1
00:00:10,680 --> 00:00:14,560
Speaker 1: Hello, and welcome to another episode of the Odd Lots podcast.

2
00:00:14,640 --> 00:00:16,919
Speaker 2: I'm Joe Wisenthal and I'm Tracy Alloway.

3
00:00:17,040 --> 00:00:20,280
Speaker 1: Tracy, it's been too long since we've done a semiconductor episode.

4
00:00:20,400 --> 00:00:23,480
Speaker 3: I know, I actually feel really behind on what's been

5
00:00:23,520 --> 00:00:26,520
Speaker 3: happening in the industry. I've seen some headlines. I mean,

6
00:00:26,560 --> 00:00:29,280
Speaker 3: obviously we talked about in Video earlier in the year

7
00:00:29,360 --> 00:00:33,440
Speaker 3: in that stock just exploded, and recently they released their earnings.

8
00:00:33,479 --> 00:00:38,760
Speaker 3: I've seen some news about additional export restrictions and things

9
00:00:38,920 --> 00:00:41,080
Speaker 3: like that. But I haven't been paying close enough attention,

10
00:00:41,120 --> 00:00:42,400
Speaker 3: and I feel really bad about that.

11
00:00:42,840 --> 00:00:47,000
Speaker 1: Do you ever read, Tracy, or have you ever ever read, Yes, Joe,

12
00:00:47,159 --> 00:00:50,120
Speaker 1: you read. Have you ever tried to read those like

13
00:00:50,280 --> 00:00:55,640
Speaker 1: really technical semiconductor publications that talk about, oh, the new

14
00:00:56,120 --> 00:01:00,120
Speaker 1: in video chip is four different cores and a thing here,

15
00:01:00,160 --> 00:01:01,960
Speaker 1: And have you ever seen those sites?

16
00:01:02,080 --> 00:01:02,320
Speaker 4: I have?

17
00:01:02,520 --> 00:01:03,560
Speaker 2: Sometimes people drop.

18
00:01:03,440 --> 00:01:05,720
Speaker 1: The links at our discord and I try to catch

19
00:01:05,800 --> 00:01:07,880
Speaker 1: up to speed on what the hot new chip is,

20
00:01:07,920 --> 00:01:10,039
Speaker 1: and I just I always give up one third of

21
00:01:10,040 --> 00:01:10,840
Speaker 1: the way through the post.

22
00:01:11,000 --> 00:01:13,280
Speaker 3: It's really difficult. I mean, even the ones where they

23
00:01:13,319 --> 00:01:16,039
Speaker 3: talk about like wafer thinness and stuff like like, I

24
00:01:16,040 --> 00:01:19,640
Speaker 3: can kind of understand it, but there's so much within

25
00:01:19,760 --> 00:01:22,959
Speaker 3: this specific subject. There's sort of the big picture things

26
00:01:23,000 --> 00:01:26,880
Speaker 3: like geopolitical tensions and trade tensions and things like that,

27
00:01:26,920 --> 00:01:29,679
Speaker 3: and then you can really cut into I don't want

28
00:01:29,680 --> 00:01:32,360
Speaker 3: to say the nuts and bolts of the specific technology,

29
00:01:32,480 --> 00:01:35,120
Speaker 3: but you know, I don't know the wafers and pins

30
00:01:35,160 --> 00:01:36,280
Speaker 3: of specific technology.

31
00:01:36,600 --> 00:01:38,600
Speaker 1: So the one thing is it seems like the lower

32
00:01:38,680 --> 00:01:42,080
Speaker 1: nanometer number the better, Yes, right, But even there, and

33
00:01:42,120 --> 00:01:44,560
Speaker 1: I know we've done some episodes in the past with

34
00:01:44,680 --> 00:01:47,680
Speaker 1: us Stacy Rasken, etcetera. Even there, that only tells you

35
00:01:47,760 --> 00:01:49,720
Speaker 1: so much about a chip's performance, and there are all

36
00:01:49,720 --> 00:01:54,840
Speaker 1: different kinds of architectures and yields. We've talked about that too.

37
00:01:54,920 --> 00:01:57,760
Speaker 1: Or in theory, you could have a really powerful chip,

38
00:01:58,000 --> 00:02:00,680
Speaker 1: but maybe it's not economical because you lose so many

39
00:02:00,760 --> 00:02:03,680
Speaker 1: chips in the process, et cetera. So wrapping one's head

40
00:02:03,720 --> 00:02:05,320
Speaker 1: around chips is tough.

41
00:02:05,520 --> 00:02:08,360
Speaker 3: Yes, agreed, But what will we be doing on this

42
00:02:08,440 --> 00:02:10,480
Speaker 3: episode China wrap our heads around chips.

43
00:02:10,800 --> 00:02:11,080
Speaker 2: Yeah.

44
00:02:11,120 --> 00:02:14,840
Speaker 1: So, you know, there's been some news particularly related to China,

45
00:02:14,919 --> 00:02:18,120
Speaker 1: and I think Huawei came out with a phone fairly

46
00:02:18,200 --> 00:02:20,720
Speaker 1: recently in the last couple of months, and what claued

47
00:02:20,760 --> 00:02:25,000
Speaker 1: people's attention was it seemed to have performance that people

48
00:02:25,120 --> 00:02:27,840
Speaker 1: didn't assume it could have given what was known about

49
00:02:27,840 --> 00:02:31,240
Speaker 1: the state of domestic Chinese semiconductor compress seven.

50
00:02:31,080 --> 00:02:34,600
Speaker 3: Nanometers, Yeah, seven nanometers. I shouldn't whisper on a podcast.

51
00:02:35,160 --> 00:02:36,880
Speaker 3: I don't know why I did that. It's not a secret.

52
00:02:37,240 --> 00:02:40,440
Speaker 3: So supposedly it had a chip made by SMI C,

53
00:02:41,200 --> 00:02:45,600
Speaker 3: a Chinese chip maker, that was seven nanometers and something

54
00:02:45,800 --> 00:02:50,200
Speaker 3: that people thought China wasn't able to produce just yet.

55
00:02:50,240 --> 00:02:53,000
Speaker 3: And yet here we are talking about the seven and

56
00:02:53,280 --> 00:02:54,239
Speaker 3: M in these phones.

57
00:02:54,440 --> 00:02:56,560
Speaker 1: And I guess the question is, is this represent a

58
00:02:56,560 --> 00:03:01,320
Speaker 1: major domestic breakthrough for China's semiconductor into Does it mean

59
00:03:01,360 --> 00:03:03,840
Speaker 1: that some technology which wasn't supposed to get into the

60
00:03:03,840 --> 00:03:06,639
Speaker 1: country somehow got in? So there's some sort of US

61
00:03:06,800 --> 00:03:08,280
Speaker 1: national security implications.

62
00:03:08,639 --> 00:03:11,000
Speaker 2: But it's a good time. You know, we've talked about this.

63
00:03:11,000 --> 00:03:13,360
Speaker 1: For years, you know, with people like Dan Wong, like

64
00:03:13,600 --> 00:03:16,799
Speaker 1: what is the state of Chinese semiconductor? Are they catching up?

65
00:03:16,960 --> 00:03:20,040
Speaker 1: So I think it's a good time to take stock

66
00:03:20,200 --> 00:03:20,919
Speaker 1: of the situation.

67
00:03:21,120 --> 00:03:21,640
Speaker 5: I agree.

68
00:03:21,680 --> 00:03:24,800
Speaker 3: And also, you didn't mention the biggest thing that people

69
00:03:24,800 --> 00:03:27,160
Speaker 3: are talking about now, which is the idea that have

70
00:03:27,400 --> 00:03:34,200
Speaker 3: the restrictions basically created the exact opposite result intended and

71
00:03:34,320 --> 00:03:37,720
Speaker 3: maybe accelerated China's semiconductor technology.

72
00:03:37,280 --> 00:03:40,560
Speaker 1: Which is something people have warned about that ultimately, sure,

73
00:03:40,720 --> 00:03:42,640
Speaker 1: maybe you set the country back a few years in

74
00:03:42,680 --> 00:03:46,080
Speaker 1: its development, but if you restrict its capacity to get

75
00:03:46,120 --> 00:03:50,080
Speaker 1: international technology, then it just builds faster its own homegrown

76
00:03:50,240 --> 00:03:52,840
Speaker 1: and all kinds of questions related to chips right now

77
00:03:52,880 --> 00:03:54,880
Speaker 1: that we need to do, we need to catch up on.

78
00:03:55,120 --> 00:03:57,680
Speaker 3: Yes, And also, can I just say that I blame

79
00:03:57,920 --> 00:04:01,920
Speaker 3: the semiconductor restriction on me having to read the Three

80
00:04:02,000 --> 00:04:02,840
Speaker 3: Body Problem?

81
00:04:03,240 --> 00:04:03,960
Speaker 5: Did you ever read that?

82
00:04:04,360 --> 00:04:06,800
Speaker 3: And now these visions. I don't know how many people

83
00:04:06,840 --> 00:04:08,760
Speaker 3: have read it, but it's a lot of people used

84
00:04:08,760 --> 00:04:12,560
Speaker 3: it as an analogy for China's technological development. But now

85
00:04:12,600 --> 00:04:16,240
Speaker 3: I have nightmares about like little shriveled up dead people.

86
00:04:16,920 --> 00:04:19,320
Speaker 3: And if you've read the book, if you've read the book,

87
00:04:19,360 --> 00:04:20,000
Speaker 3: this makes sense.

88
00:04:20,160 --> 00:04:22,000
Speaker 2: Okay, Well, let's talk chips.

89
00:04:22,040 --> 00:04:25,120
Speaker 1: And we really do have two perfect guests, one of

90
00:04:25,160 --> 00:04:27,840
Speaker 1: whom we've had on the show before, Doug O'Laughlin. He

91
00:04:27,960 --> 00:04:31,880
Speaker 1: is the chief analyst and Fabricated Knowledge, a semiconductor research

92
00:04:31,960 --> 00:04:35,919
Speaker 1: service and Dylan Fitzel, chief analyst. It's Semi Analysis, a

93
00:04:35,960 --> 00:04:39,000
Speaker 1: boutique semiconductor and AI research firms. So we are going

94
00:04:39,040 --> 00:04:42,640
Speaker 1: to pick their brands about the state of Chinese chips. So,

95
00:04:42,920 --> 00:04:45,760
Speaker 1: Doug and Dylan, thank you so much for coming on,

96
00:04:45,800 --> 00:04:46,600
Speaker 1: odd Laws.

97
00:04:46,560 --> 00:04:48,679
Speaker 5: Thank you for having us, Thanks for having me again.

98
00:04:49,320 --> 00:04:51,400
Speaker 2: Let's just start. What was it?

99
00:04:51,440 --> 00:04:54,000
Speaker 1: What's the deal with this Huawei phone that caught everyone

100
00:04:54,040 --> 00:04:54,599
Speaker 1: by surprise?

101
00:04:55,200 --> 00:04:59,200
Speaker 4: So Huawei has historically been a leader in technology. That

102
00:04:59,279 --> 00:05:02,240
Speaker 4: wasn't a surprise anyone. And a few years ago, of course,

103
00:05:02,520 --> 00:05:05,520
Speaker 4: they got banned from, you know, many aspects of the

104
00:05:05,640 --> 00:05:08,240
Speaker 4: US and Western semi conductor supply chain. They got banned

105
00:05:08,279 --> 00:05:12,440
Speaker 4: from utilizing TSMC, the world's largest chip maker, and all

106
00:05:12,440 --> 00:05:14,480
Speaker 4: of a sudden they come out with their own phone, right,

107
00:05:14,520 --> 00:05:17,279
Speaker 4: their new phone that has their own chip made in China.

108
00:05:17,600 --> 00:05:20,359
Speaker 4: The expectations, of course when they first announced it were whatever.

109
00:05:20,880 --> 00:05:23,080
Speaker 4: But then once people got their hands on the phone,

110
00:05:23,080 --> 00:05:25,920
Speaker 4: it was like, oh my god, this is actually very good.

111
00:05:26,160 --> 00:05:29,039
Speaker 4: You know, when you compare to foreign phones right with

112
00:05:29,200 --> 00:05:32,039
Speaker 4: wou'd say, Qualcom chips from like say, Samsung, it is

113
00:05:32,080 --> 00:05:34,559
Speaker 4: only you know, a year and a half behind, right,

114
00:05:34,640 --> 00:05:37,279
Speaker 4: And in some respects. It's actually just as good in

115
00:05:37,320 --> 00:05:39,760
Speaker 4: certain specifications. It depends on how how nitty gritty you

116
00:05:39,800 --> 00:05:41,760
Speaker 4: want to get into it. But at worst one point

117
00:05:41,760 --> 00:05:45,000
Speaker 4: five years behind at best on part right. And so

118
00:05:45,080 --> 00:05:47,760
Speaker 4: that was a big, big shocker for everyone, just.

119
00:05:47,760 --> 00:05:51,520
Speaker 1: Real quickly on those specifications. Like when you say the

120
00:05:51,600 --> 00:05:53,880
Speaker 1: phone is really good, because I don't know, all phones

121
00:05:53,920 --> 00:05:56,040
Speaker 1: sort of seem pretty good and I don't really notice

122
00:05:56,120 --> 00:05:58,440
Speaker 1: advances and phone. So when you say that the phone

123
00:05:58,440 --> 00:06:02,440
Speaker 1: surprised people by capacity, could you just be specific about

124
00:06:02,440 --> 00:06:03,120
Speaker 1: what shocked people?

