WEBVTT 00:00:00.000 --> 00:00:02.980 We are surrounded by complexity that we just 00:00:02.980 --> 00:00:05.839 take for granted. Think about the phone in your 00:00:05.839 --> 00:00:07.719 hand or the screen you might be looking at right 00:00:07.719 --> 00:00:12.179 now. All of that is powered by microchips. It's 00:00:12.179 --> 00:00:14.500 displayed on these incredible high -resolution 00:00:14.500 --> 00:00:16.440 screens. These things, they don't just define 00:00:16.440 --> 00:00:18.920 the modern economy, they define modern... Life 00:00:18.920 --> 00:00:22.940 itself. But here's the thing. Almost nobody stops 00:00:22.940 --> 00:00:27.460 to think about the massive, hyper precise and 00:00:27.460 --> 00:00:30.039 just astronomically expensive infrastructure 00:00:30.039 --> 00:00:33.100 it takes just to make those things. Right. We're 00:00:33.100 --> 00:00:35.600 not talking about Intel or Samsung, you know, 00:00:35.600 --> 00:00:37.640 the companies that design and sell the final 00:00:37.640 --> 00:00:39.520 chips. We're talking about the company that sells 00:00:39.520 --> 00:00:41.679 the machines that build the entire digital world. 00:00:41.840 --> 00:00:43.979 Exactly. And that's where we're going today. 00:00:44.060 --> 00:00:46.179 We're doing a deep dive into applied materials, 00:00:46.479 --> 00:00:49.520 ink or AMAT. Stock Exchange, this is the ultimate 00:00:49.520 --> 00:00:52.259 invisible giant. It really is. They're headquartered 00:00:52.259 --> 00:00:54.259 in Santa Clara, California, but their reach is 00:00:54.259 --> 00:00:56.560 completely global. They provide the absolute 00:00:56.560 --> 00:00:59.020 fundamental equipment for pretty much every piece 00:00:59.020 --> 00:01:01.439 of advanced electronics made anywhere on Earth. 00:01:01.520 --> 00:01:03.380 When you see fundamental, you really mean it. 00:01:03.619 --> 00:01:06.019 AMAD is critical infrastructure. Absolutely. 00:01:06.180 --> 00:01:08.920 They supply the tools, the services, and all 00:01:08.920 --> 00:01:10.719 the specialized software you need to manufacture 00:01:10.719 --> 00:01:13.640 integrated circuit chips, flat panel displays. 00:01:13.840 --> 00:01:15.859 I mean, everything from the tiniest smartphone 00:01:15.859 --> 00:01:18.920 to the biggest TV. And even specialized coatings 00:01:18.920 --> 00:01:20.959 for things like flexible electronics and advanced 00:01:20.959 --> 00:01:23.659 packaging. Yeah, all of it. So our mission today 00:01:23.659 --> 00:01:26.700 is to really pull back the curtain on this behemoth. 00:01:26.700 --> 00:01:28.439 We want to understand their technological core. 00:01:28.659 --> 00:01:31.359 What are these multi -million dollar machines 00:01:31.359 --> 00:01:34.390 actually doing on a silicon wafer? And then we'll 00:01:34.390 --> 00:01:36.810 track their history, which is it's a really dynamic 00:01:36.810 --> 00:01:40.569 story of growth, huge acquisitions and some pretty 00:01:40.569 --> 00:01:43.469 spectacular high profile failures. And then critically, 00:01:43.609 --> 00:01:46.129 we have to connect all of those dots to what's 00:01:46.129 --> 00:01:48.230 happening right now, because the sources show 00:01:48.230 --> 00:01:50.689 a company that is sitting right at the most critical 00:01:50.689 --> 00:01:53.170 and honestly the most dangerous intersection 00:01:53.170 --> 00:01:55.709 of the modern global economy. The stakes just 00:01:55.709 --> 00:01:57.790 couldn't be higher. And to understand why AMAT 00:01:57.790 --> 00:02:00.189 is so central to global conflict at this very 00:02:00.189 --> 00:02:02.489 moment, you have to get a sense of their sheer 00:02:02.489 --> 00:02:05.079 scale. So let's look at the numbers. Based on 00:02:05.079 --> 00:02:08.060 the 2025 financial data, this is a titan. It's 00:02:08.060 --> 00:02:12.460 pulling in US $28 .37 billion in