WEBVTT - Lots More on the Ongoing Mess That Is Intel 0:00:03.120 --> 0:00:08.520 Bloomberg Audio Studios, podcasts, radio news. 0:00:10.119 --> 0:00:12.920 I am worried about Intel, No, for real. 0:00:13.200 --> 0:00:15.920 We're recording this on October first, So I'm worried about 0:00:15.920 --> 0:00:18.360 some other things. But sure, sure, no. 0:00:18.440 --> 0:00:20.680 There's plenty to worry about in the world. But like, 0:00:20.720 --> 0:00:21.640 I'm worried about. 0:00:21.440 --> 0:00:26.040 Intel specifically, because it's all nice to talk about industrial 0:00:26.079 --> 0:00:29.960 policy and we're going to build these you know, build 0:00:29.960 --> 0:00:32.120 out all these industries, et cetera. But here we have 0:00:32.200 --> 0:00:35.800 one of the true industrial giants, one of the flagship 0:00:35.840 --> 0:00:39.479 companies of the United States getting a lot of support, 0:00:39.600 --> 0:00:42.760 are leading company in the field of semiconductors, which is 0:00:42.800 --> 0:00:45.919 a major strategic priority for the US, and they do 0:00:46.040 --> 0:00:47.840 not seem to be doing. 0:00:47.560 --> 0:00:48.520 Well at all. 0:00:48.680 --> 0:00:51.639 The stock is absolutely dismal. They're cutting jobs. 0:00:51.960 --> 0:00:55.160 If this is the flagship company like them and Boeing 0:00:55.560 --> 0:00:57.880 loomed very large in my head about I guess what 0:00:57.920 --> 0:01:00.000 I would say is like, I don't know if corporate 0:01:00.200 --> 0:01:02.240 rot is too strong of a word, but yeah, like 0:01:02.280 --> 0:01:02.920 corporate rot. 0:01:03.320 --> 0:01:06.199 Not to add to your anxiety, but did you also 0:01:06.240 --> 0:01:08.600 see some of the headlines coming out of North Carolina 0:01:08.720 --> 0:01:10.320 about quartz sand. 0:01:10.360 --> 0:01:11.600 Yeah, I don't know what that's all about. 0:01:11.680 --> 0:01:15.039 It's kind of an interesting one. So after Hurricane Helene, 0:01:15.240 --> 0:01:19.160 there's a town called Spruce Pine where apparently most of 0:01:19.200 --> 0:01:23.480 the world's supply of pure quartz sand comes from, and 0:01:23.520 --> 0:01:26.240 that could have an impact on the production of semiconductors. 0:01:26.280 --> 0:01:29.000 So we get to I guess we get to unveil 0:01:29.040 --> 0:01:31.480 the bullhorn for the bullwhip effect. Again. 0:01:32.440 --> 0:01:36.399 We're not doing this. Apparently we are doing this again. 0:01:39.400 --> 0:01:40.720 I did a deadlift. 0:01:40.360 --> 0:01:46.440 One two, Jimmy okay, uh barges. 0:01:46.600 --> 0:01:48.760 This isn't after school Special, except. 0:01:48.360 --> 0:01:50.920 I've decided I'm going to base my entire personality going 0:01:50.920 --> 0:01:54.160 forward on campaigning for a strategic pork reserve in the US. 0:01:54.280 --> 0:01:56.000 Where's the best costa? 0:01:56.160 --> 0:01:58.760 These are the important question? Is it robots taking over 0:01:58.760 --> 0:01:59.120 the world? 0:01:59.200 --> 0:02:02.040 No, I think that like in a couple of years, 0:02:02.240 --> 0:02:04.520 the AI will do a really good job of making 0:02:04.560 --> 0:02:07.840 the odd Launch podcast And people say, I don't really 0:02:07.920 --> 0:02:09.760 need to listen to Joe and Tracy anymore. 0:02:09.960 --> 0:02:15.440 We do have touching, the perfect welcome to lots More 0:02:15.480 --> 0:02:17.560 where we catch up with friends about what's going on 0:02:17.840 --> 0:02:18.560 right now. 0:02:18.639 --> 0:02:21.919 Because even when Odd Lots is over, there's always lots 0:02:21.960 --> 0:02:23.040 more And. 0:02:23.000 --> 0:02:28.880 We really do have the perfect guest stacy, what's going 0:02:28.880 --> 0:02:30.279 wrong until. 0:02:31.160 --> 0:02:33.640 What's not going wrong? At Intel? Right now? 0:02:34.280 --> 0:02:37.679 They're in a tough spot, right, I mean they are 0:02:39.040 --> 0:02:44.000 embarking on a very risky and capital intensive strategy. I 0:02:44.040 --> 0:02:46.760 say embarking, but frankly they've been along on the strategy 0:02:46.800 --> 0:02:47.480 for quite a few years. 