the same components as IBM's personal 419 00:27:53,359 --> 00:27:58,560 Speaker 1: computers and licensed essentially the same operating system, allowing any 420 00:27:58,640 --> 00:28:01,720 Speaker 1: company the chance to build their version of an IBM 421 00:28:01,720 --> 00:28:05,320 Speaker 1: PC but offer it for a much more competitive price. 422 00:28:05,720 --> 00:28:09,840 Speaker 1: IBM had effectively set its own course to ultimately withdraw 423 00:28:10,000 --> 00:28:13,520 Speaker 1: from the home PC market further down the line, though 424 00:28:13,600 --> 00:28:17,480 Speaker 1: that would take several more years. The point, however, is 425 00:28:17,560 --> 00:28:21,280 Speaker 1: that the IBM design was firmly entrenched in the market. 426 00:28:21,520 --> 00:28:25,359 Speaker 1: There were tons of options for machines, and more importantly, 427 00:28:25,600 --> 00:28:29,199 Speaker 1: there was an enormous amount of software available that had 428 00:28:29,240 --> 00:28:33,520 Speaker 1: been developed specifically for the IBM design of computers. The 429 00:28:33,560 --> 00:28:37,240 Speaker 1: our Comedies, a computer with a totally different processor and 430 00:28:37,280 --> 00:28:41,480 Speaker 1: a different operating system was just getting started in this market, 431 00:28:41,560 --> 00:28:45,000 Speaker 1: and there was no enormous library of software to support that. 432 00:28:45,080 --> 00:28:48,720 Speaker 1: System sales as a result were slow. I mean, what 433 00:28:48,800 --> 00:28:51,720 Speaker 1: good is a computer if there's no software to run 434 00:28:51,880 --> 00:28:56,440 Speaker 1: on the computer. You could program your own software, but 435 00:28:56,520 --> 00:28:59,040 Speaker 1: that sort of approach tends to appeal to, you know, 436 00:28:59,120 --> 00:29:02,840 Speaker 1: a super narrow sliver of the overall computer market. So 437 00:29:02,880 --> 00:29:05,200 Speaker 1: it would take a few years for programmers to develop 438 00:29:05,240 --> 00:29:08,880 Speaker 1: software for the ARM architecture and for the Archimedes platform 439 00:29:09,040 --> 00:29:12,240 Speaker 1: to a point where it could stand as a worthy 440 00:29:12,240 --> 00:29:15,520 Speaker 1: alternative to the IBM PC. And I want to be 441 00:29:15,560 --> 00:29:18,680 Speaker 1: clear here, I'm not saying the Archimedes was a bad computer. 442 00:29:18,840 --> 00:29:21,880 Speaker 1: It wasn't. It was just that it was starting at 443 00:29:21,920 --> 00:29:24,680 Speaker 1: a point where it was at a huge disadvantage to 444 00:29:24,760 --> 00:29:29,000 Speaker 1: the IBM PC, which had an enormous head start. Meanwhile, 445 00:29:29,360 --> 00:29:31,640 Speaker 1: the R and D team with an Acorn was hard 446 00:29:31,680 --> 00:29:34,600 Speaker 1: at work at the next generation of ARM architecture, which 447 00:29:34,600 --> 00:29:38,160 Speaker 1: would be the ARM three, and man, it is so 448 00:29:38,280 --> 00:29:41,320 Speaker 1: much easier to follow this naming convention compared to some 449 00:29:41,400 --> 00:29:45,120 Speaker 1: other technologies, but don't get used to it, because before 450 00:29:45,200 --> 00:29:48,280 Speaker 1: long things are going to get confusing again. So the 451 00:29:48,480 --> 00:29:51,960 Speaker 1: ARM three saw further improvements in design, with an on 452 00:29:52,200 --> 00:29:56,560 Speaker 1: chip data and instruction CASH and a four kilobyte capacity 453 00:29:56,720 --> 00:30:00,240 Speaker 1: of that CASH Ohana a bide is eight bits. A 454 00:30:00,320 --> 00:30:05,000 Speaker 1: kilo bite is one thousand bites, or really because the 455 00:30:05,000 --> 00:30:08,120 Speaker 1: power of two properties, it's more properly one thousand, twenty 456 00:30:08,200 --> 00:30:11,400 Speaker 1: four bites. Will get more into that later. Essentially, this 457 00:30:11,480 --> 00:30:15,240 Speaker 1: meant that more instructions could load into the pipeline for 458 00:30:15,280 --> 00:30:20,000 Speaker 1: the processors simultaneously, which sped things up considerably. In addition, 459 00:30:20,040 --> 00:30:22,400 Speaker 1: the team was able to get a much faster clock speed. 