revenue. Wow. 00:02:12.810 --> 00:02:16.110 With a net income of almost $7 billion, that's 00:02:16.110 --> 00:02:19.949 U .S. $6 .998 billion, and they have a dedicated 00:02:19.949 --> 00:02:24.030 global workforce of 36 ,500 people. And of course, 00:02:24.050 --> 00:02:26.310 they're a component of both the S &P 500 and 00:02:26.310 --> 00:02:29.349 the NASDAQ 100. This is not some small niche 00:02:29.349 --> 00:02:31.990 player. No, not at all. This is the engine room 00:02:31.990 --> 00:02:34.650 of global technology. And as we often remind 00:02:34.650 --> 00:02:36.789 listeners, you know, many people have heard of 00:02:36.789 --> 00:02:39.340 the Dutch company ASML. because of their crucial 00:02:39.340 --> 00:02:42.139 lithography machines. But Applied Materials is 00:02:42.139 --> 00:02:44.340 the second largest supplier of semiconductor 00:02:44.340 --> 00:02:46.939 equipment in the world. They are completely indispensable. 00:02:47.319 --> 00:02:49.419 And that indispensability is exactly what we 00:02:49.419 --> 00:02:51.719 need to unpack. Okay, so where do we start? Let's 00:02:51.719 --> 00:02:53.060 start with how they've structured this global 00:02:53.060 --> 00:02:54.879 operation. They basically have three pillars 00:02:54.879 --> 00:02:56.939 to their business. And then we'll dive deep into 00:02:56.939 --> 00:02:58.900 pillar one, which is the semiconductor products 00:02:58.900 --> 00:03:00.879 group. That's the heart of all their technical 00:03:00.879 --> 00:03:03.259 expertise. Let's do it. Let's jump straight into 00:03:03.259 --> 00:03:05.560 that technical core. Because I think if you want 00:03:05.560 --> 00:03:08.819 to grasp why the U .S. government cares so much 00:03:08.819 --> 00:03:11.000 about what AMAT sells to China, for example, 00:03:11.139 --> 00:03:14.000 you have to appreciate the sheer complexity of 00:03:14.000 --> 00:03:16.500 what their machines can do. Absolutely. So Applied 00:03:16.500 --> 00:03:19.360 Materials organizes its sprawling business into 00:03:19.360 --> 00:03:22.250 three major sectors. And they also have a really 00:03:22.250 --> 00:03:25.449 important venture investing arm called Applied 00:03:25.449 --> 00:03:27.930 Ventures, which is always scanning the horizon 00:03:27.930 --> 00:03:30.030 for the next big thing. But the money's in the 00:03:30.030 --> 00:03:33.280 first pillar. The vast majority of that $28 billion 00:03:33.280 --> 00:03:36.340 in revenue is generated by Pillar 1, semiconductor 00:03:36.340 --> 00:03:39.180 products. This is the division that develops 00:03:39.180 --> 00:03:41.400 and builds the highly specialized equipment used 00:03:41.400 --> 00:03:43.840 in the most intricate phases of creating a chip. 00:03:44.060 --> 00:03:46.599 We're talking about the wafer fabrication steps. 00:03:46.939 --> 00:03:49.180 I think the best way to visualize this for everyone 00:03:49.180 --> 00:03:51.939 listening is to imagine the silicon wafer not 00:03:51.939 --> 00:03:54.599 just as a flat disk, but as a three -dimensional 00:03:54.599 --> 00:03:57.020 blueprint that's built up layer by atomic layer. 00:03:57.389 --> 00:03:59.349 That's a great analogy. It's not an assembly 00:03:59.349 --> 00:04:00.770 line in the traditional sense. It's like the 00:04:00.770 --> 00:04:02.729 most sophisticated molecular construction site 00:04:02.729 --> 00:04:05.069 humanity has ever come up with. That's a perfect 00:04:05.069 --> 00:04:07.629 way to frame it. And AMAT provides the entire 00:04:07.629 --> 00:04:10.330 toolkit for that construction site. To really 00:04:10.330 --> 00:04:12.990 simplify the process, all the incredibly sophisticated 00:04:12.990 --> 00:04:15.909 machines they build fall into three basic categories 00:04:15.909 --> 00:04:19.230 of action. You're either adding material, modifying 00:04:19.230 --> 00:04:21.709 that material, or you're removing and polishing 00:04:21.709 --> 00:04:23.389 that material. Okay, let's start with adding 00:04:23.389 --> 00:04:26.310 material. The sources say AMAT calls this deposition. 