0:02:47.520 --> 0:02:51.399 Yeah, because CEO Pat Gelsinger came in in twenty twenty one, 0:02:51.560 --> 0:02:53.880 and so we're like at least three years into what 0:02:54.040 --> 0:02:58.240 is supposedly an embarkment on a new strategy. And if 0:02:58.280 --> 0:02:59.920 I just look at the stock price, it's lower than 0:03:00.120 --> 0:03:00.960 was in twenty fifteen. 0:03:01.800 --> 0:03:04.240 Yeah, in fact, you know it's I actually look, I 0:03:04.280 --> 0:03:07.440 launched on the stock in June of two thousand and nine. 0:03:07.480 --> 0:03:09.720 I think it was fifteen dollars and ninety four cents, 0:03:09.760 --> 0:03:12.040 so fifteen or sixteen years ago, and it's. 0:03:11.880 --> 0:03:13.240 Not that far off of there. Now. 0:03:13.840 --> 0:03:16.280 It's had a bit of a balance in the recent weeks. 0:03:16.280 --> 0:03:18.360 We can talk about why, but in general they're in 0:03:18.360 --> 0:03:20.560 a tough spot. Yeah, I mean the biggest issue that 0:03:20.560 --> 0:03:22.919 they're having, and we can argue, is the. 0:03:22.880 --> 0:03:25.200 Strategy the right one or the wrong one. That's a 0:03:25.200 --> 0:03:25.760 discussion we. 0:03:25.760 --> 0:03:29.120 Can have, but it might be the biggest issue is 0:03:29.160 --> 0:03:32.640 just the core business that was supposed to support them 0:03:33.360 --> 0:03:35.600 as they rolled that strategy out is no longer really 0:03:35.600 --> 0:03:37.120 sufficient to support it anymore. 0:03:37.120 --> 0:03:40.680 But they maybe too far along their path to stop. Right. 0:03:41.120 --> 0:03:44.160 The whole idea was, you know, we're gonna be doubling 0:03:44.200 --> 0:03:45.480 down on manufacturing. 0:03:45.800 --> 0:03:47.760 We're going to building lots of factories in the US 0:03:47.800 --> 0:03:48.560 and everywhere else. 0:03:49.080 --> 0:03:52.360 We're going to be building up a significant third party 0:03:52.360 --> 0:03:55.840 foundry business to make parts not just for ourselves but 0:03:55.960 --> 0:03:59.200 for other third party customers as well. We're getting a 0:03:59.200 --> 0:04:02.440 lot of external funding to do this, and you know, 0:04:02.600 --> 0:04:04.200 eventually we'll get there and it's going to be great. 0:04:04.240 --> 0:04:06.560 But that the problem is that the you know, these 0:04:06.560 --> 0:04:08.800 things are very you know, intensive, like they cost a 0:04:08.800 --> 0:04:11.640 lot of money, and because of the deterioration that we've 0:04:11.640 --> 0:04:14.720 seen in Intel's core business for a variety of reasons, 0:04:14.880 --> 0:04:17.000 the macro's not been great. You know, they were over 0:04:17.080 --> 0:04:19.960 earning as many were during COVID and and that bubble 0:04:20.040 --> 0:04:22.600 is popped. They're still losing share in many of their 0:04:22.680 --> 0:04:26.440 key markets. They've completely missed the AI revolution, which is 0:04:26.480 --> 0:04:28.760 one of the few areas of semiconductors today that's actually 0:04:28.800 --> 0:04:31.880 really on fire. And they effectively have nothing to show 0:04:32.160 --> 0:04:34.440 in that as and and those markets are starting to 0:04:34.440 --> 0:04:36.280 eat into some of the core stuff that Intel does. 0:04:36.839 --> 0:04:39.599 That core business doesn't doesn't really support that the path 0:04:39.600 --> 0:04:43.000 that they're on anymore, even after the subsidies and and 0:04:43.320 --> 0:04:46.279 frankly doing some other things that are probably not you 0:04:46.279 --> 0:04:47.680 wouldn't want to do if you didn't have to. They've 0:04:47.680 --> 0:04:49.880 been selling parts of some of their factories to private 0:04:49.920 --> 0:04:51.920 equity and getting cash for that. But even after all 0:04:51.960 --> 0:04:54.159 of that, you know, it's still costing them like a 0:04:54.160 --> 0:04:56.440 bunch of money that they don't really have anymore. They 0:04:56.480 --> 0:04:59.599 