460 00:30:22,680 --> 00:30:24,960 Speaker 1: The previous generation ran at eight Mega hurts, but the 461 00:30:25,080 --> 00:30:28,480 Speaker 1: ARM three hit twenty five Mega hurts. The first Acorn 462 00:30:28,480 --> 00:30:32,400 Speaker 1: computers running on ARM three technology would launch in nineteen nine. 463 00:30:33,160 --> 00:30:35,440 Speaker 1: The team also worked to build a version of ARMED 464 00:30:35,440 --> 00:30:38,560 Speaker 1: two tech that had a lower power requirement than the 465 00:30:38,720 --> 00:30:42,240 Speaker 1: standard armed to processors, and this became known as armed 466 00:30:42,240 --> 00:30:46,480 Speaker 1: to a S Little A Big S. This design was 467 00:30:46,520 --> 00:30:48,560 Speaker 1: aimed at filling a market need for companies that were 468 00:30:48,560 --> 00:30:53,040 Speaker 1: building lower cost portable and handheld devices like communication hand 469 00:30:53,120 --> 00:30:56,000 Speaker 1: sets or portable computers, and the team got as far 470 00:30:56,040 --> 00:30:59,360 Speaker 1: as developing working prototypes of the chip, but never got 471 00:30:59,400 --> 00:31:02,440 Speaker 1: to bring it to market. One thing that was working 472 00:31:02,480 --> 00:31:07,160 Speaker 1: really well, however, was the general dedication to risk based architecture. 473 00:31:07,440 --> 00:31:10,960 Speaker 1: The chips required less power than CISC based systems, and 474 00:31:11,000 --> 00:31:14,200 Speaker 1: with the right software they were incredibly powerful and efficient, 475 00:31:14,560 --> 00:31:17,640 Speaker 1: and they cost much less than the more complicated CISC 476 00:31:17,800 --> 00:31:20,880 Speaker 1: bay systems did. As a result, more companies were getting 477 00:31:20,920 --> 00:31:25,040 Speaker 1: interested in developing risk based technologies. The ARM family of 478 00:31:25,080 --> 00:31:28,360 Speaker 1: processors was a clear candidate for that model, but not 479 00:31:28,480 --> 00:31:30,880 Speaker 1: everyone was keen on the idea of relying on a 480 00:31:30,920 --> 00:31:36,000 Speaker 1: technology that belonged to a specific computer manufacturer, that being Acorn. 481 00:31:36,440 --> 00:31:40,320 Speaker 1: There was, however, a solution to this problem, and I'll 482 00:31:40,360 --> 00:31:44,080 Speaker 1: explain what it was after we return from this quick break. 483 00:31:51,800 --> 00:31:55,280 Speaker 1: Behind closed doors, a series of meetings had been pushing 484 00:31:55,280 --> 00:31:58,880 Speaker 1: the idea of breaking the ARM technology division out of 485 00:31:58,880 --> 00:32:04,240 Speaker 1: Acorn and into its own entity, its own company. Acorn 486 00:32:04,280 --> 00:32:07,240 Speaker 1: itself was part of these discussions, and the idea would 487 00:32:07,280 --> 00:32:09,640 Speaker 1: be that the ARM branch would spin off into a 488 00:32:09,640 --> 00:32:14,320 Speaker 1: new company, and that company would then develop new ARM technologies, 489 00:32:14,840 --> 00:32:17,959 Speaker 1: acting as a business to business enterprise. It would actually 