00:04:27.030 --> 00:04:29.149 This seems to be where their core strength really 00:04:29.149 --> 00:04:32.529 lies. What are the key processes here? So we're 00:04:32.529 --> 00:04:35.089 talking about a whole alphabet soup of acronyms, 00:04:35.129 --> 00:04:37.970 but the main ones are atomic layer deposition, 00:04:38.110 --> 00:04:42.569 or ALD, chemical vapor deposition, CVD, and physical 00:04:42.569 --> 00:04:44.750 vapor deposition, PVD. Okay, let's break those 00:04:44.750 --> 00:04:47.410 down. What are CVD and PVD? You can think of 00:04:47.410 --> 00:04:50.029 CVD and PVD as the workhorses. They've been around 00:04:50.029 --> 00:04:53.149 for a while. CVD uses chemical gas reactions 00:04:53.149 --> 00:04:55.850 inside a chamber to literally grow a thin film 00:04:55.850 --> 00:04:59.029 on the wafer surface. And PVD? PVD is a bit different. 00:04:59.949 --> 00:05:02.889 process called sputtering. Imagine blasting a 00:05:02.889 --> 00:05:05.509 solid target of, say, copper or aluminum with 00:05:05.509 --> 00:05:08.589 ions in a vacuum so that tiny particles vaporize 00:05:08.589 --> 00:05:10.750 and then coat the wafer. So they're both essential 00:05:10.750 --> 00:05:13.250 for laying down the different layers you need 00:05:13.250 --> 00:05:15.870 in a chip, the conductive wires, the insulating 00:05:15.870 --> 00:05:18.449 materials, all of that. Exactly. They are absolutely 00:05:18.449 --> 00:05:21.110 fundamental. But the real game changers, especially 00:05:21.110 --> 00:05:23.069 in the advanced chips where all the geopolitical 00:05:23.069 --> 00:05:27.410 tension is focused, seems to be ALD, atomic layer 00:05:27.410 --> 00:05:30.319 deposition. Why is that process so critical right 00:05:30.319 --> 00:05:34.259 now? This is where AMAT earns its geopolitical 00:05:34.259 --> 00:05:37.839 relevance. Modern chips are no longer flat two 00:05:37.839 --> 00:05:39.639 -dimensional structures. Right, we've moved into 00:05:39.639 --> 00:05:42.259 3D. We've moved into 3D architectures, first 00:05:42.259 --> 00:05:44.980 with FinFETs and now with what we call gate all 00:05:44.980 --> 00:05:48.839 around or GAA sisters. These designs have vertical 00:05:48.839 --> 00:05:52.019 fins, incredibly deep and narrow trenches. I 00:05:52.019 --> 00:05:54.379 mean, the topography is insane. So you can't 00:05:54.379 --> 00:05:56.620 just spray a layer of material on top and hope 00:05:56.620 --> 00:05:58.939 for the best. You absolutely can't. When you 00:05:58.939 --> 00:06:01.620 need to lay down a perfect insulating layer or 00:06:01.620 --> 00:06:03.819 a conducting gate material inside one of those 00:06:03.819 --> 00:06:07.439 tiny deep features, traditional CVD or PVD just 00:06:07.439 --> 00:06:10.040 won't work. The material won't conform uniformly 00:06:10.040 --> 00:06:12.379 to that complex 3D shape. You'll get more on 00:06:12.379 --> 00:06:14.600 top and less on the bottom. And if the material 00:06:14.600 --> 00:06:17.139 doesn't conform perfectly, the transistor fails. 00:06:17.420 --> 00:06:20.730 The whole chip could be useless. Precisely. ALD 00:06:20.730 --> 00:06:23.730 solves this problem. It uses alternating chemical 00:06:23.730 --> 00:06:27.050 pulses, where each pulse reacts with the surface, 00:06:27.310 --> 00:06:30.930 coating it one single atomic layer at a time. 00:06:31.029 --> 00:06:34.110 Imagine trying to paint the inside of a microscopic 00:06:34.110 --> 00:06:37.750 winding cave with such precision that every single 00:06:37.750 --> 00:06:41.089 internal surface, every nook and cranny, is coated 00:06:41.089 --> 00:06:43.649 with exactly the same thickness of paint. That's 00:06:43.649 --> 00:06:47.170 incredible. It is. AMBAT's ALD tools allow chip 00:06:47.170 --> 00:06:49.209 makers to build things like high -thigh electrics 00:06:49.209 --> 00:06:51.310 and other essential materials with these ultra 00:06:51.310 --> 00:06:54.709 -thin, perfectly uniform films. That guarantees 00:06:54.709 --> 00:06:56.930 the performance and the reliability of those 00:06:56.930 --> 00:06:59.750 3D transistors. So the ability to do this consistently 00:06:59.750 --> 00:07:03.129 at scale across millions of wafers is foundational 00:07:03.129 --> 00:07:06.170 to every modern logic and memory chip. It is. 00:07:06.189 --> 00:07:08.589 And this is why AMBAT's specific tool sets grant 00:07:08.589 --> 00:07:10.889 them, frankly, a chokehold power in the industry. 00:07:11.149 --> 00:07:13.009 Okay, so that's deposition or adding material. 00:07:13.769 --> 00:07:15.670 deposited the layers, we move into the refinement 00:07:15.670 --> 00:07:18.230 modification phase. The material is there, but 00:07:18.230 --> 00:07:20.569 it's not functional yet. How does AMAT's equipment 00:07:20.569 --> 00:07:23.009 actually change it? The two main processes here 00:07:23.009 --> 00:07:26.750 are rapid thermal processing, or RTP, and ion 00:07:26.750 --> 00:07:29.600 implantation. let's take rtp first what does 00:07:29.600 --> 00:07:32.519 that do rtp is a crucial step that happens after 00:07:32.519 --> 00:07:34.920 you've deposited material or etched some features 00:07:34.920 --> 00:07:37.980 those processes can be really violent at a microscopic 00:07:37.980 --> 00:07:41.240 level and they can damage or stress the silicon's 00:07:41.240 --> 00:07:43.779 crystal structure okay so you need to heal it 00:07:43.779 --> 00:07:46.199 you have to heat the wafer to very high temperatures 00:07:46.199 --> 00:07:48.819 sometimes up to a thousand degrees celsius but 00:07:49.389 --> 00:07:51.089 And this is the key. You have to do it incredibly 00:07:51.089 --> 00:07:53.149 quickly and precisely in just a few seconds. 00:07:53.370 --> 00:07:56.110 Why so fast? Because you need to repair those 00:07:56.110 --> 00:07:58.870 crystal defects or activate the dopants you've 00:07:58.870 --> 00:08:01.389 introduced without letting the heat spread and 00:08:01.389 --> 00:08:03.550 damage all the delicate surrounding layers you've 00:08:03.550 --> 00:08:06.529 already built. So AMAT's RTT tools are all about 00:08:06.529 --> 00:08:09.110 lightning fast, highly controlled heating and 00:08:09.110 --> 00:08:11.709 cooling cycles. And then there's ion implantation. 00:08:11.870 --> 00:08:14.449 This sounds like where the chip actually becomes 00:08:14.449 --> 00:08:16.509 programmable. Is that right? That's exactly right. 00:08:16.589 --> 00:08:20.269 This is really where the magic happens. raw inert 00:08:20.269 --> 00:08:23.410 silicon into a semiconductor. Ion implantation 00:08:23.410 --> 00:08:26.569 is literally a high energy particle accelerator 00:08:26.569 --> 00:08:30.009 that fires specific ions like boron or phosphorus 00:08:30.009 --> 00:08:33.470 directly into the wafer's surface. These ions 00:08:33.470 --> 00:08:36.090 are called dopants. They penetrate the silicon 00:08:36.090 --> 00:08:38.389 lattice and change its electrical properties. 00:08:38.710 --> 00:08:40.610 And that's what creates the different types of 00:08:40.610 --> 00:08:43.750 silicon. Yes. It defines whether a specific area 00:08:43.750 --> 00:08:47.750 will become p -type or n -type silicon. And without 00:08:47.950 --> 00:08:50.450 That distinction, you cannot create the diodes 00:08:50.450 --> 00:08:52.529 and transistors that form all the logic gates. 