put out some targets recently that sort of showed a 0:04:59.680 --> 0:05:02.840 path to you know, a better companies, but in twenty thirty, 0:05:03.360 --> 0:05:04.360 it's twenty twenty four. 0:05:04.440 --> 0:05:06.240 Right now, that's a long ways away. 0:05:06.240 --> 0:05:08.600 There's a lot of wood to chop between now and then, 0:05:08.680 --> 0:05:10.440 even if they can manage to execute on that. 0:05:10.480 --> 0:05:11.719 So they're in a tough spot. 0:05:12.120 --> 0:05:14.680 So we are speaking with Stacey Raskin, who is of 0:05:14.720 --> 0:05:18.640 course the semiconductor analyst over at Bernstein. We've spoken to 0:05:18.680 --> 0:05:22.240 him before. We also have McKenzie Hawkins, who is our 0:05:22.320 --> 0:05:27.920 colleague here at Bloomberg. McKenzie, how existential are the next Oh, 0:05:27.960 --> 0:05:31.520 I guess, let's see October, November, December, next three months. 0:05:31.640 --> 0:05:32.479 For Intel. 0:05:33.240 --> 0:05:36.520 This is a crucial stretch for Intel, in part because 0:05:36.520 --> 0:05:39.640 it's a crucial stretch for the Biden administration. The company 0:05:39.920 --> 0:05:44.560 is the single largest beneficiary of this massive US effort 0:05:44.600 --> 0:05:48.640 to resource semiconductor manufacturing through Call twenty twenty two Chips 0:05:48.640 --> 0:05:50.960 in Science Act. They're in line to receive eight and 0:05:51.040 --> 0:05:54.279 a half billion dollars in grants for projects across four 0:05:54.360 --> 0:05:57.240 US states for commercial partners. They're also supposed to get 0:05:57.240 --> 0:05:59.760 three billion dollars to make chips for the Pentagon and 0:06:00.000 --> 0:06:02.640 eleven billion dollars in loans, not to mention twenty five 0:06:02.680 --> 0:06:05.920 percent tax breaks on their US investments. There isn't really 0:06:05.920 --> 0:06:08.839 a fear that the broader SEPSTY program will go away 0:06:09.000 --> 0:06:13.720 after President Biden leaves office. But Intel's been negotiating this 0:06:13.839 --> 0:06:17.880 award package with officials in the Biden administration for well 0:06:17.920 --> 0:06:21.440 over a year now. It's not yet finalized, and they 0:06:21.600 --> 0:06:24.640 really want to prove two officials that yes, we need 0:06:24.680 --> 0:06:26.719 to sign these documents. We want to get money out 0:06:26.720 --> 0:06:29.559 the door, ideally by the end of this administration, because 0:06:29.600 --> 0:06:33.280 who really knows what will come after January, and they're 0:06:33.279 --> 0:06:36.080 all trying to do that before they really really prove 0:06:36.120 --> 0:06:39.120 in the technology stack on which these awards are premised. 0:06:39.839 --> 0:06:43.440 Stacy two questions in your when you think of Intel, 0:06:44.000 --> 0:06:47.719 how important is the aid from the US government to 0:06:47.839 --> 0:06:50.839 its future? And when you think of the US government, 0:06:51.200 --> 0:06:54.800 how important is Intel to the policy objectives of the 0:06:54.880 --> 0:06:55.760 United States. 0:06:56.480 --> 0:06:59.240 They're clearly both important. The aid from the government is 0:06:59.640 --> 0:07:02.479 helping them. And by the way, I will say, I 0:07:02.520 --> 0:07:05.440 don't think they're in a critical situation from a cash 0:07:05.480 --> 0:07:07.840 standpoint right now, partially because of some. 0:07:07.800 --> 0:07:08.799 Of those government funds. 