490 00:32:18,480 --> 00:32:22,160 Speaker 1: fabricate the technologies as well. It would be an original 491 00:32:22,200 --> 00:32:25,160 Speaker 1: equipment manufacturer or o e M. That's a type of 492 00:32:25,200 --> 00:32:28,840 Speaker 1: company that makes products that are used as components in 493 00:32:28,920 --> 00:32:32,800 Speaker 1: products made by other companies under their own branding. The 494 00:32:33,040 --> 00:32:35,360 Speaker 1: two other companies that were part of this discussion in 495 00:32:35,400 --> 00:32:39,120 Speaker 1: addition to Acorn were v L s I Technology that 496 00:32:39,200 --> 00:32:42,600 Speaker 1: was the company that had fabricated the original ARM one processor, 497 00:32:43,120 --> 00:32:49,840 Speaker 1: and drumroll please, Apple Computers. Presumably, Apple was keen on 498 00:32:49,920 --> 00:32:53,280 Speaker 1: making use of ARM based processors, but didn't want to 499 00:32:53,280 --> 00:32:55,880 Speaker 1: put out computers that could be said to have Acorn 500 00:32:55,920 --> 00:33:01,800 Speaker 1: computing technology inside them, spending off armwood sidestep that awkward fact. However, 501 00:33:01,840 --> 00:33:05,320 Speaker 1: there is another explanation that isn't quite so, you know, petty, 502 00:33:05,400 --> 00:33:07,800 Speaker 1: and this is that Apple had taken a keen interest 503 00:33:07,880 --> 00:33:10,920 Speaker 1: in the ARM three processors in an effort to develop 504 00:33:10,960 --> 00:33:13,920 Speaker 1: computers that could go up against the IBM compatible four 505 00:33:14,080 --> 00:33:17,880 Speaker 1: D six generation, but the ARM three lacked an integrated 506 00:33:17,960 --> 00:33:22,479 Speaker 1: memory management unit or mm U, and as such, Apple 507 00:33:22,520 --> 00:33:26,240 Speaker 1: felt that the ARM processor design wasn't quite where Apple 508 00:33:26,360 --> 00:33:30,600 Speaker 1: needed it to be. However, developing a new ARM processor 509 00:33:30,800 --> 00:33:34,720 Speaker 1: with an integrated MMU was going to be expensive and 510 00:33:34,800 --> 00:33:38,320 Speaker 1: Acorn Computers just didn't have the resources to do it itself, 511 00:33:38,680 --> 00:33:42,720 Speaker 1: so it really necessitated a move to an independent spinoff 512 00:33:42,880 --> 00:33:46,200 Speaker 1: that had more support behind it. So Acorn Computers would 513 00:33:46,200 --> 00:33:49,280 Speaker 1: supply the design and engineering behind the development of the 514 00:33:49,360 --> 00:33:52,760 Speaker 1: ARM architecture, primarily in the form of a workforce of 515 00:33:52,840 --> 00:33:58,160 Speaker 1: twelve engineers v l S. I would supply the fabrication 516 00:33:58,200 --> 00:34:01,760 Speaker 1: facilities to make physical chick and Apple would supply the 517 00:34:01,920 --> 00:34:06,400 Speaker 1: cold hard cash needed to fund the whole thing. That's 518 00:34:06,480 --> 00:34:09,200 Speaker 1: oversimplifying things a little bit, but generally that's how the 519 00:34:09,280 --> 00:34:14,560 Speaker 1: arrangement worked. The new company was Advanced Risk Machines Limited 520 00:34:14,680 --> 00:34:18,160 Speaker 1: a k a. ARM Limited. The main goal for the 521 00:34:18,160 --> 00:34:22,480 Speaker 1: new company was to advance ARM microprocessors. This new company 522 00:34:22,480 --> 00:34:28,120 Speaker 1: had its fancy schmancy headquarters in a barn in Cambridge, England. 