125
00:06:03,160 --> 00:06:04,240
Speaker 2: And they picked up this device?

126
00:06:04,720 --> 00:06:08,040
Speaker 4: Sure? So stage one is sort of the network performance, right,

127
00:06:08,080 --> 00:06:10,760
Speaker 4: so it can download and upload data just as fast

128
00:06:10,760 --> 00:06:12,680
Speaker 4: as any foreign phone, okay, which is the sort of

129
00:06:12,720 --> 00:06:14,520
Speaker 4: on par thing right. In fact, it was better than

130
00:06:14,560 --> 00:06:17,440
Speaker 4: the current iPhone. It was on par with the current

131
00:06:17,520 --> 00:06:20,640
Speaker 4: Qualcom based phones. And so then that's like one specification.

132
00:06:20,720 --> 00:06:24,080
Speaker 4: Other specification is your CPUs and GPUs and AI aspects

133
00:06:24,120 --> 00:06:27,159
Speaker 4: of the phone. Right. And there they were not only

134
00:06:27,279 --> 00:06:31,360
Speaker 4: using you know, they're again using domestic manufacturing capabilities, they

135
00:06:31,360 --> 00:06:34,839
Speaker 4: were also using domestic design capabilities and were able to

136
00:06:34,920 --> 00:06:37,320
Speaker 4: match again what folks had done a year and a

137
00:06:37,320 --> 00:06:40,479
Speaker 4: half ago or in some cases even sooner more or

138
00:06:40,480 --> 00:06:43,000
Speaker 4: more recently. Right, So this is sort of on on

139
00:06:43,080 --> 00:06:45,520
Speaker 4: a performance basis, whether it's in gaming, whether it's in

140
00:06:45,839 --> 00:06:48,520
Speaker 4: you know, uploading and downloading videos, whether it's in camera.

141
00:06:48,960 --> 00:06:51,360
Speaker 4: Every every aspect of this phone was on par, if

142
00:06:51,360 --> 00:06:51,680
Speaker 4: you will.

143
00:06:52,080 --> 00:06:55,560
Speaker 5: I think that people just probably don't appreciate how impressive

144
00:06:55,560 --> 00:06:57,920
Speaker 5: that is, given the fact that how little technology they

145
00:06:58,040 --> 00:07:00,320
Speaker 5: have access to. So this they did this without EV

146
00:07:00,480 --> 00:07:03,680
Speaker 5: which was the big October restrictions in twenty twenty two.

147
00:07:04,200 --> 00:07:06,760
Speaker 5: And I think the thing that's most impressive about this is,

148
00:07:06,800 --> 00:07:09,400
Speaker 5: like it's a really good chip with both hands tied

149
00:07:09,400 --> 00:07:11,960
Speaker 5: behind their back. And I think that like, if the

150
00:07:12,080 --> 00:07:16,200
Speaker 5: restrictions weren't there, the implication is that SMC could probably

151
00:07:16,480 --> 00:07:19,440
Speaker 5: ship a leading edge phone as good as TSMC and

152
00:07:19,520 --> 00:07:22,240
Speaker 5: maybe even better than Intel or domestic production. And I

153
00:07:22,280 --> 00:07:24,200
Speaker 5: think that that's probably the biggest takeaway that I think

154
00:07:24,200 --> 00:07:27,400
Speaker 5: people need to understand, you know, and there's definitely some

155
00:07:27,480 --> 00:07:29,640
Speaker 5: DV tools that snuck in, and we can talk about

156
00:07:29,640 --> 00:07:32,880
Speaker 5: the you know, all the mechanics of the cross. You know,

157
00:07:33,000 --> 00:07:35,880
Speaker 5: how slippery the restrictions have been and how maybe poorly

158
00:07:35,960 --> 00:07:39,040
Speaker 5: enforced it's been. But like they did an amazing phone

159
00:07:39,080 --> 00:07:40,920
Speaker 5: with both hands tied behind their back. And I think

160
00:07:40,920 --> 00:07:43,520
Speaker 5: that in the conversation that we've been having about semicuctors

161
00:07:43,680 --> 00:07:46,760
Speaker 5: in China for a long time, it's always been like, well,

162
00:07:46,760 --> 00:07:49,040
Speaker 5: they're really far behind, they'll never catch up. And I

163
00:07:49,040 --> 00:07:51,320
Speaker 5: think that this is the first time where you can say, like,

164
00:07:51,440 --> 00:07:54,000
Speaker 5: if they had what we had, they've caught up. Like

165
00:07:54,240 --> 00:07:56,360
Speaker 5: I think that's Dylan. Do you agree or disagree with that?

166
00:07:56,680 --> 00:07:58,400
Speaker 4: I would say, you know, if if you compare just

167
00:07:58,440 --> 00:08:01,239
Speaker 4: what's shipping in the market, Yeah, of course TSMC shipping better,

168
00:08:01,440 --> 00:08:04,240
Speaker 4: Samsung is shipped better. But and everyone talks about this

169
00:08:04,320 --> 00:08:07,720
Speaker 4: Intel turnaround that may or may not happen. Intel's best

170
00:08:07,800 --> 00:08:11,200
Speaker 4: chips that they ship today are seven animeter, the same

171
00:08:11,240 --> 00:08:13,640
Speaker 4: as what China has today, right, so it's it's on

172
00:08:13,840 --> 00:08:16,920
Speaker 4: par now. Of course, Intel's close to releasing their four

173
00:08:16,960 --> 00:08:19,320
Speaker 4: N animeter, and you could you could caveat that in

174
00:08:19,320 --> 00:08:20,880
Speaker 4: a hundred ways. But if you look at what's in

175
00:08:20,880 --> 00:08:24,080
Speaker 4: the market today, the dentist right again, as you mentioned earlier,

176
00:08:24,080 --> 00:08:26,440
Speaker 4: lower nanimeter ship that you can buy today from an

177
00:08:26,440 --> 00:08:29,720
Speaker 4: American manufacturer is from Intel, and the dentist is from

178
00:08:29,760 --> 00:08:33,079
Speaker 4: Smick and Huawei, and it's the same, right, It's it's

179
00:08:33,080 --> 00:08:34,200
Speaker 4: a similar capability.

180
00:08:34,720 --> 00:08:38,800
Speaker 3: Okay, so not since the invention of pringles have people

181
00:08:38,880 --> 00:08:43,160
Speaker 3: been so excited about a single chip. But you sort

182
00:08:43,160 --> 00:08:45,400
Speaker 3: of alluded to this, Doug, But can you maybe walk

183
00:08:45,480 --> 00:08:49,400
Speaker 3: us through what is needed to produce a seven and

184
00:08:49,800 --> 00:08:53,040
Speaker 3: M and what to your point about China basically doing

185
00:08:53,120 --> 00:08:55,880
Speaker 3: this with a hand or two tied behind its back,

186
00:08:56,280 --> 00:08:59,120
Speaker 3: what was actually available to them?

187
00:08:59,200 --> 00:09:01,720
Speaker 5: So this is this is a really hard question answer.

188
00:09:01,720 --> 00:09:03,760
Speaker 5: I'm gonna have to ask Dylan a lot about this

189
00:09:03,840 --> 00:09:06,720
Speaker 5: as well. But the thing that I think probably differentiates

190
00:09:06,720 --> 00:09:09,640
Speaker 5: it from let's say TSMC's process is that they did

191
00:09:09,640 --> 00:09:12,800
Speaker 5: not have access to EV. That's clearly the big delineation.

192
00:09:13,160 --> 00:09:16,000
Speaker 5: But if you remember the original the extreme yes EV

193
00:09:16,280 --> 00:09:20,200
Speaker 5: Extreme Ultraviolet that's the latest and greatest from ASML, and

194
00:09:20,240 --> 00:09:22,079
Speaker 5: they cost like, you know, three hundred million plus a

195
00:09:22,120 --> 00:09:25,960
Speaker 5: pop their extremely advanced technology to make tiny, tiny wavelengths

196
00:09:26,000 --> 00:09:28,600
Speaker 5: of light. But they managed to get around this with

197
00:09:28,840 --> 00:09:31,720
Speaker 5: something called quad patterning self a line quad patterning, which

198
00:09:31,760 --> 00:09:33,320
Speaker 5: is like we're not going to go into the details

199
00:09:33,320 --> 00:09:36,120
Speaker 5: of that, extremely technical but an extremely hard thing that

200
00:09:36,160 --> 00:09:38,720
Speaker 5: Intel got caught up on. So they managed to ship

201
00:09:38,720 --> 00:09:41,679
Speaker 5: the seven animeter Chip much quicker than Intel managed to

202
00:09:41,720 --> 00:09:44,240
Speaker 5: get through all of their problems doing the same quad

203
00:09:44,240 --> 00:09:47,360
Speaker 5: patterning DV process. And I think that that's a big deal,

204
00:09:47,440 --> 00:09:50,160
Speaker 5: Like it shows that there's a lot of technological umph

205
00:09:50,280 --> 00:09:53,120
Speaker 5: underneath the restrictions alone. And then on top of that,

206
00:09:53,120 --> 00:09:56,160
Speaker 5: the restrictions have been extremely poorly enforced. That's why they

207
00:09:56,160 --> 00:10:00,360
Speaker 5: had some, you know, some restatements this year about specificities.

208
00:10:00,400 --> 00:10:02,959
Speaker 5: But what happens is, like a good example is SMIC

209
00:10:03,200 --> 00:10:06,240
Speaker 5: has a leading edge and a lagging edge factory, right,

210
00:10:06,280 --> 00:10:09,560
Speaker 5: and they're both related entities. But my understanding is the

211
00:10:09,640 --> 00:10:13,760
Speaker 5: lagging edge entity can go by extremely advanced deposition and

212
00:10:13,840 --> 00:10:16,160
Speaker 5: etched tools, ones that are on the restrictions for the

213
00:10:16,200 --> 00:10:19,040
Speaker 5: October twenty twenty two, and then they can just kind

214
00:10:19,040 --> 00:10:21,480
Speaker 5: of shuttle those tools into the fab of the leading

215
00:10:21,559 --> 00:10:23,160
Speaker 5: edge and then effectively be able to use it.

216
00:10:23,240 --> 00:10:25,840
Speaker 3: That doesn't seem like it's doing you know, the restriction

217
00:10:25,960 --> 00:10:27,640
Speaker 3: is doing what it was intended to do in.

218
00:10:27,559 --> 00:10:30,959
Speaker 5: That case, Yeah, not at all so far. Pretty much

219
00:10:30,960 --> 00:10:33,200
Speaker 5: what's happened, and this has happened every single time we've

220
00:10:33,200 --> 00:10:37,600
Speaker 5: had restrictions on American semicap companies, is that the restrictions

221
00:10:37,640 --> 00:10:39,640
Speaker 5: come out all the companies say, oh, this is going

222
00:10:39,679 --> 00:10:42,240
Speaker 5: to impact us, and then they slowly find ways for

223
00:10:42,320 --> 00:10:45,240
Speaker 5: loopholes to be pushed through. An example is Applied material

224
00:10:45,320 --> 00:10:48,480
Speaker 5: how to South Korean factory that probably had I think

225
00:10:48,480 --> 00:10:51,560
Speaker 5: they are right now under investigation, and I think the

226
00:10:51,640 --> 00:10:54,200
Speaker 5: thing is it's pretty clear that some of the leading

227
00:10:54,320 --> 00:10:57,959
Speaker 5: edge tools X THEEV stuff is getting into China and

228
00:10:58,000 --> 00:11:01,160
Speaker 5: they're able to use clever engine hearing to make a

229
00:11:01,160 --> 00:11:03,839
Speaker 5: better chip than we thought was possible. Yeah.

230
00:11:03,880 --> 00:11:06,680
Speaker 4: I would say that while in spirit the regulations pretty

231
00:11:06,760 --> 00:11:09,160
Speaker 4: much said hey, you can't have less than fourteen nanimeters,

232
00:11:09,559 --> 00:11:12,680
Speaker 4: the specifics of what was actually banned were quite a

233
00:11:12,679 --> 00:11:14,880
Speaker 4: bit more varied. Right. The tool that you can use

234
00:11:14,880 --> 00:11:17,600
Speaker 4: for twenty eight and nanimeter, well you just use many

235
00:11:17,600 --> 00:11:19,640
Speaker 4: more of them for seven, right. I mean obviously there's

236
00:11:19,640 --> 00:11:23,520
Speaker 4: a lot more complications there, and so the government sort

237
00:11:23,559 --> 00:11:28,000
Speaker 4: of handed out export licenses like candy to companies like Smick,

238
00:11:28,080 --> 00:11:30,600
Speaker 4: saying hey, yeah, you can expand your twenty eight nanimeter

239
00:11:30,640 --> 00:11:32,720
Speaker 4: all you want. Right, that's not the spirit of the regulation.

