1 00:00:10,800 --> 00:00:14,840 Speaker 1: Hello, and welcome to another episode of the Odd Lots podcast. 2 00:00:14,920 --> 00:00:21,240 Speaker 1: I'm Joe Wisenthal and I'm Tracy Halloway. So Tracy obviously, Yeah, 3 00:00:21,280 --> 00:00:25,000 Speaker 1: we've done a lot of episodes about the semiconductor industry, 4 00:00:25,120 --> 00:00:28,560 Speaker 1: about chips. There's one specific, I guess I would say, 5 00:00:28,600 --> 00:00:31,200 Speaker 1: sub component of the story that people are like, Oh, 6 00:00:31,240 --> 00:00:32,800 Speaker 1: you guys got to do that, you guys got to 7 00:00:32,840 --> 00:00:34,800 Speaker 1: do that, which we have yet to hit so far. 8 00:00:35,520 --> 00:00:37,760 Speaker 1: You say, there's one thing that we have yet to do, 9 00:00:37,800 --> 00:00:40,320 Speaker 1: but I have a feeling like this is the Endless 10 00:00:40,360 --> 00:00:43,200 Speaker 1: Semiconductor series, and as soon as we finished this episode, 11 00:00:43,240 --> 00:00:46,440 Speaker 1: we're going to discover some other hidden component of the 12 00:00:46,479 --> 00:00:48,800 Speaker 1: semi conductor supply chain and that's going to lead to 13 00:00:48,880 --> 00:00:52,240 Speaker 1: another episode. But yes, you're right, um, there is one 14 00:00:52,360 --> 00:00:57,200 Speaker 1: sort of big elephant, big semiconductor thing in the room, 15 00:00:57,400 --> 00:01:00,600 Speaker 1: and that is a company called a s m L. 16 00:01:01,120 --> 00:01:03,120 Speaker 1: Not to be confused with a s m R, which 17 00:01:03,160 --> 00:01:06,680 Speaker 1: I always seem to do a s mL but hopefully 18 00:01:06,680 --> 00:01:10,000 Speaker 1: for like a certain kind of person, listening to an 19 00:01:10,040 --> 00:01:13,400 Speaker 1: hour of people talking about chips is a type of 20 00:01:13,880 --> 00:01:16,600 Speaker 1: for them SMR. So maybe killed two birds with one stone, 21 00:01:16,600 --> 00:01:18,480 Speaker 1: but yes, s m L, you know, one of the 22 00:01:18,560 --> 00:01:21,680 Speaker 1: things that we established in thinking about how these sort 23 00:01:21,720 --> 00:01:25,319 Speaker 1: of the chip ecosystem works. Maybe part of our characterization 24 00:01:25,440 --> 00:01:30,400 Speaker 1: is we twan semiconductor the biggest contract fabrication company in 25 00:01:30,440 --> 00:01:32,039 Speaker 1: the world. I think we sort of think of as 26 00:01:32,080 --> 00:01:35,000 Speaker 1: like they're the final boss right in chims, Like in 27 00:01:35,040 --> 00:01:37,840 Speaker 1: the end, they're like the central bank of chips. Their 28 00:01:37,920 --> 00:01:41,800 Speaker 1: capacity kind of almost dictates chip capacity overall. There's some 29 00:01:41,840 --> 00:01:46,119 Speaker 1: other companies that make chips, including Intel and Global Boundaries, 30 00:01:46,200 --> 00:01:51,000 Speaker 1: but t SMC is the big one. But TSMC has 31 00:01:51,040 --> 00:01:53,840 Speaker 1: to buy equipment from others to you know not no 32 00:01:53,880 --> 00:01:57,520 Speaker 1: one is completely self sufficient in this industry, and t 33 00:01:57,720 --> 00:02:01,720 Speaker 1: SMC is a huge client or a huge purchaser of 34 00:02:01,840 --> 00:02:06,120 Speaker 1: equipment made by this company. It's a Dutch company a SML. Yeah, 35 00:02:06,200 --> 00:02:08,360 Speaker 1: so you mentioned that it's Dutch, And this is the 36 00:02:08,400 --> 00:02:11,160 Speaker 1: other thing I mean, in addition to not really understanding 37 00:02:11,360 --> 00:02:14,120 Speaker 1: what this company does or the type of equipment it's 38 00:02:14,120 --> 00:02:18,480 Speaker 1: actually making for semiconductor manufacturers like t SMC, the other 39 00:02:18,520 --> 00:02:21,200 Speaker 1: thing I don't really get about it is why is 40 00:02:21,240 --> 00:02:23,480 Speaker 1: it a Dutch company? Because the one thing I know 41 00:02:23,480 --> 00:02:26,640 Speaker 1: about it is it has its origins in the US 42 00:02:27,280 --> 00:02:29,760 Speaker 1: UM I think in the you know, like nineteen eighties, 43 00:02:29,760 --> 00:02:31,920 Speaker 1: it sort of came out of the collapse of a 44 00:02:31,960 --> 00:02:35,520 Speaker 1: bunch of US lithography firms or something like that. And 45 00:02:35,600 --> 00:02:38,520 Speaker 1: yet now it's a Dutch company that has this enormous 46 00:02:38,720 --> 00:02:43,760 Speaker 1: role in the global supply chain. It's squarely like kind 47 00:02:43,760 --> 00:02:46,600 Speaker 1: of crowning, like it's sort of like a kingmaker for 48 00:02:46,760 --> 00:02:51,800 Speaker 1: semiconductor technology or expertise. And I don't know, I just 49 00:02:51,840 --> 00:02:55,120 Speaker 1: have so many questions already. I know we haven't even started. Yeah, 50 00:02:55,240 --> 00:02:57,040 Speaker 1: I know, I have no questions too. We'll get started 51 00:02:57,040 --> 00:02:59,120 Speaker 1: just a second. But you know, you mentioned the oddity 52 00:02:59,160 --> 00:03:03,280 Speaker 1: of it big Dutch. There's another element here, and I 53 00:03:03,280 --> 00:03:05,280 Speaker 1: I don't you know, I feel like reluctant to like 54 00:03:05,360 --> 00:03:11,760 Speaker 1: talking like cliches or stereotypes, but I don't really think. No, 55 00:03:11,880 --> 00:03:15,160 Speaker 1: it is true of like Northern Europe or Europe in 56 00:03:15,240 --> 00:03:18,919 Speaker 1: general as being like this like cutting edge high tech 57 00:03:18,960 --> 00:03:21,560 Speaker 1: hotbed for anything. When I think about tech, I think 58 00:03:21,560 --> 00:03:24,959 Speaker 1: about Silicon Valley, maybe more of the consumer and but 59 00:03:25,000 --> 00:03:27,720 Speaker 1: also like you know, obviously a long history. It means 60 00:03:27,720 --> 00:03:31,000 Speaker 1: Silicon Valley for a reason. And I think about various 61 00:03:31,000 --> 00:03:33,720 Speaker 1: parts of East Asia, and when I think about the 62 00:03:33,800 --> 00:03:36,720 Speaker 1: engineering prowess in Europe. Hey, I don't think about tech 63 00:03:36,760 --> 00:03:39,120 Speaker 1: in Europe that much, and when I do think about 64 00:03:39,200 --> 00:03:42,520 Speaker 1: engineering prowess in Europe, it's typically I'm thinking more on 65 00:03:42,640 --> 00:03:47,000 Speaker 1: these sort of like bigger industrial engineering, so a company 66 00:03:47,080 --> 00:03:51,000 Speaker 1: like Siemens or companies that are really good at public 67 00:03:51,240 --> 00:03:54,560 Speaker 1: public works or trains or whatever. And I don't think 68 00:03:54,600 --> 00:03:58,840 Speaker 1: of Europe as being a hotbed of say, semiconductor innovation, 69 00:03:59,760 --> 00:04:02,520 Speaker 1: and there's probably countless like counter examples. I'm just sort 70 00:04:02,520 --> 00:04:04,800 Speaker 1: of thinking, like it doesn't fit into my mental models 71 00:04:04,800 --> 00:04:07,160 Speaker 1: of this stuff. So it is interesting that it's Dutch. 72 00:04:07,520 --> 00:04:09,880 Speaker 1: Well also, just when you think about the European market, 73 00:04:09,960 --> 00:04:13,360 Speaker 1: like you start thinking about the biggest companies there, and yeah, sure, 74 00:04:13,400 --> 00:04:17,760 Speaker 1: stuff like Semens, LVMH like luxury good makers. But mL 75 00:04:17,920 --> 00:04:21,520 Speaker 1: is absolutely massive, and like just looking at the share 76 00:04:21,560 --> 00:04:24,760 Speaker 1: price chart, it has had a huge, huge run up 77 00:04:25,040 --> 00:04:27,719 Speaker 1: over the past year or so, I mean basically since 78 00:04:27,880 --> 00:04:31,240 Speaker 1: the global pandemic, much like a lot of other semiconductor stocks, 79 00:04:31,240 --> 00:04:35,640 Speaker 1: but I mean amazing run up, huge market cap, and 80 00:04:35,720 --> 00:04:39,359 Speaker 1: yet it's sort of like this company that outside of 81 00:04:39,400 --> 00:04:43,440 Speaker 1: the semiconductor sphere, it doesn't seem to get that much attention. Yeah, 82 00:04:43,520 --> 00:04:45,159 Speaker 1: I mean it's A. It's like a it's like a 83 00:04:45,240 --> 00:04:48,360 Speaker 1: third over three billion dollar market cap. It is. It's 84 00:04:48,640 --> 00:04:51,159 Speaker 1: it's one of the biggest companies in the world. Um, 85 00:04:51,200 --> 00:04:53,200 Speaker 1: but not many people know about it, far from my 86 00:04:53,279 --> 00:04:55,919 Speaker 1: household name. Okay, so we have a million questions. So 87 00:04:55,920 --> 00:04:57,880 Speaker 1: we got to get right into this discussion. And we 88 00:04:57,920 --> 00:05:00,480 Speaker 1: have the perfect guest to tell us a out this company. 89 00:05:00,520 --> 00:05:03,240 Speaker 1: We're gonna be speaking with Chris Miller. He's an assistant 90 00:05:03,279 --> 00:05:07,480 Speaker 1: professor of international history at the Fletcher School at Toughs University, 91 00:05:07,560 --> 00:05:10,120 Speaker 1: and he is the author of a forthcoming book that 92 00:05:10,160 --> 00:05:12,719 Speaker 1: will be out next year entitled Chip War, The Struggle 93 00:05:12,760 --> 00:05:16,440 Speaker 1: for the World's Most Critical Technology, And he can answer 94 00:05:16,520 --> 00:05:20,400 Speaker 1: all of our questions about a SML. Chris, thank you 95 00:05:20,520 --> 00:05:23,760 Speaker 1: so much for joining us. Thanks for having me. Chris, 96 00:05:23,760 --> 00:05:26,000 Speaker 1: what is lithography? You know? I think that this is 97 00:05:26,040 --> 00:05:28,360 Speaker 1: one of these questions that's like the word gets It's 98 00:05:28,360 --> 00:05:31,560 Speaker 1: probably come up on like every episode. And I pride 99 00:05:31,600 --> 00:05:34,920 Speaker 1: myself on never being too embarrassed to like ask a question, 100 00:05:35,320 --> 00:05:38,200 Speaker 1: but I think I actually was too embarrassed to ask 101 00:05:38,760 --> 00:05:42,040 Speaker 1: this on some of the other episodes. I'm like, hm, phothography, 102 00:05:42,520 --> 00:05:46,080 Speaker 1: what phithography? So if you want to make a semi 103 00:05:46,080 --> 00:05:50,080 Speaker 1: connected device, you take a slab of silicon, you cover 104 00:05:50,240 --> 00:05:54,200 Speaker 1: it with chemicals called photo resists, which are chemicals that 105 00:05:54,279 --> 00:05:58,600 Speaker 1: react with light, and then you shoot light raise or 106 00:05:58,680 --> 00:06:02,200 Speaker 1: now extreme all tra violet light rays at the silicon 107 00:06:02,200 --> 00:06:05,800 Speaker 1: wafer UH, and the shapes that you shoot at it 108 00:06:05,839 --> 00:06:09,719 Speaker 1: will form the transistors. So that's that's the simplest version. 