0:07:08.800 --> 0:07:12.000 So if you add up what they've just done, some 0:07:12.040 --> 0:07:14.520 of the were structuring that they've just announced, the CAPEX cuts, 0:07:15.120 --> 0:07:18.200 the op X cuts, the suspended dividend, they suspended their 0:07:18.200 --> 0:07:21.480 dividend right, the governmental funds both from the tax credit 0:07:21.480 --> 0:07:23.200 as well as the grants of those rolling as well 0:07:23.200 --> 0:07:26.120 as some of those private equity funds. It's something like 0:07:26.240 --> 0:07:28.920 forty billion dollars of incremental cash that lends on the 0:07:28.960 --> 0:07:32.440 balance sheet like through the end of twenty five of 0:07:32.440 --> 0:07:34.560 twenty twenty five, versus not doing anything at all. So 0:07:34.640 --> 0:07:38.080 like again, I think their cash position is okay because 0:07:38.080 --> 0:07:40.280 of this stuff. I don't think they're desperate. It's partially 0:07:40.280 --> 0:07:42.960 be due to those government funds. I will say, however, 0:07:43.040 --> 0:07:44.720 to get those grants, it's not like they don't get 0:07:44.720 --> 0:07:46.640 an eight and a half billion dollar check like they're 0:07:46.680 --> 0:07:49.240 milestone base. They have to do the projects and build 0:07:49.280 --> 0:07:51.119 them out and hit the milestones and then that money 0:07:51.120 --> 0:07:53.720 get gets apportioned. But that money is important in terms 0:07:53.760 --> 0:07:55.400 of like how important is Intel for the for the 0:07:55.440 --> 0:07:58.800 government's efforts. I mean, it's clearly very important. Although I 0:07:58.880 --> 0:08:01.200 will say, you know, if if the goal is really 0:08:01.520 --> 0:08:04.520 to get a ton of leading edge manufacturing capacity built 0:08:04.840 --> 0:08:09.400 in the US, you could argue if the government was 0:08:09.880 --> 0:08:15.400 only goal oriented, maybe picking Intel as their champion was 0:08:15.920 --> 0:08:18.040 not the best way to go. There are other players 0:08:18.560 --> 0:08:20.840 that you could have encouraged, who whether they're all building here, 0:08:20.840 --> 0:08:22.600 but you could have encouraged, for example, somebody like a 0:08:22.640 --> 0:08:24.679 TSMC to build a lot more here than they're doing. 0:08:25.080 --> 0:08:27.320 Now that being set, we're not just goal already. 0:08:27.400 --> 0:08:29.680 Other like politics is a huge piece. It's not politically 0:08:29.760 --> 0:08:33.000 viable to do something like that. Intel is the only 0:08:33.760 --> 0:08:37.000 US based company that can even hope to build out 0:08:37.000 --> 0:08:39.000 a leading edge footprint, and so I understand why they've 0:08:39.000 --> 0:08:41.160 been named the national champion, but I mean, it's just 0:08:41.200 --> 0:08:43.679 given the issues that they're having, it does sort of 0:08:43.720 --> 0:08:48.440 throw the eventual success of the program, especially especially around 0:08:48.440 --> 0:08:50.440 the leading edge side, into a little bit of a loop, 0:08:50.520 --> 0:08:52.160 like we'll have to see how things go for them. 0:08:52.640 --> 0:08:54.880 And you think it's important to remember here. Intel is 0:08:55.040 --> 0:08:57.560 the largest beneficiary of the Chips Act, but definitely not 0:08:57.600 --> 0:08:59.319 the only one. The US is in line to give 0:08:59.320 --> 0:09:02.960 more than six billion dollars to TSMC, the unambiguous industry leader. 0:09:03.000 --> 0:09:05.360 There's also more than six billion dollars lated for Samsung, 0:09:05.360 --> 0:09:08.720 which is building leading edge capacity in Texas. But Intel 0:09:08.960 --> 0:09:13.000 is really important one because the Pentagon has decided that 0:09:13.120 --> 0:09:16.319 we need an American company to make cutting edge chips 0:09:16.360 --> 0:09:20.439 for military and intelligence purposes. They've had early conversations about 0:09:20.480 --> 0:09:23.040 buying those chips from foreign owned foundries that are on 0:09:23.280 --> 0:09:26.960 US soil. It's crucial that we're having TSMC building three 0:09:26.960 --> 0:09:28.880 fabs in Arizona, and that was a commitment that the 0:09:28.960 --> 0:09:31.679 US government got from them with the carot of these subsidies. 