240
00:11:33,000 --> 00:11:35,520
Speaker 4: Applied Materials, go ahead and ship whatever tools you want

241
00:11:35,800 --> 00:11:39,800
Speaker 4: to the twenty eight nanimeter fab in Beijing for Smick,

242
00:11:39,840 --> 00:11:41,959
Speaker 4: and then Smick then gets those tools ship to Shading

243
00:11:42,040 --> 00:11:44,400
Speaker 4: High right where they are making fourteen nanaimeter, where they're

244
00:11:44,400 --> 00:11:47,640
Speaker 4: making seven nanimeter. And this applies not only to you know,

245
00:11:47,720 --> 00:11:50,520
Speaker 4: applied Materials, but also applies to ASML and every other

246
00:11:50,520 --> 00:11:53,360
Speaker 4: equipment company. Right, So when you say, like, hey, yes,

247
00:11:53,440 --> 00:11:56,199
Speaker 4: they're banned from having less than fourteen ananimeter on one hand.

248
00:11:56,280 --> 00:11:59,439
Speaker 4: On the other hand, every single tool that they used

249
00:11:59,440 --> 00:12:03,400
Speaker 4: for their seven nanimeter was the equivalent of what TSMC

250
00:12:03,480 --> 00:12:05,920
Speaker 4: had when they made their seven animeter, or an upgraded

251
00:12:06,000 --> 00:12:08,120
Speaker 4: version of it, right, So it's not like any specific

252
00:12:08,200 --> 00:12:12,000
Speaker 4: tools were banned that were required for seven nanimeter. And

253
00:12:12,080 --> 00:12:14,400
Speaker 4: so it's kind of like, you know, the regulation and

254
00:12:14,440 --> 00:12:17,480
Speaker 4: the implementation were so far away from each other, and

255
00:12:17,520 --> 00:12:20,440
Speaker 4: that's sort of what these recent regulations hopefully are going

256
00:12:20,440 --> 00:12:23,320
Speaker 4: to try and help. But there's still maybe some holes there.

257
00:12:38,160 --> 00:12:41,600
Speaker 1: So actually explain this in an abstract sense, why is

258
00:12:41,640 --> 00:12:45,640
Speaker 1: it hard to align the implementation of the law with

259
00:12:45,679 --> 00:12:47,560
Speaker 1: the letter of the law or the spirit of the law.

260
00:12:48,040 --> 00:12:49,440
Speaker 4: I think it has a lot to do with the

261
00:12:49,480 --> 00:12:52,800
Speaker 4: fact that the US doesn't want to wholesale band chip

262
00:12:52,800 --> 00:12:56,200
Speaker 4: production in China. Pyriice right, while China's automotive you know

263
00:12:56,240 --> 00:13:00,480
Speaker 4: sort of chip manufacturing is exploding bids of vertically integrated

264
00:13:00,520 --> 00:13:03,040
Speaker 4: monster that is like really taking over the world with

265
00:13:03,160 --> 00:13:06,319
Speaker 4: amazing vehicles and chips for those vehicles. I don't think

266
00:13:06,320 --> 00:13:09,240
Speaker 4: the US has really an intention to block those, but

267
00:13:09,280 --> 00:13:11,640
Speaker 4: the tools that are required there. It turns out all

268
00:13:11,679 --> 00:13:14,720
Speaker 4: the advances and tools that have happened over the last decade,

269
00:13:14,840 --> 00:13:16,920
Speaker 4: many of them would still also be applied to that say,

270
00:13:16,920 --> 00:13:18,880
Speaker 4: twenty eight animeter chip, just the same as it would

271
00:13:18,880 --> 00:13:21,520
Speaker 4: be applied to that seven animeter chip. It's just the differences,

272
00:13:21,559 --> 00:13:24,200
Speaker 4: you know, throughput versus you know, accuracy, if you will, right,

273
00:13:24,240 --> 00:13:26,719
Speaker 4: in a simplified sense. And then the other problem is

274
00:13:26,800 --> 00:13:30,240
Speaker 4: this is so incredibly technical, right like Doug basically told

275
00:13:30,240 --> 00:13:32,520
Speaker 4: me ten times, don't say, you know, like a list

276
00:13:32,520 --> 00:13:35,640
Speaker 4: of words that are too complicated for audience because it's

277
00:13:35,720 --> 00:13:38,480
Speaker 4: It's like, the problem is the government is mostly talking

278
00:13:38,559 --> 00:13:40,520
Speaker 4: to hey, like, who's a lithography expert? Well, they all

279
00:13:40,520 --> 00:13:43,600
Speaker 4: work at ASML, and what is their incentive, right, it

280
00:13:43,640 --> 00:13:46,480
Speaker 4: is to ship as many tools period as possible, and

281
00:13:46,559 --> 00:13:49,200
Speaker 4: including to China, Right, Like what is ASML care that

282
00:13:49,760 --> 00:13:51,679
Speaker 4: you know, XYZ is happening. They want to be the

283
00:13:51,720 --> 00:13:53,880
Speaker 4: monopoly and lithography and continue to ship, and if they

284
00:13:53,880 --> 00:13:55,400
Speaker 4: don't ship, then you know, there will be a Chinese

285
00:13:55,400 --> 00:13:57,600
Speaker 4: company eventually one day. Right, So their incentive is for

286
00:13:57,840 --> 00:14:00,280
Speaker 4: the government to have as week of a control around

287
00:14:00,320 --> 00:14:02,880
Speaker 4: lithography as possible, and they want to be like, well, hey,

288
00:14:02,920 --> 00:14:05,079
Speaker 4: like this tool's used heavily here, you can't just blanket

289
00:14:05,080 --> 00:14:06,880
Speaker 4: ban it. No no, no, no, just ban it for

290
00:14:07,160 --> 00:14:09,319
Speaker 4: that fab and we promise we'll make sure that we

291
00:14:09,360 --> 00:14:11,439
Speaker 4: won't ship it to that fab. But if the customer

292
00:14:11,480 --> 00:14:13,040
Speaker 4: decides to move it from Fab A to Fab B,

293
00:14:13,200 --> 00:14:14,160
Speaker 4: then you know, oh no.

294
00:14:14,520 --> 00:14:17,280
Speaker 5: And also I think there's like an implication that China's

295
00:14:17,280 --> 00:14:20,320
Speaker 5: going to play extremely fair with the regulations, and that's

296
00:14:20,680 --> 00:14:24,000
Speaker 5: just clearly like I feel like we keep making these

297
00:14:24,040 --> 00:14:26,560
Speaker 5: regulations and they're really cute and it's like, well the

298
00:14:26,600 --> 00:14:28,960
Speaker 5: spirit of the law and stuff. And then like, meanwhile,

299
00:14:29,000 --> 00:14:30,960
Speaker 5: you look at what China is doing for their domestic

300
00:14:31,000 --> 00:14:33,840
Speaker 5: semicucter production and they're like they don't care, Like they

301
00:14:33,880 --> 00:14:36,240
Speaker 5: do not care. And what they're doing is probably one

302
00:14:36,280 --> 00:14:39,480
Speaker 5: of the most aggressive industrial policies ever to ramp their

303
00:14:39,560 --> 00:14:41,160
Speaker 5: leading edge semicuacter production.

304
00:14:41,240 --> 00:14:42,720
Speaker 4: Like I think it looks like the US in the

305
00:14:42,800 --> 00:14:43,600
Speaker 4: nineteen thirties.

306
00:14:43,840 --> 00:14:46,360
Speaker 5: Yeah, it's totally different. Like, you know, the Chips Act.

307
00:14:46,440 --> 00:14:49,200
Speaker 5: I'm sure everyone is like heard of and understand the

308
00:14:49,240 --> 00:14:51,440
Speaker 5: Chips Act. Right, it's like fifty two billion dollars and

309
00:14:51,840 --> 00:14:55,479
Speaker 5: then plus like another twenty something in tax credits. That's peanuts.

310
00:14:55,520 --> 00:14:58,120
Speaker 5: Like we are talking a completely different game, and how

311
00:14:58,240 --> 00:15:01,320
Speaker 5: meaningful the incentives are there. We could go on about

312
00:15:01,320 --> 00:15:03,680
Speaker 5: this forever, but like essentially every step along the way

313
00:15:03,840 --> 00:15:08,160
Speaker 5: has massive cost and taxes and R and D credits

314
00:15:08,160 --> 00:15:10,800
Speaker 5: and rent reductions and you know, the big fun one,

315
00:15:10,880 --> 00:15:13,840
Speaker 5: two and subsequently three are about to launch and all

316
00:15:13,840 --> 00:15:16,800
Speaker 5: of these things together is probably putting hundreds of billions

317
00:15:16,840 --> 00:15:20,520
Speaker 5: of dollars in subsidies to encourage the semiconductor industry to

318
00:15:20,920 --> 00:15:23,000
Speaker 5: figure it out. And I think that, you know, with

319
00:15:23,160 --> 00:15:26,160
Speaker 5: that much push from the top, you know who cares

320
00:15:26,240 --> 00:15:27,920
Speaker 5: if a tool is not being used in the spirit

321
00:15:27,920 --> 00:15:30,000
Speaker 5: of the wall, right, Like, like China's government isn't going

322
00:15:30,080 --> 00:15:33,080
Speaker 5: to be like, well, darn you for using this for

323
00:15:33,120 --> 00:15:35,920
Speaker 5: this using this deposition tool for not its intended purpose.

324
00:15:36,120 --> 00:15:38,960
Speaker 5: It's pretty clear that they're weaponizing the split in the

325
00:15:39,000 --> 00:15:42,160
Speaker 5: East and West semicopter supply chains, and they're trying to

326
00:15:42,200 --> 00:15:44,200
Speaker 5: do it as fast as possible. And I think that

327
00:15:44,200 --> 00:15:48,160
Speaker 5: that's something that the West just continues to underestimate. How

328
00:15:48,280 --> 00:15:49,720
Speaker 5: much gump should they have toward that?

329
00:15:50,240 --> 00:15:54,080
Speaker 3: Yeah, I remember Dan Wong, another one of our favorite

330
00:15:54,120 --> 00:15:58,120
Speaker 3: Odd Lots guests. He called it China's Sputnik moment when

331
00:15:58,160 --> 00:16:01,480
Speaker 3: the tech restrictions started coming into play, because like, basically

332
00:16:01,840 --> 00:16:04,080
Speaker 3: it was a huge wake up call for China that

333
00:16:04,120 --> 00:16:06,920
Speaker 3: it could no longer count on the US to supply

334
00:16:07,120 --> 00:16:10,440
Speaker 3: its technology and that it would have to basically encourage

335
00:16:10,440 --> 00:16:14,400
Speaker 3: its own domestic alternatives. So on that note, I take

336
00:16:14,440 --> 00:16:17,360
Speaker 3: the point about industrial policy and the scale of the

337
00:16:17,360 --> 00:16:20,400
Speaker 3: way China is doing it here, and certainly China has

338
00:16:20,480 --> 00:16:24,320
Speaker 3: a lot of experience in both industrial policy and just

339
00:16:24,400 --> 00:16:28,120
Speaker 3: generally a centrally planned economy. But what sense do we

340
00:16:28,280 --> 00:16:34,840
Speaker 3: have of how efficient their semiconductor manufacturing process is so far.

341
00:16:34,880 --> 00:16:37,400
Speaker 3: I've seen bits and pieces about this, you know, I've

342
00:16:37,440 --> 00:16:40,480
Speaker 3: seen people talk about there is a huge profit drop

343
00:16:40,600 --> 00:16:43,800
Speaker 3: in SMIC earnings for the third quarter, I think something

344
00:16:43,840 --> 00:16:46,440
Speaker 3: like eighty percent. So a lot of people are going like, yeah, Okay,

345
00:16:46,480 --> 00:16:51,280
Speaker 3: they're producing these new chips. The seven nanometers, but maybe

346
00:16:51,320 --> 00:16:54,480
Speaker 3: they're spending an insane amount of money to do it.

347
00:16:54,880 --> 00:16:57,240
Speaker 3: And also I think there was an inventory shortage of

348
00:16:57,280 --> 00:16:59,800
Speaker 3: the phones, so maybe there's a sense that they're not

349
00:16:59,800 --> 00:17:02,320
Speaker 3: able to produce these at scales just yet.

