1 00:00:07,840 --> 00:00:10,360 Speaker 1: When you save that picture of your cute dog or 2 00:00:10,440 --> 00:00:13,360 Speaker 1: cat on your computer, you expect it to be there 3 00:00:13,440 --> 00:00:15,600 Speaker 1: later when you want to show it off or send 4 00:00:15,640 --> 00:00:18,840 Speaker 1: it to your favorite podcaster. Side note, Thanks for all 5 00:00:18,880 --> 00:00:21,760 Speaker 1: the pet pictures. We love them and keep them coming. 6 00:00:22,680 --> 00:00:25,439 Speaker 1: But how can you be sure that the information you 7 00:00:25,520 --> 00:00:28,680 Speaker 1: store in your computer will be there uncorrupted when you 8 00:00:28,720 --> 00:00:33,000 Speaker 1: come back for it? Is our digital information infrastructure robust 9 00:00:33,400 --> 00:00:37,240 Speaker 1: or fragile. Since this is a physics podcast, you might 10 00:00:37,280 --> 00:00:40,080 Speaker 1: be wondering if there are physics reasons that your data 11 00:00:40,240 --> 00:00:44,519 Speaker 1: might not be safe. There are in this case. Unfortunately, 12 00:00:44,640 --> 00:00:48,400 Speaker 1: it's not aliens. I'd love if it were aliens. But 13 00:00:48,520 --> 00:00:53,519 Speaker 1: the concern is radiation from space. We know that it 14 00:00:53,520 --> 00:00:56,279 Speaker 1: can hurt our bodies and rewrite our DNA, which are 15 00:00:56,360 --> 00:01:01,280 Speaker 1: essentially biological hard drives. Can it also erupt our computers 16 00:01:01,400 --> 00:01:05,560 Speaker 1: cosmic ray bitflips? What does that mean for data centers 17 00:01:05,600 --> 00:01:09,440 Speaker 1: in space or puppies in space or pictures of cats 18 00:01:09,480 --> 00:01:14,560 Speaker 1: on Mars? Oh my? Welcome to Daniel and Kelly's Extraordinary 19 00:01:14,680 --> 00:01:15,720 Speaker 1: Digital Universe. 20 00:01:28,920 --> 00:01:32,040 Speaker 2: Hello. I'm Kelly water Smith. I study parasites and space, 21 00:01:32,120 --> 00:01:34,360 Speaker 2: and if something goes wrong with my computer, I have 22 00:01:34,520 --> 00:01:35,679 Speaker 2: no clue how to deal with it. 23 00:01:38,720 --> 00:01:42,160 Speaker 1: Hi, I'm Daniel. I'm a particle physicist and I'm technically 24 00:01:42,200 --> 00:01:45,000 Speaker 1: an experimental list but really all of my sciences on 25 00:01:45,120 --> 00:01:45,679 Speaker 1: the computer. 26 00:01:46,200 --> 00:01:48,560 Speaker 2: Now, you told me that you were thinking about going 27 00:01:48,600 --> 00:01:51,800 Speaker 2: into computer science before you decided to go into particle physics. 28 00:01:52,200 --> 00:01:54,760 Speaker 2: So are you the person that everyone goes to when 29 00:01:54,760 --> 00:01:57,920 Speaker 2: they have computer problems or now in the era of 30 00:01:58,000 --> 00:02:00,560 Speaker 2: like MacBooks, is it just like two hard for you 31 00:02:00,600 --> 00:02:03,200 Speaker 2: to troubleshoot? Or yeah? Are you? Are you the go 32 00:02:03,240 --> 00:02:03,760 Speaker 2: to person? 33 00:02:04,000 --> 00:02:07,200 Speaker 1: I'm definitely the tech support person in my family. You know, 34 00:02:07,640 --> 00:02:09,800 Speaker 1: why won't this thing load? Or I want to print 35 00:02:09,800 --> 00:02:12,440 Speaker 1: from here? Why can't I? For some reason, everybody comes 36 00:02:12,480 --> 00:02:15,880 Speaker 1: to me, even though like computer science undergraduate degrees is 37 00:02:15,919 --> 00:02:18,520 Speaker 1: mostly about like red black trees and sorting and whatever, 38 00:02:18,600 --> 00:02:21,080 Speaker 1: and you don't really get expertise in that kind of stuff. 39 00:02:21,200 --> 00:02:25,720 Speaker 1: Nothing useful, nothing useful exactly. But you have very recently 40 00:02:25,760 --> 00:02:28,440 Speaker 1: become aware of how fragile computers are. 41 00:02:28,600 --> 00:02:33,120 Speaker 2: Sure, Yes, I'm coming to you from Zex's chromebook because 42 00:02:33,120 --> 00:02:36,919 Speaker 2: I have with in one year, broken my computer screen 43 00:02:36,960 --> 00:02:39,320 Speaker 2: for the second time and it is in the shop again. 44 00:02:39,560 --> 00:02:41,720 Speaker 1: And have you managed to blame this on like weird 45 00:02:41,760 --> 00:02:42,880 Speaker 1: particles from space. 46 00:02:43,600 --> 00:02:48,079 Speaker 2: No, no, what, just a lot of clumsy members in 47 00:02:48,120 --> 00:02:51,840 Speaker 2: the Wienersmith household. Fifty percent of the time it was 48 00:02:51,880 --> 00:02:54,920 Speaker 2: my fault. The other fifty percent was my adorable children. 49 00:02:57,520 --> 00:02:59,680 Speaker 1: Well, maybe today's episode is going to give somebody over 50 00:02:59,720 --> 00:03:02,560 Speaker 1: there excuse I didn't break that computer screen, even if 51 00:03:02,560 --> 00:03:04,720 Speaker 1: my footprint is on it. It was cosmic rays. 52 00:03:05,400 --> 00:03:08,280 Speaker 2: In this case, it was their teeth marks. But yes, 53 00:03:08,360 --> 00:03:11,320 Speaker 2: maybe maybe they can blame a really amazing combination of 54 00:03:11,360 --> 00:03:13,480 Speaker 2: particle rays that came in just the right pattern. 55 00:03:14,880 --> 00:03:17,000 Speaker 1: Well, today's topic really connects with a few of my 56 00:03:17,120 --> 00:03:20,280 Speaker 1: interests because, of course I like thinking about cosmic rays. 57 00:03:20,320 --> 00:03:23,400 Speaker 1: They are particles from space, and everything I do is 58 00:03:23,639 --> 00:03:26,600 Speaker 1: on the computer. Plus, this is a concept that appears 59 00:03:26,600 --> 00:03:30,639 Speaker 1: a lot in popular science, cosmic rays flipping bits on computers, 60 00:03:30,919 --> 00:03:33,160 Speaker 1: And I thought, let's make sure everybody out there really 61 00:03:33,200 --> 00:03:36,240 Speaker 1: understands how that works and how serious a problem it is, 62 00:03:36,600 --> 00:03:40,480 Speaker 1: especially with all this talk of sending data centers into space. 63 00:03:40,880 --> 00:03:42,600 Speaker 2: Does it occur a lot in pop science? I can't 64 00:03:42,640 --> 00:03:43,680 Speaker 2: think of a single example. 65 00:03:44,160 --> 00:03:46,480 Speaker 1: Wow, maybe I read the wrong pop science. 66 00:03:47,200 --> 00:03:50,880 Speaker 2: Or maybe I read the wrong pop science. Can you 67 00:03:51,000 --> 00:04:00,360 Speaker 2: name five examples? No? I'm just kidding. Name one. I'm 68 00:04:00,400 --> 00:04:01,480 Speaker 2: winny h. 69 00:04:04,240 --> 00:04:06,120 Speaker 1: You put me on the spot, and I cannot name 70 00:04:06,160 --> 00:04:08,320 Speaker 1: a single one right now off the top of my head. 71 00:04:08,560 --> 00:04:10,840 Speaker 1: But I have the feeling like this is something that 72 00:04:10,920 --> 00:04:14,000 Speaker 1: I hear a lot in popular science, or that people 73 00:04:14,080 --> 00:04:17,440 Speaker 1: in general know a little bit about, but my benefit 74 00:04:17,440 --> 00:04:20,840 Speaker 1: from understanding, like the real physics that underlies all of 75 00:04:20,880 --> 00:04:23,080 Speaker 1: this stuff, so that in the future, when there is 76 00:04:23,080 --> 00:04:25,720 Speaker 1: an onslaught of popular science articles after we launched the 77 00:04:25,760 --> 00:04:29,400 Speaker 1: first data centers into space, everyone will be well equipped. 78 00:04:29,480 --> 00:04:31,640 Speaker 2: Well, but I mean, this is a thing that could 79 00:04:31,640 --> 00:04:35,400 Speaker 2: be important for our lives personally, So like, not only 80 00:04:35,440 --> 00:04:38,240 Speaker 2: could this be important if we put data centers in space, 81 00:04:38,279 --> 00:04:40,760 Speaker 2: but for example, one of our listeners, Simon, sent us 82 00:04:40,760 --> 00:04:43,120 Speaker 2: this article where there was a plane that was flying 83 00:04:43,160 --> 00:04:46,640 Speaker 2: and all of a sudden it dropped really fast, and 84 00:04:46,680 --> 00:04:49,400 Speaker 2: nobody died. Some people did get hurt. But the thought 85 00:04:49,480 --> 00:04:51,960 Speaker 2: was that some part of the computer on the plane 86 00:04:52,000 --> 00:04:54,000 Speaker 2: got hit by a cosmic ray. It flipped a bit, 87 00:04:54,040 --> 00:04:56,799 Speaker 2: which Daniel's going to explain, and that sort of messed 88 00:04:56,839 --> 00:04:59,240 Speaker 2: up the control of the plane and the plane dropped 89 00:04:59,279 --> 00:05:02,000 Speaker 2: really quick and some people got hurt. And so this 90 00:05:02,160 --> 00:05:04,640 Speaker 2: is like a thing that could impact our lives, and 91 00:05:04,680 --> 00:05:07,000 Speaker 2: at some point I have another space related story that 92 00:05:07,040 --> 00:05:09,280 Speaker 2: I'll tell. So you know, this isn't just a thing 93 00:05:09,320 --> 00:05:10,920 Speaker 2: that might be important in the future. It's a thing 94 00:05:10,920 --> 00:05:11,640 Speaker 2: that matters. Now. 95 00:05:11,960 --> 00:05:13,960 Speaker 1: Hey, that sounds a lot to me like a popular 96 00:05:14,000 --> 00:05:15,880 Speaker 1: science article about a cosmic ray. 97 00:05:15,920 --> 00:05:20,360 Speaker 2: Bitflip, I thought you said a sci fi thing. 98 00:05:20,440 --> 00:05:25,360 Speaker 1: Oh are you too popular? No play the tape, we said, 99 00:05:25,400 --> 00:05:30,479 Speaker 1: POPSI this is a concept that appears a lot in 100 00:05:30,600 --> 00:05:31,360 Speaker 1: popular science. 101 00:05:31,400 --> 00:05:32,800 Speaker 2: Does it occur a lot in pop science? 102 00:05:32,800 --> 00:05:39,920 Speaker 1: Pop science? Popular science? Pop science, popular science, popular science articles? Oh? 103 00:05:40,040 --> 00:05:43,480 Speaker 2: I thought you said sci fi? All right? All right, anyway, 104 00:05:43,600 --> 00:05:46,000 Speaker 2: Daniel and I aren't communicating well today, But that's all right. 105 00:05:46,000 --> 00:05:48,320 Speaker 2: We're gonna move forward. I'll give you the win on 106 00:05:48,360 --> 00:05:51,920 Speaker 2: that one, all right. So let's let's lay some groundwork here. 107 00:05:52,080 --> 00:05:54,400 Speaker 1: But before we explain how everything works. I was wondering 108 00:05:54,440 --> 00:05:58,160 Speaker 1: what people knew about how cosmic rays could mess up 109 00:05:58,160 --> 00:06:00,880 Speaker 1: those pictures of your puppies. So I went out there 110 00:06:00,920 --> 00:06:04,560 Speaker 1: to ask our volunteers their thoughts, without any googling, of course, 111 00:06:04,800 --> 00:06:09,520 Speaker 1: whether cosmic rays can corrupt computers. Here's what people had 112 00:06:09,560 --> 00:06:10,040 Speaker 1: to say. 113 00:06:10,240 --> 00:06:12,800 Speaker 3: Perhaps theris magnetosphere is protecting us from that issue. 114 00:06:12,960 --> 00:06:15,200 Speaker 1: Cosmic rays are nasty. 115 00:06:15,440 --> 00:06:19,800 Speaker 3: Of course, Cosmic rays can corrupt computer data. Cosmic rays 116 00:06:19,800 --> 00:06:22,400 Speaker 3: definitely can flip bits in computer systems. 117 00:06:22,520 --> 00:06:25,240 Speaker 1: I'm sure cosmic rays can cause problems or another. I 118 00:06:25,400 --> 00:06:28,560 Speaker 1: can with some dish drawers. I'm not sure. 119 00:06:29,000 --> 00:06:33,720 Speaker 3: Cosmic rays can absolutely corrupt computer data by flipping essential bits, 120 00:06:33,920 --> 00:06:36,479 Speaker 3: even when we build in redundancy. I think there have 121 00:06:36,600 --> 00:06:41,400 Speaker 3: even been stories of car accidents attributed to cosmic ray disruptions. 122 00:06:41,640 --> 00:06:46,680 Speaker 1: I would think so because the higher frequency spectrum could 123 00:06:46,720 --> 00:06:50,800 Speaker 1: penetrate I wouldn't imagine across the computer barriers. 124 00:06:50,920 --> 00:06:54,040 Speaker 3: Cosmic rays and computers do not get along. I think 125 00:06:54,080 --> 00:06:57,600 Speaker 3: cosmic rays messing with our computers all the time, and 126 00:06:57,640 --> 00:07:00,440 Speaker 3: that's why we have checksums and stuff. Yes, I know 127 00:07:00,480 --> 00:07:03,880 Speaker 3: for certain in space, electronics that are exposed to cosmic 128 00:07:03,920 --> 00:07:06,919 Speaker 3: rays do have a risk of that bits getting flipped. 129 00:07:07,120 --> 00:07:11,800 Speaker 1: Absolutely, cosmic rays can corrupt data. Satellites can be corrupted 130 00:07:11,800 --> 00:07:16,080 Speaker 1: by cosmic rays, so they have to be hardened and redundant. 