109 00:06:09,920 --> 00:06:12,920 Speaker 1: Now today, if you buy a new iPhone, the most 110 00:06:12,960 --> 00:06:16,640 Speaker 1: events process around it will have ten billion transistors. So 111 00:06:16,720 --> 00:06:21,839 Speaker 1: you've got to shoot extraordinarily narrow wavelengths of light through 112 00:06:22,240 --> 00:06:26,880 Speaker 1: masks that create these shapes on the silicon wafer, and 113 00:06:26,960 --> 00:06:30,440 Speaker 1: the masks need to be able to project all of 114 00:06:30,480 --> 00:06:33,000 Speaker 1: these shapes onto the wafer. So making this possible at 115 00:06:33,040 --> 00:06:36,520 Speaker 1: the scale of ten billion transistors per chip is is 116 00:06:36,640 --> 00:06:39,720 Speaker 1: what a SML does. Wait, how many per chip? What 117 00:06:39,839 --> 00:06:42,560 Speaker 1: do you say? Ten billion per chip? That's right, a 118 00:06:42,640 --> 00:06:46,279 Speaker 1: new Apple processor in your iPhone will have ten billion 119 00:06:46,320 --> 00:06:49,360 Speaker 1: transistors per chip. Some chips that go into data centers 120 00:06:49,360 --> 00:06:52,600 Speaker 1: will have more than that. But the scale of transistors 121 00:06:52,600 --> 00:06:55,560 Speaker 1: that we produce at any given year is is more 122 00:06:55,600 --> 00:06:58,560 Speaker 1: than the scale of all goods produced by all companies 123 00:06:58,600 --> 00:07:00,920 Speaker 1: and all other industries and all of world history. It's 124 00:07:00,920 --> 00:07:04,359 Speaker 1: a tremendous number. So could you maybe describe where a 125 00:07:04,520 --> 00:07:08,960 Speaker 1: SML sits in the sort of ecosystem of the semiconductor industry. 126 00:07:09,000 --> 00:07:13,080 Speaker 1: So I gather it doesn't seem to have much competition, 127 00:07:13,200 --> 00:07:16,840 Speaker 1: but like who does it actually supply? And also who 128 00:07:16,880 --> 00:07:19,200 Speaker 1: does it not supply? Like are there people out there 129 00:07:19,240 --> 00:07:22,360 Speaker 1: who try to do this on their own. In the 130 00:07:22,400 --> 00:07:25,560 Speaker 1: early days of the chip industry, companies built lithography machines 131 00:07:25,600 --> 00:07:28,600 Speaker 1: in house, so Texas Instruments would have had its own 132 00:07:28,640 --> 00:07:32,240 Speaker 1: Lithography Machine Division IBM. But today the machines are so 133 00:07:32,280 --> 00:07:35,440 Speaker 1: complex and expensive that there's just a couple of companies 134 00:07:35,520 --> 00:07:39,200 Speaker 1: that make lithography machines in general, and just one company, 135 00:07:39,240 --> 00:07:42,239 Speaker 1: a SML, as able to make a u V lithography machines, 136 00:07:42,280 --> 00:07:45,600 Speaker 1: which are the most advanced type. Anyone who operates a 137 00:07:45,680 --> 00:07:48,520 Speaker 1: chip fab sility where chips are made has to buy 138 00:07:48,600 --> 00:07:51,440 Speaker 1: lithography equipment, and so for the most cutting edge chips, 139 00:07:51,480 --> 00:07:53,480 Speaker 1: you've got no choice but to buy from a s mL. 140 00:07:54,040 --> 00:07:57,200 Speaker 1: This is fasting. So whether we're talking about UNTIL doing 141 00:07:57,240 --> 00:08:00,960 Speaker 1: its own chips or TSMC or any one else. And 142 00:08:01,000 --> 00:08:03,400 Speaker 1: we've talked with other people who talked to Stacy Raskin 143 00:08:03,480 --> 00:08:07,400 Speaker 1: of Burnstein and about the nanometer wars and all of them. 144 00:08:07,600 --> 00:08:12,080 Speaker 1: If you're doing cutting edge manufacturing, you are a customer 145 00:08:12,200 --> 00:08:14,760 Speaker 1: of a a s mL. That's right, that's right. For 146 00:08:14,800 --> 00:08:17,520 Speaker 1: the most cutting edge lithography machines, a SML is the 147 00:08:17,560 --> 00:08:21,720 Speaker 1: only supplier for slightly less cutting edge machinery. Uh Night 148 00:08:21,760 --> 00:08:24,760 Speaker 1: kind of Japan is also a competitor of s mls. 149 00:08:25,000 --> 00:08:27,800 Speaker 1: They have a duopoly for anything that's not the most 150 00:08:27,840 --> 00:08:31,800 Speaker 1: cutting edge. So what is it about the technology that 151 00:08:32,040 --> 00:08:35,840 Speaker 1: makes it, I guess so proprietary to a s mL, Like, 152 00:08:36,000 --> 00:08:38,080 Speaker 1: how did they get into a position where they basically 153 00:08:38,360 --> 00:08:41,000 Speaker 1: control it? And what is it that they've been able 154 00:08:41,080 --> 00:08:44,640 Speaker 1: to do that others haven't. So the the challenge with 155 00:08:44,960 --> 00:08:47,880 Speaker 1: EUV lithography in particular and lithography in general is that 156 00:08:47,920 --> 00:08:52,080 Speaker 1: you've got to uh manage a supply chain that is 157 00:08:52,160 --> 00:08:55,560 Speaker 1: extraordinary complex. A SML has got around four thousand suppliers, 158 00:08:55,600 --> 00:08:59,120 Speaker 1: and many of these suppliers are producing equipment that only 159 00:08:59,160 --> 00:09:01,280 Speaker 1: they can produce. So, just to give you a couple 160 00:09:01,280 --> 00:09:06,800 Speaker 1: of examples them, the mirrors within s MLS lithography machines, 161 00:09:06,840 --> 00:09:10,320 Speaker 1: the eu V machines are the flattest structure that humans 162 00:09:10,360 --> 00:09:12,720 Speaker 1: have ever made, the flattest man made structure in the universe. 163 00:09:13,240 --> 00:09:16,600 Speaker 1: And when you go through the list of materials and 164 00:09:17,200 --> 00:09:20,319 Speaker 1: components that you need to produce an e V lithography machine, 165 00:09:20,360 --> 00:09:22,720 Speaker 1: there are multiple parts of the system that are the 166 00:09:22,760 --> 00:09:26,560 Speaker 1: most this or the most that, and managing that is 167 00:09:26,600 --> 00:09:29,440 Speaker 1: an extraordinary complex business. If you talk to people at 168 00:09:29,440 --> 00:09:32,679 Speaker 1: a sm L, they'll say, our biggest engineering challenge is 169 00:09:33,120 --> 00:09:36,240 Speaker 1: not actually engineering any particular part, engineering the supply chain, 170 00:09:36,320 --> 00:09:38,559 Speaker 1: making sure that all of our suppliers are producing things 171 00:09:38,840 --> 00:09:40,480 Speaker 1: so that they all fit together, they all work together, 172 00:09:40,480 --> 00:09:42,880 Speaker 1: they arrive on time. And it's hard enough to do 173 00:09:42,920 --> 00:09:46,880 Speaker 1: that with basic machinery, but when you're trying to manipulate uh, 174 00:09:46,960 --> 00:09:49,800 Speaker 1: individual atoms, which is what SML is able to do, 175 00:09:50,240 --> 00:09:53,720 Speaker 1: it's even more complex. Tracy, I already loved this episode 176 00:09:53,760 --> 00:09:55,840 Speaker 1: so much. I don't know, like how many things like 177 00:09:55,880 --> 00:09:59,280 Speaker 1: I've learned already in five minutes. And then the fact 178 00:09:59,320 --> 00:10:03,480 Speaker 1: that like it's it's also a supply Like Okay, obviously 179 00:10:03,520 --> 00:10:05,800 Speaker 1: there's a chip supply chain, but then the idea that 180 00:10:05,880 --> 00:10:10,040 Speaker 1: the most advanced technology within the chip is actually itself 181 00:10:10,080 --> 00:10:13,480 Speaker 1: a supply chain. I'm just like, I'm already obsessed with this, 182 00:10:13,600 --> 00:10:16,080 Speaker 1: but where do they So Okay, you mentioned that for 183 00:10:16,480 --> 00:10:18,720 Speaker 1: sort of like okay, for the very cutting edge, there's 184 00:10:18,760 --> 00:10:24,360 Speaker 1: just a SML for slightly less cutting edge. Uh, Nikon, 185 00:10:24,600 --> 00:10:28,240 Speaker 1: did you say, Nikon? That's right, Nikon. In Japan, they're 186 00:10:28,280 --> 00:10:32,680 Speaker 1: also in the game. Do other players aspire to be 187 00:10:32,920 --> 00:10:36,679 Speaker 1: cutting edge or is there some barrier that just basically 188 00:10:36,720 --> 00:10:39,080 Speaker 1: makes it so that no one else is really trying 189 00:10:39,120 --> 00:10:43,520 Speaker 1: to get be at that level in the which is 190 00:10:43,520 --> 00:10:47,439 Speaker 1: when investment in EUV began. Nikon made a choice not 191 00:10:47,600 --> 00:10:51,840 Speaker 1: to try to commercialize UV technology. The first physics papers 192 00:10:51,880 --> 00:10:55,079 Speaker 1: on the V actually came out of a Japanese university, 193 00:10:55,520 --> 00:10:59,280 Speaker 1: so there's plenty of optics expertise in Japan that a 194 00:10:59,440 --> 00:11:01,720 Speaker 1: SML was the only company that was willing to bet 195 00:11:01,720 --> 00:11:05,800 Speaker 1: on EUV from the forward and capable of raising the 196 00:11:05,800 --> 00:11:09,600 Speaker 1: funds and assembling the expertise. So right now, if someone 197 00:11:09,679 --> 00:11:12,800 Speaker 1: wanted to replicate what U vs what a SML has 198 00:11:12,840 --> 00:11:15,199 Speaker 1: done with UV, it would take them a decade and 199 00:11:15,480 --> 00:11:18,880 Speaker 1: billions and billions of dollars in investment. And because the 200 00:11:18,920 --> 00:11:21,920 Speaker 1: suppliers that work with a SML have exclusivity agreements with 201 00:11:21,960 --> 00:11:25,280 Speaker 1: a SML. SML has invested in some of his key suppliers. 202 00:11:26,000 --> 00:11:29,520 Speaker 1: It's just basically impossible for anyone to break into this 203 00:11:29,640 --> 00:11:32,640 Speaker 1: without replicating their entire separate supply chain. It would take 204 00:11:32,679 --> 00:11:52,320 Speaker 1: a decade to do. So, maybe this is a good 205 00:11:52,320 --> 00:11:55,400 Speaker 1: place to start talking about the history of the company 206 00:11:55,480 --> 00:11:57,400 Speaker 1: and where it actually came from. And I think that 207 00:11:57,440 --> 00:12:00,280 Speaker 1: will help us like understand some of these dynamics how 208 00:12:00,280 --> 00:12:03,679 Speaker 1: it built up a competitive edge versus it's you know, 209 00:12:04,080 --> 00:12:08,960 Speaker 1: non existent or very few competitors. But my understanding is 210 00:12:08,960 --> 00:12:13,160 Speaker 1: the sort of like sprung up out of US military 211 00:12:13,200 --> 00:12:16,520 Speaker 1: technology of some sort. Can you start like at the beginning? 212 00:12:16,679 --> 00:12:19,520 Speaker 1: How like? I guess this goes back to Joe's question, 213 00:12:19,640 --> 00:12:22,840 Speaker 1: what is lithography? Why is the U. S Army interested 214 00:12:22,880 --> 00:12:25,319 Speaker 1: in it? And how did it play into a s 215 00:12:25,440 --> 00:12:29,439 Speaker 1: ML's creation story. When the transistor was first invented in 216 00:12:29,480 --> 00:12:32,360 Speaker 1: the late nineteen forties in Bell Labs in New Jersey, 217 00:12:32,400 --> 00:12:36,240 Speaker 1: it was predominantly used in military devices for its its 218 00:12:36,240 --> 00:12:39,959 Speaker 1: first commercialization, and there was a scientist in a US 219 00:12:40,040 --> 00:12:42,600 Speaker 1: Army lab named J. Lathrop in the nineteen fifties who 220 00:12:42,640 --> 00:12:44,760 Speaker 1: was trying to find out how to minim miniature eye 221 00:12:44,800 --> 00:12:48,040 Speaker 1: transistors produced them smaller and smaller so they could be 222 00:12:48,040 --> 00:12:51,800 Speaker 1: put in smaller devices. One day, he and his his 223 00:12:51,840 --> 00:12:56,080 Speaker 1: assistant realized that they could use photo resist chemicals, these 224 00:12:56,160 --> 00:13:00,040 Speaker 1: chemicals that react with lights to create shapes on on 225 00:13:00,160 --> 00:13:02,199 Speaker 1: the silicon and germanium that they were working with, and 226 00:13:02,480 --> 00:13:04,760 Speaker 1: they turned their microscope that they were using the lab 227 00:13:04,840 --> 00:13:08,200 Speaker 1: upside down. So normally a microscope lets you see something 228 00:13:08,240 --> 00:13:10,760 Speaker 1: small and it expands the image for your eye. They 229 00:13:10,880 --> 00:13:13,960 Speaker 1: did the opposite. They had a shape that was large 230 00:13:13,960 --> 00:13:15,880 Speaker 1: and were able to project that in a smaller version 231 00:13:15,880 --> 00:13:18,200 Speaker 1: by using an upside down microscope, And with that they 232 00:13:18,200 --> 00:13:20,960 Speaker 1: filed the first patent or photo lithography and coined the 233 00:13:21,000 --> 00:13:24,960 Speaker 1: phrase in the nineteen fifties. Over the next couple of decades, 234 00:13:25,160 --> 00:13:27,880 Speaker 1: photo lithography was used both by chipmakers who were making 235 00:13:27,920 --> 00:13:30,640 Speaker 1: their own machines and house and then eventually a couple 236 00:13:30,640 --> 00:13:34,800 Speaker 1: of specialized photo lithography companies emerged in Connecticut and Massachusetts. 