350
00:17:02,720 --> 00:17:05,680
Speaker 4: So the thing about the manufacturing here is that, you know, yeah,

351
00:17:05,720 --> 00:17:08,160
Speaker 4: their volumes are limited, right, only about seven million phones

352
00:17:08,160 --> 00:17:10,360
Speaker 4: will ship this year, looking like they're going to ship

353
00:17:10,359 --> 00:17:12,920
Speaker 4: maybe forty million next year. Versus the you know one

354
00:17:12,960 --> 00:17:15,480
Speaker 4: point four billion or now one point two billion phones

355
00:17:15,520 --> 00:17:17,399
Speaker 4: that ship a year. This is, you know, kind of

356
00:17:17,400 --> 00:17:19,800
Speaker 4: like an initial drop in the bucket, but the scale

357
00:17:19,840 --> 00:17:22,119
Speaker 4: of the ramp is huge. And what's more important, I

358
00:17:22,119 --> 00:17:24,199
Speaker 4: think that's hard to recognize for most people is that

359
00:17:24,440 --> 00:17:28,800
Speaker 4: semi conductor manufacturing is literally the most complex manufacturing supply

360
00:17:28,880 --> 00:17:31,520
Speaker 4: chain in the world, bar none. There's more processed steps,

361
00:17:31,760 --> 00:17:34,280
Speaker 4: there's more complex there's more R and D in this

362
00:17:34,400 --> 00:17:36,960
Speaker 4: field than any other field. It is the most complex

363
00:17:36,960 --> 00:17:39,480
Speaker 4: supply chain, period and so when you're talking about thousands

364
00:17:39,520 --> 00:17:42,280
Speaker 4: and thousands of processed steps, each step has one hundred

365
00:17:42,280 --> 00:17:44,919
Speaker 4: different knobs that you could turn on each tool getting

366
00:17:44,960 --> 00:17:47,760
Speaker 4: good yield, right, getting the number of chips you try

367
00:17:47,800 --> 00:17:49,640
Speaker 4: to make for how many actually work in the end

368
00:17:49,960 --> 00:17:53,040
Speaker 4: is very very difficult. But the way you learn is

369
00:17:53,119 --> 00:17:57,360
Speaker 4: by ramping right by by producing and then hey, if

370
00:17:57,359 --> 00:17:59,720
Speaker 4: I'm running a thousand experiments in flight, and each one

371
00:17:59,760 --> 00:18:02,359
Speaker 4: has one knob slightly turned differently, and then I see, oh,

372
00:18:02,400 --> 00:18:02,920
Speaker 4: which one work?

373
00:18:03,119 --> 00:18:03,439
Speaker 5: Awesome?

374
00:18:03,480 --> 00:18:05,119
Speaker 4: Now I know that I leave that knob turned, and

375
00:18:05,160 --> 00:18:06,720
Speaker 4: now I move on to another piece. Right, sort of,

376
00:18:07,000 --> 00:18:10,080
Speaker 4: there's this like complex like you know, constant tweaking of

377
00:18:10,080 --> 00:18:13,199
Speaker 4: the process, right, this is what's made TSMC, you know,

378
00:18:13,280 --> 00:18:16,440
Speaker 4: so amazing, versus say a company like Intel. Is that TSMC,

379
00:18:16,840 --> 00:18:19,479
Speaker 4: their engineers are constantly doing this even on fifteen year

380
00:18:19,520 --> 00:18:22,560
Speaker 4: old process technology, still getting they old better, still getting

381
00:18:22,560 --> 00:18:25,480
Speaker 4: the performance better on say ninety to animeter, whereas Intel

382
00:18:25,520 --> 00:18:28,320
Speaker 4: would move on to work on the newest technology every

383
00:18:28,359 --> 00:18:31,399
Speaker 4: single generation. And so Smick by you know, yes, they

384
00:18:31,440 --> 00:18:34,600
Speaker 4: are not profitable, especially if you strip away the subsidies, right,

385
00:18:34,600 --> 00:18:38,240
Speaker 4: the Smick Shanghai joint venture, the subsidies are getting from Beijing.

386
00:18:38,640 --> 00:18:40,880
Speaker 4: These are massive. If you were stripped those out, they'd

387
00:18:40,880 --> 00:18:43,840
Speaker 4: be losing billions a year, right, But you know that's

388
00:18:43,960 --> 00:18:46,600
Speaker 4: that's the thing is like, hey, let's skip forward now,

389
00:18:46,720 --> 00:18:49,440
Speaker 4: is there twenty eight animeter profitable probably, Yeah, it's their

390
00:18:49,440 --> 00:18:51,760
Speaker 4: seven animeter that's not okay. But we skip forward a

391
00:18:51,800 --> 00:18:54,080
Speaker 4: couple of years and it's like seven animeter will be

392
00:18:54,160 --> 00:18:56,879
Speaker 4: very profitable. And guess what, over eighty five percent of

393
00:18:56,880 --> 00:19:01,000
Speaker 4: the world's chip value is not under seven animeter today, right,

394
00:19:01,760 --> 00:19:04,359
Speaker 4: And so like there is a long tail of like, hey,

395
00:19:04,359 --> 00:19:06,719
Speaker 4: my car has zero seven animeter chips, right, But there

396
00:19:06,720 --> 00:19:08,440
Speaker 4: are a ton of chips that are made by Texas

397
00:19:08,440 --> 00:19:11,000
Speaker 4: Instruments and analog devices and microchip and you go down

398
00:19:11,040 --> 00:19:12,720
Speaker 4: the list and it's like, well, China's going to compete

399
00:19:12,760 --> 00:19:14,239
Speaker 4: with these companies and they're going to compete with them

400
00:19:14,320 --> 00:19:14,880
Speaker 4: very strongly.

401
00:19:15,080 --> 00:19:15,280
Speaker 2: Yeah.

402
00:19:15,320 --> 00:19:17,320
Speaker 5: And I think that also just looking at as in

403
00:19:17,320 --> 00:19:21,280
Speaker 5: a pure profitability perspective maybe loses some of the context

404
00:19:21,280 --> 00:19:24,600
Speaker 5: of history. Taiwan Korea did this exact same thing, where

405
00:19:25,119 --> 00:19:26,840
Speaker 5: you can look at the early days of TSMC, and

406
00:19:26,840 --> 00:19:29,399
Speaker 5: they lost a lot of money along the way. But

407
00:19:29,840 --> 00:19:33,320
Speaker 5: the important thing is that by pushing their self sufficiency,

408
00:19:33,359 --> 00:19:34,640
Speaker 5: what they're going to be able to do is they're

409
00:19:34,640 --> 00:19:36,720
Speaker 5: going to be able to ride up the experience curves

410
00:19:36,720 --> 00:19:40,200
Speaker 5: on every single technology as they're pushing for full self sufficiency.

411
00:19:40,400 --> 00:19:42,720
Speaker 5: And I think that this story continues to evolve, and

412
00:19:42,760 --> 00:19:44,200
Speaker 5: I think that, like that's one of the reasons why

413
00:19:44,200 --> 00:19:46,159
Speaker 5: we're here to be excited to talk about it. Is

414
00:19:46,200 --> 00:19:49,520
Speaker 5: like CXMT today actually I think is literally today or yesterday.

415
00:19:49,680 --> 00:19:53,199
Speaker 5: Which company c x MT. It's their d RAM company.

416
00:19:53,240 --> 00:19:56,520
Speaker 5: So they have a nand company which is YMTC, which

417
00:19:56,520 --> 00:19:59,760
Speaker 5: for context, is shipping the most dense memory in the world.

418
00:19:59,760 --> 00:20:01,880
Speaker 5: I think that's correct, Dylan, like.

419
00:20:01,600 --> 00:20:03,800
Speaker 4: Most dense memory. Maybe the yields aren't great, but they've

420
00:20:03,880 --> 00:20:07,920
Speaker 4: they've been effectively banned from tooling like properly, like no loopholes,

421
00:20:07,960 --> 00:20:10,280
Speaker 4: and so they've been they've been stumbling. But six MT

422
00:20:10,760 --> 00:20:14,200
Speaker 4: company found it in twenty sixteen, Yeah, and they are.

423
00:20:14,440 --> 00:20:17,359
Speaker 4: They're looking to spend about seven to eight billion dollars

424
00:20:17,400 --> 00:20:19,560
Speaker 4: on equipment next year. Yeah, which is more than Micron

425
00:20:19,680 --> 00:20:20,040
Speaker 4: by the way.

426
00:20:20,160 --> 00:20:22,760
Speaker 5: Yeah. And so so like they're shipping these memories that

427
00:20:22,800 --> 00:20:24,919
Speaker 5: are like maybe two to four years old. But like

428
00:20:25,320 --> 00:20:27,520
Speaker 5: this conversation, if we had this five years ago, it

429
00:20:27,520 --> 00:20:30,119
Speaker 5: would seem insane, Like truly, I think the rate of

430
00:20:30,240 --> 00:20:32,439
Speaker 5: change is something that people continue to miss and I

431
00:20:32,440 --> 00:20:36,080
Speaker 5: think that as this continues, it's going to be you know,

432
00:20:36,160 --> 00:20:39,800
Speaker 5: after they have their own domestic phone or domestic five

433
00:20:39,840 --> 00:20:42,600
Speaker 5: G modem, the domestic CPU for the phone or like whatever,

434
00:20:42,640 --> 00:20:45,199
Speaker 5: the NPU. Then they have their own domestic nand then

435
00:20:45,200 --> 00:20:47,520
Speaker 5: they have their own domestic DRAM. Then all of a sudden,

436
00:20:47,640 --> 00:20:50,160
Speaker 5: you just like look at it, and that's the entire

437
00:20:50,200 --> 00:20:52,760
Speaker 5: semicapter tool chain, Like you know, so this is this

438
00:20:52,800 --> 00:20:55,359
Speaker 5: is and so yeah, is it capital efficient? Hell no,

439
00:20:55,600 --> 00:20:59,479
Speaker 5: Like Big fun two, for example, has allegations of massive

440
00:20:59,480 --> 00:21:02,679
Speaker 5: amounts of and that definitely is happening. And there's hundreds

441
00:21:02,720 --> 00:21:05,200
Speaker 5: and hundreds of companies that have started probably this year

442
00:21:05,280 --> 00:21:09,240
Speaker 5: to take advantage of the subsidies for semicopters. And yes,

443
00:21:09,320 --> 00:21:12,880
Speaker 5: there are lots of misses, but in terms of being

444
00:21:12,960 --> 00:21:15,800
Speaker 5: able to get the talent there and then also bring

445
00:21:15,840 --> 00:21:17,840
Speaker 5: them all up the learning experience, I think that that's

446
00:21:17,880 --> 00:21:19,800
Speaker 5: a huge thing. And I think that's also a conversation

447
00:21:20,080 --> 00:21:22,359
Speaker 5: that is completely missed about this as well, is that

448
00:21:22,400 --> 00:21:25,520
Speaker 5: like Dylan and I we go to you know, trade

449
00:21:25,520 --> 00:21:29,000
Speaker 5: shows in Japan or Taiwan or Korea, and like the

450
00:21:29,040 --> 00:21:31,080
Speaker 5: amount of young people there versus the amount of young

451
00:21:31,119 --> 00:21:33,600
Speaker 5: people in the West that are making chips is just

452
00:21:34,000 --> 00:21:36,840
Speaker 5: drastically different. And I think that you fast forward that

453
00:21:36,880 --> 00:21:39,439
Speaker 5: same equation, and yes, maybe they don't have ev tools,

454
00:21:39,480 --> 00:21:41,680
Speaker 5: but there are a lot of ways to make better

455
00:21:41,720 --> 00:21:44,960
Speaker 5: phones using packaging or other clever engineering tricks. And so

456
00:21:45,480 --> 00:21:47,720
Speaker 5: what's going to happen is that not only do they

457
00:21:47,760 --> 00:21:50,600
Speaker 5: have the technological know how, they also have the talent too.

458
00:21:51,040 --> 00:21:53,560
Speaker 5: And we could be talking about in you know, ten

459
00:21:53,640 --> 00:21:58,000
Speaker 5: twenty years that America's chip dominance is very you know,

460
00:21:58,040 --> 00:22:00,960
Speaker 5: backwards looking, and I think people need to watch this closer.

461
00:22:01,000 --> 00:22:04,080
Speaker 5: The rate of change has just constantly shocked me at least,

462
00:22:04,119 --> 00:22:21,000
Speaker 5: and I'm sure Dylan, but like the progress is staggering.

463
00:22:22,240 --> 00:22:24,359
Speaker 3: This kind of gets me to I don't know if

464
00:22:24,400 --> 00:22:26,919
Speaker 3: it's a strange question or theoretical question, but when it

465
00:22:26,920 --> 00:22:31,399
Speaker 3: comes to semiconductors, you hear a lot about the importance

466
00:22:31,520 --> 00:22:34,360
Speaker 3: of the overall supply chain. So, you know, people will

467
00:22:34,400 --> 00:22:37,240
Speaker 3: say that in Video is really good at managing its

468
00:22:37,240 --> 00:22:40,280
Speaker 3: own supply chain and that's been part of its success story.

469
00:22:41,000 --> 00:22:45,800
Speaker 3: If China is now manufacturing more and more of its

470
00:22:45,840 --> 00:22:49,720
Speaker 3: own chips on a domestic basis, I guess like maybe

471
00:22:49,720 --> 00:22:53,440
Speaker 3: it'll take them longer to develop some things, But does

472
00:22:53,480 --> 00:22:56,400
Speaker 3: that in the end kind of lead to a more

473
00:22:56,480 --> 00:23:00,480
Speaker 3: resilient or reliable supply chain for that technology. How do

474
00:23:00,520 --> 00:23:02,720
Speaker 3: you see those two things interacting?