00:08:52.769 --> 00:08:55.409 This is a fundamental step. And the sources note 00:08:55.409 --> 00:08:58.190 that AMAP made a huge move in this area with 00:08:58.190 --> 00:09:00.649 a major acquisition. They did. They bought Varian 00:09:00.649 --> 00:09:03.330 Semiconductor back in 2011, which was a huge 00:09:03.330 --> 00:09:05.450 deal, and made them a full -service provider 00:09:05.450 --> 00:09:07.950 of these critical modification tools and really 00:09:07.950 --> 00:09:10.029 solidified their market position. Which brings 00:09:10.029 --> 00:09:12.389 us to the final category, shaping and finishing. 00:09:12.879 --> 00:09:15.500 Once you've layered, doped, and healed the wafer, 00:09:15.639 --> 00:09:17.720 you have to actually carve the circuits and then 00:09:17.720 --> 00:09:19.659 polish the surface to get ready for the next 00:09:19.659 --> 00:09:22.100 layer. Right. This involves two key processes, 00:09:22.539 --> 00:09:25.279 etch and chemical mechanical polishing, or CMP. 00:09:25.580 --> 00:09:28.460 Etch is the carving part. Exactly. AMAT provides 00:09:28.460 --> 00:09:31.720 plasma etch tools that selectively remove layers 00:09:31.720 --> 00:09:34.580 of material. They usually follow a pattern that's 00:09:34.580 --> 00:09:38.080 been laid down by a lithography machine. So this 00:09:38.080 --> 00:09:39.899 is how you create the actual circuit pathways. 00:09:40.299 --> 00:09:43.460 It is. Etch is how you create those deep vertical 00:09:43.460 --> 00:09:46.679 trenches and complex high aspect ratio patterns 00:09:46.679 --> 00:09:49.779 that define the wiring of the chip. It's incredibly 00:09:49.779 --> 00:09:53.039 precise subtraction. And then there's CMP, chemical 00:09:53.039 --> 00:09:55.720 mechanical polishing. Now, this sounds like the 00:09:55.720 --> 00:09:57.580 simplest step, you know, just polishing something. 00:09:57.759 --> 00:10:00.600 But in a 3D chip environment, that must be an 00:10:00.600 --> 00:10:03.200 immense engineering challenge. It is arguably 00:10:03.200 --> 00:10:05.679 one of the most critical processes for maintaining 00:10:05.679 --> 00:10:08.720 yield or the percentage of working chips in modern 00:10:08.720 --> 00:10:11.210 manufacturing. Why is it so important? Well think 00:10:11.210 --> 00:10:12.830 about it. By the time you're done, you might 00:10:12.830 --> 00:10:15.789 have dozens or even hundreds of layers on a wafer 00:10:15.789 --> 00:10:18.690 alternating between tiny copper wires and insulating 00:10:18.690 --> 00:10:21.789 materials. The surface inevitably develops tiny 00:10:21.789 --> 00:10:24.129 bumps and valleys. It's not perfectly flat. And 00:10:24.129 --> 00:10:26.990 if that surface isn't absolutely perfectly flat 00:10:26.990 --> 00:10:29.570 before you try to pattern the next layer, your 00:10:29.570 --> 00:10:32.750 lithography step won't focus correctly. the light 00:10:32.750 --> 00:10:35.409 will be distorted and the entire multi -million 00:10:35.409 --> 00:10:38.230 dollar wafer will fail so to use our earlier 00:10:38.230 --> 00:10:41.210 analogy you have to polish the surface of a microscopic 00:10:41.210 --> 00:10:44.169 skyscraper perfectly flat every hundred feet 00:10:44.169 --> 00:10:46.429 so you can build the next hundred feet without 00:10:46.429 --> 00:10:48.769 the whole thing collapsing that's a fantastic 00:10:48.769 --> 00:10:51.830 way to put it that's exactly right cmp uses a 00:10:51.830 --> 00:10:54.580 chemical slurry combined with a high -tech polishing 00:10:54.580 --> 00:10:57.759 pad to polish the wafer surface down to a near 00:10:57.759 --> 00:11:00.799 atomic level of flatness across the entire 300 00:11:00.799 --> 00:11:03.940 millimeter disc across the entire disc AMT's 00:11:03.940 --> 00:11:06.700 CMT equipment ensures that uniformity, which 00:11:06.700 --> 00:11:08.980 is absolutely essential for maximizing the number 00:11:08.980 --> 00:11:11.519 of working chips you get from every single wafer. 