237 00:13:35,160 --> 00:13:38,200 Speaker 1: They were defense contractors primarily, but realized they could use 238 00:13:38,240 --> 00:13:40,960 Speaker 1: their specialized optics things that they had honed in spy 239 00:13:41,040 --> 00:13:44,880 Speaker 1: satellites and military equipment like that for the semiconductor industry, 240 00:13:45,040 --> 00:13:47,720 Speaker 1: and so until the mid nineteen eighties, the center of 241 00:13:47,760 --> 00:13:50,560 Speaker 1: the photo lithography industry was in New England, but those 242 00:13:50,600 --> 00:13:52,880 Speaker 1: companies faced hard times in the nineteen eighties. They were 243 00:13:53,360 --> 00:13:56,120 Speaker 1: poorly managed, and the nineteen eighties were a time when 244 00:13:56,120 --> 00:13:58,960 Speaker 1: the Japanese ship industry in general was rising. A Nikon 245 00:13:59,320 --> 00:14:01,280 Speaker 1: as well as can and then the two camera companies 246 00:14:01,320 --> 00:14:04,240 Speaker 1: began investing in photo lithography. For a time in the 247 00:14:04,240 --> 00:14:08,160 Speaker 1: eighties and nineties, they were the dominant companies in the industry, 248 00:14:08,280 --> 00:14:11,400 Speaker 1: which the US was quite worried about. Worried about being 249 00:14:11,400 --> 00:14:14,240 Speaker 1: too reliant on Japan at a time of commercial and 250 00:14:14,280 --> 00:14:19,000 Speaker 1: also geopolitical tension, and so US chip makers began turning 251 00:14:19,000 --> 00:14:22,640 Speaker 1: to a s mL, both to diversify their supplier base, 252 00:14:22,680 --> 00:14:25,680 Speaker 1: but also because a SML was able to produce very 253 00:14:25,760 --> 00:14:29,240 Speaker 1: high quality equipment as well. In the nineteen nineties in 254 00:14:29,320 --> 00:14:32,080 Speaker 1: the mid as well, was the only company willing to 255 00:14:32,120 --> 00:14:34,520 Speaker 1: take the gamble on EUV, and since that point it's 256 00:14:34,520 --> 00:14:38,240 Speaker 1: become the dominant firm in the industry. We just have 257 00:14:39,160 --> 00:14:42,480 Speaker 1: extreme ultra violet. I don't know if we just want 258 00:14:42,520 --> 00:14:46,440 Speaker 1: to make sure we've established what UV stands for? Can 259 00:14:46,480 --> 00:14:49,120 Speaker 1: I just ask so, can you explain again? What's the 260 00:14:49,160 --> 00:14:53,320 Speaker 1: difference between extreme ultra violet and I guess non extreme 261 00:14:53,480 --> 00:14:56,800 Speaker 1: ultra violet. So over the past couple of decades is 262 00:14:56,840 --> 00:15:00,320 Speaker 1: we've tried to make ever smaller device is in ever 263 00:15:00,440 --> 00:15:04,640 Speaker 1: smaller features on silicon wafers. We've begun to use different 264 00:15:04,720 --> 00:15:09,080 Speaker 1: and smaller wavelengths of light, and so extreme multi violet 265 00:15:09,200 --> 00:15:11,640 Speaker 1: has a wavelength of thirteen point five nanometers. Is the 266 00:15:11,680 --> 00:15:14,680 Speaker 1: smallest wavelength of light that we've been able to use 267 00:15:14,760 --> 00:15:18,240 Speaker 1: in in mass productions. So if you rewind several decades ago, 268 00:15:18,280 --> 00:15:21,440 Speaker 1: we were using larger wavelengths of light that we're uncapable 269 00:15:21,480 --> 00:15:24,440 Speaker 1: of producing the small feature sizes on silicon wafers that 270 00:15:24,480 --> 00:15:27,560 Speaker 1: we demand today. One of the reasons why I think 271 00:15:28,040 --> 00:15:32,200 Speaker 1: the chip episodes, well, why we've done so many chip episodes, 272 00:15:32,240 --> 00:15:34,880 Speaker 1: and why they're why they keep resonating. I think there's 273 00:15:34,880 --> 00:15:38,040 Speaker 1: a few things. I mean, one is there's the chip shortage, 274 00:15:38,200 --> 00:15:40,280 Speaker 1: and it should be noted that the shortage is actually 275 00:15:40,400 --> 00:15:42,960 Speaker 1: wore at the is not really at the advanced level. 276 00:15:42,960 --> 00:15:45,160 Speaker 1: It's a lot of cheap chips, etcetera. But we're starting 277 00:15:45,200 --> 00:15:50,480 Speaker 1: to the chip shortage that relates to automobiles, etcetera. Has 278 00:15:50,520 --> 00:15:53,200 Speaker 1: sort of brought people a lot of awareness about lack 279 00:15:53,280 --> 00:15:57,200 Speaker 1: of domestic US manufacturing capacity. I think another reason people 280 00:15:57,240 --> 00:16:00,800 Speaker 1: care about chips is obviously just the general explosion of 281 00:16:00,880 --> 00:16:04,600 Speaker 1: chip demand. Even where there isn't an acute shortage, there's 282 00:16:04,680 --> 00:16:07,200 Speaker 1: chips h and everything. And then I think the other 283 00:16:07,240 --> 00:16:10,400 Speaker 1: thing that makes it sort of an interesting story right 284 00:16:10,400 --> 00:16:13,200 Speaker 1: now is that, at least in the US and probably 285 00:16:13,240 --> 00:16:16,400 Speaker 1: elsewhere around the world, there is a rethinking about the 286 00:16:16,480 --> 00:16:21,400 Speaker 1: role of state capacity in and um state investment into 287 00:16:21,480 --> 00:16:23,800 Speaker 1: certain space. And of course, as we all know, the 288 00:16:23,880 --> 00:16:26,080 Speaker 1: chip industry and as you just talked about, the chip 289 00:16:26,120 --> 00:16:29,320 Speaker 1: industry overall really was sort of born out of defense, 290 00:16:29,360 --> 00:16:32,520 Speaker 1: so like sort of the ultimate in UH state investing 291 00:16:32,680 --> 00:16:38,320 Speaker 1: and government spending, and at various times throughout US history 292 00:16:38,360 --> 00:16:40,880 Speaker 1: at least we seem to go in waves of how 293 00:16:41,000 --> 00:16:43,200 Speaker 1: much the government wants to get in to protect the 294 00:16:43,280 --> 00:16:46,040 Speaker 1: chip sector, to invest in the chip sector, to build 295 00:16:46,160 --> 00:16:50,000 Speaker 1: and bolster a homegrown chip sector. You mentioned the sort 296 00:16:50,040 --> 00:16:53,480 Speaker 1: of stress and tension with the Japanese or reliance on 297 00:16:53,560 --> 00:16:56,760 Speaker 1: Japanese companies in the eighties and nineties that seemed to 298 00:16:57,120 --> 00:17:00,160 Speaker 1: produce a wave of um sort of defensive and use. 299 00:17:00,400 --> 00:17:03,720 Speaker 1: Perhaps it could be characterized. Talk to us about how 300 00:17:03,760 --> 00:17:07,080 Speaker 1: a SML fits into that in terms of you know, 301 00:17:07,119 --> 00:17:09,159 Speaker 1: when we talked about to say, the history of t SMC, 302 00:17:09,800 --> 00:17:12,159 Speaker 1: that was clearly in part it was a very like 303 00:17:12,240 --> 00:17:15,840 Speaker 1: public sort of private venture. There was the government backed 304 00:17:15,840 --> 00:17:18,680 Speaker 1: it up under the condition that it could raise private 305 00:17:18,960 --> 00:17:21,520 Speaker 1: foreign money as well. Talk to us about the role 306 00:17:21,560 --> 00:17:25,240 Speaker 1: of like public money in the creation of a SML. 307 00:17:25,880 --> 00:17:29,040 Speaker 1: So a sm L emerged first as a division of 308 00:17:29,040 --> 00:17:31,840 Speaker 1: of Phillips, the Dutch electronic company UM and it was 309 00:17:32,000 --> 00:17:36,280 Speaker 1: spun out in at a time when the Europeanship industry 310 00:17:36,400 --> 00:17:39,240 Speaker 1: was relatively small as a player on the world stage. 311 00:17:39,240 --> 00:17:40,720 Speaker 1: It was the U S and Japan at the time 312 00:17:40,720 --> 00:17:43,280 Speaker 1: that were the biggest players, and there were a variety 313 00:17:43,320 --> 00:17:46,879 Speaker 1: of Dutch and European Union programs to support R and 314 00:17:47,000 --> 00:17:49,960 Speaker 1: D at s mL. But for a SML in particular, 315 00:17:50,400 --> 00:17:53,320 Speaker 1: actually the most important government support was from the U 316 00:17:53,359 --> 00:17:56,399 Speaker 1: S government because in in the nineteen nineties, when the 317 00:17:56,480 --> 00:18:00,240 Speaker 1: investments and EUV were being made, Intel, which at the 318 00:18:00,240 --> 00:18:03,880 Speaker 1: time was the the industry leader in hip making, decided 319 00:18:03,880 --> 00:18:05,800 Speaker 1: to take a big bet on eu V being the 320 00:18:05,840 --> 00:18:11,119 Speaker 1: next lithography technology and established a consortium of a number 321 00:18:11,119 --> 00:18:13,960 Speaker 1: of private ship firms and a number of US national labs, 322 00:18:14,480 --> 00:18:17,359 Speaker 1: uh the Lawrence Livermore for example, that would work together 323 00:18:17,480 --> 00:18:21,720 Speaker 1: to produce prototype UV machines. And so the technology that 324 00:18:22,200 --> 00:18:25,199 Speaker 1: we use today an SML Systems really comes from this 325 00:18:25,280 --> 00:18:28,320 Speaker 1: work with US national labs. It was largely funded by 326 00:18:28,359 --> 00:18:31,760 Speaker 1: industry but using the scientists there. And at the time 327 00:18:32,080 --> 00:18:34,879 Speaker 1: there was some interest in trying to turn the technology 328 00:18:34,920 --> 00:18:37,440 Speaker 1: over to a US company to producing commercialized on the 329 00:18:37,440 --> 00:18:40,200 Speaker 1: grounds that it came largely out of US national labs, 330 00:18:40,240 --> 00:18:42,760 Speaker 1: but there was no US lithography firm at the time 331 00:18:42,800 --> 00:18:45,919 Speaker 1: that was seen as a credible candidate to commercialize. If 332 00:18:45,920 --> 00:18:48,919 Speaker 1: the options were NICON or a SML. Given the tensions 333 00:18:48,920 --> 00:18:52,280 Speaker 1: with Japan, a SML was seen as the least risky option, 334 00:18:52,680 --> 00:18:55,240 Speaker 1: and also they had a long track record of producing 335 00:18:55,359 --> 00:18:58,640 Speaker 1: quality machines. And so we've got this strange situation now 336 00:18:58,680 --> 00:19:02,160 Speaker 1: where a lot of the core technology in this machinery 337 00:19:02,200 --> 00:19:05,480 Speaker 1: that's assembled in the Netherlands actually comes out of California. 338 00:19:05,600 --> 00:19:08,360 Speaker 1: And indeed SML has actually bought a number of companies 339 00:19:08,359 --> 00:19:10,760 Speaker 1: over the course of the past couple of decades in 340 00:19:10,800 --> 00:19:13,320 Speaker 1: California as well. So there's a lot of US technology 341 00:19:13,840 --> 00:19:16,840 Speaker 1: in a SML Systems partially funded by the US government. 