475
00:23:03,600 --> 00:23:07,439
Speaker 4: You know, it depends. In capital intensive industries, what you

476
00:23:07,600 --> 00:23:10,960
Speaker 4: tend to see is a very big sort of winner

477
00:23:11,000 --> 00:23:14,840
Speaker 4: takes all vibe, which is why in prior you know

478
00:23:14,880 --> 00:23:17,960
Speaker 4: sort of technologies, if you will, right solar for example, Hey,

479
00:23:18,000 --> 00:23:20,920
Speaker 4: Germany in the US invented all of it, and they

480
00:23:20,920 --> 00:23:23,800
Speaker 4: were great until China dumped way more capital in a

481
00:23:23,840 --> 00:23:26,800
Speaker 4: very capital inefficient way, put way more people at work

482
00:23:26,840 --> 00:23:29,479
Speaker 4: at it, right, smart engineers working on it, and all

483
00:23:29,520 --> 00:23:31,480
Speaker 4: of a sudden, you know, hey, ninety plus percent of

484
00:23:31,520 --> 00:23:34,239
Speaker 4: solar cells come from China. Right. And likewise, you know,

485
00:23:34,680 --> 00:23:38,000
Speaker 4: US companies were the first to do electric vehicles, right,

486
00:23:38,200 --> 00:23:41,800
Speaker 4: like GM initially than Tesla. But hey, China makes like

487
00:23:41,800 --> 00:23:44,600
Speaker 4: three times as many electric vehicles as the West does now,

488
00:23:44,760 --> 00:23:46,679
Speaker 4: and so you know, sort of you see these technology

489
00:23:46,680 --> 00:23:49,520
Speaker 4: curves and there's a significant amount of winner takes all, right.

490
00:23:49,600 --> 00:23:51,320
Speaker 4: You know, if you look at Latin American markets, I

491
00:23:51,359 --> 00:23:53,640
Speaker 4: mean even even Europe had to put up tariffs recently,

492
00:23:54,000 --> 00:23:56,760
Speaker 4: Chinese evs are killing it, right. The US has had

493
00:23:56,760 --> 00:23:59,399
Speaker 4: to put up massive tariffs and then has had the

494
00:23:59,560 --> 00:24:02,280
Speaker 4: sort of the flat reduction act with the huge subsidies

495
00:24:02,280 --> 00:24:04,600
Speaker 4: for battery production and cathodes and anodes and all these

496
00:24:04,640 --> 00:24:07,480
Speaker 4: sorts of things. There is a big winner takes all

497
00:24:07,600 --> 00:24:10,600
Speaker 4: sort of element of capital intensive industries, and the semi

498
00:24:10,600 --> 00:24:13,440
Speaker 4: conductor industry is not different. In fact, it is even

499
00:24:13,520 --> 00:24:16,480
Speaker 4: more salted that way. Every single vertical you look in,

500
00:24:16,520 --> 00:24:18,560
Speaker 4: and there's hundreds of verticals, people think, oh, it's such

501
00:24:18,600 --> 00:24:20,680
Speaker 4: a complex supply chain. Well, every single spot where you

502
00:24:20,760 --> 00:24:24,600
Speaker 4: zone in, there's two, maybe three competitors right in any

503
00:24:24,640 --> 00:24:27,600
Speaker 4: specific technology and any specific chip, and many times one

504
00:24:27,880 --> 00:24:30,280
Speaker 4: and everyone makes a good profit. And it's very like,

505
00:24:30,359 --> 00:24:34,359
Speaker 4: you know, strong industry. But what happens when someone comes

506
00:24:34,359 --> 00:24:36,359
Speaker 4: with breakthrough innovation and now all of a sudden, you know,

507
00:24:36,800 --> 00:24:39,880
Speaker 4: it's capital intensive, and those players start to fall off

508
00:24:40,000 --> 00:24:42,800
Speaker 4: or they exit markets. That's something that happens, right, and

509
00:24:42,840 --> 00:24:44,760
Speaker 4: the winner does take all. And so it's not just

510
00:24:44,800 --> 00:24:48,040
Speaker 4: necessarily a hey, it's a it's a domestic supply chain.

511
00:24:48,119 --> 00:24:50,840
Speaker 4: It's hey, well what about all of the vehicles that

512
00:24:50,880 --> 00:24:53,760
Speaker 4: are being shipped out of China into Latin Southeast Asia

513
00:24:53,880 --> 00:24:55,919
Speaker 4: and you know, maybe even Europe if they don't have

514
00:24:56,000 --> 00:24:59,119
Speaker 4: tariffs or you know, the US already blocks Chinese vehicles mostly,

515
00:24:59,200 --> 00:25:01,520
Speaker 4: or or how about you know solar right, like battery,

516
00:25:01,640 --> 00:25:04,040
Speaker 4: you know, solar inverters to convert the power from sort

517
00:25:04,080 --> 00:25:05,960
Speaker 4: of the solar panels to what's acceptable for your home

518
00:25:06,040 --> 00:25:08,000
Speaker 4: or for the grid. That's going to become a Chinese

519
00:25:08,000 --> 00:25:09,760
Speaker 4: supply chain. Why wouldn't it, Right, they already have the

520
00:25:09,800 --> 00:25:12,960
Speaker 4: solar panels, So it's not just like hey, domestic versus not.

521
00:25:13,080 --> 00:25:16,639
Speaker 4: It's actually there's a very strong element of hey, this

522
00:25:16,720 --> 00:25:18,919
Speaker 4: is going to be a Chinese supply chain, or this

523
00:25:19,000 --> 00:25:20,679
Speaker 4: is going to be a Japanese supply chain, or this

524
00:25:20,760 --> 00:25:22,840
Speaker 4: is going to be a Korean supply chain, which is

525
00:25:22,840 --> 00:25:25,119
Speaker 4: already the case in sort of semiconductors, right, or a

526
00:25:25,200 --> 00:25:26,040
Speaker 4: US supply chain.

527
00:25:26,160 --> 00:25:28,880
Speaker 5: Yeah, and I think there's also like an important thing

528
00:25:28,960 --> 00:25:30,920
Speaker 5: to note there. At least I don't think this will

529
00:25:30,920 --> 00:25:32,760
Speaker 5: happen in the United States, but Europe is where this

530
00:25:32,800 --> 00:25:35,160
Speaker 5: is happening. The hardest is so we've been talking about

531
00:25:35,160 --> 00:25:37,760
Speaker 5: the leading edge which they put out this new Huawei phone,

532
00:25:37,840 --> 00:25:41,520
Speaker 5: really amazing technology, but there's problems with ramping, there's problems

533
00:25:41,520 --> 00:25:44,479
Speaker 5: with yield. We have some cutoffs. But like you know,

534
00:25:45,000 --> 00:25:47,840
Speaker 5: China at the same time is also pressing their foots

535
00:25:47,840 --> 00:25:49,679
Speaker 5: of the gas in the lagging edge. And we've been

536
00:25:49,680 --> 00:25:51,760
Speaker 5: talking about automotive, and I think that that's probably where

537
00:25:51,800 --> 00:25:55,960
Speaker 5: it's scariest. Frankly. So, if you are a European automotive

538
00:25:55,960 --> 00:25:59,199
Speaker 5: OEM and you are trying to sell more ebs as

539
00:25:59,240 --> 00:26:03,600
Speaker 5: everyone is, and you're buying your semicopters from other semicapter companies,

540
00:26:03,680 --> 00:26:05,440
Speaker 5: and then you're buying the battery and then you're you're

541
00:26:05,480 --> 00:26:07,600
Speaker 5: putting it all together and you're trying to have this margin.

542
00:26:07,680 --> 00:26:09,240
Speaker 5: You know, it's like a it's a single digit or

543
00:26:09,240 --> 00:26:11,400
Speaker 5: you know, maybe double digit margin. It's it's a very

544
00:26:11,440 --> 00:26:14,480
Speaker 5: low dollar value added industry, and then you're trying to

545
00:26:14,480 --> 00:26:18,239
Speaker 5: compete against bid who and BYD in this conversation is

546
00:26:18,280 --> 00:26:20,320
Speaker 5: making their own batteries. They're buying their own ships to

547
00:26:20,400 --> 00:26:23,119
Speaker 5: ship the cars from China to Wow. Yeah they are.

548
00:26:23,440 --> 00:26:24,280
Speaker 4: They have their own fab.

549
00:26:24,400 --> 00:26:26,359
Speaker 5: Yeah, they have their own fab They make all the semicopters,

550
00:26:26,359 --> 00:26:29,480
Speaker 5: and like the semiconductors are because they're becoming larger and

551
00:26:29,560 --> 00:26:32,720
Speaker 5: larger parts of cars, and they are the most profitable.

552
00:26:32,880 --> 00:26:36,359
Speaker 5: They are the largest gross dollar profit pool within cars.

553
00:26:36,880 --> 00:26:38,400
Speaker 5: And so if you just do it, talk.

554
00:26:38,320 --> 00:26:40,760
Speaker 1: Called about batteries all the time. But the real profits

555
00:26:40,760 --> 00:26:41,760
Speaker 1: are in the chips.

556
00:26:41,560 --> 00:26:44,800
Speaker 5: Well well batteries as well, batteries and batteries and semicopters,

557
00:26:44,800 --> 00:26:48,040
Speaker 5: and they're doing both completely vertically integrated. So what they're

558
00:26:48,080 --> 00:26:50,199
Speaker 5: gonna do is they're going to collapse all that margin

559
00:26:50,280 --> 00:26:52,080
Speaker 5: that the suppliers make and then they're going to take

560
00:26:52,119 --> 00:26:53,600
Speaker 5: it out of the price of the car and they'll

561
00:26:53,600 --> 00:26:57,320
Speaker 5: make the full stack margin. But if you're a European

562
00:26:57,680 --> 00:27:00,679
Speaker 5: or American company competing against this, you will never compete

563
00:27:00,680 --> 00:27:03,520
Speaker 5: on price. And I think that this is and you see.

564
00:27:03,320 --> 00:27:06,040
Speaker 4: It with Tesla, right, Tesla has the highest gross margins

565
00:27:06,040 --> 00:27:08,280
Speaker 4: in the industry. You can say what you want about

566
00:27:08,320 --> 00:27:11,120
Speaker 4: their cars, right, but they have the highest gross margins,

567
00:27:11,119 --> 00:27:13,920
Speaker 4: and especially after the like the seventh price cuts this year,

568
00:27:13,960 --> 00:27:16,600
Speaker 4: they're cheap, yeah, right. And Bidea is the same, except

569
00:27:16,680 --> 00:27:19,600
Speaker 4: even better in many regards it even more vertically and integrated.

570
00:27:19,760 --> 00:27:20,000
Speaker 2: Yep.

571
00:27:20,280 --> 00:27:22,439
Speaker 5: And it's happening a lot quicker than people realize, and

572
00:27:22,480 --> 00:27:25,080
Speaker 5: I think that that's something that it's another example of

573
00:27:25,119 --> 00:27:28,640
Speaker 5: where the industrial policy is just going as fast as

574
00:27:28,640 --> 00:27:31,200
Speaker 5: possible with this hundreds of billions of dollars to subsidies,

575
00:27:31,240 --> 00:27:33,800
Speaker 5: and people are not really paying attention to the story

576
00:27:33,800 --> 00:27:34,959
Speaker 5: that's happening right in front of us.

577
00:27:35,040 --> 00:27:38,439
Speaker 1: I'm glad you brought up the car chips because it

578
00:27:38,520 --> 00:27:42,280
Speaker 1: does seem as though the disruptions to car chips that

579
00:27:42,320 --> 00:27:45,080
Speaker 1: we saw in twenty twenty is sort of what brought

580
00:27:45,119 --> 00:27:47,760
Speaker 1: the semiconductor supply chain to the forefront. It's what sort

581
00:27:47,800 --> 00:27:50,160
Speaker 1: of got everyone's consciousness. So we can't make cars because

582
00:27:50,200 --> 00:27:52,639
Speaker 1: we can't get the chips. But those are, you know,

583
00:27:52,680 --> 00:27:56,400
Speaker 1: the lagging edge chips, So we sort of it feels

584
00:27:56,440 --> 00:27:59,399
Speaker 1: like there is this sort of incoherence that we maybe

585
00:27:59,440 --> 00:28:01,639
Speaker 1: tell about the story here where it says, okay, we

586
00:28:01,680 --> 00:28:03,880
Speaker 1: need to invest in chips, but then we talk about

587
00:28:03,920 --> 00:28:06,040
Speaker 1: we need to invest in the leading age, whereas actually

588
00:28:06,080 --> 00:28:07,560
Speaker 1: the real disruption was at.

589
00:28:07,440 --> 00:28:08,200
Speaker 2: The lagging edge.

590
00:28:08,280 --> 00:28:09,640
Speaker 1: Can you talk a little bit more what you said

591
00:28:09,680 --> 00:28:12,840
Speaker 1: about how the US companies gave up on the lagging edge,

592
00:28:12,840 --> 00:28:16,359
Speaker 1: Whereas you say TSMC and others continue to improve their

593
00:28:16,440 --> 00:28:17,560
Speaker 1: yields on the lagging edge.

594
00:28:17,600 --> 00:28:19,399
Speaker 2: We don't have that here with say Intel.

595
00:28:19,640 --> 00:28:22,800
Speaker 4: So in some cases, right with companies like Analog Devices,

596
00:28:22,840 --> 00:28:25,399
Speaker 4: who used to vertically manufacture all their chips, they're moving

597
00:28:25,440 --> 00:28:27,760
Speaker 4: more and more to having TSMC make for them, but

598
00:28:27,840 --> 00:28:30,800
Speaker 4: in other cases, Texas Instruments, right, both of these companies

599
00:28:30,840 --> 00:28:33,080
Speaker 4: are over one hundred million dollar companies, but Texas Instruments,

600
00:28:33,280 --> 00:28:36,320
Speaker 4: they are actually investing hugely in making more fabs and

601
00:28:36,800 --> 00:28:39,640
Speaker 4: having more than eighty percent of their chips manufactured by themselves.