131 00:07:16,200 --> 00:07:18,360 Speaker 1: I'm not sure about on the surface of the Earth, though, 132 00:07:18,680 --> 00:07:22,280 Speaker 1: a cosmic ray flipped a bit inside the game as 133 00:07:22,320 --> 00:07:27,320 Speaker 1: the player was playing and basically cheated for them. They 134 00:07:27,440 --> 00:07:31,240 Speaker 1: would be able to alter the bit in a data stream. 135 00:07:31,280 --> 00:07:35,040 Speaker 2: So yes, great answers. As always, I think when someone 136 00:07:35,080 --> 00:07:39,400 Speaker 2: says can X mess up y, my answer is usually yeah, 137 00:07:39,440 --> 00:07:43,120 Speaker 2: probably as like I'm negative. But I don't think I 138 00:07:43,200 --> 00:07:45,280 Speaker 2: really knew that this was something to worry about until 139 00:07:45,320 --> 00:07:48,120 Speaker 2: I was researching a city on Mars and we came 140 00:07:48,160 --> 00:07:51,520 Speaker 2: across some stories about like space radiation messing up computers. 141 00:07:52,560 --> 00:07:55,560 Speaker 1: Yeah. People think of their computers as robust, like you 142 00:07:55,560 --> 00:07:58,160 Speaker 1: put a one in memory somewhere, it's going to be there. 143 00:07:58,560 --> 00:08:01,400 Speaker 1: It's not like somethuzz if you a paper that can 144 00:08:01,440 --> 00:08:05,040 Speaker 1: get overwritten or destroyed or whatever. But in reality, these 145 00:08:05,040 --> 00:08:07,760 Speaker 1: are physical systems and they live in the universe, and 146 00:08:07,800 --> 00:08:10,480 Speaker 1: the universe is not always a friendly place to store 147 00:08:10,520 --> 00:08:11,240 Speaker 1: to your data. 148 00:08:11,360 --> 00:08:13,160 Speaker 2: Yeah, and you're not supposed to bite the screen either. 149 00:08:15,600 --> 00:08:19,120 Speaker 2: So let's start with understanding how computers store data because 150 00:08:19,160 --> 00:08:22,040 Speaker 2: I still find this like slightly magical, which which is 151 00:08:22,120 --> 00:08:25,280 Speaker 2: kind of great. So how are computers storing information? 152 00:08:25,760 --> 00:08:29,120 Speaker 1: Yeah, it's a great question, and fundamentally computer store information 153 00:08:29,240 --> 00:08:33,000 Speaker 1: using quantum mechanics. Right, you could store information on just 154 00:08:33,040 --> 00:08:35,360 Speaker 1: like a long tape, like the way a Turing machine does. 155 00:08:35,400 --> 00:08:37,679 Speaker 1: You're like writing ones and zeros and principle, it can 156 00:08:37,760 --> 00:08:40,240 Speaker 1: be anything, but you want it to be compact, so 157 00:08:40,320 --> 00:08:42,079 Speaker 1: you could have like lots and lots of ones and 158 00:08:42,160 --> 00:08:44,880 Speaker 1: zeros and not have it take up a whole room. 159 00:08:45,120 --> 00:08:46,839 Speaker 1: And you want it to be fast so you can 160 00:08:46,880 --> 00:08:49,160 Speaker 1: read it out really quickly and you're not waiting an 161 00:08:49,240 --> 00:08:52,040 Speaker 1: hour for that picture of your puppies to load. And 162 00:08:52,120 --> 00:08:55,599 Speaker 1: so we make it small and fast by using atomic systems. 163 00:08:56,080 --> 00:08:58,400 Speaker 1: And at its core, all these atomic systems are built 164 00:08:58,400 --> 00:09:01,360 Speaker 1: on semiconductors, which is why you hear about silicon so much. 165 00:09:01,400 --> 00:09:05,360 Speaker 1: Silicon is a semiconductor. Well, what does that mean semiconductors, Well, 166 00:09:05,360 --> 00:09:10,840 Speaker 1: semiconductors sit between insulators which don't conduct electricity and conductors 167 00:09:11,160 --> 00:09:14,400 Speaker 1: like metals, which conduct a lot of electricity. And understand 168 00:09:14,400 --> 00:09:16,200 Speaker 1: why that is. You have to understand something about the 169 00:09:16,240 --> 00:09:19,360 Speaker 1: atomic structure here, and we're used to thinking about atoms 170 00:09:19,400 --> 00:09:23,400 Speaker 1: by themselves. You have like energy levels, and a silicon 171 00:09:23,440 --> 00:09:25,679 Speaker 1: atom has like where electrons can be. There's like a 172 00:09:25,760 --> 00:09:29,079 Speaker 1: ladder of them, right, and they can absorb and emit photons. 173 00:09:29,120 --> 00:09:31,040 Speaker 1: This kind of stuff is like a very discrete set 174 00:09:31,080 --> 00:09:34,280 Speaker 1: of energy levels, and that's true for atoms. And if 175 00:09:34,280 --> 00:09:36,640 Speaker 1: you have, like a gas of silicon, it's going to 176 00:09:36,720 --> 00:09:39,360 Speaker 1: emit photons at certain energy levels and absorb at those 177 00:09:39,480 --> 00:09:42,240 Speaker 1: energy levels, and that's all cool. So that's atomic physics. 178 00:09:42,440 --> 00:09:45,640 Speaker 2: Is atomic physics another word for chemistry, because it kind 179 00:09:45,679 --> 00:09:48,200 Speaker 2: of sounds like it's another word for chemistry. 180 00:09:48,400 --> 00:09:51,200 Speaker 1: It's the foundation of all chemistry. Yet everything in chemistry 181 00:09:51,240 --> 00:09:53,760 Speaker 1: comes out of those energy levels. And then it gets 182 00:09:53,800 --> 00:09:56,520 Speaker 1: more complicated because you have atoms binding with each other 183 00:09:56,600 --> 00:10:00,520 Speaker 1: and sharing electrons. And if you take that too, you 184 00:10:00,559 --> 00:10:02,880 Speaker 1: get things like solids where you have a whole lattice 185 00:10:02,920 --> 00:10:05,200 Speaker 1: of atoms, and now you don't have to solve these 186 00:10:05,200 --> 00:10:07,000 Speaker 1: problems just for one electron. You have to think about 187 00:10:07,040 --> 00:10:11,000 Speaker 1: what happens to an electron that's shared among many, many atoms. 188 00:10:11,440 --> 00:10:14,480 Speaker 1: And so instead of thinking about a lattice of silicon, 189 00:10:14,520 --> 00:10:17,240 Speaker 1: atoms is like each a bunch of individual atoms with 190 00:10:17,320 --> 00:10:21,520 Speaker 1: their own energy levels. These things overlap because they're so close, 191 00:10:21,559 --> 00:10:26,280 Speaker 1: and so they spread these sharp atomic energy levels into bands. 192 00:10:26,520 --> 00:10:29,240 Speaker 1: So we talked about bands of electron energy levels. So 193 00:10:29,240 --> 00:10:31,720 Speaker 1: you shouldn't think about electron in a piece of silicon 194 00:10:31,880 --> 00:10:34,240 Speaker 1: is like this one belongs to that atom, or belongs 195 00:10:34,240 --> 00:10:36,719 Speaker 1: to this atom, or belongs to this other atom. It 196 00:10:36,840 --> 00:10:41,600 Speaker 1: moves smoothly right the ions the nuclei are bound together, 197 00:10:41,640 --> 00:10:43,520 Speaker 1: they're sort of like fixed in a lattice, and the 198 00:10:43,520 --> 00:10:47,000 Speaker 1: electrons are all moving around, and there's different energy levels there, 199 00:10:47,120 --> 00:10:49,720 Speaker 1: which roughly get grouped into two different kinds of bands. 200 00:10:50,120 --> 00:10:53,040 Speaker 1: There's the valence bands. Those are the low energy ones 201 00:10:53,040 --> 00:10:55,800 Speaker 1: where the electrons are mostly local and bound to a 202 00:10:55,840 --> 00:10:59,480 Speaker 1: specific ion. But then there's above that, there's the conduction band, 203 00:11:00,200 --> 00:11:03,720 Speaker 1: are usually empty and the electrons can flow around. So 204 00:11:04,000 --> 00:11:06,640 Speaker 1: if you have, for example, a metal, then there's a 205 00:11:06,800 --> 00:11:10,160 Speaker 1: very small gap between the valiance band and the conduction band, 206 00:11:10,520 --> 00:11:13,240 Speaker 1: and the electrons that fill up the valance band can 207 00:11:13,320 --> 00:11:15,520 Speaker 1: jump up to the conduction band like getting on the 208 00:11:15,559 --> 00:11:19,040 Speaker 1: highway really easily, and flowing all around the metal. So 209 00:11:19,080 --> 00:11:21,560 Speaker 1: if you put an electric field near a conductor, the 210 00:11:21,640 --> 00:11:24,559 Speaker 1: electrons will jump up to the conduction band and flow 211 00:11:24,600 --> 00:11:27,560 Speaker 1: all around. If you have a big gap there, like 212 00:11:27,679 --> 00:11:30,400 Speaker 1: no entrances to your freeway, then the electrons are all 213 00:11:30,440 --> 00:11:32,600 Speaker 1: stuck in their local area, and even if you put 214 00:11:32,600 --> 00:11:35,120 Speaker 1: an electric field over it, they're really not going to move. 215 00:11:35,679 --> 00:11:39,320 Speaker 1: So that's how to understand an insulator and a conductor 216 00:11:39,440 --> 00:11:42,920 Speaker 1: or something we call a metal. Usually a semiconductor like silicon, 217 00:11:43,240 --> 00:11:45,680 Speaker 1: is something where that gap is sort of like in between. 218 00:11:45,800 --> 00:11:49,120 Speaker 1: It's not really big, it's not really small. All that 219 00:11:49,280 --> 00:11:53,000 Speaker 1: silicon lithography helps you do some fancy local chemistry to 220 00:11:53,080 --> 00:11:57,800 Speaker 1: make a silicon dioxide or silicon nitride that changes its conductivity, 221 00:11:57,840 --> 00:12:00,920 Speaker 1: which means you can now separate components and dielectrics and 222 00:12:00,920 --> 00:12:03,559 Speaker 1: whatever you need to build circuits. And so you can 223 00:12:03,600 --> 00:12:08,240 Speaker 1: design really really small electrical circuits just by doping the 224 00:12:08,240 --> 00:12:09,560 Speaker 1: silicon in various ways. 225 00:12:10,000 --> 00:12:13,360 Speaker 2: Okay, so the reason you want to use silicon is 226 00:12:13,400 --> 00:12:17,280 Speaker 2: because it's helpful to have something that's in between, because 227 00:12:17,320 --> 00:12:21,520 Speaker 2: you well, so, like, why wouldn't you just want to 228 00:12:21,640 --> 00:12:25,600 Speaker 2: have a full insulator or a full conductor on your 229 00:12:25,679 --> 00:12:29,600 Speaker 2: chip as opposed to just using silicon everywhere and then 230 00:12:29,640 --> 00:12:33,400 Speaker 2: tinkering with it. Tinkering sounds possibly more complicated than just 231 00:12:33,800 --> 00:12:35,440 Speaker 2: putting conductors in some places. 232 00:12:35,840 --> 00:12:38,720 Speaker 1: Yeah, you could construct it out of just conductors wrapped 233 00:12:38,720 --> 00:12:42,280 Speaker 1: in insulators, and that's basically how you build circuits on 234 00:12:42,320 --> 00:12:45,680 Speaker 1: your bench, Right, you have like copper wires wrapped in rubber, right, 235 00:12:45,720 --> 00:12:47,680 Speaker 1: and that works, But it's really hard to do that 236 00:12:47,720 --> 00:12:51,120 Speaker 1: at a micro scale. To manufacture that stuff, and to 237 00:12:51,200 --> 00:12:54,400 Speaker 1: do it cheaply and at high volume. It's easier to 238 00:12:54,440 --> 00:12:57,160 Speaker 1: take silicon, which can be tweaked in either direction. It 239 00:12:57,200 --> 00:12:58,880 Speaker 1: turns out to be a lot easier to tweak it 240 00:12:58,960 --> 00:13:01,679 Speaker 1: than to just like build it from scratch, okay. 241 00:13:01,400 --> 00:13:03,400 Speaker 2: And so to tweak it, you just like flick some 242 00:13:03,480 --> 00:13:06,240 Speaker 2: germanium on it, and now you're good. 243 00:13:06,640 --> 00:13:09,560 Speaker 1: Yeah. Essentially we have a whole episode about how silicon 244 00:13:09,600 --> 00:13:12,120 Speaker 1: lithography works, and people should dig into the details there. 245 00:13:12,240 --> 00:13:14,679 Speaker 1: There's a lot of nuance that's missed in this summary here, 246 00:13:14,920 --> 00:13:17,360 Speaker 1: but roughly, that's why you want to start with silicon, 247 00:13:17,400 --> 00:13:20,000 Speaker 1: because you can easily change it to be a conductor 248 00:13:20,160 --> 00:13:20,920 Speaker 1: or an insulator. 249 00:13:21,120 --> 00:13:24,640 Speaker 2: I'll go check that out. My memory is just the 250 00:13:24,760 --> 00:13:27,760 Speaker 2: pits lately, Okay. So now we've got our semiconductors. 251 00:13:27,840 --> 00:13:29,400 Speaker 1: Yeah, and so you can use that to build all 252 00:13:29,400 --> 00:13:32,120 Speaker 1: sorts of microscopic bits of your circuits out of silicon. 