0:09:32.280 --> 0:09:35.199 But Intel, I mean, Gelsinger kind of hinged his entire 0:09:35.280 --> 0:09:38.960 turnaround plan on this US manufacturing expansion was a major 0:09:39.000 --> 0:09:41.920 lobbyist pushing for the Ships Act, riding the coattails of 0:09:41.920 --> 0:09:45.280 that subsidy program. And yes, you know this says really 0:09:45.360 --> 0:09:48.320 high political stakes for the Biden administration and for industrial 0:09:48.320 --> 0:09:51.120 policy writ large. You know, I think it's a feature, 0:09:51.200 --> 0:09:53.760 not a bug of the program that Intel hasn't gotten 0:09:53.800 --> 0:09:57.640 any money yet. As Stacy mentioned, these are milestone based awards. 0:09:57.679 --> 0:10:00.160 You know, companies will get a tranch of money and 0:10:00.200 --> 0:10:02.400 they have you know, first wafers out of their facilities. 0:10:02.440 --> 0:10:04.520 Look at a trants of money when facilities are completed. 0:10:04.960 --> 0:10:08.000 But for a company that is, you know, not acutely 0:10:08.080 --> 0:10:10.120 running out of cash, but not in a great financial 0:10:10.160 --> 0:10:13.320 position like Intel, it really matters when that money actually flows. 0:10:13.679 --> 0:10:16.320 And the US government is still vetting the viability of 0:10:16.360 --> 0:10:19.160 these manufacturing plans in the first place before they're willing 0:10:19.200 --> 0:10:21.400 to sign a final agreement. All they've announced is a 0:10:21.440 --> 0:10:26.640 preliminary award. 0:10:36.040 --> 0:10:38.800 Is there a viable alternative to Intel? I get the 0:10:38.840 --> 0:10:40.680 sense that maybe there isn't. 0:10:41.400 --> 0:10:44.920 I mean, there are three companies in theory that can 0:10:44.920 --> 0:10:49.400 do leading edge semiconductor manufacturing global It's TSMC, Samsung, and Intel. 0:10:50.240 --> 0:10:53.440 And of those three, only really TSMC is hitting out 0:10:53.480 --> 0:10:56.960 of the park in terms of executing to their roadmaps. 0:10:57.000 --> 0:10:58.800 Like Intel, cole is having problems and again we can 0:10:58.880 --> 0:11:01.120 talk about their roadmap and where that may be going, 0:11:01.120 --> 0:11:03.960 but they've been having issues and Samsung as well as 0:11:03.960 --> 0:11:07.319 also having issues, Like this stuff is not easy, Like yeah, I. 0:11:07.320 --> 0:11:11.960 Always Samsung by the way, by the way, by the way, 0:11:11.960 --> 0:11:15.400 on the day we're recording this October first, Bloomberg reporting 0:11:15.520 --> 0:11:18.760 some substantial layoffs at Samsung and they're losing some shared 0:11:18.800 --> 0:11:22.800 to competitors like Skhnix, et cetera in certain lines. So yeah, 0:11:22.840 --> 0:11:26.040 just to talk about it's not just Intel having the struggles. Okay, 0:11:26.080 --> 0:11:28.360 keep going, Yeah. 0:11:27.679 --> 0:11:29.600 Yeah, so it's the stuff is hard to do. Again, 0:11:29.720 --> 0:11:31.440 I'm always amazed that any of it works at all. 0:11:32.000 --> 0:11:34.360 But to McKenzie's probably one of the three. 0:11:34.720 --> 0:11:37.440 I mean, the administration has said they want an American 0:11:37.480 --> 0:11:40.600 company to do this, and so I guess I'm wondering, like, 0:11:40.840 --> 0:11:44.240 are there ways that they could structure something with TSMC 0:11:44.440 --> 0:11:47.360 or Samsung or does it have to be an American company, 0:11:47.360 --> 0:11:49.640 in which case it seems like Intel is the only option. 0:11:50.559 --> 0:11:52.560 Yeah, if if you want an American company to do 0:11:52.600 --> 0:11:54.360 this stuff, they are the only option. It may be 0:11:54.400 --> 0:11:57.560 helpful to talk about their roadmap. So Patt has talked 0:11:57.559 --> 0:11:59.920 a lot about what he calls five nodes in four years, 0:12:00.120 --> 0:12:02.520 and according to Intel, they're executing, although if you kind 0:12:02.559 --> 0:12:05.760 of look at it, it's kind of kind of questionable. 0:12:05.760 --> 0:12:08.480 I mean, they're delivering what they're calling seven animeter today 0:12:08.480 --> 0:12:11.120 in volume. That's the old what they used to call ten. 0:12:11.640 --> 0:12:14.000 They are now starting to ship Intel four animeter. It's 0:12:14.000 --> 0:12:16.160 a single product that's called Meteor Lake, and in fact 0:12:16.360 --> 0:12:20.079 is it's a client PC product. It has a tiled architecture. 