342 00:19:17,040 --> 00:19:19,760 Speaker 1: Could you imagine something like that happening today? Like, I 343 00:19:19,840 --> 00:19:23,520 Speaker 1: just think the environment is so different, and the idea 344 00:19:23,560 --> 00:19:25,879 Speaker 1: of like the U S government funding a technology and 345 00:19:25,920 --> 00:19:28,880 Speaker 1: then deciding like, well, okay, I guess the best company 346 00:19:29,000 --> 00:19:32,119 Speaker 1: to actually make this stuff is over in the Netherlands, 347 00:19:32,119 --> 00:19:33,960 Speaker 1: so we'll just let them do it and give up 348 00:19:34,040 --> 00:19:37,560 Speaker 1: like a key component of a highly competitive supply chain. 349 00:19:37,600 --> 00:19:43,320 Speaker 1: It just seems so so unlikely in the current environment. Yeah, 350 00:19:43,359 --> 00:19:45,240 Speaker 1: it's it's an interesting question. On the one hand, you 351 00:19:45,280 --> 00:19:47,840 Speaker 1: do hear a lot of conversation in Washington, d C. 352 00:19:48,000 --> 00:19:50,760 Speaker 1: About joint R and D project with Allied countries, and 353 00:19:50,840 --> 00:19:52,800 Speaker 1: in some ways this is a perfect example of this. 354 00:19:52,960 --> 00:19:55,200 Speaker 1: I think. The other thing is that a SML is 355 00:19:55,240 --> 00:19:58,040 Speaker 1: a Dutch company, But if you look at the components 356 00:19:58,080 --> 00:20:01,560 Speaker 1: of their EUV machines, for example, they're sourcing from all 357 00:20:01,600 --> 00:20:05,040 Speaker 1: around Europe, around the US and really worldwide. So to 358 00:20:05,119 --> 00:20:08,000 Speaker 1: describe them as as a Dutch company misses the fact 359 00:20:08,040 --> 00:20:11,080 Speaker 1: that you can't produce an EUV system was for example, 360 00:20:11,119 --> 00:20:13,639 Speaker 1: the light source, which is produced by an S and 361 00:20:13,720 --> 00:20:16,080 Speaker 1: L subsidiary in San Diego, so that there are a 362 00:20:16,119 --> 00:20:19,400 Speaker 1: Dutch company S, but they're really a global supply chain 363 00:20:19,480 --> 00:20:22,280 Speaker 1: that's focused on on the U S and Europe. So 364 00:20:22,359 --> 00:20:25,679 Speaker 1: this is interesting because you mentioned that, Okay, at the 365 00:20:25,800 --> 00:20:29,560 Speaker 1: time that the technology was sort of they decided a 366 00:20:29,640 --> 00:20:32,800 Speaker 1: SML would be the most credible entity to commercialize this 367 00:20:33,119 --> 00:20:37,040 Speaker 1: sort of US funded technology. There was this view that, Okay, 368 00:20:37,160 --> 00:20:41,840 Speaker 1: it was better them than a Japanese player, in part 369 00:20:41,920 --> 00:20:45,679 Speaker 1: because we already had anxiety about our reliance on Japanese 370 00:20:45,720 --> 00:20:49,160 Speaker 1: chips at the time for other other chips, including DAM. 371 00:20:49,200 --> 00:20:52,160 Speaker 1: How much are the same dynamics essentially in play when 372 00:20:52,160 --> 00:20:56,800 Speaker 1: people think about the geopolitics of chips. Obviously one of 373 00:20:56,840 --> 00:20:59,640 Speaker 1: the things that you know, we talked about anxiety about 374 00:20:59,680 --> 00:21:03,399 Speaker 1: how my we rely on Taiwan. There's perhaps some anxiety 375 00:21:03,560 --> 00:21:07,240 Speaker 1: about the domestic homegrown chip industry. Although China seems to 376 00:21:07,280 --> 00:21:10,840 Speaker 1: be a several years behind in terms of mainland chip technology. 377 00:21:11,280 --> 00:21:13,920 Speaker 1: How much does it still sort of benefit everyone this 378 00:21:14,000 --> 00:21:18,640 Speaker 1: idea that this crucial component player is not part of 379 00:21:18,840 --> 00:21:22,080 Speaker 1: either in US or Asia. That's an interesting question. I 380 00:21:22,080 --> 00:21:25,480 Speaker 1: think certainly if you're a Chinese customer of SML, you're 381 00:21:25,560 --> 00:21:28,159 Speaker 1: pleased that it's not a U S company, But the 382 00:21:28,240 --> 00:21:30,760 Speaker 1: reality is that if if the US wanted to use 383 00:21:30,800 --> 00:21:34,119 Speaker 1: actually poor controls to constrict SML sales to China, that 384 00:21:34,160 --> 00:21:36,879 Speaker 1: wouldn't be very difficult to do. A SML already doesn't 385 00:21:36,920 --> 00:21:40,800 Speaker 1: send it's eu V machines to China. In theory, that's 386 00:21:40,840 --> 00:21:44,200 Speaker 1: because of Dutch restrictions. In reality, it's because of US 387 00:21:44,240 --> 00:21:47,760 Speaker 1: pressure on the Netherlands to impose these restrictions. And there's 388 00:21:47,800 --> 00:21:52,160 Speaker 1: discussion um in in Washington and Japan elsewhere about whether 389 00:21:52,200 --> 00:21:54,840 Speaker 1: there ought to be stricter limits um the type of 390 00:21:54,840 --> 00:21:58,080 Speaker 1: lithography machines you can sell to China, and legally there's 391 00:21:58,160 --> 00:22:00,800 Speaker 1: there's nothing that would really stop the US from posting 392 00:22:00,800 --> 00:22:05,720 Speaker 1: those restrictions unilaterally. Is the concerning that if those machines 393 00:22:05,920 --> 00:22:09,919 Speaker 1: were shipped to China, that they would be able to 394 00:22:10,200 --> 00:22:15,320 Speaker 1: UH that would accelerate China's semiconductor capabilities, or that literally 395 00:22:15,359 --> 00:22:18,040 Speaker 1: having them in Chinese hands would then maybe allow them 396 00:22:18,080 --> 00:22:21,760 Speaker 1: to be more easily sort of deconstructed and reverse engineered 397 00:22:22,320 --> 00:22:25,360 Speaker 1: and that would be a big knowledge transfer. No, it's 398 00:22:25,400 --> 00:22:28,439 Speaker 1: the former. I think if you just receive an SMaL machine, 399 00:22:28,480 --> 00:22:30,960 Speaker 1: you have no idea how to produce it. Okay, Okay, 400 00:22:31,119 --> 00:22:33,600 Speaker 1: it's that the more advanced cocoric machines you have, the 401 00:22:33,640 --> 00:22:36,480 Speaker 1: more advanced shipmaker you have, the stronger the Chinese ecosystem is. 402 00:22:37,240 --> 00:22:40,960 Speaker 1: So how big of an impediment is not having access 403 00:22:41,040 --> 00:22:44,840 Speaker 1: to a sml S eu V technology to Beijing's like 404 00:22:44,960 --> 00:22:50,520 Speaker 1: overall semiconductor development drive, Like, is it such an essential 405 00:22:50,520 --> 00:22:53,080 Speaker 1: piece of technology that it basically means they're on a 406 00:22:53,119 --> 00:22:58,200 Speaker 1: completely different footing to something like TSMC. That's right. For now, 407 00:22:58,240 --> 00:23:02,280 Speaker 1: there's there's no viable of producing the most advanced chips 408 00:23:02,280 --> 00:23:06,639 Speaker 1: with the smallest features without using UV. There are some 409 00:23:06,640 --> 00:23:09,000 Speaker 1: some scientists who think there might theoretically be ways to 410 00:23:09,000 --> 00:23:12,040 Speaker 1: get around it, but for the next decade there's there's 411 00:23:12,080 --> 00:23:14,280 Speaker 1: just no choice but to use a SMLS machinery if 412 00:23:14,320 --> 00:23:18,240 Speaker 1: you want to produce the smallest chips. So what are 413 00:23:18,320 --> 00:23:20,560 Speaker 1: you know, let's talk a little bit more about a 414 00:23:20,760 --> 00:23:24,440 Speaker 1: s MLS constraints. Everyone this year is becoming aware of, like, 415 00:23:24,600 --> 00:23:28,159 Speaker 1: you know, constraints, and there's only so much boundary capacity 416 00:23:28,520 --> 00:23:31,960 Speaker 1: in the world at any given time. The entities that 417 00:23:32,040 --> 00:23:34,680 Speaker 1: wanted to buy cheap chips that go into cars sort 418 00:23:34,680 --> 00:23:36,920 Speaker 1: of got shut out because they canceled their order for 419 00:23:36,920 --> 00:23:38,600 Speaker 1: a while. And now they're scrambling and it might be 420 00:23:38,680 --> 00:23:41,399 Speaker 1: years before they could catch up again. So we know 421 00:23:41,560 --> 00:23:45,680 Speaker 1: that that's constraints. How strained is a s mls own 422 00:23:46,080 --> 00:23:49,000 Speaker 1: capacity to grow and where do they face Is it 423 00:23:49,160 --> 00:23:51,680 Speaker 1: just in the complexity of the supply chain? Is it 424 00:23:51,800 --> 00:23:55,280 Speaker 1: in raw materials? Like what are their constraints? It's mostly 425 00:23:55,280 --> 00:23:58,600 Speaker 1: in the supply chain complexity. So SML last year's shipped 426 00:23:58,680 --> 00:24:02,439 Speaker 1: thirty one EUV machine sans, so we're talking getting one 427 00:24:02,520 --> 00:24:04,240 Speaker 1: or two more machines out of the out of their 428 00:24:04,280 --> 00:24:07,240 Speaker 1: production process is something that's hard to do because each 429 00:24:07,280 --> 00:24:10,760 Speaker 1: of their suppliers is similarly constrained. And the ability to 430 00:24:10,840 --> 00:24:14,320 Speaker 1: ramp up manufacturing, you know, this isn't high volume manufacturing 431 00:24:14,320 --> 00:24:17,240 Speaker 1: when you're producing thirty one machines a year, and because 432 00:24:17,280 --> 00:24:20,040 Speaker 1: their supply chain has so many specialized parts solely for 433 00:24:20,080 --> 00:24:23,199 Speaker 1: their machines, their suppliers are producing thirty one or so 434 00:24:23,560 --> 00:24:26,439 Speaker 1: of the components needed each year, and so there's just 435 00:24:26,480 --> 00:24:28,880 Speaker 1: no way to ramp How many three hundred billion dollar 436 00:24:29,000 --> 00:24:33,679 Speaker 1: companies in the world make produce thirty one make thirty 437 00:24:33,680 --> 00:24:35,720 Speaker 1: one units a year, so we're talking like each one 438 00:24:35,800 --> 00:24:38,119 Speaker 1: is like half a goon or something. There's thirty one 439 00:24:38,160 --> 00:24:41,199 Speaker 1: of the of EUV machines. They will also sell some 440 00:24:41,240 --> 00:24:45,120 Speaker 1: of the older equipment, but they sold units in total 441 00:24:45,200 --> 00:24:47,119 Speaker 1: last year, so it's still a tiny number of units. 442 00:24:47,119 --> 00:24:49,240 Speaker 1: It's still not very much. How much is it if 443 00:24:49,280 --> 00:24:51,439 Speaker 1: you or I wanted to pull together and by a 444 00:24:51,600 --> 00:24:53,919 Speaker 1: EUV machine? Like what are they? What are they retail for? 445 00:24:54,320 --> 00:24:57,480 Speaker 1: Average average revenue per UV machine last year was around 446 00:24:57,480 --> 00:25:01,880 Speaker 1: a hundred forty million euros. Got it? Okay, So this 447 00:25:01,920 --> 00:25:03,879 Speaker 1: is something that comes up a lot in our supply 448 00:25:04,040 --> 00:25:08,320 Speaker 1: chain discussions. But like how does ordering actually work and 449 00:25:08,480 --> 00:25:12,080 Speaker 1: is there preference given to certain customers over others? Like, 450 00:25:12,200 --> 00:25:14,720 Speaker 1: you know, if one company wants to buy an EUV machine, 451 00:25:14,760 --> 00:25:16,760 Speaker 1: I imagine there there are plenty of other companies that 452 00:25:16,800 --> 00:25:18,639 Speaker 1: want to do the same thing. For a limited supply 453 00:25:19,400 --> 00:25:22,320 Speaker 1: How does a SML actually make the decisions about who 454 00:25:22,359 --> 00:25:25,400 Speaker 1: gets allocated what? Also, how long does it take? Like 455 00:25:25,800 --> 00:25:30,159 Speaker 1: what's the waiting time to actually get one of these things? Yeah, 456 00:25:30,440 --> 00:25:33,480 Speaker 1: we don't really know the details as to how SML 457 00:25:33,560 --> 00:25:36,560 Speaker 1: decides to allocate the number of potential customers for a 458 00:25:36,640 --> 00:25:39,800 Speaker 1: hundred fifty million dollar machine. Is is limited? It's a 459 00:25:39,840 --> 00:25:43,160 Speaker 1: half dozen potential customers in the world, you be realistically 460 00:25:43,160 --> 00:25:45,080 Speaker 1: looking to buy one. But if you look at the 461 00:25:45,119 --> 00:25:48,600 Speaker 1: main customers, it's t SMC, which has around half of 462 00:25:48,600 --> 00:25:52,680 Speaker 1: operating the UV machines in its own fabs, Samsung, Intel, 463 00:25:52,760 --> 00:25:56,800 Speaker 1: a handful of others, and there's almost certainly a premium 464 00:25:56,800 --> 00:26:00,520 Speaker 1: that t SMC is paid for getting so many machines valuable. 