253 00:13:32,360 --> 00:13:34,839 Speaker 1: But keep that in mind for later when we're talking 254 00:13:34,840 --> 00:13:37,320 Speaker 1: about what happens when a cosmic ray tears through a 255 00:13:37,360 --> 00:13:40,640 Speaker 1: piece of silicon. So, now you have the silicon, and 256 00:13:40,640 --> 00:13:42,120 Speaker 1: then you want to build a computer, how do you 257 00:13:42,160 --> 00:13:44,920 Speaker 1: actually use it to store memory? So your computer has 258 00:13:44,960 --> 00:13:48,480 Speaker 1: several different kinds of memory depending on your need. So 259 00:13:48,520 --> 00:13:51,320 Speaker 1: the memory you might think about are things like RAM. Right, 260 00:13:51,360 --> 00:13:54,400 Speaker 1: These are like things your computer remembers when it's turned on, 261 00:13:54,960 --> 00:13:57,440 Speaker 1: but if your turn off, the computer is empty. So 262 00:13:57,640 --> 00:14:00,520 Speaker 1: like loading a program into memory so that your computer 263 00:14:00,559 --> 00:14:02,880 Speaker 1: can run it. Right. So this is you know, random 264 00:14:02,920 --> 00:14:06,679 Speaker 1: access memory, and these days it's called d RAM dynamic 265 00:14:07,080 --> 00:14:10,360 Speaker 1: random access memory. And random access just means that you 266 00:14:10,400 --> 00:14:13,000 Speaker 1: can like get a piece of informmation from anywhere. You 267 00:14:13,000 --> 00:14:15,439 Speaker 1: don't have to like read it in order, like a book. 268 00:14:15,480 --> 00:14:17,640 Speaker 1: You can just have an index. You can say, tell 269 00:14:17,679 --> 00:14:20,520 Speaker 1: me what's stored here, tell me what's stored there. Song. 270 00:14:20,560 --> 00:14:22,480 Speaker 1: As you know the index, you can look it up. 271 00:14:22,960 --> 00:14:26,600 Speaker 2: Okay, So I've got a semiconductor which allows me to 272 00:14:26,680 --> 00:14:29,040 Speaker 2: move electrons around. 273 00:14:29,560 --> 00:14:32,560 Speaker 1: And to build circuits. And so now the question is 274 00:14:32,720 --> 00:14:35,880 Speaker 1: how do you use little electrical circuits to store information? 275 00:14:36,600 --> 00:14:40,440 Speaker 1: And so DRAM does this by using a capacitor connected 276 00:14:40,440 --> 00:14:43,520 Speaker 1: to a transistor. So a capacitor physically, if you build 277 00:14:43,600 --> 00:14:46,720 Speaker 1: a large one, is just like two surfaces insulated from 278 00:14:46,760 --> 00:14:49,400 Speaker 1: each other that are separated, and so you can have 279 00:14:49,560 --> 00:14:52,160 Speaker 1: like a charge stored across them. You can put a 280 00:14:52,160 --> 00:14:54,760 Speaker 1: bunch of electrons on one side and then you have 281 00:14:54,880 --> 00:14:58,360 Speaker 1: like a voltage across the capacitor. And so capacitor is 282 00:14:58,560 --> 00:15:01,160 Speaker 1: very useful for all sorts of things circuits. But in 283 00:15:01,200 --> 00:15:03,280 Speaker 1: this case you can just ask like, well, is there 284 00:15:03,360 --> 00:15:04,880 Speaker 1: voltage on it? If so, I'm going to call that 285 00:15:04,880 --> 00:15:07,280 Speaker 1: a one. If there's no voltage across it, I'm going 286 00:15:07,320 --> 00:15:09,560 Speaker 1: to call that a zero. You just need some sort 287 00:15:09,600 --> 00:15:12,800 Speaker 1: of microscopic state which you can control and you can 288 00:15:12,880 --> 00:15:15,160 Speaker 1: read out, and then you assign the meaning of one 289 00:15:15,440 --> 00:15:17,520 Speaker 1: to one of the states and zero to the other state. 290 00:15:17,720 --> 00:15:19,920 Speaker 2: Okay, but so what could actually be happening in the 291 00:15:19,960 --> 00:15:22,840 Speaker 2: capacitor is there's a lot of different ways that you 292 00:15:22,840 --> 00:15:25,440 Speaker 2: could have more charge on one side and less charge 293 00:15:25,480 --> 00:15:28,080 Speaker 2: on the other, but it all gets sort of summarized 294 00:15:28,120 --> 00:15:29,120 Speaker 2: down to one in zero. 295 00:15:29,400 --> 00:15:32,480 Speaker 1: Yeah, exactly, because this is digital logic. In the end, 296 00:15:32,560 --> 00:15:35,880 Speaker 1: everything is analog, like the universe is mostly analog, but 297 00:15:35,920 --> 00:15:38,840 Speaker 1: we assign digital meaning to it. And as these things 298 00:15:38,920 --> 00:15:42,520 Speaker 1: get smaller and faster than the amount of charge to 299 00:15:42,560 --> 00:15:45,600 Speaker 1: capacitor holds gets really really small, so that it's quick 300 00:15:45,600 --> 00:15:48,200 Speaker 1: to read out, it's quick to assign, and it's easy 301 00:15:48,240 --> 00:15:50,240 Speaker 1: to make small. So in this case, it's like a 302 00:15:50,240 --> 00:15:53,960 Speaker 1: few tens of thousands of electrons hold the charge. And 303 00:15:54,000 --> 00:15:56,120 Speaker 1: you can imagine more complicated systems. If you don't want 304 00:15:56,160 --> 00:15:58,800 Speaker 1: to use binary logic, you want to use trinary, you 305 00:15:58,840 --> 00:16:01,480 Speaker 1: could have like two different threashs. You could say zero 306 00:16:01,680 --> 00:16:04,400 Speaker 1: is anywhere from zero to ten thousand electrons, and a 307 00:16:04,440 --> 00:16:07,000 Speaker 1: one is ten thousand to twenty thousand electrons, and anything 308 00:16:07,000 --> 00:16:09,280 Speaker 1: above twenty thousand I'm gonna call it two if you 309 00:16:09,320 --> 00:16:12,200 Speaker 1: have trinary logic, but we use binary. So it's like 310 00:16:12,360 --> 00:16:16,000 Speaker 1: zero or somewhere above ten thousand electrons or ten to 311 00:16:16,000 --> 00:16:19,880 Speaker 1: thirty fempto coulums. And so there's a capacitor that holds 312 00:16:19,880 --> 00:16:23,000 Speaker 1: it and then a transistor which allows you access to it. 313 00:16:23,240 --> 00:16:26,720 Speaker 1: And so that's the basis of DRAM. And this requires 314 00:16:26,800 --> 00:16:29,960 Speaker 1: constant power. When you turn off the computer, that charge 315 00:16:29,960 --> 00:16:32,520 Speaker 1: dissipates and it's gone, which is why when you turn 316 00:16:32,560 --> 00:16:35,080 Speaker 1: off your computer, it doesn't remember anymore what was in 317 00:16:35,120 --> 00:16:35,720 Speaker 1: its memory. 318 00:16:36,160 --> 00:16:39,920 Speaker 2: Okay, so you've got the semiconductor, and those electrons are 319 00:16:40,120 --> 00:16:44,440 Speaker 2: jumping up into the highway and they're getting moved around 320 00:16:44,720 --> 00:16:47,320 Speaker 2: in the capacitor so that they can store up these 321 00:16:47,320 --> 00:16:49,040 Speaker 2: ones in these zeros. 322 00:16:48,840 --> 00:16:51,280 Speaker 1: And it's critical there that you have conductors, Like the 323 00:16:51,320 --> 00:16:53,960 Speaker 1: plates on the capacity are made of conductors and insulators 324 00:16:54,000 --> 00:16:57,240 Speaker 1: between the plates so that the chargers don't just flow across. 325 00:16:58,240 --> 00:17:00,080 Speaker 2: Okay, all right, so you want them to move and 326 00:17:00,080 --> 00:17:01,400 Speaker 2: then you want them to stay once you've. 327 00:17:01,240 --> 00:17:05,400 Speaker 1: Moved there, exactly, Okay, exactly, But these things don't work great. 328 00:17:05,600 --> 00:17:07,880 Speaker 1: Like there's a leakage there because these things are super 329 00:17:07,960 --> 00:17:10,919 Speaker 1: duper small and the insulator isn't perfect, and so every 330 00:17:11,040 --> 00:17:14,840 Speaker 1: sixty four milliseconds or so, it just like leaks away. 331 00:17:15,359 --> 00:17:18,840 Speaker 1: So your computer has to constantly refresh your RAM every 332 00:17:18,880 --> 00:17:22,040 Speaker 1: sixty milliseconds or so, it rewrites the number onto it. 333 00:17:22,040 --> 00:17:23,120 Speaker 1: It's not very stable. 334 00:17:23,200 --> 00:17:25,479 Speaker 2: That doesn't sound like a great system. I mean, I 335 00:17:25,480 --> 00:17:28,120 Speaker 2: mean it seems to work great. I love my computer 336 00:17:28,240 --> 00:17:29,280 Speaker 2: when it's not in the shop. 337 00:17:29,359 --> 00:17:30,960 Speaker 1: It's kind of amazing that it works as well as 338 00:17:31,000 --> 00:17:32,840 Speaker 1: it does. You don't just like call up a picture 339 00:17:33,280 --> 00:17:34,840 Speaker 1: and it's like totally corrupted. 340 00:17:35,040 --> 00:17:35,280 Speaker 2: Yeah. 341 00:17:35,359 --> 00:17:38,560 Speaker 1: Then there's another kind of memory, which is in your CPU. 342 00:17:38,640 --> 00:17:41,359 Speaker 1: So your computer basically has like a hard drive to 343 00:17:41,359 --> 00:17:44,440 Speaker 1: store information. And then there's the CPU, the central processing 344 00:17:44,520 --> 00:17:47,440 Speaker 1: unit that does all the actual calculations. Then it has 345 00:17:47,520 --> 00:17:50,080 Speaker 1: access to the main memory that we just talked about, 346 00:17:50,119 --> 00:17:52,600 Speaker 1: the RAM, But when it's doing the calculations, it slurps 347 00:17:52,600 --> 00:17:56,560 Speaker 1: stuff from the RAM into a special super fast memory 348 00:17:56,600 --> 00:17:59,400 Speaker 1: called the CASH, which is close to the CPU so 349 00:17:59,440 --> 00:18:01,000 Speaker 1: that it doesn't have to go all the way to 350 00:18:01,040 --> 00:18:03,359 Speaker 1: memory which turns out to be a little bit slow. 351 00:18:03,680 --> 00:18:06,000 Speaker 1: So you have these hierarchies of memory in your computer 352 00:18:06,400 --> 00:18:09,760 Speaker 1: which are smaller and faster or bigger and slower. So 353 00:18:09,920 --> 00:18:13,360 Speaker 1: SRAM is the fastest, smallest memory. It's like right there 354 00:18:13,440 --> 00:18:15,800 Speaker 1: next to the CPU. It's like if you're trying to 355 00:18:15,800 --> 00:18:18,040 Speaker 1: do your household budget, you don't keep all of the 356 00:18:18,080 --> 00:18:19,959 Speaker 1: information right in front of you. You have like a 357 00:18:19,960 --> 00:18:22,000 Speaker 1: sheet of paper or a spreadsheet with like the critical 358 00:18:22,040 --> 00:18:24,480 Speaker 1: details right in front of you that you're calculating right now. 359 00:18:24,880 --> 00:18:27,240 Speaker 1: The CPU's CASH is essentially like its. 360 00:18:27,119 --> 00:18:32,159 Speaker 2: Little worksheet and sm SRAM stands for is that slow 361 00:18:32,320 --> 00:18:34,240 Speaker 2: random access memory? Is that what it stands for. 362 00:18:35,920 --> 00:18:39,600 Speaker 1: It stands for a static random access memory, okay, as 363 00:18:39,600 --> 00:18:42,359 Speaker 1: opposed to dynamic, which is the kind you have for 364 00:18:42,880 --> 00:18:46,680 Speaker 1: your most of your memory. And this is cool because 365 00:18:46,720 --> 00:18:50,000 Speaker 1: it's built by a pair of logic gates usually, so 366 00:18:50,040 --> 00:18:53,520 Speaker 1: it's built using not gates. Sont gits are something where 367 00:18:53,520 --> 00:18:55,680 Speaker 1: if you take in a one, you output a zero, 368 00:18:56,000 --> 00:18:57,960 Speaker 1: if you take in a zero, you output a one. 369 00:18:58,280 --> 00:19:00,560 Speaker 1: So it's like a basic logic gate in the same 370 00:19:00,600 --> 00:19:03,960 Speaker 1: categories like and gates you know that require two ones 371 00:19:04,000 --> 00:19:06,800 Speaker 1: to output a one, or an ore gate that requires 372 00:19:07,080 --> 00:19:09,840 Speaker 1: either of the inputs to be one to output a one. 373 00:19:09,920 --> 00:19:10,720 Speaker 2: Oh yes, of course. 374 00:19:11,720 --> 00:19:14,640 Speaker 1: And these again are built on top of transistors which 375 00:19:14,680 --> 00:19:17,040 Speaker 1: are made of silicon, So you can tie a few 376 00:19:17,040 --> 00:19:19,600 Speaker 1: transistors together to make a nand gate or an and 377 00:19:19,880 --> 00:19:22,959 Speaker 1: gate or a knot gate. All these basic digital logics 378 00:19:22,960 --> 00:19:25,200 Speaker 1: are built on top of the same fundamental technology, which 379 00:19:25,200 --> 00:19:28,679 Speaker 1: are silicon transistors which have these various levels of doping 380 00:19:28,760 --> 00:19:31,439 Speaker 1: so they can do what they need to do. And 381 00:19:31,480 --> 00:19:33,840 Speaker 1: so the way SRAM works is you have two knock 382 00:19:33,840 --> 00:19:36,679 Speaker 1: gates looped on each other. It's kind of silly, but 383 00:19:36,720 --> 00:19:39,159 Speaker 1: it works really amazingly well. Like, say I have two 384 00:19:39,240 --> 00:19:41,920 Speaker 1: knock gates. If the first one has an output value 385 00:19:41,960 --> 00:19:44,639 Speaker 1: of one, and then you feed that into the second one, 386 00:19:44,760 --> 00:19:46,240 Speaker 1: the second one is going to have an output value 387 00:19:46,280 --> 00:19:48,760 Speaker 1: of zero, right, feed the output of the second one 388 00:19:48,800 --> 00:19:52,120 Speaker 1: back into the first one, and it's going to keep 389 00:19:52,160 --> 00:19:55,120 Speaker 1: the first one having an output value of one because 390 00:19:55,160 --> 00:19:57,199 Speaker 1: it's input with zero. And so it's just sort of 391 00:19:57,200 --> 00:19:59,560 Speaker 1: like if you have two of these things in series 392 00:19:59,680 --> 00:20:02,240 Speaker 1: and then looped on each other, they support each other, 393 00:20:02,840 --> 00:20:05,760 Speaker 1: and there's two stable states here. Either the first one 394 00:20:05,800 --> 00:20:07,800 Speaker 1: is one and the second one is zero, or the 395 00:20:07,800 --> 00:20:09,560 Speaker 1: first one is zero and the second one is one, 396 00:20:10,080 --> 00:20:13,120 Speaker 1: and either way they just constantly support each other, confirming 397 00:20:13,119 --> 00:20:13,520 Speaker 1: each other. 