0:12:20.120 --> 0:12:22.240 There was one tile on there that's on Intel four 0:12:22.600 --> 0:12:24.640 and the other three tiles are made it actually outsourced. 0:12:24.640 --> 0:12:25.600 They made it TSMC, and. 0:12:25.520 --> 0:12:27.880 The margins of that product are not very good and 0:12:27.920 --> 0:12:30.640 it's actually it impacted their margins in the most recent quarter. 0:12:30.880 --> 0:12:33.199 They're getting ready to now to ship another client product 0:12:33.200 --> 0:12:35.480 which is called Lunar Lake, which actually from some of 0:12:35.480 --> 0:12:37.520 the ben shorts looks like a decent product, but it's 0:12:37.559 --> 0:12:40.680 one hundred percent outsource. There's no Intel technology and process 0:12:40.720 --> 0:12:43.400 technology and it all it's all made at TSMC, and 0:12:43.640 --> 0:12:46.280 they've talked about that further impacting their margins as we 0:12:46.360 --> 0:12:48.839 go into next year. The big one and I think 0:12:48.840 --> 0:12:50.679 the one that Pat is kind of betting the company 0:12:50.679 --> 0:12:52.200 on it is what's called eighteen A. 0:12:52.600 --> 0:12:54.600 So there were two flavors. There was a twenty eight. 0:12:54.480 --> 0:12:56.400 Process that was supposed to come out with a product 0:12:56.400 --> 0:13:00.400 called arrow Lake that they have canceled. So the five 0:13:00.440 --> 0:13:01.839 nodes in four years of the canceled they're going to 0:13:01.920 --> 0:13:02.560 jump straight. 0:13:02.320 --> 0:13:02.840 To eighteen A. 0:13:03.320 --> 0:13:05.400 Their characterization of that is an eighteen A is so 0:13:05.480 --> 0:13:07.439 good they don't need twenty get twenty eight, So I 0:13:07.440 --> 0:13:09.880 guess we'll have to see eighteen A in theory is 0:13:09.880 --> 0:13:13.120 when they start bringing volume back from the foundries back 0:13:13.120 --> 0:13:15.240 into their own fabs, which in theory is supposed to 0:13:15.240 --> 0:13:17.560 make their cost structure better. And if you think about 0:13:17.559 --> 0:13:20.800 the dream case that they laid out, the idea was, 0:13:20.840 --> 0:13:23.520 we do this, we bring all this stuff back. It 0:13:23.559 --> 0:13:26.040 makes our margins better because the cost are better because 0:13:26.040 --> 0:13:28.880 we're now internal. It's a better process, so we can 0:13:28.920 --> 0:13:31.440 take back market share and we can charge more for 0:13:31.480 --> 0:13:33.920 our product because we can design better products. And it 0:13:33.960 --> 0:13:36.800 gives customers like lots of confidence that things are going good, 0:13:36.840 --> 0:13:39.120 so they're confident to put a lot of volume with us, 0:13:39.400 --> 0:13:41.240 and we grow from there. And they talked about hitting 0:13:41.280 --> 0:13:43.520 the manufacturing business kind of hitting break even around twenty 0:13:43.520 --> 0:13:47.160 twenty seven and then growing to like very sizeable amounts 0:13:47.160 --> 0:13:48.559 by twenty thirty, so that. 0:13:48.360 --> 0:13:51.640 That's the dream. It's all riding on the success of 0:13:51.679 --> 0:13:52.120 eighteen A. 0:13:53.240 --> 0:13:54.840 Many of the sort of you know, kind of like 0:13:55.520 --> 0:13:57.360 supply change ets and channels that can pick up for 0:13:57.360 --> 0:13:59.360 eight and right, they're kind of mixed at best, I 0:13:59.400 --> 0:14:02.640 would say. On the positive, they just announced Amazon as 0:14:02.640 --> 0:14:04.600 a customer, and I guess Amazon was willing to have 0:14:04.640 --> 0:14:06.560 their name attached to it, so I guess that's good. 0:14:06.800 --> 0:14:08.480 It doesn't look like it's a high volume part. 0:14:08.520 --> 0:14:09.839 I think it's going to take time for this to 0:14:09.920 --> 0:14:12.600 ramp and I guess we're just waiting around and seeing 0:14:12.600 --> 0:14:14.440 and that's probably about how long they have. You know, 0:14:14.640 --> 0:14:17.520 in theory, eighteen