465 00:26:00,600 --> 00:26:02,320 Speaker 1: If if if a new company came online and wanted to 466 00:26:02,359 --> 00:26:04,800 Speaker 1: buy machines, they'd face a weight of at least a 467 00:26:04,800 --> 00:26:08,679 Speaker 1: couple of years because capacity has been purchased and advanced. 468 00:26:09,160 --> 00:26:11,600 Speaker 1: Intel has said it's going to be the first customer 469 00:26:11,680 --> 00:26:15,000 Speaker 1: of the next generation EUV machine, which would be online 470 00:26:15,040 --> 00:26:18,960 Speaker 1: around Presumably it's paid something for the right to get 471 00:26:18,960 --> 00:26:21,760 Speaker 1: the first generation, but we don't know any of the details. Yeah, 472 00:26:21,880 --> 00:26:23,840 Speaker 1: speaking of Intel, and I want to back up to 473 00:26:23,880 --> 00:26:26,640 Speaker 1: something we talked about earlier. Why was this never part 474 00:26:26,680 --> 00:26:30,800 Speaker 1: of Intel's own ambitions? Because you know, over the years, 475 00:26:31,080 --> 00:26:34,760 Speaker 1: I guess the degree to which Intel has wanted to 476 00:26:34,840 --> 00:26:37,679 Speaker 1: be an i P first company or a manufacturing company 477 00:26:37,680 --> 00:26:39,920 Speaker 1: at Waxes and wind, so at one point it was 478 00:26:39,960 --> 00:26:43,560 Speaker 1: a manufacturing powerhouse, then has sort of scaled. Back then 479 00:26:43,680 --> 00:26:45,400 Speaker 1: it was more of an I P company, and that's 480 00:26:45,400 --> 00:26:49,080 Speaker 1: sally the anxiety these days now they seem to want 481 00:26:49,200 --> 00:26:52,680 Speaker 1: to get back into being uh manufacturing, and the new 482 00:26:52,760 --> 00:26:55,320 Speaker 1: CEO has made a point of like, we're not going 483 00:26:55,400 --> 00:27:00,879 Speaker 1: to give up on being a manufacturing powerhouse. Why was 484 00:27:01,000 --> 00:27:07,359 Speaker 1: lithography or advanced lithography never part of the Intel strategy? Well, 485 00:27:07,359 --> 00:27:09,920 Speaker 1: when INTIL was founded, it was. It was founded at 486 00:27:09,920 --> 00:27:12,560 Speaker 1: a time where you could already buy lithography equipment on 487 00:27:12,600 --> 00:27:15,040 Speaker 1: the open market, so they always decided they were going 488 00:27:15,119 --> 00:27:19,480 Speaker 1: to produce ships, but by lithography machines from suppliers. Over 489 00:27:19,520 --> 00:27:22,040 Speaker 1: their fifty years, they've been one of the biggest buyers 490 00:27:22,040 --> 00:27:25,119 Speaker 1: of lithography equipment in the industry, and the development of 491 00:27:25,160 --> 00:27:28,200 Speaker 1: e V actually wouldn't have happened without Intel. When Andy 492 00:27:28,240 --> 00:27:31,080 Speaker 1: Grove was still the Intel CEO in the early nine nineties, 493 00:27:31,560 --> 00:27:33,920 Speaker 1: he made the first big bet on the development of 494 00:27:33,960 --> 00:27:36,760 Speaker 1: EV lithography, putting up two or million dollars in nearly 495 00:27:37,160 --> 00:27:39,959 Speaker 1: nines to begin to develop this on the grounds not 496 00:27:39,960 --> 00:27:42,760 Speaker 1: that Intel was ever going to produce lithography equipment, but 497 00:27:42,800 --> 00:27:46,200 Speaker 1: that it would eventually need EUV to produce the most 498 00:27:46,240 --> 00:27:50,320 Speaker 1: advanced ships. And even as as recently as when a 499 00:27:50,440 --> 00:27:53,560 Speaker 1: SML needed to raise more capital to to fund its 500 00:27:53,560 --> 00:27:57,200 Speaker 1: development of UV. It went to Intel, t SMC, and Samsung, 501 00:27:57,480 --> 00:27:59,640 Speaker 1: and Intel was the biggest investor in a SML at 502 00:27:59,640 --> 00:28:02,480 Speaker 1: the time, putting in a several billion dollars to help 503 00:28:02,520 --> 00:28:06,080 Speaker 1: fund SML's development. So until quite recently, it seemed like 504 00:28:06,119 --> 00:28:09,960 Speaker 1: Intel would be the biggest user of e V lithography machines. 505 00:28:10,440 --> 00:28:13,560 Speaker 1: It's only in the past couple of years that Intel decided, 506 00:28:14,080 --> 00:28:16,560 Speaker 1: in what looks to be a mistake in hindsight, that 507 00:28:16,920 --> 00:28:20,160 Speaker 1: EV wouldn't be ready by around now, where t SMC 508 00:28:20,359 --> 00:28:23,080 Speaker 1: maybe opposite that that EUV would be ready for high 509 00:28:23,160 --> 00:28:26,240 Speaker 1: value manufacturing. And t SMC was proved right, which is 510 00:28:26,240 --> 00:28:28,159 Speaker 1: why it's done so well the past couple of years, 511 00:28:28,200 --> 00:28:31,480 Speaker 1: and Intel has proven wrong, which I think most observers 512 00:28:31,480 --> 00:28:34,119 Speaker 1: can explain some of the manufacturing problems it's had in 513 00:28:34,160 --> 00:28:36,960 Speaker 1: recent years by not using EUV and trying to use 514 00:28:37,000 --> 00:28:40,000 Speaker 1: older versions of lithography to produce its most advanced ships. 515 00:28:40,480 --> 00:28:43,920 Speaker 1: Now Intel is changing it's it's it's plans, it's it 516 00:28:43,960 --> 00:28:46,440 Speaker 1: says it's investing very heavily in UV, but it's gonna 517 00:28:46,480 --> 00:28:49,160 Speaker 1: take a couple of years for them to learn how 518 00:28:49,200 --> 00:28:52,720 Speaker 1: to actually use the UV and high volume manufacturing. Wait, 519 00:28:52,760 --> 00:28:55,520 Speaker 1: so I have a slightly related question, although maybe it's 520 00:28:55,560 --> 00:28:59,800 Speaker 1: sort of reversed, I guess, but like, given a SML 521 00:29:00,840 --> 00:29:05,880 Speaker 1: competitive edge in producing a key technological component for semiconductors, 522 00:29:07,000 --> 00:29:11,120 Speaker 1: could they ever have just gone into making semiconductors themselves? 523 00:29:11,240 --> 00:29:13,480 Speaker 1: Like if they have a monopoly on this technology, no 524 00:29:13,520 --> 00:29:15,840 Speaker 1: one can do it as well as them, Like why 525 00:29:15,880 --> 00:29:20,440 Speaker 1: not just start making the finished product yourself. To make 526 00:29:20,440 --> 00:29:22,880 Speaker 1: a chip, you not only need lithography, which is one 527 00:29:22,880 --> 00:29:25,360 Speaker 1: of the key steps, but there are other steps as well. 528 00:29:25,440 --> 00:29:28,880 Speaker 1: You need to be able to deposit films of materials 529 00:29:28,960 --> 00:29:32,360 Speaker 1: um with with atomic level precision. There's there are different 530 00:29:32,360 --> 00:29:35,720 Speaker 1: companies Applied Materials, for example, in California that make that 531 00:29:35,800 --> 00:29:38,959 Speaker 1: type of equipment. You need measurement equipment that can measure 532 00:29:39,360 --> 00:29:42,840 Speaker 1: individual atomic level errors in your final chips to make 533 00:29:42,840 --> 00:29:46,360 Speaker 1: sure you understand what errors you have. That's a different 534 00:29:46,360 --> 00:29:48,800 Speaker 1: set of companies that makes that equipment. UM. So there's 535 00:29:48,920 --> 00:29:51,720 Speaker 1: a lot of different specialized equipment that you need to 536 00:29:52,160 --> 00:29:55,080 Speaker 1: make chips. The cost of a new fab that for example, 537 00:29:55,120 --> 00:29:57,480 Speaker 1: t SMC will build more than half of that cost 538 00:29:57,560 --> 00:29:59,440 Speaker 1: is the equipment that goes in it and a SML 539 00:29:59,720 --> 00:30:02,040 Speaker 1: and as ptography machines are a critical part of that, 540 00:30:02,120 --> 00:30:05,320 Speaker 1: but that's far from enough to make chips, and SML 541 00:30:05,440 --> 00:30:08,719 Speaker 1: specialty is really solely in lithography UH and not at 542 00:30:08,720 --> 00:30:11,000 Speaker 1: all in deposition or matching or the other types of 543 00:30:11,240 --> 00:30:14,360 Speaker 1: equipment that you need to actually make finished chips. This 544 00:30:14,440 --> 00:30:16,320 Speaker 1: at this point is so fascinating to me, like to 545 00:30:16,400 --> 00:30:19,280 Speaker 1: think about, like, okay, a s mL among the many 546 00:30:19,360 --> 00:30:22,840 Speaker 1: extraordinary things. They also lay claim to having the flattest 547 00:30:22,880 --> 00:30:26,400 Speaker 1: service in the world, and presumably in order to get 548 00:30:26,400 --> 00:30:28,959 Speaker 1: the flattest service of the world, there's some technology, as 549 00:30:29,000 --> 00:30:31,040 Speaker 1: you sort of just mentioned, that has to be able 550 00:30:31,040 --> 00:30:34,680 Speaker 1: to measure flatness and actually measure if the services not 551 00:30:34,800 --> 00:30:37,320 Speaker 1: flat and it sort of speaks to like you know, 552 00:30:37,400 --> 00:30:39,640 Speaker 1: we think about like in the US these days, and 553 00:30:39,760 --> 00:30:42,800 Speaker 1: there is UH. There's a bill in d C that's 554 00:30:42,880 --> 00:30:48,440 Speaker 1: designed to invest in UH, designed to bolster US capacity. 555 00:30:48,840 --> 00:30:51,680 Speaker 1: And again that's part of why you keep having these discussions, 556 00:30:51,720 --> 00:30:54,880 Speaker 1: because there's this effort on the way. It's kind of bipartisan. 557 00:30:54,960 --> 00:30:57,440 Speaker 1: The bill might pass, it might not pass, but there's 558 00:30:57,440 --> 00:31:01,440 Speaker 1: this sort of bipartisan interest to bolster domestic capacity. But 559 00:31:01,520 --> 00:31:04,440 Speaker 1: I don't even know what that means sometimes because obviously, 560 00:31:04,440 --> 00:31:08,280 Speaker 1: as you describe, the internationalization and complexity of the chip 561 00:31:08,480 --> 00:31:13,120 Speaker 1: supply chain is so extreme, and as we've talked about 562 00:31:13,120 --> 00:31:16,240 Speaker 1: with other guests, chips across borders a million times before 563 00:31:16,240 --> 00:31:20,520 Speaker 1: they arrive in your xbox or your iPhone or whatever. 564 00:31:20,560 --> 00:31:23,160 Speaker 1: It is, Like, what does it even mean in your 565 00:31:23,240 --> 00:31:25,160 Speaker 1: view just to think about like this question of like 566 00:31:25,440 --> 00:31:30,840 Speaker 1: expanding domestic capacity in an industry that but that just 567 00:31:31,080 --> 00:31:36,240 Speaker 1: is so extremely fragmented and international. Yeah, I think the 568 00:31:36,280 --> 00:31:38,000 Speaker 1: first thing is you've got to be specific as to 569 00:31:38,080 --> 00:31:41,520 Speaker 1: what type of capacity you wanted to spand expand domestically. 