523 00:34:28,680 --> 00:34:31,080 Speaker 1: Typically with tech companies, I talk about starting out in 524 00:34:31,080 --> 00:34:34,239 Speaker 1: a garage. But with ARM it was a barn. And 525 00:34:34,320 --> 00:34:37,680 Speaker 1: so while our story started in the late nineteen seventies 526 00:34:37,719 --> 00:34:42,160 Speaker 1: with Acorn Computers, some ARM histories really point to nineteen 527 00:34:42,239 --> 00:34:45,480 Speaker 1: nine as the beginning of ARM. I think that ends 528 00:34:45,520 --> 00:34:48,480 Speaker 1: up skipping some important early work. However, that's just my 529 00:34:48,520 --> 00:34:53,040 Speaker 1: own personal opinion. Hermann Hauser of Acorn Computers slash Cambridge 530 00:34:53,080 --> 00:34:57,000 Speaker 1: Processing Unit reached out to Robin Saxby to serve as 531 00:34:57,080 --> 00:35:00,239 Speaker 1: the CEO of this new company. Saxby had come from 532 00:35:00,239 --> 00:35:03,839 Speaker 1: Motorola and it worked closely with Acorn Computers back when 533 00:35:03,920 --> 00:35:07,560 Speaker 1: the PCs the company made we're running on Motorola based chips. 534 00:35:08,040 --> 00:35:11,880 Speaker 1: The first processor this new company developed was called wait 535 00:35:11,960 --> 00:35:16,759 Speaker 1: for it, ARMS six. Wait I'm sorry, wait, hang on, 536 00:35:16,840 --> 00:35:20,879 Speaker 1: that can't be right. Six? Hang up? Wasn't the last 537 00:35:20,920 --> 00:35:24,839 Speaker 1: full processor the ARM three? What the heck happened to four? 538 00:35:24,840 --> 00:35:28,600 Speaker 1: And five? Why did we jump to six? What is 539 00:35:28,640 --> 00:35:31,439 Speaker 1: it with tech companies and the desire to leap over 540 00:35:31,680 --> 00:35:35,520 Speaker 1: entire numbers when releasing new versions of products? You know, 541 00:35:35,600 --> 00:35:39,080 Speaker 1: I I wish I had answers for these questions, but 542 00:35:39,360 --> 00:35:42,920 Speaker 1: my research didn't pull up anything definitive. Now that's not 543 00:35:43,000 --> 00:35:46,240 Speaker 1: to say there aren't answers out there. It's entirely possible 544 00:35:46,280 --> 00:35:49,000 Speaker 1: that there is, and I just missed it. But based 545 00:35:49,000 --> 00:35:52,040 Speaker 1: on what I could find, there was never any announcement 546 00:35:52,239 --> 00:35:56,719 Speaker 1: for ARMED four or ARM five as planned commercial products, 547 00:35:57,200 --> 00:36:00,399 Speaker 1: nor any record of an ARMED four or ARE five 548 00:36:00,440 --> 00:36:03,680 Speaker 1: processor being produced, either as a potential product or even 549 00:36:03,719 --> 00:36:07,520 Speaker 1: as just an internal prototype. Based on the information I 550 00:36:07,560 --> 00:36:12,280 Speaker 1: can find, the fourth generation ARM processor was in fact 551 00:36:12,760 --> 00:36:15,759 Speaker 1: the ARMS six, and the new company skipped four and 552 00:36:15,880 --> 00:36:19,560 Speaker 1: five for reasons that are beyond my ken. As it were, 553 00:36:20,320 --> 00:36:23,560 Speaker 1: one thing that definitely shaped the development of the ARM 554 00:36:23,640 --> 00:36:27,160 Speaker 1: six was an intended use for the tech within an 555 00:36:27,200 --> 00:36:33,080 Speaker 1: ambitious Apple product, the Apple Newton. Now a lot has 556 00:36:33,120 --> 00:36:36,879 Speaker 1: been said of the Newton, much of it unkind and 557 00:36:37,160 --> 00:36:41,520 Speaker 1: for arguably justifiable reasons. The Newton was meant to be 558 00:36:41,719 --> 00:36:46,279 Speaker 1: a defining example of personal digital assistance or p d 559 00:36:46,440 --> 00:36:48,840 Speaker 1: a s. In fact, the story goes that the Apple 560 00:36:48,920 --> 00:36:53,480 Speaker 1: CEO of the time, John Scully, coined the phrase personal 561 00:36:53,560 --> 00:36:58,400 Speaker 1: digital assistant