398 00:20:13,760 --> 00:20:16,000 Speaker 2: Okay, I'm following, but I feel like the big picture 399 00:20:16,040 --> 00:20:19,640 Speaker 2: here for anyone who is maybe having a little trouble following, 400 00:20:20,119 --> 00:20:22,560 Speaker 2: is that in all of these cases, what's happening is 401 00:20:22,560 --> 00:20:26,359 Speaker 2: we are very carefully keeping track of where electrons are going, 402 00:20:26,560 --> 00:20:29,280 Speaker 2: and they need to be going in particular places and 403 00:20:29,359 --> 00:20:30,080 Speaker 2: staying there. 404 00:20:30,440 --> 00:20:32,680 Speaker 1: And so you have your DRAM, which is a capacity 405 00:20:32,680 --> 00:20:34,840 Speaker 1: and a transistor. You have your s RAM, which has 406 00:20:34,880 --> 00:20:37,840 Speaker 1: this pair of knock gates which you're keeping your information. 407 00:20:38,440 --> 00:20:41,520 Speaker 1: And these are more robust than your DRAM because they 408 00:20:41,560 --> 00:20:44,320 Speaker 1: don't leak the same way. But these transistors are super 409 00:20:44,440 --> 00:20:47,320 Speaker 1: duper small and so like a little bit of tweaking 410 00:20:47,400 --> 00:20:50,040 Speaker 1: can corrupt them. And then the last kind of memory 411 00:20:50,040 --> 00:20:52,320 Speaker 1: in your computer is the hard drive. This is the 412 00:20:52,320 --> 00:20:54,240 Speaker 1: most robust, the thing where if you turn it off, 413 00:20:54,320 --> 00:20:56,440 Speaker 1: it will stay there. You can put your picture of 414 00:20:56,480 --> 00:20:58,959 Speaker 1: your puppy on your hard drive and turn your computer 415 00:20:59,000 --> 00:21:01,680 Speaker 1: off and on it's still there. And this is also 416 00:21:01,720 --> 00:21:04,200 Speaker 1: the slowest kind of memory and the biggest and these 417 00:21:04,240 --> 00:21:07,680 Speaker 1: days this uses flash technology, which essentially like an insulated 418 00:21:07,680 --> 00:21:10,440 Speaker 1: box and you try to trap some electrons in there, 419 00:21:10,840 --> 00:21:13,439 Speaker 1: or you let the electrons out, and it's insulated so 420 00:21:13,440 --> 00:21:16,879 Speaker 1: the electrons can't leave. And essentially you say, if the 421 00:21:16,920 --> 00:21:18,920 Speaker 1: electrons are trapped in there, then it's going to be 422 00:21:18,960 --> 00:21:21,040 Speaker 1: a one. If the electrons are not trapped, it's going 423 00:21:21,080 --> 00:21:22,919 Speaker 1: to be a zero. And so this is like a 424 00:21:22,960 --> 00:21:26,400 Speaker 1: really big energy barrier that's very robust and it's very 425 00:21:26,400 --> 00:21:28,600 Speaker 1: hard to pertur, but it's kind of slow, which is 426 00:21:28,600 --> 00:21:31,000 Speaker 1: why you use it for your biggest, slowest memory, like 427 00:21:31,040 --> 00:21:31,760 Speaker 1: your hard drives. 428 00:21:31,880 --> 00:21:34,720 Speaker 2: Does your computer start with all the electrons that it 429 00:21:34,760 --> 00:21:36,960 Speaker 2: needs and just move them around or does it pull 430 00:21:36,960 --> 00:21:38,960 Speaker 2: electrons out of the environment when it needs them? 431 00:21:39,600 --> 00:21:42,760 Speaker 1: Why are you Oh, great question? No, I love this 432 00:21:42,880 --> 00:21:49,160 Speaker 1: question because it touches on like, you know, electrons versus electricity, right, 433 00:21:49,720 --> 00:21:53,600 Speaker 1: and mostly electricity is the motion of electrons, So you're 434 00:21:53,600 --> 00:21:56,800 Speaker 1: not like running out of electrons in a sense. But 435 00:21:56,800 --> 00:21:58,680 Speaker 1: people confuse these two things and they like to talk 436 00:21:58,680 --> 00:22:02,080 Speaker 1: about electrons when they mean electricity. And remember that these 437 00:22:02,280 --> 00:22:05,800 Speaker 1: numbers of electrons are super due per tiny, right, Like 438 00:22:05,840 --> 00:22:08,760 Speaker 1: the number of electrons in silicon is like, you know, 439 00:22:08,840 --> 00:22:11,679 Speaker 1: ten to the twenty nine per gram or some crazy number. 440 00:22:12,200 --> 00:22:14,320 Speaker 1: So if you're talking about circuits that need like ten 441 00:22:14,359 --> 00:22:17,879 Speaker 1: thousand electrons, it's really just you're never going to run 442 00:22:17,920 --> 00:22:21,399 Speaker 1: out of electrons. So electrons are everywhere. We're all just 443 00:22:21,480 --> 00:22:24,160 Speaker 1: living in a vast ocean of electrons. It's about controlling 444 00:22:24,440 --> 00:22:27,359 Speaker 1: where they are and where they aren't and where they go. Okay, 445 00:22:27,440 --> 00:22:29,160 Speaker 1: you don't need to recharge your electrons, but it sounds 446 00:22:29,200 --> 00:22:31,719 Speaker 1: to me like a really fun scam we can sell people, 447 00:22:31,760 --> 00:22:33,920 Speaker 1: like let's top up your electrons. 448 00:22:34,640 --> 00:22:36,120 Speaker 2: All right, Well, Daniel and I are going to think 449 00:22:36,119 --> 00:22:39,320 Speaker 2: of how we can monetize electrons. See if you can 450 00:22:39,359 --> 00:22:42,280 Speaker 2: hear are commercial for it during the break, and we'll 451 00:22:42,280 --> 00:23:04,200 Speaker 2: be back in a moment. We're back. Daniel just gave 452 00:23:04,240 --> 00:23:08,040 Speaker 2: us a great explanation for how computers store information, and 453 00:23:08,080 --> 00:23:10,360 Speaker 2: now we're going to hear about how cosmic rays can 454 00:23:10,600 --> 00:23:14,280 Speaker 2: mess that all up. So, Daniel, what is a cosmic ray. 455 00:23:14,320 --> 00:23:17,119 Speaker 2: I'm sure physicists have this absolutely nailed down. 456 00:23:21,359 --> 00:23:25,960 Speaker 1: Cosmic rays are basically just particles from space. Right. Rays 457 00:23:26,119 --> 00:23:28,080 Speaker 1: is just like a generic name for a particle, or 458 00:23:28,119 --> 00:23:30,640 Speaker 1: sort of an old fashioned name, and cosmic just means 459 00:23:30,720 --> 00:23:34,200 Speaker 1: it comes from space somewhere, and the Earth is constantly 460 00:23:34,240 --> 00:23:37,680 Speaker 1: being hit by particles from space. Because space is not empty. 461 00:23:37,760 --> 00:23:39,600 Speaker 1: You might think of it as like a grand vacuum, 462 00:23:39,640 --> 00:23:42,840 Speaker 1: but really it's filled with particles, just much lower density 463 00:23:43,119 --> 00:23:46,080 Speaker 1: than here on Earth. So in our Solar system, for example, 464 00:23:46,320 --> 00:23:49,040 Speaker 1: there's a massive producer of cosmic rays in the Sun. 465 00:23:49,400 --> 00:23:52,680 Speaker 1: It pumps out protons and electrons and photons and all 466 00:23:52,720 --> 00:23:55,239 Speaker 1: sorts of stuff, and when the protons and electrons hit 467 00:23:55,280 --> 00:23:58,960 Speaker 1: the Earth, we consider those cosmic rays. And Jupiter also 468 00:23:59,040 --> 00:24:01,919 Speaker 1: puts out a lot of radio contributes to cosmic rays. 469 00:24:02,240 --> 00:24:04,639 Speaker 1: The center of our galaxy is a massive source of 470 00:24:04,720 --> 00:24:07,560 Speaker 1: radiation because there's a lot of crazy stuff going on 471 00:24:07,640 --> 00:24:10,760 Speaker 1: in there. It's hot, it's dense, it's nasty. Things around 472 00:24:10,800 --> 00:24:13,919 Speaker 1: a black hole also emit cosmic rays because there's a 473 00:24:13,920 --> 00:24:16,480 Speaker 1: lot of energy there. And so basically a lot of 474 00:24:16,520 --> 00:24:19,320 Speaker 1: things in the universe are emitting cosmic rays and they 475 00:24:19,359 --> 00:24:22,159 Speaker 1: go from fairly low energy all the way up to 476 00:24:22,280 --> 00:24:26,080 Speaker 1: like the craziest highest energy things ever seen in the universe, 477 00:24:26,200 --> 00:24:29,560 Speaker 1: which are what which are created by things we don't know. So, 478 00:24:29,640 --> 00:24:31,600 Speaker 1: for example, the large hadron collide, which is like the 479 00:24:31,600 --> 00:24:35,640 Speaker 1: pinnacle of energy the humanity's ever created. We accelerate things 480 00:24:35,640 --> 00:24:39,480 Speaker 1: to have like thirteen terra electron bolts. That's thirteen trillion 481 00:24:39,560 --> 00:24:43,080 Speaker 1: electron bolts of energy. But out in space we see 482 00:24:43,080 --> 00:24:46,200 Speaker 1: things that have a million times as much energy. Whoa 483 00:24:46,280 --> 00:24:49,960 Speaker 1: yeah wow, And we don't know what causes that. Like 484 00:24:50,040 --> 00:24:52,040 Speaker 1: we know that the center of the galaxy emits some 485 00:24:52,320 --> 00:24:55,199 Speaker 1: but not that high energy. We know supernovas creates some 486 00:24:55,359 --> 00:24:57,640 Speaker 1: but not that high energy. We know if you take 487 00:24:57,680 --> 00:25:00,879 Speaker 1: particles from supernovas whiz them around um black hole, that 488 00:25:00,920 --> 00:25:03,080 Speaker 1: gives them a little bit more energy, but still not 489 00:25:03,240 --> 00:25:06,399 Speaker 1: that high energy. So we see these particles from space 490 00:25:06,720 --> 00:25:09,239 Speaker 1: that have crazy high energy, and nothing we know in 491 00:25:09,280 --> 00:25:12,399 Speaker 1: the universe can make particles that high energy. It's like 492 00:25:12,440 --> 00:25:15,760 Speaker 1: a huge mystery. Maybe it's a glitch in the simulation. 493 00:25:16,000 --> 00:25:20,040 Speaker 1: Maybe it's alien particle physicists shooting their beams at us. 494 00:25:20,240 --> 00:25:22,239 Speaker 1: Maybe there's information there we don't know. 495 00:25:22,720 --> 00:25:25,560 Speaker 2: Okay, go ahead and mark that spot on your dKu 496 00:25:25,640 --> 00:25:28,399 Speaker 2: Bingo cards. Friends, I knew that was coming. 497 00:25:28,840 --> 00:25:33,400 Speaker 1: Aliens, Yeah, yep. And so they impact life on Earth 498 00:25:33,440 --> 00:25:36,600 Speaker 1: here because they hit us, but they hit the atmosphere, 499 00:25:36,760 --> 00:25:39,399 Speaker 1: and the atmosphere is a great shield from cosmic rays. 500 00:25:39,520 --> 00:25:41,760 Speaker 1: But it doesn't just like delete them. What happens when 501 00:25:41,760 --> 00:25:43,879 Speaker 1: a particle hits the top of the atmosphere is it 502 00:25:43,920 --> 00:25:46,879 Speaker 1: collides into the particles in the atmosphere. And so you 503 00:25:46,960 --> 00:25:49,959 Speaker 1: go from having one very energetic particle to two that 504 00:25:50,000 --> 00:25:52,119 Speaker 1: have half as much energy, to four with a quarter 505 00:25:52,160 --> 00:25:55,359 Speaker 1: as much energy, to a trillions of particles, each with 506 00:25:55,440 --> 00:25:58,679 Speaker 1: a small slice of the energy, but still enough to 507 00:25:58,720 --> 00:26:00,399 Speaker 1: make it all the way down to the star of 508 00:26:00,400 --> 00:26:03,600 Speaker 1: the Earth. And so we have a constant flux of 509 00:26:03,640 --> 00:26:05,919 Speaker 1: cosmic rays on the surface of the Earth. It's like 510 00:26:06,080 --> 00:26:08,760 Speaker 1: one hundred and fifty muons per meter squared wow per 511 00:26:08,880 --> 00:26:12,439 Speaker 1: second at sea level. So like every fingernail you have 512 00:26:13,040 --> 00:26:16,880 Speaker 1: has hundreds of muons passing through it every second. They're everywhere. 513 00:26:17,240 --> 00:26:19,560 Speaker 2: So you know what I'm wondering, do I need to 514 00:26:19,600 --> 00:26:20,159 Speaker 2: worry about that? 515 00:26:22,320 --> 00:26:25,280 Speaker 1: Well, it contributes to you know, changes in your DNA 516 00:26:25,520 --> 00:26:29,160 Speaker 1: because cosmic grays carry energy and if they hit your DNA, 517 00:26:29,840 --> 00:26:33,840 Speaker 1: they can change essentially the information stored there in great 518 00:26:33,840 --> 00:26:36,760 Speaker 1: analogy to what happens when a cosmic gray hits a computer. Right, 519 00:26:37,240 --> 00:26:40,800 Speaker 1: this is how biology stores information, but it's not totally robust. 