570 00:31:41,720 --> 00:31:44,760 Speaker 1: Certainly the US could expand production of chips domestically if 571 00:31:44,800 --> 00:31:47,800 Speaker 1: we wanted to, but that wouldn't have any effect on 572 00:31:47,880 --> 00:31:51,080 Speaker 1: the reality that there's no way to buy lithography equipment, 573 00:31:51,120 --> 00:31:55,560 Speaker 1: for example, except from foreign suppliers, either Nicon or a SML. 574 00:31:56,160 --> 00:31:58,480 Speaker 1: I think domestic production is a is a great thing 575 00:31:58,600 --> 00:32:01,160 Speaker 1: to support UM, but thesis that we're going to have 576 00:32:01,360 --> 00:32:05,160 Speaker 1: a entirely domestically produced supply chain is a fantasy. The 577 00:32:05,200 --> 00:32:08,160 Speaker 1: only reason that we're able to produce ships with such 578 00:32:08,160 --> 00:32:10,680 Speaker 1: small features is because we're able to take advantage of 579 00:32:11,000 --> 00:32:14,400 Speaker 1: expertise from companies in many different countries around the world, 580 00:32:14,800 --> 00:32:17,000 Speaker 1: and there's no one in the industry who thinks there's 581 00:32:17,040 --> 00:32:19,520 Speaker 1: any conceivable future even if you're to spend a trillion 582 00:32:19,520 --> 00:32:22,400 Speaker 1: dollars over a decade where you'd get a domestically produced 583 00:32:22,400 --> 00:32:24,760 Speaker 1: supply chain as anywhere near as efficient. Is what we've 584 00:32:24,760 --> 00:32:27,840 Speaker 1: got now, you know. I think the supply chain risk 585 00:32:28,240 --> 00:32:31,400 Speaker 1: discussion is often takes place at a thirty foot level, 586 00:32:31,440 --> 00:32:33,040 Speaker 1: when what you really need to look at is what 587 00:32:33,120 --> 00:32:36,200 Speaker 1: of the specific components you're worried about, Are there specific 588 00:32:36,280 --> 00:32:39,959 Speaker 1: suppliers you're worried about, and how can you mitigate those 589 00:32:40,000 --> 00:32:43,440 Speaker 1: specific risks? But just talking about domestic versus foreign production 590 00:32:43,480 --> 00:32:45,800 Speaker 1: is not nearly specific enough to have any sort of 591 00:32:46,000 --> 00:32:49,160 Speaker 1: real meaning. So we kind of mentioned this in the intro. 592 00:32:49,400 --> 00:32:52,600 Speaker 1: But again, one of the themes that comes up repeatedly 593 00:32:52,760 --> 00:32:56,440 Speaker 1: on these episodes is the idea of supply chains all 594 00:32:56,440 --> 00:32:59,600 Speaker 1: the way down. So if there's a bottleneck in one thing, 595 00:32:59,840 --> 00:33:03,760 Speaker 1: there's probably a bottleneck um and a component and even 596 00:33:03,840 --> 00:33:06,560 Speaker 1: smaller component that leads into that one thing. So if 597 00:33:06,560 --> 00:33:11,720 Speaker 1: there's a bottleneck in lithography equipment that's impacting semiconductors, I 598 00:33:11,720 --> 00:33:14,640 Speaker 1: guess my question is is there a bottleneck in something 599 00:33:14,680 --> 00:33:17,840 Speaker 1: that goes into the lithography machines or the eu V 600 00:33:18,080 --> 00:33:22,360 Speaker 1: machines that is preventing a s m L from expanding production. 601 00:33:23,800 --> 00:33:26,800 Speaker 1: It certainly could be there's not enough public information about 602 00:33:26,920 --> 00:33:30,320 Speaker 1: SML supply chain to know, and it's very plausible that 603 00:33:30,400 --> 00:33:33,560 Speaker 1: a SML, despite being real experts at management supply chain, 604 00:33:33,600 --> 00:33:38,040 Speaker 1: doesn't always uh No. They report having around four thousands suppliers, 605 00:33:38,520 --> 00:33:40,640 Speaker 1: and all of their suppliers who you speak to will 606 00:33:40,680 --> 00:33:44,160 Speaker 1: say they ask lots of questions about their suppliers suppliers, 607 00:33:44,960 --> 00:33:48,320 Speaker 1: But the reality is that there are you know, multiple 608 00:33:48,400 --> 00:33:51,160 Speaker 1: orders of magnitude more suppliers of their suppliers and so 609 00:33:51,240 --> 00:33:54,320 Speaker 1: tracing them all down the chain is is basically impossible. 610 00:33:54,440 --> 00:33:56,720 Speaker 1: So what a SML tries to do is understand what 611 00:33:56,760 --> 00:33:59,600 Speaker 1: other biggest risks. They've even gone so far as to 612 00:33:59,600 --> 00:34:02,920 Speaker 1: purchase some of their suppliers to gain more detailed control 613 00:34:03,080 --> 00:34:08,200 Speaker 1: over managing those risks. But they simply can't know every 614 00:34:08,280 --> 00:34:11,840 Speaker 1: ultimate component that goes into all of all their suppliers systems, 615 00:34:12,160 --> 00:34:14,239 Speaker 1: and so that's that's where the supply chain management just 616 00:34:14,280 --> 00:34:17,000 Speaker 1: becomes extraordinarily difficult. Now what they've been really good at 617 00:34:17,320 --> 00:34:20,399 Speaker 1: I think better than their competitors over the past couple 618 00:34:20,440 --> 00:34:23,840 Speaker 1: of decades is managing that so it hasn't generally caused 619 00:34:24,760 --> 00:34:27,560 Speaker 1: any sort of serious delays. And one of the things 620 00:34:27,560 --> 00:34:30,319 Speaker 1: that they're known for with their customers is being able 621 00:34:30,360 --> 00:34:33,240 Speaker 1: to deliver mostly on time when they promise a machine 622 00:34:33,360 --> 00:34:35,399 Speaker 1: and managing this, which is something that no one else 623 00:34:35,440 --> 00:34:37,879 Speaker 1: has been able to manage. I think. The other thing 624 00:34:38,200 --> 00:34:40,200 Speaker 1: to note is that you know, when you've got this 625 00:34:40,280 --> 00:34:45,279 Speaker 1: equipment that is is manipulating individual atoms or trying to 626 00:34:46,120 --> 00:34:50,480 Speaker 1: trying to control the movement of light with extraordinary precision. 627 00:34:51,000 --> 00:34:52,960 Speaker 1: It's it's one thing to have a machine that will 628 00:34:53,000 --> 00:34:55,640 Speaker 1: do this once or twice or sporadically. It's another thing 629 00:34:55,680 --> 00:34:57,279 Speaker 1: to have a machine that will do this day and 630 00:34:57,360 --> 00:35:00,880 Speaker 1: day out, operating twenty four hours a day. And that's 631 00:35:00,920 --> 00:35:03,440 Speaker 1: that's the other thing that a SML has done very 632 00:35:03,480 --> 00:35:06,319 Speaker 1: successfully over the past couple of years. It was clear 633 00:35:06,360 --> 00:35:08,480 Speaker 1: as early as the nine nineties that it was possible 634 00:35:08,520 --> 00:35:13,040 Speaker 1: to make a chip with evy lithography. What's been difficult 635 00:35:13,160 --> 00:35:15,880 Speaker 1: is making millions of ships with the euvy lithography and 636 00:35:15,920 --> 00:35:18,719 Speaker 1: doing in a cost effective way. And that's what a 637 00:35:18,840 --> 00:35:22,360 Speaker 1: smlls really stood out is that their machines are rarely broken, 638 00:35:22,840 --> 00:35:26,879 Speaker 1: always functioning um They need less um, less tuning, less 639 00:35:26,920 --> 00:35:29,839 Speaker 1: cleaning than their competitors. That's what a sml has done 640 00:35:30,200 --> 00:35:32,680 Speaker 1: quite well. So it's not simply managing the physics, which 641 00:35:32,719 --> 00:35:35,120 Speaker 1: is very hard, but it's also making sure that you've 642 00:35:35,160 --> 00:35:38,319 Speaker 1: got this extraordinary precise physics that's always operating in these 643 00:35:38,360 --> 00:35:41,120 Speaker 1: act way you expected to operate. Yeah, I'm thinking back 644 00:35:41,160 --> 00:35:45,000 Speaker 1: to one of our discussions with HBS professor. Really, she 645 00:35:45,680 --> 00:35:48,279 Speaker 1: and I don't remember the math exactly, but if like 646 00:35:48,400 --> 00:35:52,160 Speaker 1: chip making is like a seven thousand step process, then 647 00:35:52,239 --> 00:35:56,480 Speaker 1: even you know, ninety nine point nine nine percent execution 648 00:35:56,600 --> 00:36:00,759 Speaker 1: in each step is insufficient in many cases because by 649 00:36:00,800 --> 00:36:04,000 Speaker 1: the time you're down to the seventh seven thousand, you've 650 00:36:04,080 --> 00:36:06,480 Speaker 1: like basically lost all your chips. I don't remember the 651 00:36:06,520 --> 00:36:09,640 Speaker 1: exact math, but it is interesting to think about like 652 00:36:09,960 --> 00:36:14,239 Speaker 1: building up that comp that competent now just in can 653 00:36:14,280 --> 00:36:16,640 Speaker 1: the machine make the chip, but can make it over 654 00:36:16,760 --> 00:36:20,920 Speaker 1: and over and over again without without many errors. If 655 00:36:20,920 --> 00:36:24,520 Speaker 1: you look at a s mls revenue statements, what you'll 656 00:36:24,520 --> 00:36:26,760 Speaker 1: find is they've got a growing share of their revenue 657 00:36:26,800 --> 00:36:29,600 Speaker 1: coming from services which is servicing the machines that they operate. 658 00:36:29,600 --> 00:36:34,160 Speaker 1: They've got staff in t SMCS, facilities in in Samsung, etcetera, 659 00:36:34,280 --> 00:36:35,960 Speaker 1: making sure that not only the machines are operating, but 660 00:36:36,000 --> 00:36:39,720 Speaker 1: they're operating exactly according to plan, they're not breaking down. 661 00:36:40,160 --> 00:36:42,640 Speaker 1: The other thing that SML is is doing more of 662 00:36:42,800 --> 00:36:46,360 Speaker 1: is managing the software for their machines. And the way 663 00:36:46,400 --> 00:36:49,280 Speaker 1: that lithography works at at the scale that we're talking 664 00:36:49,320 --> 00:36:52,520 Speaker 1: about is that if you want to print a certain pictures, 665 00:36:52,520 --> 00:36:55,360 Speaker 1: so you want to print an AX, you don't reflect 666 00:36:55,480 --> 00:36:58,080 Speaker 1: the shape of the ax on your silicon because the 667 00:36:58,160 --> 00:37:00,680 Speaker 1: way that light reflects if you print and actually get 668 00:37:00,719 --> 00:37:03,600 Speaker 1: something different, so you actually learn over time all of 669 00:37:03,640 --> 00:37:06,800 Speaker 1: the unexpected errors and light refraction and errors in the 670 00:37:06,800 --> 00:37:10,560 Speaker 1: way the chemicals react, and you print the error version 671 00:37:10,640 --> 00:37:11,960 Speaker 1: and then it will give you an act. And so 672 00:37:12,000 --> 00:37:15,880 Speaker 1: there's extradinally complex software that now tries to understand how 673 00:37:15,920 --> 00:37:18,399 Speaker 1: all these different effects work. And you can actually look 674 00:37:18,400 --> 00:37:21,399 Speaker 1: at the images that smls producing to produce a straight 675 00:37:21,440 --> 00:37:23,239 Speaker 1: line and it looks nothing like a straight line. And 676 00:37:23,239 --> 00:37:25,759 Speaker 1: so that that software as well as something that SML 677 00:37:25,840 --> 00:37:44,319 Speaker 1: has been focusing on. So one of the things that 678 00:37:44,600 --> 00:37:46,680 Speaker 1: we've talked about. You know, it's like the sort of 679 00:37:46,719 --> 00:37:50,239 Speaker 1: the nanometer wars and the obvious. You know, people talking 680 00:37:50,239 --> 00:37:54,680 Speaker 1: about Moore's law and whether it's t SMC or Intel 681 00:37:54,760 --> 00:37:57,040 Speaker 1: or anyone else, so they're always or a MG. Maybe 682 00:37:57,200 --> 00:38:01,120 Speaker 1: they're always bragging about like making smaller and smaller chips. 