to refer specifically to the Newton. It was 562 00:36:58,680 --> 00:37:02,279 Speaker 1: in many ways a incursor to the iPhone, which would 563 00:37:02,320 --> 00:37:06,440 Speaker 1: debut twenty years after the company had first started working 564 00:37:06,520 --> 00:37:09,800 Speaker 1: on the Newton. So you could say that the Newton 565 00:37:09,960 --> 00:37:12,960 Speaker 1: came out twenty years too early, and I think I 566 00:37:12,960 --> 00:37:14,960 Speaker 1: think a lot of people would agree with you. The 567 00:37:14,960 --> 00:37:18,759 Speaker 1: Newton had a tablet style form factor and it used 568 00:37:18,760 --> 00:37:22,880 Speaker 1: a touch screen input with a stylus. Apple was pushing 569 00:37:22,920 --> 00:37:26,480 Speaker 1: really hard for a device that could actually interpret handwriting. 570 00:37:26,920 --> 00:37:29,839 Speaker 1: So theoretically you would be able to write on the 571 00:37:29,840 --> 00:37:34,000 Speaker 1: tablet in normal handwriting and the Newton would interpret each 572 00:37:34,080 --> 00:37:37,640 Speaker 1: letter and capture it in text on screen. And that 573 00:37:37,760 --> 00:37:40,720 Speaker 1: was a super cool and innovative idea. And Apple really 574 00:37:40,719 --> 00:37:44,200 Speaker 1: needed a processor that could power operations without requiring too 575 00:37:44,239 --> 00:37:48,239 Speaker 1: much juice, because a handheld computing device isn't really that 576 00:37:48,360 --> 00:37:50,640 Speaker 1: useful if it can only operate for an hour or 577 00:37:50,680 --> 00:37:53,400 Speaker 1: so before it needs a recharge. Would that in mind? 578 00:37:53,760 --> 00:37:57,000 Speaker 1: The ARM six micro architecture began to take shape with 579 00:37:57,160 --> 00:37:59,959 Speaker 1: lots of decisions in the development guided by the knee 580 00:38:00,000 --> 00:38:03,160 Speaker 1: eads of the Newton. The name of the family of 581 00:38:03,239 --> 00:38:08,040 Speaker 1: ARM six microprocessors, because there were a few chips that 582 00:38:08,160 --> 00:38:12,439 Speaker 1: fell under this designation, was the ARMS six macro cell. 583 00:38:12,920 --> 00:38:15,960 Speaker 1: And I'll give a few of the changes that happened 584 00:38:16,000 --> 00:38:19,319 Speaker 1: between the ARM three generation and the ARM six. For 585 00:38:19,440 --> 00:38:23,000 Speaker 1: one thing, the process had become more precise. The ARM 586 00:38:23,040 --> 00:38:27,640 Speaker 1: three micro architecture used a one point five micron process, 587 00:38:27,680 --> 00:38:31,640 Speaker 1: whereas the ARM six shrank that down to point eight microns. 588 00:38:31,680 --> 00:38:33,759 Speaker 1: So what does that mean, Well, it means that the 589 00:38:33,800 --> 00:38:37,960 Speaker 1: individual components on the chip could be made much smaller, 590 00:38:38,320 --> 00:38:41,560 Speaker 1: which also means you could fit more components onto a 591 00:38:41,640 --> 00:38:45,720 Speaker 1: microprocessor without having to increase the size of the actual 592 00:38:45,760 --> 00:38:49,279 Speaker 1: processor chip. This falls in line with an observation that 593 00:38:49,360 --> 00:38:52,920 Speaker 1: Gordon Moore had made decades earlier, where he observed that 594 00:38:53,000 --> 00:38:57,840 Speaker 1: market influences incentivized companies to develop new ways to cram 595 00:38:58,000 --> 00:39:02,600 Speaker 1: smaller and smaller components onto a square inch of silicon wafer. 