602
00:28:39,880 --> 00:28:41,520
Speaker 4: And you know, I think, you know, just a little

603
00:28:41,520 --> 00:28:43,440
Speaker 4: bit of a chime in on the whole term lagging

604
00:28:43,520 --> 00:28:45,160
Speaker 4: edge is that it take a bit of a misnomer

605
00:28:45,240 --> 00:28:47,760
Speaker 4: because it's not like they're selling the same chip that

606
00:28:47,800 --> 00:28:49,959
Speaker 4: the designed in the nineteen eight in nineties, right, well,

607
00:28:50,000 --> 00:28:53,000
Speaker 4: they are, but they're also selling many, many specialized chips

608
00:28:53,000 --> 00:28:55,040
Speaker 4: that just aren't low nanimeters, but they have some material

609
00:28:55,080 --> 00:28:57,920
Speaker 4: innovations or you know, they have some different properties with

610
00:28:57,960 --> 00:29:01,680
Speaker 4: them that aren't necessarily the smallest post but definitely specialized.

611
00:29:01,960 --> 00:29:05,200
Speaker 4: And there's a humongous variety of these chips, right, So

612
00:29:05,240 --> 00:29:07,960
Speaker 4: it's not like, oh, there's one chip, right, it's it's

613
00:29:08,200 --> 00:29:10,480
Speaker 4: you know, phone chips are quite easy to understand, and

614
00:29:10,520 --> 00:29:12,960
Speaker 4: it's a it's a single chip that it has a

615
00:29:13,000 --> 00:29:16,320
Speaker 4: tens of billions of dollars market, right. AI chips same thing,

616
00:29:16,440 --> 00:29:18,320
Speaker 4: tens of billions of dollars of market, right, so we

617
00:29:18,320 --> 00:29:20,480
Speaker 4: can point to that one in video chip and be like, wow,

618
00:29:20,520 --> 00:29:23,560
Speaker 4: look at it. But when you look at Texas Instruments catalogs,

619
00:29:23,720 --> 00:29:27,040
Speaker 4: the catalog is like so thick. It's like it's incredible

620
00:29:27,280 --> 00:29:28,960
Speaker 4: of how many different chips they sell.

621
00:29:29,040 --> 00:29:32,240
Speaker 5: It's thousands of products. It's thousands and thousands of products.

622
00:29:32,240 --> 00:29:35,040
Speaker 5: And I think Texas Instruments will say this that like

623
00:29:35,120 --> 00:29:37,800
Speaker 5: half of their products have been from nineteen ninety in earlier.

624
00:29:38,400 --> 00:29:40,360
Speaker 5: So there there's a lot of lagging edge chips. But

625
00:29:40,360 --> 00:29:43,720
Speaker 5: the difference here is that Texas Instruments is making like

626
00:29:43,760 --> 00:29:46,560
Speaker 5: a sixty five seventy percent gross margin. The business is

627
00:29:46,800 --> 00:29:50,560
Speaker 5: managed extremely like a mature semicopter company, it is. But

628
00:29:50,600 --> 00:29:53,760
Speaker 5: then you have the new entrants in China that are

629
00:29:53,760 --> 00:29:56,280
Speaker 5: just saying, let's like completely screw up the industry structure,

630
00:29:56,280 --> 00:29:59,680
Speaker 5: and that's that's something that's very new MCUs micro controller

631
00:29:59,720 --> 00:30:02,440
Speaker 5: units and China on that is like the hottest topic

632
00:30:02,560 --> 00:30:06,840
Speaker 5: of disruption and what people are trying to do. And MCUs, Yeah,

633
00:30:06,880 --> 00:30:08,600
Speaker 5: it might be a twenty eight animeter chip or maybe

634
00:30:08,600 --> 00:30:10,720
Speaker 5: it's like a you know, fourteen animeter chip or something

635
00:30:10,760 --> 00:30:12,800
Speaker 5: like that. It's not the sexiest thing, but that is

636
00:30:12,840 --> 00:30:14,640
Speaker 5: a you know, that's a tens of billions of dollar

637
00:30:14,680 --> 00:30:17,200
Speaker 5: market there. And I think for the reason why it's

638
00:30:17,200 --> 00:30:19,800
Speaker 5: maybe not as viewed as strategically important is, let's put

639
00:30:19,800 --> 00:30:21,880
Speaker 5: this in the context of AI. Right, it is not

640
00:30:21,960 --> 00:30:24,920
Speaker 5: a leading edge AI chip that you know, and the

641
00:30:25,360 --> 00:30:28,160
Speaker 5: original restrictions that we put on China was to stop

642
00:30:28,360 --> 00:30:32,160
Speaker 5: AI progress, right, So that's not exactly you know, strategically

643
00:30:32,200 --> 00:30:35,520
Speaker 5: quote unquote important, but in terms of for the businesses,

644
00:30:35,600 --> 00:30:38,840
Speaker 5: for economics, it's it's a big deal. So what they've

645
00:30:38,840 --> 00:30:41,080
Speaker 5: done is they've kind of refocused on the lagging edge

646
00:30:41,080 --> 00:30:43,600
Speaker 5: where they can, and they're throwing as much money into there,

647
00:30:43,640 --> 00:30:45,840
Speaker 5: and it's going to probably create crazy price.

648
00:30:46,040 --> 00:30:48,280
Speaker 4: Going back to this this sort of idea of that, hey,

649
00:30:48,280 --> 00:30:50,160
Speaker 4: this catalog is thousands of chips. This is why no

650
00:30:50,200 --> 00:30:52,960
Speaker 4: one competes with each other, right, Like Texas instruments, they're

651
00:30:53,000 --> 00:30:55,840
Speaker 4: the only player in many markets, right, or analog devices,

652
00:30:55,840 --> 00:30:57,880
Speaker 4: And maybe the chip is only eight cents to sell,

653
00:30:58,400 --> 00:30:59,840
Speaker 4: but it maybe it only cost them two cents to

654
00:30:59,880 --> 00:31:03,080
Speaker 4: me manufacturer. The thing about China is they've they've literally

655
00:31:03,120 --> 00:31:06,360
Speaker 4: sighoped their entire generation into wanting to work in semiconductors.

656
00:31:06,400 --> 00:31:09,280
Speaker 4: It's literally the coolest job. There are two different dramas

657
00:31:09,280 --> 00:31:09,760
Speaker 4: that I know of.

658
00:31:10,080 --> 00:31:12,440
Speaker 3: I was about to ask, like, are there any TV

659
00:31:12,640 --> 00:31:16,000
Speaker 3: characters that are working in semiconductors, because that's when you know.

660
00:31:16,280 --> 00:31:18,880
Speaker 4: Yeah, there's there's one of like this kid who's in

661
00:31:18,960 --> 00:31:20,920
Speaker 4: college and he's laying but then he like goes and

662
00:31:20,920 --> 00:31:23,120
Speaker 4: works in the semiconductor industry and now he's super cool.

663
00:31:23,200 --> 00:31:25,400
Speaker 4: And then there's another one where it's like a love

664
00:31:25,480 --> 00:31:27,479
Speaker 4: story and they both work in a fab and they

665
00:31:27,560 --> 00:31:29,960
Speaker 4: like fall in love and that's it's like a drama, right,

666
00:31:30,040 --> 00:31:30,600
Speaker 4: It's it's.

667
00:31:30,520 --> 00:31:32,960
Speaker 3: They're like gazing at each other through the glass and

668
00:31:33,000 --> 00:31:35,080
Speaker 3: stuff like that. That'd be pretty funny.

669
00:31:35,240 --> 00:31:36,720
Speaker 5: Yeah, both wearing bunny suits.

670
00:31:38,440 --> 00:31:41,880
Speaker 3: But I don't know if anyone heard, but I think

671
00:31:42,240 --> 00:31:45,200
Speaker 3: the sound earlier was both Joe and I scrambling to

672
00:31:45,240 --> 00:31:48,920
Speaker 3: pull up the Texas Instruments.

673
00:31:46,960 --> 00:31:49,400
Speaker 2: A literally we both how did you know what? I

674
00:31:49,480 --> 00:31:49,920
Speaker 2: was typing?

675
00:31:51,480 --> 00:31:51,560
Speaker 3: That?

676
00:31:51,600 --> 00:31:53,880
Speaker 2: We both googled Texas Instruments catalog.

677
00:31:53,960 --> 00:31:56,320
Speaker 3: But I can tell you I'm looking at the page

678
00:31:56,600 --> 00:31:59,200
Speaker 3: just for MCUs and it looks like it's more than

679
00:31:59,240 --> 00:32:02,480
Speaker 3: a thousand, two hundred that was just one category product.

680
00:32:02,720 --> 00:32:06,040
Speaker 1: Yeah, yeah, twelve hundred and thirty eight micro controllers and processors.

681
00:32:06,160 --> 00:32:07,640
Speaker 2: Can we just order them online?

682
00:32:08,280 --> 00:32:10,000
Speaker 3: I guess you can. I mean, click the box, see

683
00:32:10,000 --> 00:32:10,560
Speaker 3: what happens.

684
00:32:10,640 --> 00:32:12,520
Speaker 4: Each of those has a hundred couple hundred pages of

685
00:32:12,520 --> 00:32:15,200
Speaker 4: PDFs associated with how to design them into a product.

686
00:32:15,240 --> 00:32:17,760
Speaker 4: And it's like, there's a reason no one redesigns this stuff.

687
00:32:17,800 --> 00:32:21,200
Speaker 4: But well, now when you have thousands, hundreds of thousands

688
00:32:21,240 --> 00:32:23,920
Speaker 4: of people coming into the supply chain and like, hey,

689
00:32:23,920 --> 00:32:26,440
Speaker 4: we can't order this from TXN anymore. We need to

690
00:32:26,520 --> 00:32:28,240
Speaker 4: order this, we need to get do this domestically, it's

691
00:32:28,240 --> 00:32:30,720
Speaker 4: like and then and then you also have subsidies that say, hey,

692
00:32:30,720 --> 00:32:33,240
Speaker 4: if every time you design a chip of you know,

693
00:32:33,320 --> 00:32:36,760
Speaker 4: certain ages, you get pure tax credits.

694
00:32:36,400 --> 00:32:38,480
Speaker 5: You get you get you get just straight up a

695
00:32:38,480 --> 00:32:41,640
Speaker 5: payout local governments. Essentially, each tape out will give you

696
00:32:41,680 --> 00:32:42,520
Speaker 5: like a million.

697
00:32:42,200 --> 00:32:43,320
Speaker 4: Bucks tape out as a design of it.

698
00:32:43,480 --> 00:32:45,080
Speaker 5: Yeah, yeah, a tape out all the way to design.

699
00:32:45,160 --> 00:32:46,920
Speaker 5: So if you make one of these, it doesn't matter

700
00:32:46,920 --> 00:32:48,760
Speaker 5: what it is, you just get money. And you know,

701
00:32:48,760 --> 00:32:52,000
Speaker 5: if you're engaged in doing anything related to a semicopter,

702
00:32:52,040 --> 00:32:54,120
Speaker 5: you have five years for every company, ten years for

703
00:32:54,200 --> 00:32:57,480
Speaker 5: leading edge companies tax free like two hundred percent R

704
00:32:57,480 --> 00:32:59,680
Speaker 5: and D credits, meaning that like for every ten dollars

705
00:32:59,720 --> 00:33:01,840
Speaker 5: you spend or let's saye hundred dollars you spend in

706
00:33:02,200 --> 00:33:04,040
Speaker 5: R and D, you get two hundred and twenty dollars

707
00:33:04,120 --> 00:33:07,400
Speaker 5: back for certain leading edge categories, like the subsidies here,

708
00:33:07,680 --> 00:33:09,240
Speaker 5: you know, they want you to work in it. They

709
00:33:09,240 --> 00:33:11,560
Speaker 5: want to give you as much money as possible to

710
00:33:12,000 --> 00:33:13,960
Speaker 5: you know, be moderately successful. You don't have to be

711
00:33:14,000 --> 00:33:15,840
Speaker 5: that great at it. Honestly, if you if you're just

712
00:33:15,880 --> 00:33:18,000
Speaker 5: taping out bad chips, you can probably make a living

713
00:33:18,040 --> 00:33:21,040
Speaker 5: doing this. Now that that probably creates some pretty messed

714
00:33:21,120 --> 00:33:26,240
Speaker 5: up IT incentives there, but the focus and the desire

715
00:33:26,360 --> 00:33:29,160
Speaker 5: to you know, to get to Chinese domestic by twenty

716
00:33:29,200 --> 00:33:32,000
Speaker 5: thirty or whatever, and that number, they're meaningfully below that.

717
00:33:32,120 --> 00:33:35,160
Speaker 5: But this is a long journey, and it's pretty clear

718
00:33:35,240 --> 00:33:37,320
Speaker 5: how they feel about it, right like this is this

719
00:33:37,400 --> 00:33:38,400
Speaker 5: is like splitting the interest.