683 00:38:01,160 --> 00:38:02,920 Speaker 1: And one of the things that we learned is that 684 00:38:03,360 --> 00:38:06,359 Speaker 1: actually the chip companies all defined these measures a little 685 00:38:06,360 --> 00:38:09,000 Speaker 1: bit differently, so to some extent it's made up. But 686 00:38:09,120 --> 00:38:13,160 Speaker 1: how much are the chip manufacturers themselves as they advertised, 687 00:38:13,160 --> 00:38:15,080 Speaker 1: like Okay, we're gonna be able to make a seven 688 00:38:15,160 --> 00:38:18,120 Speaker 1: NIM chip or maybe a five NIM chip or whatever. 689 00:38:18,719 --> 00:38:24,120 Speaker 1: How much are there timelines dependent on a s m 690 00:38:24,320 --> 00:38:27,560 Speaker 1: l's I guess I would say a smls own learning curve. 691 00:38:28,000 --> 00:38:31,640 Speaker 1: And what are as a sort of monopoly provider, I 692 00:38:31,640 --> 00:38:34,080 Speaker 1: don't want to say monopoly, but is the sole provider 693 00:38:34,160 --> 00:38:37,480 Speaker 1: of the most cutting edge technology. What are the forces 694 00:38:37,920 --> 00:38:43,440 Speaker 1: that drive technological gains for a sm L itself. If 695 00:38:43,480 --> 00:38:46,640 Speaker 1: you look at how a SML has begun to roll 696 00:38:46,680 --> 00:38:49,719 Speaker 1: out at CV machines in into high volume manufacturing, which 697 00:38:49,760 --> 00:38:53,080 Speaker 1: has mostly been at t SMC, the learning has happened 698 00:38:53,120 --> 00:38:56,200 Speaker 1: collectively with a sm L and t SMC, so there's 699 00:38:56,239 --> 00:38:58,879 Speaker 1: been lots of SMA personnel that spend tons of time 700 00:38:58,880 --> 00:39:02,479 Speaker 1: in Taiwan and vice versa. So you really wouldn't have 701 00:39:02,800 --> 00:39:04,440 Speaker 1: the rollout of e V over the past couple of 702 00:39:04,480 --> 00:39:07,160 Speaker 1: years had you not have this collective effort between t 703 00:39:07,360 --> 00:39:10,879 Speaker 1: s MC and a SML. And that's that puts other 704 00:39:10,880 --> 00:39:14,239 Speaker 1: companies at a disadvantage because t SMC knows better than 705 00:39:14,280 --> 00:39:18,360 Speaker 1: anyone how a s MLS machines actually work in practice, 706 00:39:18,840 --> 00:39:22,000 Speaker 1: and and the highlight manufacturing is really crucial to understanding 707 00:39:22,080 --> 00:39:24,920 Speaker 1: how how these machines work, because you don't really know 708 00:39:25,040 --> 00:39:27,600 Speaker 1: until you've got them in production. And once you've got 709 00:39:27,600 --> 00:39:30,600 Speaker 1: them in production, you've got thousands and thousands and thousands 710 00:39:30,600 --> 00:39:34,560 Speaker 1: of iterations every single day where you can understand what 711 00:39:34,680 --> 00:39:38,280 Speaker 1: the errors are, what the idiosyncrasies are of a given machine, 712 00:39:38,360 --> 00:39:41,520 Speaker 1: and begin to correct for them. We talked about technological 713 00:39:41,560 --> 00:39:44,000 Speaker 1: progress and that that's important, but in some ways the 714 00:39:44,600 --> 00:39:48,000 Speaker 1: real challenge here is actually manufacturing progress. Understanding what the 715 00:39:48,360 --> 00:39:51,000 Speaker 1: what what the ADO syncrasies are at the manufacturing stage 716 00:39:51,040 --> 00:39:52,600 Speaker 1: and then learning to correct for them. And that's that's 717 00:39:52,600 --> 00:39:54,640 Speaker 1: what t SMC has done extraordinarly well ad and it's 718 00:39:54,640 --> 00:39:58,719 Speaker 1: been hand in hand with with a SML. So I 719 00:39:58,760 --> 00:40:02,040 Speaker 1: have a sort of market oriented question. But you know, 720 00:40:02,080 --> 00:40:05,960 Speaker 1: we think of semiconductors as this highly cyclical industry that 721 00:40:06,120 --> 00:40:09,440 Speaker 1: usually moves in line with whatever is going on with 722 00:40:09,480 --> 00:40:12,680 Speaker 1: GDP and economic growth, and that hasn't really been the 723 00:40:12,719 --> 00:40:16,040 Speaker 1: case since the pandemic because we've had, you know, this 724 00:40:16,160 --> 00:40:20,319 Speaker 1: big boom in demand for electronic goods and it's been 725 00:40:20,320 --> 00:40:22,960 Speaker 1: a struggle to keep up. But I imagine for a 726 00:40:22,960 --> 00:40:25,640 Speaker 1: company like a s mL, it has also traditionally been 727 00:40:25,640 --> 00:40:28,680 Speaker 1: considered cyclical and its fortunes are sort of tied to 728 00:40:28,760 --> 00:40:32,960 Speaker 1: what's going on with the actual semiconductor manufacturers. But just 729 00:40:33,000 --> 00:40:37,440 Speaker 1: looking at SML's most recent results, they're forecasting basically like 730 00:40:37,800 --> 00:40:41,680 Speaker 1: a boom in revenue for the next decade, something they 731 00:40:41,719 --> 00:40:44,960 Speaker 1: expect to last for ten years. Is there anything that 732 00:40:45,000 --> 00:40:48,560 Speaker 1: could sort of knock that revenue cycle at this point 733 00:40:48,600 --> 00:40:51,120 Speaker 1: in time or is this business like, is there such 734 00:40:51,120 --> 00:40:54,360 Speaker 1: a steep moat around the eu V technology that is 735 00:40:54,360 --> 00:40:57,400 Speaker 1: just going to be impossible for anything to um to 736 00:40:57,520 --> 00:41:02,120 Speaker 1: hit it. There's there's definitely at mote around SMILS technology, 737 00:41:02,160 --> 00:41:04,360 Speaker 1: and they won't be overtaken into UV in a decade. 738 00:41:04,520 --> 00:41:08,800 Speaker 1: I should say I'm mixing my metaphors and steep and 739 00:41:09,040 --> 00:41:12,080 Speaker 1: impossible to overcome mote. There's no real competition that a 740 00:41:12,200 --> 00:41:14,600 Speaker 1: small faces when it comes to UV. That the question 741 00:41:14,680 --> 00:41:17,800 Speaker 1: is what is going to be demand for UV machines, 742 00:41:17,920 --> 00:41:20,440 Speaker 1: and that's a question ultimately of final demand for the 743 00:41:20,440 --> 00:41:25,040 Speaker 1: most advanced chips. A SMILS projections are are based on 744 00:41:25,719 --> 00:41:28,799 Speaker 1: the assumption that we've reached a point where there's a 745 00:41:28,880 --> 00:41:32,040 Speaker 1: secular increase in demand for chips, as we have more 746 00:41:32,360 --> 00:41:35,200 Speaker 1: demand for data center capacity, um as we have the 747 00:41:35,200 --> 00:41:38,080 Speaker 1: five G rollout and new devices that are taking advantage 748 00:41:38,120 --> 00:41:40,000 Speaker 1: of five G networks. Their bat is that we're going 749 00:41:40,080 --> 00:41:42,000 Speaker 1: to have more chips per g d P and therefore 750 00:41:42,000 --> 00:41:44,640 Speaker 1: more demand for a SML. That's a bat that you know, 751 00:41:44,719 --> 00:41:46,200 Speaker 1: it's not clear what that's going to play out. What 752 00:41:46,320 --> 00:41:49,239 Speaker 1: is clear is that anyone who's producing advanced ships will 753 00:41:49,239 --> 00:41:51,360 Speaker 1: have no choice but to turn to sm L and 754 00:41:51,400 --> 00:41:53,920 Speaker 1: so in some ways they're perfectly exposed to the fluctuations 755 00:41:53,920 --> 00:41:56,640 Speaker 1: in the semiconductor industry. The more chips that are produced, 756 00:41:56,640 --> 00:41:59,160 Speaker 1: the more machines you need. But the opposite is also true. 757 00:41:59,800 --> 00:42:03,839 Speaker 1: You kind of hinted at this early on. But the 758 00:42:03,880 --> 00:42:08,040 Speaker 1: idea that at uh seven nnimms in below at cutting edge, 759 00:42:08,160 --> 00:42:13,960 Speaker 1: maybe EUV in theory isn't the only technology. So it's like, okay, 760 00:42:14,000 --> 00:42:17,720 Speaker 1: if if if EUV is the only technology that can 761 00:42:17,760 --> 00:42:20,600 Speaker 1: perform the test, then there's no one who can attack 762 00:42:20,640 --> 00:42:24,200 Speaker 1: as well. Are there other theoretical approaches for accomplishing the 763 00:42:24,239 --> 00:42:26,120 Speaker 1: same thing besides the UV, I feel like you hinted 764 00:42:26,160 --> 00:42:29,719 Speaker 1: at in the beginning, there are It's really not a 765 00:42:29,800 --> 00:42:33,200 Speaker 1: question of science but of of manufacturing efficiency. So if 766 00:42:33,239 --> 00:42:36,560 Speaker 1: you if you take the previous generation of lithography machines, 767 00:42:36,560 --> 00:42:39,760 Speaker 1: which rather than thirteen point five nanometer light, we're working 768 00:42:39,800 --> 00:42:43,200 Speaker 1: with one. It became it was possible over time to 769 00:42:43,440 --> 00:42:47,839 Speaker 1: produce ever smaller feature sizes on silicon wafers by using 770 00:42:47,880 --> 00:42:50,080 Speaker 1: a number of tricks. So, for example, you can shoot 771 00:42:50,120 --> 00:42:52,520 Speaker 1: the light through water, and if you think back to 772 00:42:52,560 --> 00:42:56,400 Speaker 1: high school physics, when light refracts differently through water, that 773 00:42:56,520 --> 00:42:59,799 Speaker 1: same principle lets you shoot lithography machines through water and 774 00:43:00,000 --> 00:43:03,800 Speaker 1: and carve more specific shapes. You can also use multiple 775 00:43:03,960 --> 00:43:09,120 Speaker 1: steps of lithography to carve specific shapes that are more detailed. 776 00:43:09,239 --> 00:43:12,239 Speaker 1: The challenges just can you do this efficiently? Um So, 777 00:43:12,320 --> 00:43:15,560 Speaker 1: every step of lithography you need adds to the time 778 00:43:15,560 --> 00:43:17,640 Speaker 1: it takes to produce a way for ads to your costs. 