596 00:39:02,920 --> 00:39:06,000 Speaker 1: The effect of this is that the number of transistors 597 00:39:06,000 --> 00:39:09,839 Speaker 1: you could find on microprocessors would effectively double every two 598 00:39:09,920 --> 00:39:13,600 Speaker 1: years or so. Now these days we tend to reinterpret 599 00:39:13,719 --> 00:39:17,360 Speaker 1: this to say that a computer's processing power doubles every 600 00:39:17,440 --> 00:39:20,719 Speaker 1: two years or so due to Moore's law, and it's 601 00:39:20,719 --> 00:39:23,719 Speaker 1: really more of an observation, but that's a matter for 602 00:39:23,760 --> 00:39:26,400 Speaker 1: another episode. The point I really want to make is 603 00:39:26,440 --> 00:39:29,759 Speaker 1: that moving from a one point five micron process to 604 00:39:29,880 --> 00:39:33,560 Speaker 1: a point eight micron process is pretty much in line 605 00:39:33,600 --> 00:39:36,560 Speaker 1: with that observation, as the point eight micron components were 606 00:39:36,600 --> 00:39:40,480 Speaker 1: just a little over half the size of the one 607 00:39:40,520 --> 00:39:46,040 Speaker 1: point five micron version found in ARMED three microprocessors. In addition, 608 00:39:46,239 --> 00:39:49,719 Speaker 1: the ARMS six increased the address space from twenty six 609 00:39:49,719 --> 00:39:53,160 Speaker 1: bits to thirty two bits. Address space means the amount 610 00:39:53,200 --> 00:39:57,040 Speaker 1: of memory that's set aside for a particular computational component, 611 00:39:57,520 --> 00:40:00,960 Speaker 1: like a file or a connected device. Essentially, a computers 612 00:40:01,000 --> 00:40:05,480 Speaker 1: processor uses memory addresses in order to access information stored 613 00:40:05,520 --> 00:40:08,759 Speaker 1: within the computer's actual memory. It's how a processor can 614 00:40:08,800 --> 00:40:13,040 Speaker 1: pull relevant information for whatever process it needs to perform. 615 00:40:13,120 --> 00:40:16,080 Speaker 1: The term twenty six bit or thirty two bit tells 616 00:40:16,120 --> 00:40:20,359 Speaker 1: us how much memory the system can address. Now, remember 617 00:40:20,400 --> 00:40:23,759 Speaker 1: that a bit is a unit of binary information, either 618 00:40:23,840 --> 00:40:26,600 Speaker 1: a zero or a one. So each bit can have 619 00:40:26,760 --> 00:40:31,760 Speaker 1: one of two states or two values zero one, two states, 620 00:40:32,320 --> 00:40:34,720 Speaker 1: and you can have twenty six bits with the twenty 621 00:40:34,760 --> 00:40:38,840 Speaker 1: six bit system with that older are in three address space, 622 00:40:38,880 --> 00:40:41,320 Speaker 1: and that meant that you had a maximum of two 623 00:40:41,520 --> 00:40:46,960 Speaker 1: to the twenty six power number of address spaces. So 624 00:40:47,000 --> 00:40:49,120 Speaker 1: that meant you can have a maximum of two to 625 00:40:49,200 --> 00:40:52,920 Speaker 1: the twenty six power number of address spaces. That translates 626 00:40:52,960 --> 00:40:56,600 Speaker 1: to more than sixty seven million values. However, a thirty 627 00:40:56,640 --> 00:40:59,200 Speaker 1: two bit address space knocks this up to two to 628 00:40:59,320 --> 00:41:02,360 Speaker 1: the thirties second power values, and that goes up to 629 00:41:02,440 --> 00:41:06,520 Speaker 1: nearly four point three billion values. So you see how 630 00:41:06,880 --> 00:41:10,480 Speaker 1: a relatively small increase in bit size can have a 631 00:41:10,600 --> 00:41:15,839 Speaker 1: much bigger effect. It's not doubling or quadrupling, it's much 632 00:41:15,880 --> 00:41:19,040 Speaker 1: bigger than that. This meant that the ARM six could 633 00:41:19,080 --> 00:41:23,000 Speaker 1: map up to four Gibby bytes of memory. Gibby bytes. 