A is supposed to be here next year. 0:14:18.160 --> 0:14:20.840 They've got a low volume server part called Clearwater Force 0:14:20.920 --> 0:14:23.920 that launches in the first half of next year. Panther Lake, 0:14:23.960 --> 0:14:26.200 which is a higher volume client product, launches in the 0:14:26.200 --> 0:14:27.000 second half of next year. 0:14:27.000 --> 0:14:29.600 I guess it'll be volume, real volume in twenty twenty six. 0:14:30.200 --> 0:14:31.640 So you got about a year a year and a 0:14:31.680 --> 0:14:33.760 half to see whether or not they can actually manage 0:14:33.760 --> 0:14:37.640 to make stuff in a cost effective manner for themselves 0:14:37.640 --> 0:14:39.720 on eighteen A and then we'll see if that drives 0:14:39.760 --> 0:14:41.400 volume or not. But that's about how long they have 0:14:41.640 --> 0:14:44.640 I think twelve months, maybe twelve to eighteen months, and 0:14:45.080 --> 0:14:46.920 then we'll see if this works or not. But at 0:14:46.920 --> 0:14:47.680 this point we don't know. 0:14:48.240 --> 0:14:51.480 And you know, Intel's ability to secure major customer commitments 0:14:51.560 --> 0:14:53.760 is of course a really key metric for US government 0:14:53.760 --> 0:14:56.320 officials that are trying to evaluate the viability of there 0:14:56.320 --> 0:14:58.680 should we be giving them eight and a half billion 0:14:58.760 --> 0:15:01.520 dollars of tax payer money for commercial manufacturing. I mean, 0:15:01.560 --> 0:15:03.480 they already have three billion dollars set aside for the 0:15:03.520 --> 0:15:05.480 Pentagon as a customer. But if you talk to anybody 0:15:05.480 --> 0:15:08.040 in the chip industry, the Defense Department is a really 0:15:08.080 --> 0:15:12.000 difficult customer. Foundaries that have been in a secure Pentagon 0:15:12.000 --> 0:15:13.760 supply chain in the past of struggle to make a 0:15:13.800 --> 0:15:18.200 return on products manufactured. Those fabs and Commerce Secretary Gina Raimondo, 0:15:18.440 --> 0:15:21.120 in many ways, her legacy and her image and her 0:15:21.360 --> 0:15:23.120 the kind of main purpose of her department is riding 0:15:23.120 --> 0:15:25.520 a lot of the success of this program, and therefore, 0:15:25.560 --> 0:15:27.400 in part on the success of Intel not in a 0:15:27.440 --> 0:15:30.360 whole She's been calling up chip makers and saying, would 0:15:30.360 --> 0:15:34.320 you consider using Intel fabs? She asked in Vidia, she 0:15:34.440 --> 0:15:38.440 asked AMD to consider manufacturing at Intel's ohio facilities, and 0:15:38.480 --> 0:15:40.840 you know, in Vidias in the earliest stages of evaluating 0:15:40.880 --> 0:15:44.120 whether that's a possibility. AMD is happy with their current supplier, 0:15:44.120 --> 0:15:47.000 which is TSMC. And so you know, as much as 0:15:47.000 --> 0:15:49.640 the government really wants intels to succeed, the industry wants 0:15:49.640 --> 0:15:52.240 intels to succeed, even TSMC does. Intel's a four to 0:15:52.280 --> 0:15:55.480 five billion dollar customer of TSMC's. The custom comments just 0:15:55.560 --> 0:15:57.720 aren't there right now, and. 0:15:58.080 --> 0:16:01.000 There can't really you can't anything infirment. So you know, 0:16:01.000 --> 0:16:03.800 if they can deliver, yeah, right, So I think anyone 0:16:03.840 --> 0:16:06.240 that wants leading edge is going to evaluate them, like 0:16:06.280 --> 0:16:08.600 why wouldn't you, Like you want second sources, you want 0:16:08.600 --> 0:16:11.520 geographically secure supply. But I mean, if they can't deliver, 0:16:12.040 --> 0:16:14.400 they can't deliver. If they can deliver, they can And 0:16:14.720 --> 0:16:17.440 that's what we're waiting for. Throwing whatever money you want 0:16:17.440 --> 0:16:19.880 to throw at it, it isn't going to solve that problem. 