779 00:43:18,120 --> 00:43:20,840 Speaker 1: And so there's no doubt that if you wanted to 780 00:43:20,880 --> 00:43:24,000 Speaker 1: produce an equivalent of one of Apple's new iPhone ships 781 00:43:24,080 --> 00:43:26,440 Speaker 1: using older generation lithography, you could do it in a 782 00:43:26,560 --> 00:43:28,560 Speaker 1: lab and do one of them. The question is can 783 00:43:28,600 --> 00:43:30,520 Speaker 1: you do a million of them as skin? And that 784 00:43:30,560 --> 00:43:34,760 Speaker 1: seems pretty implausible right now. There's no really credible pathway 785 00:43:34,800 --> 00:43:37,680 Speaker 1: of of how you could do that efficiently today, and 786 00:43:37,760 --> 00:43:40,919 Speaker 1: especially when you project forward five or ten years, we're 787 00:43:40,960 --> 00:43:45,080 Speaker 1: expecting to be making ever smaller transistors with more complex 788 00:43:45,080 --> 00:43:48,400 Speaker 1: shapes on them, and it seems really implausible that you'll 789 00:43:48,440 --> 00:43:51,239 Speaker 1: be able to do that using anything besides UV. Chris, 790 00:43:51,400 --> 00:43:54,040 Speaker 1: is there anything any other sort of least key things 791 00:43:54,040 --> 00:43:55,600 Speaker 1: do you think we've missed? And I mean, I'm sure 792 00:43:55,640 --> 00:43:58,480 Speaker 1: there's a million things, but other key ideas that I 793 00:43:58,480 --> 00:44:02,160 Speaker 1: think we need to get across. I think if you're 794 00:44:02,160 --> 00:44:05,680 Speaker 1: interested in US China dynamics, obviously it's a big. One 795 00:44:05,719 --> 00:44:09,680 Speaker 1: of the key reasons why t SMC cut off huaweih 796 00:44:09,719 --> 00:44:14,600 Speaker 1: In was because the US could restrict t SMCS access 797 00:44:14,640 --> 00:44:19,160 Speaker 1: to machinery, of which lithography machinery was a It was 798 00:44:19,200 --> 00:44:22,160 Speaker 1: a key example. So when when the Chinese ship industry 799 00:44:22,200 --> 00:44:23,920 Speaker 1: looks out and says we're we're gonna get the tools 800 00:44:23,960 --> 00:44:26,800 Speaker 1: that we need, the impact of the US support controls 801 00:44:26,880 --> 00:44:31,080 Speaker 1: on companies like a SML, even foreign companies that nevertheless 802 00:44:31,239 --> 00:44:34,640 Speaker 1: use US technology in their systems, is a pretty fundamental 803 00:44:34,719 --> 00:44:37,520 Speaker 1: roadblock that China faces, and that the big concern that 804 00:44:37,880 --> 00:44:40,080 Speaker 1: SML has right now is that the US is going 805 00:44:40,120 --> 00:44:42,759 Speaker 1: to expand its restrictions on what you can send to 806 00:44:42,880 --> 00:44:45,719 Speaker 1: China in terms of lithography machinery, and China has been 807 00:44:45,719 --> 00:44:47,680 Speaker 1: a big growth market the past couple of years for 808 00:44:47,719 --> 00:44:50,560 Speaker 1: older generational lithography machines, and so SML does face a 809 00:44:50,640 --> 00:44:55,440 Speaker 1: risk that the US it expands these restrictions. What inspired 810 00:44:55,480 --> 00:44:57,560 Speaker 1: you to write a book all about a s m 811 00:44:57,640 --> 00:45:00,239 Speaker 1: L because it is like it's not something that ms 812 00:45:00,320 --> 00:45:04,240 Speaker 1: up necessarily in daily conversation. UM, so I'm just wondering 813 00:45:04,280 --> 00:45:06,160 Speaker 1: how it sort of came on your radar and what 814 00:45:06,360 --> 00:45:09,359 Speaker 1: is it that piqued your interest? We like the two 815 00:45:09,360 --> 00:45:11,719 Speaker 1: of you. I spent the past couple of years realizing 816 00:45:11,760 --> 00:45:15,760 Speaker 1: that semiconductors were vastly more important than the average person, 817 00:45:16,160 --> 00:45:20,680 Speaker 1: including myself realized, and also far cooler. The technology needed 818 00:45:20,719 --> 00:45:23,319 Speaker 1: to make them as extraordinary. The fact that we're able 819 00:45:23,400 --> 00:45:27,480 Speaker 1: to manipulate individual atoms in some cases is extraordinary, and 820 00:45:27,480 --> 00:45:30,320 Speaker 1: to do it at the scale of trillions and trillions 821 00:45:30,320 --> 00:45:33,360 Speaker 1: and transistors, I thought was was was really just wild 822 00:45:33,400 --> 00:45:36,080 Speaker 1: in terms of what what what was possible? Uh? And 823 00:45:36,120 --> 00:45:39,160 Speaker 1: it seemed to me that I took my iPhone for granted, 824 00:45:39,200 --> 00:45:41,760 Speaker 1: I took my computer for granted. I took the cloud, 825 00:45:41,800 --> 00:45:43,919 Speaker 1: which is just a bunch of silicon in big data 826 00:45:43,920 --> 00:45:46,480 Speaker 1: centers in IO. I took all that for granted without 827 00:45:46,480 --> 00:45:50,880 Speaker 1: thinking through how complex it was too actually make these 828 00:45:50,920 --> 00:45:53,759 Speaker 1: tools work. And I think for for a long time 829 00:45:53,760 --> 00:45:55,800 Speaker 1: we thought of the Internet is something out there. We 830 00:45:55,880 --> 00:45:58,640 Speaker 1: thought of data processing is something that happens somewhere else. 831 00:45:58,640 --> 00:46:01,520 Speaker 1: But it's all actually very physical, all things being carved 832 00:46:02,000 --> 00:46:06,000 Speaker 1: onto silicon by shooting light at them and depositing uh, 833 00:46:06,440 --> 00:46:09,560 Speaker 1: layers of atoms and using different chemicals. And the reality 834 00:46:09,600 --> 00:46:12,440 Speaker 1: that our entire digital world is in fact existing on 835 00:46:12,880 --> 00:46:15,000 Speaker 1: millions and millions of silicon wafers is something I don't 836 00:46:15,000 --> 00:46:17,560 Speaker 1: think we think enough about and we're just having to 837 00:46:17,600 --> 00:46:19,880 Speaker 1: come directon with that with the some innunctory shortage right 838 00:46:19,920 --> 00:46:22,600 Speaker 1: now that you can't just imagine an increase in computing 839 00:46:22,600 --> 00:46:24,880 Speaker 1: power and increase in memory, You've actually got to carve 840 00:46:24,920 --> 00:46:28,440 Speaker 1: it onto silicon in billions and billions of tiny transistors. 841 00:46:29,400 --> 00:46:32,560 Speaker 1: You know, it's interesting, I mean, your assistant professor at 842 00:46:32,560 --> 00:46:37,680 Speaker 1: the Fletchers School, which I associate with diplomacy and government, 843 00:46:37,800 --> 00:46:41,560 Speaker 1: and that seems like another sort of like fascinating dynamic here, 844 00:46:41,600 --> 00:46:44,799 Speaker 1: which is like maybe it's it feels again, or maybe 845 00:46:44,800 --> 00:46:47,719 Speaker 1: it goes in cycles, but there's appreciation. And you sort 846 00:46:47,719 --> 00:46:49,799 Speaker 1: of said it for your last point about yours trying 847 00:46:49,840 --> 00:46:53,400 Speaker 1: to like that this particular industry is sort of inseparable 848 00:46:53,600 --> 00:46:57,239 Speaker 1: from thinking about how governments relate to each other. That's right. 849 00:46:57,280 --> 00:47:01,080 Speaker 1: It's it's it's crucial for for military systems, for example, 850 00:47:01,160 --> 00:47:05,560 Speaker 1: it's crucial for controlling computing power in the future. And 851 00:47:05,640 --> 00:47:09,400 Speaker 1: it's been a prominent tool of of of geopolitics for 852 00:47:09,440 --> 00:47:12,000 Speaker 1: the past three quarters of a century. And I think 853 00:47:12,320 --> 00:47:14,560 Speaker 1: we're seeing that more to the four today. But in fact, 854 00:47:14,560 --> 00:47:16,600 Speaker 1: when you look at the history of lithography and of 855 00:47:16,640 --> 00:47:19,239 Speaker 1: semiconductor is more generally you find that it's constantly been 856 00:47:19,880 --> 00:47:22,399 Speaker 1: something that governments have thought about in in political terms 857 00:47:22,400 --> 00:47:25,000 Speaker 1: as well as in economic terms, and constantly been an 858 00:47:25,000 --> 00:47:27,360 Speaker 1: area of dispute between different governments as they tried to 859 00:47:27,760 --> 00:47:33,840 Speaker 1: vy for a bigger chunk of the semiconductor ecosystem. Chris Miller, 860 00:47:34,040 --> 00:47:37,680 Speaker 1: thank you so much for coming on. That was the 861 00:47:37,680 --> 00:47:40,040 Speaker 1: the a s m L episode We needed to do it, 862 00:47:40,080 --> 00:47:41,520 Speaker 1: and you are the perfect guest for it, and I 863 00:47:41,640 --> 00:47:44,200 Speaker 1: just learned a lot, So thank you so much for 864 00:47:44,280 --> 00:47:48,560 Speaker 1: coming out on Oblock Bain's invitation. Thanks Chris, that was 865 00:47:48,600 --> 00:48:01,759 Speaker 1: so interesting. Obviously, I love that conversation, and you know, 866 00:48:01,840 --> 00:48:05,239 Speaker 1: I sortaty interrupted like seven minutes in because like this 867 00:48:05,320 --> 00:48:09,120 Speaker 1: idea of like thinking of like a component as in 868 00:48:09,200 --> 00:48:13,560 Speaker 1: itself a supply chain story. Like the idea that really 869 00:48:13,640 --> 00:48:18,400 Speaker 1: the breakthrough is how do you coordinate four thousand different 870 00:48:18,440 --> 00:48:23,440 Speaker 1: suppliers of them of highly specific raw materials and machines 871 00:48:23,840 --> 00:48:26,880 Speaker 1: into one thing that forms a cohesive whole. The idea 872 00:48:26,920 --> 00:48:31,120 Speaker 1: that that is what the thing is is pretty fascinating 873 00:48:31,160 --> 00:48:35,040 Speaker 1: to me. Here's the important question, which is, what, like 874 00:48:35,200 --> 00:48:37,480 Speaker 1: what idea did you get out of that conversation For 875 00:48:37,480 --> 00:48:40,920 Speaker 1: the next Semiconductor episode, so I'm sure there is one. 876 00:48:41,520 --> 00:48:44,480 Speaker 1: What is the next one? Now we we we got 877 00:48:44,520 --> 00:48:48,080 Speaker 1: to the end. Now it would be like really interesting actually, okay, 878 00:48:48,120 --> 00:48:51,719 Speaker 1: so like it all seriousness, like I would like to 879 00:48:52,120 --> 00:48:56,680 Speaker 1: learn more about that process, like the actual like the 880 00:48:56,760 --> 00:48:58,920 Speaker 1: coordination the heart. It's almost like you think of like 881 00:48:58,960 --> 00:49:01,319 Speaker 1: a conductor of an orchestra. Is sort of like the 882 00:49:01,320 --> 00:49:04,000 Speaker 1: mental model I use for a company that has to 883 00:49:04,080 --> 00:49:07,439 Speaker 1: like have four thousand parts all coming together to form 884 00:49:07,560 --> 00:49:11,120 Speaker 1: thirty one units or units or whatever. It is like 885 00:49:11,400 --> 00:49:13,840 Speaker 1: thinking about like how do you do that from like 886 00:49:13,880 --> 00:49:17,560 Speaker 1: a management perspective even beyond this sort of tech perspective 887 00:49:17,719 --> 00:49:21,080 Speaker 1: is like a super fascinating thing to explore, especially at 888 00:49:21,080 --> 00:49:24,520 Speaker 1: its especially right now. I mean this is almost verging 889 00:49:24,600 --> 00:49:27,279 Speaker 1: on state secrets. But we've got to get you know, 890 00:49:27,320 --> 00:49:28,680 Speaker 1: we have to try to get the a s m 891 00:49:28,800 --> 00:49:31,360 Speaker 1: L supply chain manager on all thoughts, so you know, 892 00:49:31,480 --> 00:49:34,120 Speaker 1: a s m L hit us up. We're interested in 893 00:49:34,160 --> 00:49:36,799 Speaker 1: how you're doing it. Um, and we have to keep 894 00:49:36,840 --> 00:49:42,280 Speaker 1: the semiconductor series going. Yeah, No, that was fascinating, and um, 895 00:49:42,440 --> 00:49:46,360 Speaker 1: Chris was the Chris was the perfect guest for that one. Yeah, definitely, 896 00:49:46,440 --> 00:49:49,960 Speaker 1: shall we leave it there, Let's leave it there alright. 897 00:49:50,000 --> 00:49:52,760 Speaker 1: This has been another episode of the All Thoughts podcast. 898 00:49:52,840 --> 00:49:55,640 Speaker 1: I'm Tracy Alloway. You can follow me on Twitter at 899 00:49:55,640 --> 00:49:58,920 Speaker 1: Tracy Alloway. And I'm Joe Wisenthal. You can follow me 900 00:49:59,080 --> 00:50:02,480 Speaker 1: on Twitter at the Stalwart. Follow our guest Chris Miller. 901 00:50:02,520 --> 00:50:04,200 Speaker 1: He has a book coming out next year on the 902 00:50:04,280 --> 00:50:07,360 Speaker 1: chip industry. Assistant professor at the Fletcher School, he is 903 00:50:07,480 --> 00:50:12,279 Speaker 1: at cr Miller One. Follow our producer Laura Carlson. She's 904 00:50:12,360 --> 00:50:15,600 Speaker 1: at Laura M. Carlson. Follow the Bloomberg head of podcast, 905 00:50:15,640 --> 00:50:19,080 Speaker 1: Francesca Levi at Francesca Today, and check out all of 906 00:50:19,080 --> 00:50:23,279 Speaker 1: our podcasts at Bloomberg under the handle at podcasts. Thanks 907 00:50:23,320 --> 00:50:23,840 Speaker 1: for listening.