00:11:11.620 --> 00:11:13.360 It's clear that their semiconductor product sector 00:11:13.360 --> 00:11:16.100 isn't just about making one -off machines. It's 00:11:16.100 --> 00:11:18.779 about providing the entire integrated capability 00:11:18.779 --> 00:11:21.059 for the most advanced manufacturing possible. 00:11:21.259 --> 00:11:23.659 Right. And you can't make a chip if you can't 00:11:23.659 --> 00:11:26.240 verify the quality at every single step. That 00:11:26.240 --> 00:11:28.840 must be a huge part of their business, too. Yield 00:11:28.840 --> 00:11:32.120 is everything in this business. A single microscopic 00:11:32.120 --> 00:11:35.179 defect can ruin a wafer worth a fortune. So way 00:11:35.179 --> 00:11:38.419 back in 1996, AMAP made a strategic move and 00:11:38.419 --> 00:11:41.100 acquired two companies, Opal Technologies and 00:11:41.100 --> 00:11:44.820 Orbot Instruments, for a combined $285 million. 00:11:45.360 --> 00:11:48.139 And what was the goal there? The goal was specifically 00:11:48.139 --> 00:11:50.539 to establish leadership in yield enhancement. 00:11:51.070 --> 00:11:53.129 These acquisitions gave them the tools for high 00:11:53.129 --> 00:11:55.169 -speed metrology, which is just a fancy word 00:11:55.169 --> 00:11:57.970 for measuring critical dimensions, and for inspecting 00:11:57.970 --> 00:12:00.730 the patterned silicon wafers and masks. It's 00:12:00.730 --> 00:12:02.789 the insurance policy for the entire fabrication 00:12:02.789 --> 00:12:05.129 process. So their strategy in this pillar is 00:12:05.129 --> 00:12:06.789 just aggressive growth through consolidation. 00:12:07.309 --> 00:12:09.649 Yes. And that brings us to the planned purchase 00:12:09.649 --> 00:12:13.370 of Kokusai Electric, the $2 .2 billion deal that 00:12:13.370 --> 00:12:15.620 ultimately fell apart. That planned acquisition 00:12:15.620 --> 00:12:18.100 shows AMAT's constant desire to consolidate their 00:12:18.100 --> 00:12:20.440 deposition capabilities even further. And we 00:12:20.440 --> 00:12:22.679 will definitely come back to why that deal failed. 00:12:22.740 --> 00:12:25.539 It's a huge part of the story. It is. Kokusai 00:12:25.539 --> 00:12:29.090 Electric is a very strong player in... batch 00:12:29.090 --> 00:12:30.970 process equipment, especially for deposition, 00:12:31.090 --> 00:12:34.049 and absorbing them would have dramatically strengthened 00:12:34.049 --> 00:12:36.649 AMAT's market share. But the failure of that 00:12:36.649 --> 00:12:39.809 deal is a potent indicator of the global regulatory 00:12:39.809 --> 00:12:42.750 forces that were already at play years ago. Okay, 00:12:42.809 --> 00:12:44.350 so that's pillar one, the heart of the company. 00:12:44.509 --> 00:12:47.429 Let's move to pillar two, applied global services, 00:12:47.470 --> 00:12:51.029 or AGS. If pillar one is the engine, AGS sounds 00:12:51.029 --> 00:12:53.070 like the indispensable support and maintenance 00:12:53.070 --> 00:12:55.529 crew that keeps the entire global system running. 00:12:55.929 --> 00:12:57.950 That's a great way to think about it. AGS is 00:12:57.950 --> 00:12:59.850 the company's moat. It's what protects their 00:12:59.850 --> 00:13:02.450 business. Also. Well, chip fabrication, plants, 00:13:02.490 --> 00:13:05.750 fabs, they run 204 .7. They never stop. And every 00:13:05.750 --> 00:13:07.850 single piece of equipment we just discussed is 00:13:07.850 --> 00:13:10.769 hugely complex and prone to microscopic failures. 00:13:11.090 --> 00:13:13.950 The AGS sector is entirely focused on maintaining, 00:13:14.250 --> 00:13:16.549 upgrading and servicing that massive, expensive 00:13:16.549 --> 00:13:18.850 installed base of equipment all over the world. 