520 00:26:41,280 --> 00:26:43,119 Speaker 1: You know, I'm not an expert in biology, but my 521 00:26:43,200 --> 00:26:45,920 Speaker 1: understanding is that cosmic ray can like either hit the 522 00:26:46,000 --> 00:26:49,440 Speaker 1: DNA directly or it can ionize water, which makes free 523 00:26:49,520 --> 00:26:52,520 Speaker 1: radicals that react with the DNA. And in the end 524 00:26:52,600 --> 00:26:55,360 Speaker 1: you can have something which was originally like a cytosine 525 00:26:55,400 --> 00:26:58,080 Speaker 1: and now behaves like thiamine, and that gets read out 526 00:26:58,080 --> 00:27:01,280 Speaker 1: differently by your molecular machine. Now you're building a different 527 00:27:01,280 --> 00:27:04,760 Speaker 1: protein and like, oops, now you have cancer, or oops, 528 00:27:04,800 --> 00:27:07,399 Speaker 1: now you have like laser rays out of your eyeballs. Right, 529 00:27:07,560 --> 00:27:09,359 Speaker 1: all sorts of mutations, And so I think this is 530 00:27:09,359 --> 00:27:12,480 Speaker 1: a big part of our evolutionary history that cosmic creates 531 00:27:12,560 --> 00:27:16,639 Speaker 1: contribute to changes in our DNA. At least, it's my 532 00:27:16,720 --> 00:27:20,000 Speaker 1: favorite explanation for how me and Katrina have an athletic. 533 00:27:19,640 --> 00:27:23,239 Speaker 2: Sun Wait, wait, when I got skin cancer, you mean 534 00:27:23,320 --> 00:27:25,920 Speaker 2: I could have gotten lasers out of my eyes because 535 00:27:25,960 --> 00:27:29,320 Speaker 2: that would have been way cooler. Anyway, sunscreen and broad 536 00:27:29,400 --> 00:27:32,360 Speaker 2: rimmed hats are your friends friends. 537 00:27:31,960 --> 00:27:35,360 Speaker 1: Yes, exactly. And in that case, I think mostly skin 538 00:27:35,440 --> 00:27:38,679 Speaker 1: cancer is caused by ultraviolet radiation and not by like 539 00:27:38,800 --> 00:27:43,080 Speaker 1: more deeply penetrating muons or neutrinos. But it's the same principle, 540 00:27:43,160 --> 00:27:46,840 Speaker 1: you know, particles in that case, photons from space corrupting 541 00:27:46,880 --> 00:27:49,280 Speaker 1: the information. And so think of these things as just 542 00:27:49,359 --> 00:27:53,119 Speaker 1: high energy particles from space to come down through the atmosphere. Okay, 543 00:27:53,240 --> 00:27:55,040 Speaker 1: so you have these particles from space that carry a 544 00:27:55,040 --> 00:27:58,600 Speaker 1: lot of energy. Mostly it's muons. They're also electrons in there. 545 00:27:58,600 --> 00:28:01,679 Speaker 1: They don't penetrate as deeply. Most the electrons are blocked 546 00:28:01,720 --> 00:28:05,040 Speaker 1: by stuff like skin or you like a small sheet 547 00:28:05,080 --> 00:28:08,440 Speaker 1: of metal. Muons have more mass and so they don't 548 00:28:08,480 --> 00:28:11,000 Speaker 1: interact as much and so they can go deeper into 549 00:28:11,080 --> 00:28:14,720 Speaker 1: an object. They are also neutrons. Neutrons are created in 550 00:28:14,760 --> 00:28:17,480 Speaker 1: these showers, and because they're neutral, they pass through a 551 00:28:17,520 --> 00:28:21,280 Speaker 1: lot of material. They're penetrating as well. Neutrons are less common. 552 00:28:21,280 --> 00:28:24,480 Speaker 1: They're like one hundred times less dense than muons, but 553 00:28:24,600 --> 00:28:26,600 Speaker 1: they have a lot more mass to them, so they 554 00:28:26,600 --> 00:28:29,080 Speaker 1: can really do a lot more damage. So you have 555 00:28:29,119 --> 00:28:31,000 Speaker 1: these particles from space, and we told you that we 556 00:28:31,080 --> 00:28:35,000 Speaker 1: build all of our digital infrastructure out of semiconductors, which 557 00:28:35,040 --> 00:28:38,600 Speaker 1: rely on like the delicate balance between a conductor and 558 00:28:38,640 --> 00:28:42,600 Speaker 1: an insulator based on how exactly the electron bands are arranged, 559 00:28:42,640 --> 00:28:46,240 Speaker 1: and essentially build everything out of these building blocks. Well, 560 00:28:46,480 --> 00:28:50,080 Speaker 1: what happens when a cosmic ray passes through a lattice 561 00:28:50,120 --> 00:28:53,440 Speaker 1: of silicon. It can smash right into one of those atoms, 562 00:28:53,560 --> 00:28:56,640 Speaker 1: breaking up the nucleus, or it can push that atom, 563 00:28:57,160 --> 00:29:00,080 Speaker 1: or it can just ionize a bunch of silicon. The 564 00:29:00,120 --> 00:29:03,040 Speaker 1: silicon atoms that it kicked off can like flow through 565 00:29:03,080 --> 00:29:06,640 Speaker 1: the lattice, ionizing it. And so essentially, if you build 566 00:29:06,800 --> 00:29:09,920 Speaker 1: your whole digital infrastructure out of this stuff, it's like 567 00:29:10,000 --> 00:29:13,000 Speaker 1: taking a sharpie and just like scrawling a ride across 568 00:29:13,040 --> 00:29:13,760 Speaker 1: the page. 569 00:29:13,960 --> 00:29:16,600 Speaker 2: Okay, so it's not like it's so I'm thinking about 570 00:29:16,640 --> 00:29:19,000 Speaker 2: it as a highway now with cars moving along it, 571 00:29:19,080 --> 00:29:22,280 Speaker 2: the cars being the electrons and the silicon being the highway. 572 00:29:22,720 --> 00:29:25,480 Speaker 2: It's not that the cars are getting moved around. It's 573 00:29:25,480 --> 00:29:28,600 Speaker 2: that you're essentially like blowing up the highway and now 574 00:29:28,920 --> 00:29:31,520 Speaker 2: the cars like they don't know where to go anymore. 575 00:29:31,840 --> 00:29:34,800 Speaker 1: Yeah, essentially you're blowing up the highway. And think about 576 00:29:34,840 --> 00:29:38,080 Speaker 1: for example, your capacitor, right, your basic unit of memory 577 00:29:38,480 --> 00:29:41,600 Speaker 1: in your computer. What's going to happen if you suddenly 578 00:29:41,600 --> 00:29:44,440 Speaker 1: deposit a whole bunch of charge, or you like tear 579 00:29:44,520 --> 00:29:46,920 Speaker 1: through that and destroy the insulator so the charge leaks out. 580 00:29:47,160 --> 00:29:48,680 Speaker 1: You're going to go from a zero to a one, 581 00:29:48,880 --> 00:29:51,360 Speaker 1: or you're going to go from a one to a zero. Right, 582 00:29:51,760 --> 00:29:54,240 Speaker 1: You've built this careful thing out of these bricks and 583 00:29:54,240 --> 00:29:55,720 Speaker 1: then you have to suddenly just tear right through it, 584 00:29:55,760 --> 00:29:58,000 Speaker 1: and you still have the machinery around it that's going 585 00:29:58,040 --> 00:30:00,640 Speaker 1: to try to interpret what's going on the one or 586 00:30:00,680 --> 00:30:04,400 Speaker 1: a zero. Right, Your transistor that's next to the conductor 587 00:30:04,400 --> 00:30:06,720 Speaker 1: is gonna be like, oh, excuse me, by the way, 588 00:30:06,800 --> 00:30:09,000 Speaker 1: this doesn't seem to be like it's behaving itself. It's 589 00:30:09,040 --> 00:30:11,720 Speaker 1: just gonna be like, is there charge zero? Is there 590 00:30:11,760 --> 00:30:12,680 Speaker 1: no charge one? 591 00:30:13,200 --> 00:30:13,960 Speaker 2: That sounds bad? 592 00:30:14,360 --> 00:30:16,840 Speaker 1: Yeah, And the same is true for your s RAM 593 00:30:16,880 --> 00:30:19,320 Speaker 1: for example. This was built out of knot gates, and 594 00:30:19,360 --> 00:30:22,960 Speaker 1: these again are built out of transistors, and transistors have 595 00:30:23,200 --> 00:30:25,680 Speaker 1: various components of silicon that are dope in various ways 596 00:30:25,680 --> 00:30:27,840 Speaker 1: so they can do their logic. If you have a 597 00:30:27,840 --> 00:30:30,200 Speaker 1: cosmic ray plow through that, it's going to deposit a 598 00:30:30,200 --> 00:30:31,960 Speaker 1: bunch of charge. It's going to change the way those 599 00:30:32,000 --> 00:30:35,640 Speaker 1: gates operate in the same way. Now, hard drives, which 600 00:30:35,680 --> 00:30:39,040 Speaker 1: are the most robust and the most difficult to change, 601 00:30:39,120 --> 00:30:43,080 Speaker 1: have that like insulated box. So this requires like stronger radiation. 602 00:30:43,520 --> 00:30:45,760 Speaker 1: But essentially you want to have like charge in there 603 00:30:45,880 --> 00:30:47,880 Speaker 1: or not charge in there. If you have a cosmic 604 00:30:47,960 --> 00:30:49,680 Speaker 1: ray that comes in and tears through it and essentially 605 00:30:49,760 --> 00:30:52,280 Speaker 1: deposits a bunch of charge, you're gonna go from a 606 00:30:52,360 --> 00:30:55,080 Speaker 1: zero to a one. Or if it breaks the containment 607 00:30:55,120 --> 00:30:57,200 Speaker 1: and lets the charge out, then you're gonna go from 608 00:30:57,200 --> 00:31:00,280 Speaker 1: a one to a zero. So it's possible for every 609 00:31:00,280 --> 00:31:03,680 Speaker 1: one of these components in your computer to be susceptible 610 00:31:03,680 --> 00:31:06,600 Speaker 1: to cause and graze, very similar to how it happens 611 00:31:06,600 --> 00:31:07,160 Speaker 1: with DNA. 612 00:31:07,640 --> 00:31:11,520 Speaker 2: Okay, so that all sounds bad, but can we get 613 00:31:11,520 --> 00:31:13,959 Speaker 2: a little bit more concrete about, Like does that mean 614 00:31:14,000 --> 00:31:16,680 Speaker 2: your computer is just gonna like one hundred percent crash 615 00:31:16,760 --> 00:31:19,720 Speaker 2: every time this happens, or like what's the range of 616 00:31:19,760 --> 00:31:20,440 Speaker 2: outcomes here? 617 00:31:21,000 --> 00:31:24,120 Speaker 1: So it depends on where it hits, right, it might 618 00:31:24,240 --> 00:31:26,320 Speaker 1: hit your computer and it just hits a place where 619 00:31:26,360 --> 00:31:28,840 Speaker 1: like nothing is being used anyway, Like you have an 620 00:31:28,840 --> 00:31:30,800 Speaker 1: empty part of your hard drive and it goes from 621 00:31:30,920 --> 00:31:33,440 Speaker 1: zero to one. Whatever it was noise, it doesn't really matter. 622 00:31:33,960 --> 00:31:35,840 Speaker 1: Or maybe you were about to store a picture of 623 00:31:35,840 --> 00:31:38,320 Speaker 1: your puppy there and so it goes from zero to 624 00:31:38,360 --> 00:31:41,440 Speaker 1: one and then you overwrite your puppy on top of it, 625 00:31:41,520 --> 00:31:44,160 Speaker 1: so it doesn't matter. Right, So those are the best 626 00:31:44,160 --> 00:31:47,480 Speaker 1: case scenarios. You've overwritten your garbage with a puppy. 627 00:31:47,760 --> 00:31:49,920 Speaker 2: Oh good, I thought you were gonna say, overrode my puppy, 628 00:31:49,920 --> 00:31:51,520 Speaker 2: and I was gonna get really upset. 629 00:31:52,520 --> 00:31:53,480 Speaker 1: No, that's a disaster. 630 00:31:53,680 --> 00:31:54,840 Speaker 2: Yeah, meal is very cute. 631 00:31:54,960 --> 00:31:57,560 Speaker 1: Yeah. So in the sort of next stage of like 632 00:31:57,800 --> 00:32:01,480 Speaker 1: problematic thing, it's like something goes wrong and it's crucial 633 00:32:01,480 --> 00:32:05,000 Speaker 1: and your computer crashes. You know, like your computer follows 634 00:32:05,000 --> 00:32:07,680 Speaker 1: some instruction in its memory and the instruction isn't what 635 00:32:07,720 --> 00:32:09,560 Speaker 1: it was supposed to be because it got changed by 636 00:32:09,600 --> 00:32:12,200 Speaker 1: a cosmic ray and it does this instead of doing that, 637 00:32:12,520 --> 00:32:14,680 Speaker 1: and it doesn't work and it gets stuck and boom, 638 00:32:14,720 --> 00:32:18,440 Speaker 1: your computer has crashed. That's actually a good outcome because 639 00:32:18,560 --> 00:32:20,680 Speaker 1: you notice it and it stops you from using it, 640 00:32:20,720 --> 00:32:23,280 Speaker 1: and then you like restart your computer and you're. 641 00:32:23,160 --> 00:32:25,680 Speaker 2: Fine, so your computer can recover. 642 00:32:25,800 --> 00:32:28,840 Speaker 1: Yeah, if your computer crashes, you just reboot it, right, 643 00:32:29,000 --> 00:32:31,160 Speaker 1: Like it happens all the time, especially if you use Windows. 