634 00:41:23,040 --> 00:41:25,680 Speaker 1: I said that correctly. This is a peculiar measurement, right, 635 00:41:26,000 --> 00:41:29,000 Speaker 1: because I'm sure you've heard of gigabytes, but this is 636 00:41:29,200 --> 00:41:34,120 Speaker 1: gibby bytes. Gibby g I b I means too to 637 00:41:34,239 --> 00:41:38,360 Speaker 1: the power of thirties, So a gimba byte means one billion, 638 00:41:38,560 --> 00:41:43,360 Speaker 1: seventy three million, seven one eight hundred four bytes. You 639 00:41:43,400 --> 00:41:46,160 Speaker 1: can see how saying one give byte is more efficient. 640 00:41:46,680 --> 00:41:49,839 Speaker 1: Isn't that helpful? Anyway? The ARM six micro architecture can 641 00:41:49,880 --> 00:41:53,040 Speaker 1: map up to four of those bad boys in memory. 642 00:41:53,280 --> 00:41:56,080 Speaker 1: A gimme byte, in case you're curious, is equal to 643 00:41:56,120 --> 00:41:59,640 Speaker 1: about one point oh seven four gigabytes. The whole story 644 00:41:59,640 --> 00:42:04,040 Speaker 1: behind the various binary prefixes, because there's also kidby, maybe 645 00:42:04,040 --> 00:42:07,040 Speaker 1: and tabby and more. All that is really interesting, but 646 00:42:07,080 --> 00:42:10,400 Speaker 1: I'll save that for some other episode. The ARM six 647 00:42:10,680 --> 00:42:14,360 Speaker 1: was backwards compatible with the old ARM three architecture. It 648 00:42:14,440 --> 00:42:17,200 Speaker 1: had a twenty six bit mode of operation that it 649 00:42:17,239 --> 00:42:19,719 Speaker 1: could switch to instead of its thirty two bit. This 650 00:42:19,800 --> 00:42:22,920 Speaker 1: helped avoid making the older software that had been designed 651 00:42:22,920 --> 00:42:26,440 Speaker 1: for ARM three systems from going totally obsolete with the 652 00:42:26,480 --> 00:42:29,880 Speaker 1: release of the new micro architecture. It had the integrated 653 00:42:30,000 --> 00:42:34,160 Speaker 1: memory management unit that Apple wanted. It also had some 654 00:42:34,280 --> 00:42:37,239 Speaker 1: new processor instructions, but I'm not going to go too 655 00:42:37,280 --> 00:42:39,479 Speaker 1: far into the details, as I feel like it would 656 00:42:39,520 --> 00:42:42,560 Speaker 1: largely be lost and we've got a lot more to 657 00:42:42,600 --> 00:42:46,239 Speaker 1: say about ARM coming up anyway. The first Newton model 658 00:42:46,320 --> 00:42:52,280 Speaker 1: launched in with an ARM six ten risk microprocessor, and unfortunately, 659 00:42:52,560 --> 00:42:56,040 Speaker 1: it would ultimately be something of a clunker. The chief 660 00:42:56,040 --> 00:42:58,319 Speaker 1: problem with the Newton was not the fall of the 661 00:42:58,400 --> 00:43:03,400 Speaker 1: ARM processor. It was that the most anticipated feature, the 662 00:43:03,520 --> 00:43:08,760 Speaker 1: handwriting recognition capability, just wasn't very good. There were lots 663 00:43:08,760 --> 00:43:12,320 Speaker 1: of reviews that criticized the implementation of this feature, documenting 664 00:43:12,360 --> 00:43:15,600 Speaker 1: times when the system performed poorly and just got stuff wrong. 