0:16:20.360 --> 0:16:22.880 This gets back to like some of our first ever 0:16:23.000 --> 0:16:27.560 episodes that we did on semiconductors years ago. We're just 0:16:27.600 --> 0:16:31.880 about this basic challenge that like it's brutal to get 0:16:32.520 --> 0:16:35.240 hot yields. That there's all these processes and if you 0:16:35.280 --> 0:16:37.760 get it, if you get every process at ninety nine 0:16:37.800 --> 0:16:41.280 point nine to nine percent, right, that still doesn't get 0:16:41.320 --> 0:16:44.120 you the yields that you need to get. And so 0:16:44.160 --> 0:16:47.360 their strategy is like, Okay, you're building for this product, 0:16:47.360 --> 0:16:50.080 you're building for this end consumer, you're building, you're doing 0:16:50.080 --> 0:16:54.320 in house, and then there's execution and the big question 0:16:54.480 --> 0:16:57.280 mark is just this execution question and we just still 0:16:57.280 --> 0:16:57.680 don't know. 0:16:58.640 --> 0:17:00.840 Yeah, we don't know this. This again, even under the 0:17:00.880 --> 0:17:03.600 best of circumstances is very difficult. You know, they're going 0:17:03.600 --> 0:17:05.359 through a lot of other turmoil right now. So they 0:17:05.440 --> 0:17:08.080 have clearly in the midst of a big layoff. Yeah, 0:17:08.119 --> 0:17:10.639 they're actually like canceling some of these projects. So remember 0:17:10.880 --> 0:17:12.560 they weren't just trying to build in the US. They 0:17:12.560 --> 0:17:14.600 had a big Germany project with other subsidies that they 0:17:14.760 --> 0:17:17.320 just paused. They had a packaging plant in Poland. I 0:17:17.320 --> 0:17:19.960 think they just paused. So they don't need the capacity, 0:17:20.760 --> 0:17:21.960 right and they don't have the money. 0:17:21.720 --> 0:17:23.000 To pay for it. Even with the subsidies. 0:17:23.080 --> 0:17:26.240 So there's a lot of like other like internal turmoil 0:17:26.280 --> 0:17:28.520 that is going on there right now while they are 0:17:28.560 --> 0:17:30.959 still trying to execute on everything else that they have 0:17:31.000 --> 0:17:31.239 to do. 0:17:31.320 --> 0:17:32.520 Like it doesn't make it easier. 0:17:32.680 --> 0:17:35.720 You know, this is an intensely cyclical industry, right and 0:17:35.760 --> 0:17:37.560 you know, when we think about what does you know, 0:17:37.640 --> 0:17:40.440 Intel's success or failures say about the design of US 0:17:40.480 --> 0:17:43.520 industrial policy, you have to remember the ship sacked. Biden 0:17:43.560 --> 0:17:45.800 signed the ship sacked two years ago. That was sort 0:17:45.840 --> 0:17:48.199 of at the earliest stages of the AI boom, and 0:17:48.320 --> 0:17:51.439 were there were maybe warning signs about Intel, certainly, but 0:17:51.480 --> 0:17:53.520 not kind of the flashing red light that we see now. 0:17:53.560 --> 0:17:57.040 And so government officials not known necessarily for being super 0:17:57.119 --> 0:18:00.040 nimble or having to evaluate that in real time. We 0:18:00.080 --> 0:18:02.520 had a lot of preliminary announcements earlier this year. We 0:18:02.560 --> 0:18:05.680 saw basically a multi billion dollar announcement a week in 0:18:05.680 --> 0:18:10.359 March and April, and markets change, company dynamics change. And 0:18:10.400 --> 0:18:12.800 I think the challenge is that, you know, there's a 0:18:12.840 --> 0:18:15.399 sort of lack of understanding even among folks who pay 0:18:15.440 --> 0:18:17.600 attention to this sometimes in Washington, about whether the money 0:18:17.640 --> 0:18:20.359 has actually gone out the door. It hasn't but the 0:18:20.359 --> 0:18:22.760 public is expecting that. Intel is expecting that. I mean, 0:18:22.760 --> 0:18:25.080 the state of Ohio is investing two billion dollars in 0:18:25.160 --> 0:18:28.600 Intel's Ohio facility. Those plans are contingent on getting money 0:18:28.640 --> 0:18:31.080 from the US government, but officials also have to evaluate, 0:18:31.119 --> 0:18:34.480 you know, if a company, you know, Intel or anybody else, 0:18:34.600 --> 0:18:37.680 isn't able to execute on the timeline that we set out, 0:18:37.760 --> 0:18:40.040