644 00:32:32,080 --> 00:32:36,080 Speaker 2: Oh ouch, but isn't your silicon chip still like toast 645 00:32:36,160 --> 00:32:36,920 Speaker 2: because it got hit. 646 00:32:37,200 --> 00:32:39,320 Speaker 1: No, No, you could just refresh. It's like it will 647 00:32:39,360 --> 00:32:41,520 Speaker 1: change it from zero to one. But it's not like 648 00:32:41,560 --> 00:32:44,719 Speaker 1: it's destroyed it, right, Okay. Usually these things, you know, 649 00:32:44,760 --> 00:32:48,040 Speaker 1: are robust. It's about the charge that's stored in that circuit, 650 00:32:48,400 --> 00:32:50,800 Speaker 1: and so they can like release it or right into 651 00:32:50,840 --> 00:32:53,920 Speaker 1: it accidentally. But you restart all this stuff, You're gonna 652 00:32:53,920 --> 00:32:56,880 Speaker 1: be fine. It's not permanent damage. If it happens occasionally. 653 00:32:56,960 --> 00:32:59,320 Speaker 1: We'll talk later about what happens to electronics inside the 654 00:32:59,360 --> 00:33:01,760 Speaker 1: large hadron lighter and why that needs to be much 655 00:33:01,800 --> 00:33:04,520 Speaker 1: more radiation hard. But you know, if it happens once 656 00:33:04,600 --> 00:33:06,680 Speaker 1: or twice, you're fine. You have like enough silicon there 657 00:33:06,680 --> 00:33:07,760 Speaker 1: for these things to still work. 658 00:33:07,880 --> 00:33:10,240 Speaker 2: Okay, So then what is a bad result. 659 00:33:10,480 --> 00:33:13,479 Speaker 1: The bad result is when it happens silently, like you 660 00:33:13,520 --> 00:33:16,600 Speaker 1: have stored some data on your computer and the bit 661 00:33:16,720 --> 00:33:19,440 Speaker 1: gets flipped and you don't notice, and later you read 662 00:33:19,480 --> 00:33:21,800 Speaker 1: that in you do some analysis and you're like maybe 663 00:33:21,840 --> 00:33:24,560 Speaker 1: you're doing some science or you know, your puppy's eye 664 00:33:24,680 --> 00:33:28,600 Speaker 1: color has changed or something. And this is when it's dangerous, 665 00:33:28,600 --> 00:33:32,560 Speaker 1: when it's a silent mistake. Right. Or for example, you're 666 00:33:32,600 --> 00:33:37,080 Speaker 1: storing voting results from some precinct in Virginia and like, oops, 667 00:33:37,080 --> 00:33:40,520 Speaker 1: a bit flips and somebody now has more or fewer votes. Right, 668 00:33:40,800 --> 00:33:43,320 Speaker 1: This is when we really rely on this stuff. So 669 00:33:43,680 --> 00:33:46,520 Speaker 1: not knowing that it's happened and having it happen silently 670 00:33:46,560 --> 00:33:49,280 Speaker 1: in a way that doesn't cause any errors, that's the 671 00:33:49,280 --> 00:33:50,240 Speaker 1: worst case scenario. 672 00:33:50,480 --> 00:33:53,960 Speaker 2: Okay, this is getting scary. How often does this happen? 673 00:33:54,400 --> 00:33:56,800 Speaker 1: So there's bad news and then there's good news. Okay, 674 00:33:57,000 --> 00:34:00,280 Speaker 1: So bad news is that it's not that rare, you know, 675 00:34:00,320 --> 00:34:04,200 Speaker 1: because there are cosmic rays everywhere, So you should expect 676 00:34:04,280 --> 00:34:09,160 Speaker 1: like a few bit flips per year per gigabyte of RAM, right, 677 00:34:09,239 --> 00:34:12,080 Speaker 1: and I have like thirty gigabytes of RAM on my computer, 678 00:34:12,520 --> 00:34:15,600 Speaker 1: and so I'm expecting like dozens of times a year 679 00:34:16,200 --> 00:34:18,000 Speaker 1: a bit it's going to be flipped in my RAM. 680 00:34:18,200 --> 00:34:20,920 Speaker 1: Oh yikes, exactly, And so that seems like an issue. 681 00:34:21,200 --> 00:34:23,920 Speaker 1: But of course the nerds know this and they have 682 00:34:23,920 --> 00:34:26,520 Speaker 1: built protection for us. And sound like we're just running 683 00:34:26,520 --> 00:34:29,279 Speaker 1: out there totally exposed. If you have a computer in 684 00:34:29,320 --> 00:34:32,080 Speaker 1: an environment where you absolutely need it to be robust, 685 00:34:32,239 --> 00:34:35,319 Speaker 1: then there are clever systems to detect these errors and 686 00:34:35,520 --> 00:34:36,200 Speaker 1: correct them. 687 00:34:36,600 --> 00:34:39,440 Speaker 2: I find myself wondering, was there an era when we 688 00:34:39,560 --> 00:34:43,399 Speaker 2: had computers but we didn't know this and there were 689 00:34:43,440 --> 00:34:45,319 Speaker 2: like disasters because we didn't know this. 690 00:34:45,920 --> 00:34:48,680 Speaker 1: Yeah. One of the ways this was originally discovered is 691 00:34:49,239 --> 00:34:54,480 Speaker 1: that electronics near nuclear testing in the fifties had anomalies. 692 00:34:54,480 --> 00:34:56,640 Speaker 1: People were like, whoa, what's going on with our electronics? 693 00:34:57,040 --> 00:34:59,439 Speaker 1: And then they realized, of course that the radiation from 694 00:34:59,440 --> 00:35:03,040 Speaker 1: the blast was interfering with the electronics. It was like 695 00:35:03,120 --> 00:35:06,480 Speaker 1: in the seventies that the detailed physics involved was understood 696 00:35:06,520 --> 00:35:09,800 Speaker 1: and like demonstrated in particle beams, and so it was 697 00:35:09,840 --> 00:35:12,879 Speaker 1: really like in the eighties people understood what's going on here, 698 00:35:12,960 --> 00:35:14,080 Speaker 1: and can we build protection? 699 00:35:14,480 --> 00:35:17,080 Speaker 2: Oh wow? Okay, well I'm glad we've got protection now, 700 00:35:17,320 --> 00:35:19,279 Speaker 2: all right, So tell me how the protection works, please. 701 00:35:19,320 --> 00:35:21,760 Speaker 1: So the simplest and the most expensive way to protect 702 00:35:21,800 --> 00:35:24,880 Speaker 1: your data is just to have copies of it, right, Like, 703 00:35:24,880 --> 00:35:27,160 Speaker 1: if you have really sensitive data stored on a hard drive. 704 00:35:27,360 --> 00:35:29,520 Speaker 1: Hard drives used to be much less stable. You could 705 00:35:29,520 --> 00:35:31,640 Speaker 1: just like lose a drive, like the head dropped on it. 706 00:35:31,680 --> 00:35:34,440 Speaker 1: Back when there were magnetic spinning discs, and so we 707 00:35:34,480 --> 00:35:36,480 Speaker 1: had these things called like raidar rays where you just 708 00:35:36,520 --> 00:35:39,160 Speaker 1: had like everything was stored in duplicate or in triplicate 709 00:35:39,280 --> 00:35:42,319 Speaker 1: even and in triplicate is cool because if one of 710 00:35:42,320 --> 00:35:45,279 Speaker 1: them gets corrupted, you have two other copies, and so 711 00:35:45,320 --> 00:35:47,080 Speaker 1: you can vote. You're like, is this bit supposed to 712 00:35:47,120 --> 00:35:49,400 Speaker 1: be a zero or a one? And if two of 713 00:35:49,440 --> 00:35:51,920 Speaker 1: them agree, then you know, then that's the way you go. 714 00:35:52,040 --> 00:35:54,000 Speaker 1: I think that's how like the flight systems work on 715 00:35:54,040 --> 00:35:57,200 Speaker 1: the Space Shuttle, right, they all vote, the three independent 716 00:35:57,239 --> 00:36:00,120 Speaker 1: systems or something like that, and they all vote, so 717 00:36:00,160 --> 00:36:02,520 Speaker 1: the you always have two the degree. So that's the 718 00:36:02,560 --> 00:36:05,600 Speaker 1: most complicated and expensive way. And that's quite expensive because 719 00:36:05,640 --> 00:36:08,600 Speaker 1: now you're like tripling the cost of your memory. If 720 00:36:08,640 --> 00:36:10,319 Speaker 1: you want to store a gigabyte, you really have to 721 00:36:10,360 --> 00:36:13,239 Speaker 1: have three gigabytes of memory in there just to make 722 00:36:13,280 --> 00:36:16,799 Speaker 1: it protected, Yanks. And so since that's so expensive, people 723 00:36:16,880 --> 00:36:20,920 Speaker 1: came up with cheaper, faster, more clever ways, things like checksums. Right, 724 00:36:20,960 --> 00:36:23,160 Speaker 1: Like if you send a message across the internet and 725 00:36:23,200 --> 00:36:25,680 Speaker 1: you want to make sure the message arrived uncorrupted, then 726 00:36:25,719 --> 00:36:28,239 Speaker 1: none of it got like garbled along the way. You 727 00:36:28,280 --> 00:36:30,840 Speaker 1: can also send a checksum, which is like the output 728 00:36:31,040 --> 00:36:34,040 Speaker 1: of a little calculation you do on the data. As 729 00:36:34,040 --> 00:36:35,799 Speaker 1: simple way to think of a checksums to be like 730 00:36:35,960 --> 00:36:37,919 Speaker 1: take the whole data set and treat it like a number. 731 00:36:38,080 --> 00:36:41,000 Speaker 1: Was it even or odd? And that way, if I 732 00:36:41,040 --> 00:36:43,319 Speaker 1: send you a message, and I also send you by 733 00:36:43,320 --> 00:36:44,920 Speaker 1: the way it should have been even, then you can 734 00:36:45,000 --> 00:36:47,120 Speaker 1: check to make sure it's even. That doesn't catch all 735 00:36:47,200 --> 00:36:50,440 Speaker 1: possible errors, like if you flip two bits, for example, 736 00:36:50,680 --> 00:36:52,600 Speaker 1: But it's one simple way to say, like were there 737 00:36:52,640 --> 00:36:56,839 Speaker 1: any mistakes? So in modern computers they do something which 738 00:36:56,880 --> 00:36:59,120 Speaker 1: is more sophisticated than just a simple like is it 739 00:36:59,160 --> 00:37:02,600 Speaker 1: even or odd? They have these parody checkers that take 740 00:37:02,719 --> 00:37:05,600 Speaker 1: like groups of four data words. A word is like 741 00:37:05,640 --> 00:37:08,479 Speaker 1: a bunch of bits altogether. And you have parity bits 742 00:37:08,480 --> 00:37:10,719 Speaker 1: which check how many events there are and how many 743 00:37:10,719 --> 00:37:12,960 Speaker 1: odds there are. And if you overlap these things in 744 00:37:13,000 --> 00:37:16,000 Speaker 1: a clever way, and you have this voting system, you 745 00:37:16,000 --> 00:37:18,799 Speaker 1: can have like one parody bit checks three words and 746 00:37:18,840 --> 00:37:22,399 Speaker 1: another parody bit checks three other words. And so through 747 00:37:22,440 --> 00:37:25,640 Speaker 1: some clever algorithms and clever math, you can detect any 748 00:37:25,680 --> 00:37:30,480 Speaker 1: bitflip in any of those words without having complete duplicates. 749 00:37:30,920 --> 00:37:32,799 Speaker 1: So this is like ten percent of the data cost. 750 00:37:32,880 --> 00:37:35,440 Speaker 1: You have a gigabyte of memory, you build a gigabyte 751 00:37:35,520 --> 00:37:38,600 Speaker 1: plus ten percent, and you have enough data storage to 752 00:37:38,680 --> 00:37:42,359 Speaker 1: detect if there were any bits flipped in your core gigabyte. 753 00:37:42,400 --> 00:37:45,399 Speaker 2: But we didn't have this clever and cheap version during 754 00:37:45,440 --> 00:37:48,360 Speaker 2: the Shuttle era, and that's why the Shuttle was so expensive. 755 00:37:50,480 --> 00:37:53,359 Speaker 1: I don't know if the memory costs contributed a lot 756 00:37:53,480 --> 00:37:57,480 Speaker 1: to the Shuttle. And many computer systems have this already. 757 00:37:58,200 --> 00:38:02,000 Speaker 1: So for example, your hard drive very likely has error 758 00:38:02,040 --> 00:38:05,480 Speaker 1: correction on it. Your RAM probably has error correction on it. 759 00:38:05,760 --> 00:38:08,920 Speaker 1: CPU caches usually do not, because this is the s 760 00:38:09,000 --> 00:38:11,400 Speaker 1: RAM stuff because it has to be super duper fast. 761 00:38:11,400 --> 00:38:13,320 Speaker 1: It really will slow down your computer if you have 762 00:38:13,480 --> 00:38:16,279 Speaker 1: error correction on that CPU, like the worksheet that the 763 00:38:16,320 --> 00:38:20,120 Speaker 1: CPU is using. But if you have a really important computers, 764 00:38:20,200 --> 00:38:22,640 Speaker 1: like you're operating a data center and you need to 765 00:38:22,640 --> 00:38:26,760 Speaker 1: provide really robust calculations for your consumers, or you're running 766 00:38:26,760 --> 00:38:29,760 Speaker 1: something in space, then you need to have error correction 767 00:38:30,160 --> 00:38:33,000 Speaker 1: on your caches as well. Although I was reading that 768 00:38:33,120 --> 00:38:37,960 Speaker 1: often in space applications they take consumer PCs and they 769 00:38:38,040 --> 00:38:40,840 Speaker 1: just turn off the CPU cache because they'd rather be 770 00:38:41,000 --> 00:38:44,000 Speaker 1: robust than fast, and so they're like, this part is 771 00:38:44,040 --> 00:38:46,160 Speaker 1: not robust in space, so let's just turn it off. 772 00:38:46,640 --> 00:38:48,160 Speaker 2: All right, Well, let's take a break, and when we 773 00:38:48,200 --> 00:38:50,919 Speaker 2: get back, Kelly will tell a fun story about how 774 00:38:51,040 --> 00:39:13,120 Speaker 2: galactic cosmic rays messed up computers in space. All right, 775 00:39:13,400 --> 00:39:17,400 Speaker 2: we're back. We're on the International Space Station and an 776 00:39:17,440 --> 00:39:22,640 Speaker 2: alarm has just gone off. Samantha Christafaretti, who's an Italian astronaut, 777 00:39:23,320 --> 00:39:26,080 Speaker 2: was amazing at her ability to memorize all of the 778 00:39:26,239 --> 00:39:28,239 Speaker 2: different alarms on the station. And even though this is 779 00:39:28,280 --> 00:39:30,520 Speaker 2: the first time she's ever heard this alarm, she immediately 780 00:39:30,560 --> 00:39:33,960 Speaker 2: knows what it is and she yells ammonia leak. And 781 00:39:34,080 --> 00:39:37,799 Speaker 2: so she and Terry Vertz have memorized the protocol for 782 00:39:37,880 --> 00:39:41,080 Speaker 2: ammonia leaks. And so what they do is they run 783 00:39:41,160 --> 00:39:44,360 Speaker 2: over towards the Russian segment and they close the airlock 784 00:39:44,440 --> 00:39:47,120 Speaker 2: for the American side. And so what happens if there's 785 00:39:47,160 --> 00:39:51,239 Speaker 2: an ammonia leak is that the coolant system for keeping 786 00:39:51,239 --> 00:39:54,840 Speaker 2: the American side of the International Space Station is leaking. 