665 00:43:16,200 --> 00:43:20,080 Speaker 1: Fans of the cartoon sitcom The Simpsons might remember an 666 00:43:20,080 --> 00:43:24,000 Speaker 1: episode where they made fun of this. The school bully, 667 00:43:24,160 --> 00:43:27,280 Speaker 1: Nelson had one of his cronies take down the note 668 00:43:27,480 --> 00:43:31,759 Speaker 1: beat up Martin on his Newton, but the device interpreted 669 00:43:31,800 --> 00:43:34,799 Speaker 1: it as eat up Martha. So Nelson then grabs the 670 00:43:34,800 --> 00:43:39,360 Speaker 1: Newton and throws it at Martin, thus fulfilling the prophecy. Anyway, 671 00:43:39,480 --> 00:43:43,680 Speaker 1: the Newton had a troubled launch, which is putting it mildly, 672 00:43:44,000 --> 00:43:47,640 Speaker 1: and it would transition into a troubled life cycle. The 673 00:43:47,680 --> 00:43:50,960 Speaker 1: device failed to get a really good hold in the marketplace, 674 00:43:51,040 --> 00:43:54,359 Speaker 1: even as new versions of the hardware were released, and 675 00:43:54,440 --> 00:43:57,839 Speaker 1: upon his return to Apple, Steve Jobs would discontinue the 676 00:43:57,840 --> 00:44:03,440 Speaker 1: Newton in n eight. Meanwhile, over at ARM, CEO Robin 677 00:44:03,520 --> 00:44:06,600 Speaker 1: Saxby had a brilliant idea. He saw that depending on 678 00:44:06,760 --> 00:44:10,400 Speaker 1: being a single source of fabrication for the ARM microprocessors 679 00:44:11,040 --> 00:44:15,000 Speaker 1: is way too limiting. ARM needed more customers and would 680 00:44:15,000 --> 00:44:18,600 Speaker 1: also need to meet the production needs of those customers. 681 00:44:18,960 --> 00:44:22,040 Speaker 1: But being a small operation, this was a tough problem 682 00:44:22,080 --> 00:44:25,920 Speaker 1: to be in you couldn't easily scale up. The solution 683 00:44:26,040 --> 00:44:29,840 Speaker 1: was an interesting one. Saxby led the company toward moving 684 00:44:29,840 --> 00:44:34,400 Speaker 1: to a more intellectual property approach to micro architecture. So 685 00:44:34,600 --> 00:44:40,000 Speaker 1: rather than produce the chips themselves. ARM Limited would license 686 00:44:40,160 --> 00:44:44,759 Speaker 1: the design and instruction sets of these chips out to 687 00:44:44,960 --> 00:44:49,360 Speaker 1: other fabricators. They would become what is called a fabless 688 00:44:49,560 --> 00:44:53,359 Speaker 1: chip designer. They didn't produce the hardware themselves. The other 689 00:44:53,440 --> 00:44:58,120 Speaker 1: chip manufacturers could produce their own ARM microprocessors built on 690 00:44:58,200 --> 00:45:02,359 Speaker 1: the license designs coming from ARM itself. This move would 691 00:45:02,400 --> 00:45:06,160 Speaker 1: prove to be a game changer for the company. We're 692 00:45:06,160 --> 00:45:09,360 Speaker 1: gonna leave off there for this episode. In our next episode, 693 00:45:09,360 --> 00:45:12,160 Speaker 1: I'll pick up from that point forward, and we'll talk 694 00:45:12,200 --> 00:45:17,200 Speaker 1: about how ARM evolved over the years and cemented itself 695 00:45:17,239 --> 00:45:21,640 Speaker 1: as a huge player in the microprocessor space, as well 696 00:45:21,680 --> 00:45:25,080 Speaker 1: as talk about the acquisition by in video, at least 697 00:45:25,080 --> 00:45:28,120 Speaker 1: the proposed acquisition at the time of this recording, and 698 00:45:28,200 --> 00:45:31,640 Speaker 1: what that means. If you guys have suggestions for future 699 00:45:31,719 --> 00:45:34,080 Speaker 1: topics that I should cover on tech Stuff, reach out 700 00:45:34,080 --> 00:45:36,760 Speaker 1: to me. You can do so on Twitter. The handle 701 00:45:36,920 --> 00:45:40,040 Speaker 1: is text stuff H s W and I'll talk to 702 00:45:40,040 --> 00:45:48,640 Speaker 1: you again really soon. Text Stuff is an I Heart 703 00:45:48,719 --> 00:45:52,479 Speaker 1: Radio production. For more podcasts from my Heart Radio, visit 704 00:45:52,520 --> 00:45:55,560 Speaker 1: the I Heart Radio app, Apple Podcasts, or wherever you 705 00:45:55,640 --> 00:46:00,239 Speaker 1: listen to your favorite shows. Zero