720
00:33:38,560 --> 00:33:41,200
Speaker 4: One of the most interesting like observations of this sort

721
00:33:41,200 --> 00:33:44,400
Speaker 4: of like warped incentive structure is that the company that

722
00:33:44,440 --> 00:33:48,080
Speaker 4: makes railcars in China has made it their like national

723
00:33:48,200 --> 00:33:50,360
Speaker 4: goal to like because they make profit from that.

724
00:33:50,440 --> 00:33:54,360
Speaker 2: Right which company a cr RC Yeah, yeah.

725
00:33:54,240 --> 00:33:57,160
Speaker 4: Cr r C Yeah, So one interesting thing about this

726
00:33:57,240 --> 00:34:01,680
Speaker 4: warped incentive structure is that stayed on enterprises or very

727
00:34:01,720 --> 00:34:04,320
Speaker 4: successful companies that are doing really well and some some

728
00:34:04,480 --> 00:34:07,680
Speaker 4: market are expanding to places that make no sense like

729
00:34:07,840 --> 00:34:10,200
Speaker 4: you know, logically, right, because they have profits and they're like,

730
00:34:10,200 --> 00:34:13,000
Speaker 4: as Doug mentioned, right, you don't have taxes for a

731
00:34:13,000 --> 00:34:14,920
Speaker 4: certain amount of time. So hey, let me take all

732
00:34:14,960 --> 00:34:18,120
Speaker 4: the profits from this business and throw them into another sector.

733
00:34:18,160 --> 00:34:22,359
Speaker 4: And so CRRC is a national railway car company, and

734
00:34:22,560 --> 00:34:25,440
Speaker 4: they're plowing all of their money into making their national

735
00:34:25,440 --> 00:34:30,359
Speaker 4: goal to basically disrupt Infinian, which is Germany's largest chip maker, right,

736
00:34:30,360 --> 00:34:33,560
Speaker 4: who makes chips for power? Nothing sexy, right, but hey,

737
00:34:34,080 --> 00:34:36,480
Speaker 4: converting from one form of power to another is is

738
00:34:36,600 --> 00:34:39,080
Speaker 4: very important job for chips, and that's what that's what

739
00:34:39,120 --> 00:34:42,360
Speaker 4: Infinian's chips do, right, And this is what crr c's

740
00:34:42,719 --> 00:34:45,359
Speaker 4: sort of goal is to do, is to do that,

741
00:34:45,440 --> 00:34:48,399
Speaker 4: which is nothing at all related to railway cars. But hey,

742
00:34:48,440 --> 00:34:49,799
Speaker 4: I might as well do it because now if I

743
00:34:49,840 --> 00:34:52,720
Speaker 4: make any money on the semiconductor business, I don't get taxed.

744
00:34:52,800 --> 00:34:55,200
Speaker 4: And what I was getting taxed on my profitable, stable

745
00:34:55,200 --> 00:34:57,560
Speaker 4: business is now being funneled into this new business.

746
00:34:57,960 --> 00:35:01,319
Speaker 3: This is the irony of Shesh and Ping's crackdown on

747
00:35:01,480 --> 00:35:05,440
Speaker 3: disorderly capital in like e commerce and consumer tech, and

748
00:35:05,680 --> 00:35:09,120
Speaker 3: now there's just like money kind of flowing indiscriminately in

749
00:35:09,200 --> 00:35:10,040
Speaker 3: other sectors.

750
00:35:10,600 --> 00:35:10,799
Speaker 5: You know.

751
00:35:11,040 --> 00:35:13,120
Speaker 3: Joe mentioned in the intro that it's been a while

752
00:35:13,200 --> 00:35:15,640
Speaker 3: since we've done a semiconductor episode, and I think the

753
00:35:15,719 --> 00:35:18,400
Speaker 3: last one that we did was actually on in video.

754
00:35:18,520 --> 00:35:18,960
Speaker 3: Is that right?

755
00:35:19,160 --> 00:35:19,399
Speaker 2: Yeah?

756
00:35:19,400 --> 00:35:21,440
Speaker 3: I think I think that's right. And since then the

757
00:35:21,480 --> 00:35:26,160
Speaker 3: stock has absolutely exploded and there's been this frenzy over AI.

758
00:35:26,960 --> 00:35:30,400
Speaker 1: In video is just the absolute leader here. Obviously, China

759
00:35:30,440 --> 00:35:33,200
Speaker 1: wants to develop its own AI models. It has its

760
00:35:33,200 --> 00:35:35,840
Speaker 1: own AM models, its own large language models, its own

761
00:35:35,840 --> 00:35:39,840
Speaker 1: open AI competitor, et cetera. Where is it in terms

762
00:35:39,840 --> 00:35:43,880
Speaker 1: of its own access to the quality of chips that

763
00:35:43,920 --> 00:35:46,560
Speaker 1: it needs to make cutting edge AM models?

764
00:35:47,400 --> 00:35:50,600
Speaker 4: So that's probably the most effectively enforced part of the

765
00:35:50,680 --> 00:35:53,680
Speaker 4: China restrictions so far, right in twenty twenty two now

766
00:35:53,680 --> 00:35:57,239
Speaker 4: twenty twenty three, is is cutting back on Hey, you

767
00:35:57,280 --> 00:35:59,719
Speaker 4: can't get the chips that in video makes has been

768
00:35:59,719 --> 00:36:02,719
Speaker 4: the most successful part basically, and in Videos tried to

769
00:36:02,760 --> 00:36:05,760
Speaker 4: circumvent this by you know, releasing new China specific versions

770
00:36:06,080 --> 00:36:09,120
Speaker 4: and China's able to get some. But really this is

771
00:36:09,200 --> 00:36:13,000
Speaker 4: this is where China's ecosystem looks the most interesting for

772
00:36:13,360 --> 00:36:16,160
Speaker 4: breakthrough innovation, right, which is, hey, you know, you guys

773
00:36:16,200 --> 00:36:18,360
Speaker 4: are going in this route, which was awesome. You know,

774
00:36:18,400 --> 00:36:20,080
Speaker 4: we can't really go that route, but we have more

775
00:36:20,080 --> 00:36:22,120
Speaker 4: people working on this, so why don't we try and

776
00:36:22,160 --> 00:36:26,399
Speaker 4: do different things that will actually generate outcomes that are

777
00:36:26,800 --> 00:36:29,279
Speaker 4: in the same vein but not on the same path. Right.

778
00:36:29,360 --> 00:36:32,080
Speaker 4: So Huawei of course is making chips. They're making this

779
00:36:32,160 --> 00:36:34,360
Speaker 4: mobile chip on seven nanimeter, but they also have this

780
00:36:34,480 --> 00:36:37,560
Speaker 4: AI chip on seven nanimeter domestically made in videos. AI

781
00:36:37,640 --> 00:36:40,080
Speaker 4: chips are on seven nanimeters, only one generation behind non

782
00:36:40,080 --> 00:36:43,319
Speaker 4: processed technology. But more importantly, what Huawei and Smicker are

783
00:36:43,360 --> 00:36:46,600
Speaker 4: doing is that they're investing heavily in technologies that are

784
00:36:46,600 --> 00:36:49,399
Speaker 4: sort of a few generations out for everyone else because

785
00:36:49,400 --> 00:36:50,880
Speaker 4: they're sort of not necessary.

786
00:36:50,920 --> 00:36:51,080
Speaker 3: Right.

787
00:36:51,120 --> 00:36:55,400
Speaker 4: So things like bringing optical fibers directly to the AI chip, right,

788
00:36:55,440 --> 00:36:58,600
Speaker 4: so it's called copackage optics is the technology. Other things

789
00:36:58,640 --> 00:37:01,120
Speaker 4: are called like hybrid bonding, just like stacking chips on

790
00:37:01,160 --> 00:37:03,960
Speaker 4: top of each other. They're doing really interesting things there

791
00:37:04,239 --> 00:37:07,799
Speaker 4: to enable breakthrough innovation to enable performance that is on

792
00:37:08,000 --> 00:37:10,640
Speaker 4: par with the US chips like from in Video or

793
00:37:10,640 --> 00:37:13,080
Speaker 4: AMD right, and they aren't there yet, but they're going

794
00:37:13,160 --> 00:37:15,360
Speaker 4: to be there. And one interesting like sort of phenomenon

795
00:37:15,440 --> 00:37:17,520
Speaker 4: is that, like you know, because they can't order all

796
00:37:17,520 --> 00:37:19,520
Speaker 4: these Invidia chips, there's really only two places in the

797
00:37:19,520 --> 00:37:22,640
Speaker 4: world that will let you build data centers and have

798
00:37:22,760 --> 00:37:24,759
Speaker 4: cheap power. Right. East Asia can't do it because they

799
00:37:24,760 --> 00:37:27,600
Speaker 4: have to import all their power europates, you know, natural gas,

800
00:37:27,600 --> 00:37:29,520
Speaker 4: and natural gas is basically how you have to power

801
00:37:29,600 --> 00:37:31,680
Speaker 4: data centers. And so the US and China are really

802
00:37:31,680 --> 00:37:33,919
Speaker 4: the only place that can build data centers, and China

803
00:37:33,960 --> 00:37:36,160
Speaker 4: has been blocked off. So now China has like these

804
00:37:36,200 --> 00:37:39,399
Speaker 4: companies flooding into Malaysia and Indonesia trying to build data

805
00:37:39,400 --> 00:37:42,880
Speaker 4: centers that they can install foreign chips in. And meanwhile

806
00:37:42,880 --> 00:37:45,760
Speaker 4: they're also building these chips trying to do this breakthrough

807
00:37:45,760 --> 00:37:47,000
Speaker 4: innovation domestically.

808
00:37:47,320 --> 00:37:49,400
Speaker 5: Yeah, and I believe there is also a loophole that

809
00:37:49,480 --> 00:37:52,239
Speaker 5: might have been priorily closed talking about how you can

810
00:37:52,320 --> 00:37:54,880
Speaker 5: essentially rent a chip. So for example, a Chinese company

811
00:37:54,880 --> 00:37:58,399
Speaker 5: could rent something you know, in AWS, you can rent

812
00:37:58,440 --> 00:38:01,160
Speaker 5: a GPU that's in a cloud and say Signapore or

813
00:38:01,200 --> 00:38:03,520
Speaker 5: something like that. But I think that the conversation about

814
00:38:03,560 --> 00:38:05,799
Speaker 5: hybrid bonding and co package optics is really interesting and

815
00:38:05,880 --> 00:38:07,799
Speaker 5: kind of where we you know, at the beginning, we

816
00:38:07,800 --> 00:38:10,680
Speaker 5: were talking about how by cutting them off we would

817
00:38:10,680 --> 00:38:13,440
Speaker 5: effectively force them on their own roadmap, and that's kind

818
00:38:13,480 --> 00:38:16,920
Speaker 5: of what's already happened. The Smick roadmap is completely different

819
00:38:16,960 --> 00:38:19,640
Speaker 5: than what is sort of the industry standard roadmap for

820
00:38:19,719 --> 00:38:22,279
Speaker 5: leading edge, and as we continue to cut them off,

821
00:38:22,600 --> 00:38:25,640
Speaker 5: they're going to have to kind of create a new

822
00:38:25,840 --> 00:38:29,720
Speaker 5: roadmap for semicupters going forward, and the especially co package optics,

823
00:38:29,719 --> 00:38:32,680
Speaker 5: which is always five years ahead. And the reason why

824
00:38:32,719 --> 00:38:34,799
Speaker 5: it's never adopted in the United States or in the

825
00:38:34,840 --> 00:38:38,000
Speaker 5: western semicopter world is because it's just too expensive. But

826
00:38:38,120 --> 00:38:40,120
Speaker 5: if you don't have an option and this is your

827
00:38:40,160 --> 00:38:43,960
Speaker 5: only option for scaling out chips, they'll pursue the expensive option.

828
00:38:44,200 --> 00:38:46,840
Speaker 5: So there's going to be no how created because of

829
00:38:46,840 --> 00:38:48,160
Speaker 5: the necessity.

830
00:38:47,560 --> 00:38:49,040
Speaker 2: Of this Tracy.

831
00:38:49,160 --> 00:38:51,640
Speaker 1: I think when we released this episode, we should publish

832
00:38:51,640 --> 00:38:54,800
Speaker 1: a glossary too in all these terms, with like little

833
00:38:54,800 --> 00:38:56,600
Speaker 1: like definitions. I think that'd be very helpful.

834
00:38:56,680 --> 00:39:00,960
Speaker 3: Okay, I have fun doing that. Jo. We actually speaking

835
00:39:01,000 --> 00:39:03,360
Speaker 3: of new terms that will no doubt need to be

836
00:39:03,400 --> 00:39:07,200
Speaker 3: added to the glossary. Who makes the best chips for

837
00:39:07,560 --> 00:39:10,279
Speaker 3: q Star that's been in the news. We don't know

838
00:39:10,320 --> 00:39:11,720
Speaker 3: what it is yet, but I'm going to go ahead

839
00:39:11,719 --> 00:39:12,080
Speaker 3: and ask.

840
00:39:12,280 --> 00:39:14,239
Speaker 2: Trying to kind of explain it to me last night.