787 00:39:54,920 --> 00:39:58,120 Speaker 2: So ammonia is toxic and so they can't stay over 788 00:39:58,160 --> 00:40:00,799 Speaker 2: there anymore, So they close off their side so that 789 00:40:00,840 --> 00:40:04,000 Speaker 2: the toxic stuff can't get out. Then they're supposed to 790 00:40:04,000 --> 00:40:07,319 Speaker 2: get naked. Then they're supposed to go through to the 791 00:40:07,400 --> 00:40:11,240 Speaker 2: Russian side, and then they're supposed to close the hatch 792 00:40:11,440 --> 00:40:14,400 Speaker 2: and hope that the Russians have extra undies. And I 793 00:40:14,440 --> 00:40:18,280 Speaker 2: saw Daniel look confused, and so, Daniel, are you wondering 794 00:40:19,560 --> 00:40:21,160 Speaker 2: why they closed to the second hatch. 795 00:40:21,360 --> 00:40:24,359 Speaker 1: Why did they close the second hatch, Kelly, just to. 796 00:40:24,280 --> 00:40:25,719 Speaker 2: Make sure they didn't get any more moonia on the 797 00:40:25,760 --> 00:40:28,239 Speaker 2: other side. But you're probably wondering actually why they were 798 00:40:28,239 --> 00:40:29,160 Speaker 2: supposed to get naked. 799 00:40:29,040 --> 00:40:32,040 Speaker 1: Right, yes, exactly. I was having fun imagining it though. 800 00:40:32,400 --> 00:40:35,920 Speaker 2: Okay, well, they apparently were not having fun imagining what 801 00:40:35,960 --> 00:40:38,600 Speaker 2: it would be like because they skipped that step. But 802 00:40:38,640 --> 00:40:40,440 Speaker 2: they were supposed to take the clothes off because the 803 00:40:40,480 --> 00:40:44,160 Speaker 2: concern was ammonia might stick to their clothes and they 804 00:40:44,200 --> 00:40:47,400 Speaker 2: might end up bringing ammonia over to the Russian side. 805 00:40:47,880 --> 00:40:51,320 Speaker 2: And the Russians use Glai call to cool their side 806 00:40:51,320 --> 00:40:53,440 Speaker 2: of the International Space Station so they don't have this 807 00:40:53,560 --> 00:40:54,400 Speaker 2: toxic stuff. 808 00:40:54,480 --> 00:40:56,960 Speaker 1: So ammonia sticks to close but not to like skin. 809 00:40:57,280 --> 00:40:57,600 Speaker 1: I mean it. 810 00:40:57,640 --> 00:41:00,600 Speaker 2: Probably it could stick to everything, but the goal is 811 00:41:00,680 --> 00:41:03,600 Speaker 2: to bring as little as possible over to the Russian side. 812 00:41:03,680 --> 00:41:05,880 Speaker 1: Do you ever think the engineers are just like, let's 813 00:41:05,880 --> 00:41:07,799 Speaker 1: add getting naked to this protocol and see if the 814 00:41:07,800 --> 00:41:08,880 Speaker 1: astronauts will really do it. 815 00:41:08,960 --> 00:41:12,319 Speaker 2: Yeah, well, so you know, you've got to imagine that 816 00:41:12,360 --> 00:41:14,680 Speaker 2: at some point they were like, do I really want 817 00:41:14,719 --> 00:41:19,000 Speaker 2: to see my middle aged colleague naked and zero gravity? 818 00:41:18,719 --> 00:41:22,279 Speaker 2: And I think, and they must have both decided no, 819 00:41:22,880 --> 00:41:27,240 Speaker 2: And so they went over to the Russian side fully dressed. 820 00:41:27,280 --> 00:41:30,240 Speaker 2: They locked the second airlock, and then they got permission 821 00:41:30,280 --> 00:41:33,480 Speaker 2: to go back, and so they go back, and the 822 00:41:33,560 --> 00:41:36,360 Speaker 2: alarm goes off again, and they go through the whole 823 00:41:36,360 --> 00:41:38,759 Speaker 2: process the second time, and yet again they decide to 824 00:41:38,800 --> 00:41:41,680 Speaker 2: not take their clothes off. And I believe they ended 825 00:41:41,719 --> 00:41:43,560 Speaker 2: up saying they decided to not take their clothes off 826 00:41:43,560 --> 00:41:46,600 Speaker 2: because they couldn't smell ammonia. Ammonia has this like distinct 827 00:41:46,640 --> 00:41:49,480 Speaker 2: rotten egg smell. And it turned out that they were right. 828 00:41:49,560 --> 00:41:50,880 Speaker 2: It was a false alarm. 829 00:41:51,320 --> 00:41:53,040 Speaker 1: Was this the universe trying to get them naked? Is 830 00:41:53,040 --> 00:41:53,760 Speaker 1: that what was happening? 831 00:41:53,880 --> 00:41:57,000 Speaker 2: It could be, could be, But the best guess that 832 00:41:57,040 --> 00:42:00,080 Speaker 2: folks came up with afterwards for what happened is that 833 00:42:00,160 --> 00:42:05,040 Speaker 2: galactic cosmic radiation had hit their computers and had inadvertently 834 00:42:05,080 --> 00:42:08,040 Speaker 2: turned on the ammonia leak alarm, and there wasn't actually 835 00:42:08,080 --> 00:42:09,920 Speaker 2: an ammonia leak. They never were able to find an 836 00:42:09,920 --> 00:42:12,640 Speaker 2: ammonia leak, but it, you know, they started doing this 837 00:42:12,719 --> 00:42:15,200 Speaker 2: emergency protocol and they had to eventually do it twice 838 00:42:15,239 --> 00:42:18,279 Speaker 2: before they were able to fix their computers. Well, so 839 00:42:18,320 --> 00:42:20,680 Speaker 2: we're sort of talking about these data centers today, and 840 00:42:20,680 --> 00:42:22,160 Speaker 2: I guess one thing I want to point out is 841 00:42:22,160 --> 00:42:25,680 Speaker 2: that even though the International Space Station is to a 842 00:42:25,760 --> 00:42:30,279 Speaker 2: large extent protected by Earth's strong magnetosphere, which sends a 843 00:42:30,320 --> 00:42:33,480 Speaker 2: lot of these charge particles, you know, to the poles 844 00:42:33,520 --> 00:42:36,200 Speaker 2: and protects the International Space Station and the equipment and 845 00:42:36,239 --> 00:42:39,600 Speaker 2: people within it, you still get hit sometimes, and so 846 00:42:39,640 --> 00:42:41,520 Speaker 2: you still have to worry about the equipment. Where are 847 00:42:41,560 --> 00:42:44,440 Speaker 2: people thinking of putting these data centers? Are they within 848 00:42:44,520 --> 00:42:46,799 Speaker 2: the protection of the magnetic field or out past it? 849 00:42:46,960 --> 00:42:49,240 Speaker 1: I don't think they want to go very far because 850 00:42:49,239 --> 00:42:50,880 Speaker 1: the further you go, the harder it is to access 851 00:42:50,920 --> 00:42:53,440 Speaker 1: and to repair. And I'm not on record is thinking 852 00:42:53,440 --> 00:42:56,120 Speaker 1: that any of this is a good idea, and you know, 853 00:42:56,160 --> 00:42:58,719 Speaker 1: the cooling is are the radiation is tricky because you 854 00:42:58,719 --> 00:43:00,680 Speaker 1: don't have the benefit of the Earth's sphere, even if 855 00:43:00,719 --> 00:43:03,520 Speaker 1: you still have some of the magnetosphere. And so I 856 00:43:03,520 --> 00:43:07,160 Speaker 1: think it's a crazy idea pushed by tech bros who 857 00:43:07,160 --> 00:43:10,120 Speaker 1: think it's gonna be fun and sci fi. And you know, 858 00:43:10,200 --> 00:43:11,800 Speaker 1: I don't know about that. We'll have a whole another 859 00:43:11,800 --> 00:43:15,280 Speaker 1: episode digging into the economics of data centers in space. 860 00:43:15,680 --> 00:43:18,480 Speaker 1: But yeah, out there in space there is more radiation, 861 00:43:19,040 --> 00:43:22,920 Speaker 1: and so you need more protection and you need like 862 00:43:23,200 --> 00:43:25,560 Speaker 1: shielding for these things, or you need to make extra 863 00:43:25,680 --> 00:43:28,640 Speaker 1: error correction, or you need to use technologies that are 864 00:43:28,640 --> 00:43:33,320 Speaker 1: more radiation hard it actually on Earth. Sometimes I'm frustrated 865 00:43:33,320 --> 00:43:36,280 Speaker 1: by the amount of error correction. There are scenarios where 866 00:43:36,600 --> 00:43:39,240 Speaker 1: it would be better for physics, even if it's worse 867 00:43:39,400 --> 00:43:42,919 Speaker 1: for computers and for Amazon or whatever, to have less 868 00:43:43,040 --> 00:43:46,239 Speaker 1: error correction. Because all of these things are evidence of 869 00:43:46,280 --> 00:43:49,160 Speaker 1: cosmic rays. And if you're curious about cosmic rays, where 870 00:43:49,160 --> 00:43:51,279 Speaker 1: do they hit, what's going on? Then one of the 871 00:43:51,280 --> 00:43:54,800 Speaker 1: biggest challenge is seeing them. The really high energy cosmic 872 00:43:54,800 --> 00:43:58,040 Speaker 1: grays are really rare. It's like one per square kilometer 873 00:43:58,120 --> 00:44:01,560 Speaker 1: per century, either in a lot of square kilometers or 874 00:44:01,600 --> 00:44:04,440 Speaker 1: a lot of centuries. And these massive data centers that 875 00:44:04,480 --> 00:44:07,640 Speaker 1: are being built out are basically huge cosmic ray detectors. 876 00:44:08,000 --> 00:44:10,200 Speaker 1: And imagine if you could use all of Google and 877 00:44:10,239 --> 00:44:13,799 Speaker 1: Amazon's data centers to observe cosmic rays from space. It 878 00:44:13,800 --> 00:44:16,440 Speaker 1: would be incredible, right. I actually looked into this and 879 00:44:16,480 --> 00:44:18,439 Speaker 1: connected with folks at Google to see if we could 880 00:44:18,520 --> 00:44:21,319 Speaker 1: use it. But their air correction is so good they 881 00:44:21,320 --> 00:44:23,200 Speaker 1: don't even store it. They don't log it, and be like, 882 00:44:23,239 --> 00:44:24,560 Speaker 1: by the way, we flipped a bit here and we 883 00:44:24,640 --> 00:44:26,879 Speaker 1: corrected it, They just correct it and move on. 884 00:44:28,120 --> 00:44:30,120 Speaker 2: How hard would it be for them to collect that data? 885 00:44:30,440 --> 00:44:32,920 Speaker 1: Oh? Man, it would require a change in pretty low 886 00:44:33,000 --> 00:44:36,719 Speaker 1: level operating system stuff, unfortunately. And you know, for a 887 00:44:36,760 --> 00:44:38,480 Speaker 1: while they were going to be willing to share the 888 00:44:38,560 --> 00:44:40,719 Speaker 1: data with us if it was like something they already had, 889 00:44:40,840 --> 00:44:42,600 Speaker 1: they could just ship to us and we could analyze it. 890 00:44:42,640 --> 00:44:44,960 Speaker 1: And that was so excited. But then it turns out 891 00:44:45,120 --> 00:44:47,719 Speaker 1: it's not storage and getting them to do anything just 892 00:44:47,760 --> 00:44:50,120 Speaker 1: for fundamental physics is frustratingly difficult. 893 00:44:50,600 --> 00:44:52,320 Speaker 2: Google. I thought you were trying to not be evil 894 00:44:52,400 --> 00:44:54,000 Speaker 2: or did they like do away with that a long 895 00:44:54,040 --> 00:44:54,439 Speaker 2: time ago. 896 00:44:54,719 --> 00:44:56,880 Speaker 1: Yeah, I don't know. I mean, it's not their obligation 897 00:44:56,920 --> 00:44:59,839 Speaker 1: to fund basic research. But it's frustrating to me because 898 00:44:59,880 --> 00:45:03,399 Speaker 1: it means that these cosmic grays are there and they're 899 00:45:03,560 --> 00:45:07,160 Speaker 1: actually observed by our technology, and then that information is 900 00:45:07,600 --> 00:45:11,480 Speaker 1: just thrown away. You know, one person's error is another 901 00:45:11,480 --> 00:45:16,920 Speaker 1: person's gold maybe Nobel Prize gold. Right, So that's frustrating. 902 00:45:17,200 --> 00:45:19,800 Speaker 1: But there are other times when we think cosmic grays 903 00:45:19,920 --> 00:45:22,920 Speaker 1: could have affected things on Earth. The voting example I 904 00:45:22,960 --> 00:45:25,960 Speaker 1: mentioned earlier was not a hypothetical. There was an election 905 00:45:26,080 --> 00:45:29,480 Speaker 1: in two thousand and three in Brussels where one candidate 906 00:45:29,520 --> 00:45:32,040 Speaker 1: had a number of votes that was suspiciously off by 907 00:45:32,200 --> 00:45:35,600 Speaker 1: four ninety six. They're like, that's a weird number, Like, 908 00:45:35,640 --> 00:45:38,200 Speaker 1: how did this number just appear in the computers? It's 909 00:45:38,239 --> 00:45:41,879 Speaker 1: off by this specific power of two, right, fenty ninety 910 00:45:41,920 --> 00:45:43,880 Speaker 1: six is the power of two, So it could be 911 00:45:43,920 --> 00:45:46,759 Speaker 1: caused by a single bitflip. And you know they went 912 00:45:46,760 --> 00:45:48,960 Speaker 1: through and nobody tinkered with it. There's no evidence that 913 00:45:49,040 --> 00:45:51,800 Speaker 1: anybody like infected it or hacked it, or there was 914 00:45:51,840 --> 00:45:55,160 Speaker 1: a copyer or anything. They think probably was a bitflip 915 00:45:55,200 --> 00:45:59,200 Speaker 1: that caused this weird result in an election in Brussels, 916 00:45:59,239 --> 00:46:01,719 Speaker 1: and so it's certainly happened, right, It's something we need 917 00:46:01,760 --> 00:46:03,560 Speaker 1: to be aware of. I mean we need to know about. 918 00:46:03,800 --> 00:46:05,440 Speaker 2: Did it change the results of the election? 919 00:46:05,600 --> 00:46:07,200 Speaker 1: No, they figured out I think they knew how many 920 00:46:07,280 --> 00:46:09,919 Speaker 1: votes have been cast in total, and so they knew 921 00:46:09,920 --> 00:46:12,520 Speaker 1: that there was an excess and affected it. And I 922 00:46:12,560 --> 00:46:15,279 Speaker 1: don't think it flipped the answer either way. So it's 923 00:46:15,280 --> 00:46:18,839 Speaker 1: just sort of weird and mixed people suspicious, and you know, 924 00:46:18,920 --> 00:46:21,720 Speaker 1: it highlights something else, which is that our whole digital 925 00:46:21,719 --> 00:46:25,879 Speaker 1: infrastructure is sensitive to what's going on in space. We 926 00:46:25,960 --> 00:46:29,719 Speaker 1: recently had a really nice cosmic light show because there 927 00:46:29,760 --> 00:46:32,400 Speaker 1: was an extra amount of solar radiation which led to 928 00:46:32,840 --> 00:46:35,480 Speaker 1: northern lights all over the place on Earth. But that's 929 00:46:35,520 --> 00:46:38,520 Speaker 1: extra solar radiation and that can come at any time. 930 00:46:38,920 --> 00:46:43,080 Speaker 1: In the eighteen fifties, there was a really spectacular coronal 931 00:46:43,160 --> 00:46:47,040 Speaker 1: mass ejection, an eruption of charged particles from the Sun 932 00:46:47,080 --> 00:46:50,080 Speaker 1: which basically like a huge loop of plasma which heads 933 00:46:50,120 --> 00:46:52,680 Speaker 1: towards the Earth, which is just high energy cosmic rayse 934 00:46:52,719 --> 00:46:55,719 Speaker 1: that's what plasma is, right, and it can be like 935 00:46:55,880 --> 00:46:59,359 Speaker 1: billions of tons of materials. Most of these things missed 936 00:46:59,360 --> 00:47:01,359 Speaker 1: the Earth because they're it's pretty small and far from 937 00:47:01,400 --> 00:47:03,640 Speaker 1: the sun, but when it washes over the Earth, it 938 00:47:03,680 --> 00:47:07,280 Speaker 1: can devastate our digital infrastructure more than just like corrupting 939 00:47:07,320 --> 00:47:10,480 Speaker 1: a single bit on an airplane or in a voting machine. 940 00:47:11,000 --> 00:47:15,160 Speaker 1: This in the eighteen fifties, like set telegraph networks on fire, 941 00:47:15,160 --> 00:47:18,600 Speaker 1: like you know, sparks were flying. Yeah, it's crazy. And 942 00:47:18,680 --> 00:47:21,160 Speaker 1: if that happened today, like our digital infrastructure is not 943 00:47:21,280 --> 00:47:24,359 Speaker 1: protected from that kind of incursion, you wouldn't be able 944 00:47:24,360 --> 00:47:27,400 Speaker 1: to like shop for stuff online for days and days 945 00:47:27,560 --> 00:47:28,719 Speaker 1: while people fixed it. 946 00:47:29,040 --> 00:47:31,520 Speaker 2: I mean, can you imagine, like if something took out 947 00:47:31,920 --> 00:47:37,880 Speaker 2: all of our satellites, how absolutely pned we would be Yeah, 948 00:47:37,920 --> 00:47:41,160 Speaker 2: like nope, credit cards, I wouldn't have starlink Internet, there'd 949 00:47:41,160 --> 00:47:43,480 Speaker 2: be no dKu anymore everyone. 950 00:47:43,800 --> 00:47:46,760 Speaker 1: Can you imagine dot civilization crumbles? 951 00:47:46,880 --> 00:47:47,080 Speaker 3: Yes? 952 00:47:47,200 --> 00:47:48,640 Speaker 2: Right, yep, I agree. 953 00:47:48,880 --> 00:47:51,080 Speaker 1: And so in particle physics, we've thought about this a 954 00:47:51,080 --> 00:47:54,680 Speaker 1: lot because we're operating computers in a very high radiation 955 00:47:54,760 --> 00:47:58,239 Speaker 1: environment constantly, like we're creating that radiation. When you smash 956 00:47:58,280 --> 00:48:01,319 Speaker 1: two particles together, you generate huge numbers of protons and 957 00:48:01,360 --> 00:48:04,080 Speaker 1: neutrons and muons and all kinds of stuff, And then 958 00:48:04,160 --> 00:48:07,000 Speaker 1: we have silicon near the collision in order to measure 959 00:48:07,040 --> 00:48:10,239 Speaker 1: these things to detect these particles in exactly that way. 960 00:48:10,280 --> 00:48:12,399 Speaker 1: But if you run it for a while, eventually your 961 00:48:12,400 --> 00:48:15,520 Speaker 1: silicon gets trashed, you know, so every few years we 962 00:48:15,520 --> 00:48:17,439 Speaker 1: have to pull the thing out and send in new 963 00:48:17,520 --> 00:48:21,280 Speaker 1: silicon basically, And so people are doing stuff like exploring 964 00:48:21,320 --> 00:48:24,040 Speaker 1: whether we can use diamond instead. You know, we talk 965 00:48:24,080 --> 00:48:26,239 Speaker 1: about silicon as a semiconductor, but it's not the only 966 00:48:26,280 --> 00:48:29,000 Speaker 1: semi conductor out there. You can build these chips out 967 00:48:29,040 --> 00:48:32,960 Speaker 1: of diamond, which makes them more radiation hard which is 968 00:48:33,160 --> 00:48:38,680 Speaker 1: very expensive because there isn't a huge consumer diamond lithography 969 00:48:38,719 --> 00:48:41,560 Speaker 1: industry the way there is for silicon. So particle physics 970 00:48:41,560 --> 00:48:45,000 Speaker 1: tends to make its advances by like piggybacking on huge 971 00:48:45,040 --> 00:48:48,600 Speaker 1: trillion dollar consumer trends, and that hasn't happened for particle 972 00:48:48,600 --> 00:48:51,480 Speaker 1: physics that we don't have, you know, diamond chips yet, 973 00:48:51,800 --> 00:48:53,520 Speaker 1: but it's something that people are looking at that we 974 00:48:53,600 --> 00:48:55,680 Speaker 1: might have to do in the future in order to 975 00:48:55,719 --> 00:48:59,600 Speaker 1: protect our computers from cosmic grades or radiation more generally. 976 00:48:59,400 --> 00:49:00,920 Speaker 2: Can we take a us step back to like so 977 00:49:00,960 --> 00:49:04,320 Speaker 2: if we'd had another Carrington event type of thing, happen. 978 00:49:05,160 --> 00:49:08,359 Speaker 2: Is there something we could have done to protect all 979 00:49:08,400 --> 00:49:11,320 Speaker 2: of our satellites from an event like that? Like, should 980 00:49:11,360 --> 00:49:13,840 Speaker 2: we be encasing all of our satellites in diamonds? Is 981 00:49:13,840 --> 00:49:14,919 Speaker 2: what I'm asking. 982 00:49:17,400 --> 00:49:22,040 Speaker 1: It would make them so glittery and beautiful, that's right. Well, 983 00:49:22,120 --> 00:49:24,240 Speaker 1: essentially what you need is mass, Like either you should 984 00:49:24,239 --> 00:49:27,239 Speaker 1: build the chips themselves out of diamond to make them 985 00:49:27,280 --> 00:49:30,239 Speaker 1: radiation hard or is you just shield them, in which 986 00:49:30,280 --> 00:49:32,800 Speaker 1: case you need mass, and mass, as you know, is 987 00:49:32,840 --> 00:49:35,520 Speaker 1: expensive to put in space, and that's why it's so 988 00:49:35,600 --> 00:49:38,000 Speaker 1: difficult to have things like computing in space because it 989 00:49:38,000 --> 00:49:40,920 Speaker 1: needs the thing which is expensive to put. So I 990 00:49:40,960 --> 00:49:42,400 Speaker 1: think it would look awesome, but I don't think we 991 00:49:42,440 --> 00:49:44,360 Speaker 1: should bling up all of our satellites. 992 00:49:44,719 --> 00:49:48,480 Speaker 2: Okay, So for the foreseeable future, we're probably not going 993 00:49:48,560 --> 00:49:50,640 Speaker 2: to do anything about the fact that if we had 994 00:49:50,640 --> 00:49:54,040 Speaker 2: a Carrington event type of thing happen again, we would 995 00:49:54,040 --> 00:49:54,880 Speaker 2: be in a lot of trouble. 996 00:49:55,200 --> 00:49:58,640 Speaker 1: Yeah, that would be a massive investment in infrastructure, and 997 00:49:58,840 --> 00:50:00,880 Speaker 1: I don't see anybody doing it, though I think it 998 00:50:00,880 --> 00:50:03,319 Speaker 1: would be a good idea. But until then, we do 999 00:50:03,360 --> 00:50:05,719 Speaker 1: need to be aware that cosmic rays are everywhere. They 1000 00:50:05,719 --> 00:50:08,680 Speaker 1: are streaming through your room right now, and they are 1001 00:50:08,680 --> 00:50:11,719 Speaker 1: streaming through your computer. And sometimes they can flip a 1002 00:50:11,760 --> 00:50:13,839 Speaker 1: bit a zero to one. They can plow their way 1003 00:50:13,880 --> 00:50:16,160 Speaker 1: through a bunch of silicon and change the way those 1004 00:50:16,200 --> 00:50:19,360 Speaker 1: delicate electronics work. Most of the time this doesn't affect 1005 00:50:19,400 --> 00:50:22,279 Speaker 1: your computer, or it does it in an obvious way, or 1006 00:50:22,360 --> 00:50:26,360 Speaker 1: it's caught by error correction and fixed. But sometimes it isn't. 1007 00:50:26,600 --> 00:50:29,120 Speaker 1: And so next time your computer goes hey wire, you 1008 00:50:29,120 --> 00:50:31,239 Speaker 1: could blame it on your kids, or you could blame 1009 00:50:31,239 --> 00:50:32,560 Speaker 1: it on particles from space. 1010 00:50:32,840 --> 00:50:35,760 Speaker 2: Or the next time you get that super awesome science result, 1011 00:50:36,120 --> 00:50:39,520 Speaker 2: you should question yourself even more. That anxiety should keep 1012 00:50:39,560 --> 00:50:41,920 Speaker 2: you up all night, because maybe it was a cosmic 1013 00:50:42,000 --> 00:50:43,080 Speaker 2: ray and you were wrong. 1014 00:50:45,080 --> 00:50:50,080 Speaker 1: Maybe when you counted parasites you're off by one. All right, Well, 1015 00:50:50,120 --> 00:50:53,319 Speaker 1: thanks for taking this cosmic journey with us into how 1016 00:50:53,400 --> 00:50:56,960 Speaker 1: silicon electronics work and how they are vulnerable to particles 1017 00:50:56,960 --> 00:51:01,240 Speaker 1: from space. As always, we are grateful for your shared curiosity. 1018 00:51:01,440 --> 00:51:11,320 Speaker 2: Thanks everyone, have a good one. Daniel and Kelly's Extraordinary 1019 00:51:11,400 --> 00:51:14,640 Speaker 2: Universe is produced by iHeartRadio. We would love to hear 1020 00:51:14,680 --> 00:51:15,120 Speaker 2: from you. 1021 00:51:15,239 --> 00:51:18,200 Speaker 1: We really would. We want to know what questions you 1022 00:51:18,400 --> 00:51:21,040 Speaker 1: have about this Extraordinary Universe. 1023 00:51:21,120 --> 00:51:24,080 Speaker 2: We want to know your thoughts on recent shows, suggestions 1024 00:51:24,080 --> 00:51:27,080 Speaker 2: for future shows. If you contact us, we will get 1025 00:51:27,120 --> 00:51:27,520 Speaker 2: back to you. 1026 00:51:27,800 --> 00:51:31,319 Speaker 1: We really mean it. We answer every message. Email us 1027 00:51:31,360 --> 00:51:34,560 Speaker 1: at Questions at Danielankelly. 1028 00:51:33,640 --> 00:51:35,720 Speaker 2: Dot org, or you can find us on social media. 1029 00:51:35,800 --> 00:51:39,600 Speaker 2: We have accounts on x, Instagram, Blue Sky and on 1030 00:51:39,680 --> 00:51:41,640 Speaker 2: all of those platforms. You can find us at d 1031 00:51:42,080 --> 00:51:43,600 Speaker 2: and kuniverse. 1032 00:51:43,800 --> 00:51:45,359 Speaker 1: Don't be shy, write to us