841
00:39:14,280 --> 00:39:17,640
Speaker 5: Actually, I don't even know what. My favorite take on this,

842
00:39:17,719 --> 00:39:19,520
Speaker 5: and this is Dylan's take, is that q Star is

843
00:39:19,520 --> 00:39:22,879
Speaker 5: a sieop. It's oh, yeah, So I'm in the bay

844
00:39:22,920 --> 00:39:23,120
Speaker 5: a lot.

845
00:39:23,120 --> 00:39:25,839
Speaker 4: A lot of my friends are researchers, and we sort

846
00:39:25,880 --> 00:39:29,439
Speaker 4: of have this idea that anytime Opening Eye tweets about

847
00:39:29,480 --> 00:39:32,520
Speaker 4: something that's related to research, they're actually just trying to

848
00:39:32,640 --> 00:39:36,400
Speaker 4: siop everyone into wasting their time on a path of

849
00:39:36,520 --> 00:39:39,799
Speaker 4: mL research that is actually not like, yeah, not going

850
00:39:39,880 --> 00:39:43,520
Speaker 4: to result in a better model because there's limited amounts

851
00:39:43,520 --> 00:39:45,400
Speaker 4: of GPU, limited amounts of time. So if you're wasting

852
00:39:45,400 --> 00:39:48,080
Speaker 4: your time doing that instead of actually working on, you know,

853
00:39:48,080 --> 00:39:50,160
Speaker 4: what's the correct path at least what they think. Because

854
00:39:50,160 --> 00:39:52,040
Speaker 4: they've done all the experiments as sort of ahead of

855
00:39:52,080 --> 00:39:55,800
Speaker 4: everyone else, it's easier to innovate what's already been innovated

856
00:39:56,080 --> 00:39:58,279
Speaker 4: than it is to innovate something completely new, and so

857
00:39:58,400 --> 00:40:00,800
Speaker 4: sort of the thought is, hey, they're doing these syops

858
00:40:00,840 --> 00:40:02,239
Speaker 4: and Q Star is one of those.

859
00:40:03,080 --> 00:40:03,600
Speaker 5: We'll see.

860
00:40:03,760 --> 00:40:06,080
Speaker 2: There's a million more questions we could probably ask me.

861
00:40:06,160 --> 00:40:07,879
Speaker 2: We gotta wrap up, so we'll have to have you back.

862
00:40:08,160 --> 00:40:11,000
Speaker 1: Doga Laughlin and Dylan Patel, thank you so much for

863
00:40:11,040 --> 00:40:11,680
Speaker 1: coming on odd Law.

864
00:40:11,719 --> 00:40:12,920
Speaker 2: It's amazing conversation.

865
00:40:13,080 --> 00:40:15,400
Speaker 5: Thanks for having us. Always love to chat with you guys,

866
00:40:15,440 --> 00:40:17,319
Speaker 5: and always loves the chat chips for sure.

867
00:40:17,440 --> 00:40:19,520
Speaker 4: Yeah, yeah, this was a blast. I didn't realize you

868
00:40:19,520 --> 00:40:21,680
Speaker 4: two were even more funny when you're sitting in the same.

869
00:40:21,520 --> 00:40:25,720
Speaker 3: Room when we're pointing at each other and making faces. Yeah,

870
00:40:25,760 --> 00:40:26,600
Speaker 3: all right, thank you.

871
00:40:26,640 --> 00:40:28,320
Speaker 2: That was so much fun. That was amazing.

872
00:40:28,360 --> 00:40:42,600
Speaker 1: Thank you so much, Tracy. I thought that was an

873
00:40:42,640 --> 00:40:45,920
Speaker 1: amazing conversation. I guess it freaked me out a little bit.

874
00:40:46,239 --> 00:40:47,239
Speaker 5: Yeah, you know.

875
00:40:47,480 --> 00:40:50,160
Speaker 3: You know what's bothering me enormously now is I've realized

876
00:40:50,200 --> 00:40:53,440
Speaker 3: if you step away from semiconductors for like a month,

877
00:40:54,120 --> 00:40:57,480
Speaker 3: there's some new breakthrough. Yeah, and there's like five new

878
00:40:57,520 --> 00:41:02,640
Speaker 3: technologies and probably like but there are more companies with

879
00:41:02,760 --> 00:41:05,880
Speaker 3: like four letter acronyms. Yeah involved somehow is that like

880
00:41:05,880 --> 00:41:06,279
Speaker 3: a rule?

881
00:41:06,480 --> 00:41:09,120
Speaker 1: I remember having the same thought about stepping away, Like

882
00:41:09,400 --> 00:41:12,160
Speaker 1: every time I would ignore crypto for six months, there

883
00:41:12,160 --> 00:41:14,560
Speaker 1: would be some new scheme, except in the end, none

884
00:41:14,600 --> 00:41:16,719
Speaker 1: of that mattered because it was all it didn't matter

885
00:41:16,719 --> 00:41:18,880
Speaker 1: any This stuff actually matters.

886
00:41:18,880 --> 00:41:19,600
Speaker 2: And so you do.

887
00:41:19,680 --> 00:41:21,960
Speaker 1: It does feel like you really have to pay attention

888
00:41:22,400 --> 00:41:24,719
Speaker 1: constantly to know what's going on.

889
00:41:24,800 --> 00:41:25,960
Speaker 2: There was so much in there, you know what.

890
00:41:26,080 --> 00:41:28,600
Speaker 1: One thing just that struck me when I was thinking

891
00:41:28,600 --> 00:41:32,759
Speaker 1: about US first China industrial policy and here we freak

892
00:41:32,840 --> 00:41:34,960
Speaker 1: out politically do it to a waste. So we're still

893
00:41:35,000 --> 00:41:37,919
Speaker 1: talking about say Cylindra, so one company that was doing

894
00:41:37,960 --> 00:41:40,680
Speaker 1: tech and it ended up being scandal that for years

895
00:41:40,800 --> 00:41:44,359
Speaker 1: cast dispersions on the idea of the government investing. And

896
00:41:44,440 --> 00:41:47,759
Speaker 1: it sounds like listening to Dylan and Doug that yeah,

897
00:41:47,800 --> 00:41:52,360
Speaker 1: they probably have their equivalent of plenty of Cylindras in China.

898
00:41:52,440 --> 00:41:55,520
Speaker 1: There's probably plenty of fraud, plenty of ways, plenty of disasters,

899
00:41:55,640 --> 00:41:57,920
Speaker 1: plenty of endeavors that don't have the chance. It just

900
00:41:57,960 --> 00:42:00,719
Speaker 1: seems that they have the appetite to continue you investing

901
00:42:00,880 --> 00:42:03,840
Speaker 1: in those areas, even with some level of fraud.

902
00:42:03,920 --> 00:42:07,800
Speaker 3: But also like they are so explicit about what they're doing,

903
00:42:07,840 --> 00:42:09,719
Speaker 3: and I can't remember I'm gonna have to go back

904
00:42:09,760 --> 00:42:12,120
Speaker 3: and look who said this, but someone was basically like

905
00:42:12,680 --> 00:42:15,560
Speaker 3: the equivalent of follow the money, like China is telling

906
00:42:15,640 --> 00:42:19,040
Speaker 3: you where they're going to put the money in semiconductors.

907
00:42:19,080 --> 00:42:21,440
Speaker 3: This would have been around twenty twenty one, when they

908
00:42:21,480 --> 00:42:24,759
Speaker 3: first started cracking down on like for profit education and

909
00:42:24,880 --> 00:42:27,880
Speaker 3: the e commerce platforms and things like that. And this

910
00:42:27,960 --> 00:42:30,839
Speaker 3: person I remember them saying, like, China's telling you what

911
00:42:31,000 --> 00:42:34,879
Speaker 3: it wants you to do. As a domestic investor or

912
00:42:34,920 --> 00:42:37,880
Speaker 3: someone who's working in tech. They don't want you building,

913
00:42:38,520 --> 00:42:40,799
Speaker 3: you know, video games or something like that. They want

914
00:42:40,840 --> 00:42:43,680
Speaker 3: you to make chips that might power those video games

915
00:42:43,719 --> 00:42:48,280
Speaker 3: or something much more important. And here it's like, yes,

916
00:42:48,680 --> 00:42:52,319
Speaker 3: we have messaging, but it's never it's never quite that

917
00:42:52,400 --> 00:42:55,160
Speaker 3: way for obvious reasons. I think there's like still a

918
00:42:55,160 --> 00:42:58,360
Speaker 3: lot of discomfort here about industrial policy in general. But

919
00:42:58,719 --> 00:43:01,239
Speaker 3: of course since COVID that seems to be changing a

920
00:43:01,239 --> 00:43:01,680
Speaker 3: little bit.

921
00:43:02,400 --> 00:43:04,880
Speaker 1: So many other details I want to we need to

922
00:43:04,960 --> 00:43:07,440
Speaker 1: find one of those shows the love story of the

923
00:43:07,480 --> 00:43:11,000
Speaker 1: two people who meet in this semiconductor fab. I also

924
00:43:11,040 --> 00:43:14,080
Speaker 1: think it's just really interesting this idea of a railcar company,

925
00:43:14,440 --> 00:43:17,920
Speaker 1: yeah building, say hey, here's an opportunity to build a

926
00:43:18,040 --> 00:43:20,320
Speaker 1: chip that we have to buy from a German company,

927
00:43:20,360 --> 00:43:22,640
Speaker 1: and because of the tax structure, it makes sense for

928
00:43:22,800 --> 00:43:25,319
Speaker 1: us to invest and see if we can build it domestically.

929
00:43:25,760 --> 00:43:28,480
Speaker 1: This idea that you know we talked about bid before

930
00:43:28,520 --> 00:43:30,400
Speaker 1: on the show a couple months ago with Corey Kanter,

931
00:43:30,719 --> 00:43:33,160
Speaker 1: but that it's also a chip powerhouse, and that there's

932
00:43:33,160 --> 00:43:35,160
Speaker 1: a lot of a lot of mergin. There so many,

933
00:43:35,320 --> 00:43:37,319
Speaker 1: there's too many interesting things to go over.

934
00:43:37,520 --> 00:43:40,719
Speaker 3: Yeah, I gotta say. I went on one of crrcs,

935
00:43:41,200 --> 00:43:43,040
Speaker 3: the high speed trains. I think it must have been

936
00:43:43,040 --> 00:43:47,640
Speaker 3: done them between Beijing and Hong Kong. That's a nice train.

937
00:43:48,000 --> 00:43:51,279
Speaker 3: I really enjoyed that those so I look forward to

938
00:43:51,280 --> 00:43:54,280
Speaker 3: seeing the semiconductors too. I'm sure they'll also be nice.

939
00:43:54,560 --> 00:43:56,879
Speaker 3: Maybe shall we leave it there, Let's leave it there.

940
00:43:57,040 --> 00:43:57,400
Speaker 1: Okay.

941
00:43:57,560 --> 00:44:00,200
Speaker 3: This has been another episode of the Odd Thoughts podcast.

942
00:44:00,239 --> 00:44:03,480
Speaker 3: I'm Tracy Alloway. You can follow me at Tracy Alloway and.

943
00:44:03,440 --> 00:44:06,400
Speaker 1: I'm Joe Wisenthal. You can follow me at The Stalwart.

944
00:44:06,680 --> 00:44:08,960
Speaker 1: Follow Dylan Patel He's at Dylan five to two to

945
00:44:08,960 --> 00:44:13,880
Speaker 1: two p. Follow Doug o Laughlin at Underscore fab Knowledge Underscore.

946
00:44:14,160 --> 00:44:17,719
Speaker 1: Follow our producers Carmen Rodriguez at Carman Arman, dash Ol

947
00:44:17,760 --> 00:44:21,520
Speaker 1: Bennett at Dashbot and kel Brooks at kel Brooks. And

948
00:44:21,560 --> 00:44:24,319
Speaker 1: thank you to our producer Moses Ondam. For more odd

949
00:44:24,360 --> 00:44:27,520
Speaker 1: Lots content, go to Bloomberg dot com slash odd lots,

950
00:44:27,520 --> 00:44:30,879
Speaker 1: where we have transcripts, a blog, and a newsletter and

951
00:44:31,360 --> 00:44:33,880
Speaker 1: you can chat twenty four to seven with fellow listeners

952
00:44:33,920 --> 00:44:37,640
Speaker 1: in the discord discord dot gg slash odd lots. There's

953
00:44:37,680 --> 00:44:40,040
Speaker 1: even a room there just for semiconductors where people are

954
00:44:40,040 --> 00:44:43,200
Speaker 1: posting links to highly technical things that I don't understand

955
00:44:43,239 --> 00:44:44,520
Speaker 1: but seem very fascinating.

956
00:44:44,760 --> 00:44:45,520
Speaker 2: Go check it out.

957
00:44:45,640 --> 00:44:48,520
Speaker 3: And if you enjoy odd Lots, if you want us

958
00:44:48,520 --> 00:44:51,440
Speaker 3: to get back into semiconductors on a regular basis, then

959
00:44:51,480 --> 00:44:55,400
Speaker 3: please leave us a positive review on your favorite podcast platform.

960
00:44:55,520 --> 00:45:04,720
Speaker 3: Thanks for listening

961
00:45:13,520 --> 00:45:13,560
Speaker 1: It