00:13:18.970 --> 00:13:21.529 And I'm guessing this is the high margin recurring 00:13:21.529 --> 00:13:24.450 revenue stream. It must stabilize the business 00:13:24.450 --> 00:13:27.250 when equipment sales fluctuate. Absolutely. When 00:13:27.250 --> 00:13:30.889 a fab like TSMC or Intel spends tens or hundreds 00:13:30.889 --> 00:13:33.250 of millions of dollars on a machine, they need 00:13:33.250 --> 00:13:36.490 a guarantee that it will run 99 .9 % of the time. 00:13:37.070 --> 00:13:39.509 AGS provides that guarantee through installation 00:13:39.509 --> 00:13:42.669 support, warranty extensions, and crucially, 00:13:42.750 --> 00:13:44.870 long -term maintenance contracts. So they're 00:13:44.870 --> 00:13:48.090 locked in. They're locked in. And AGS also operates 00:13:48.090 --> 00:13:50.649 a significant business selling new and refurbished 00:13:50.649 --> 00:13:52.769 equipment components, which is another steady 00:13:52.769 --> 00:13:55.289 source of income. But the sources say the real 00:13:55.289 --> 00:13:57.470 value add isn't just fixing things when they 00:13:57.470 --> 00:14:00.210 break. It's the constant cycle of improvement. 00:14:00.809 --> 00:14:03.409 That's the key differentiator. They provide crucial 00:14:03.409 --> 00:14:05.649 upgrades and enhancements for existing machines. 00:14:06.009 --> 00:14:08.210 They can take a machine that was installed five 00:14:08.210 --> 00:14:10.870 or even 10 years ago and update its software, 00:14:11.129 --> 00:14:13.970 retrofit it with new components, or even modify 00:14:13.970 --> 00:14:16.509 its process kit to enable new manufacturing recipes 00:14:16.509 --> 00:14:19.049 for newer chip designs. So that extends the life 00:14:19.049 --> 00:14:21.570 and the capability of the client's multi -billion 00:14:21.570 --> 00:14:24.210 dollar fab investment. Exactly. And it ties the 00:14:24.210 --> 00:14:26.889 client even tighter to AMAT's ecosystem. The 00:14:26.889 --> 00:14:28.909 outline also mentions that AGS includes automation 00:14:28.909 --> 00:14:31.360 software. What's that about? Running a modern 00:14:31.360 --> 00:14:34.179 cleanroom is a logistical nightmare. You have 00:14:34.179 --> 00:14:37.220 thousands of precise process steps handling millions 00:14:37.220 --> 00:14:39.620 of valuable wafers moving between hundreds of 00:14:39.620 --> 00:14:42.779 machines. AGS provides the automation software 00:14:42.779 --> 00:14:45.940 necessary for scheduling, logistics, yield management, 00:14:46.179 --> 00:14:48.860 and process control. So it's a brain of the factory. 00:14:49.320 --> 00:14:52.460 It is. This software layer is what ensures maximum 00:14:52.460 --> 00:14:55.039 efficiency and allows the manufacturer to run 00:14:55.039 --> 00:14:57.179 with minimal human intervention, which reduces 00:14:57.179 --> 00:15:00.320 error and contamination. It's a huge value add. 00:15:00.480 --> 00:15:02.340 So we have the cutting -edge technology in Pillar 00:15:02.340 --> 00:15:05.240 1, supported globally by the high -margin services 00:15:05.240 --> 00:15:09.500 of Pillar 2. That leaves Pillar 3. Display and 00:15:09.500 --> 00:15:11.639 adjacent markets, the screens we all look at 00:15:11.639 --> 00:15:14.220 every single day. Right. This segment really 00:15:14.220 --> 00:15:16.539 came together after they acquired a company called 00:15:16.539 --> 00:15:19.460 Applied Films Corporation in mid -2006, and they 00:15:19.460 --> 00:15:21.360 combined that with an existing unit they had 00:15:21.360 --> 00:15:23.940 that was focused on large area processing. What 00:15:23.940 --> 00:15:26.220 makes AMAD a natural fit for the display market? 00:15:26.360 --> 00:15:29.000 I mean, huge glass panels seem so different from 00:15:29.000 --> 00:15:31.860 tiny silicon wafers. On the surface, yes, but 00:15:31.860 --> 00:15:33.980 the underlying processes are remarkably similar, 00:15:34.100 --> 00:15:39.899 just scaled up to a massive size. thin film transistor 00:15:39.899 --> 00:15:43.100 liquid crystal displays, which form the backbone 00:15:43.100 --> 00:15:46.460 of monitors, tablets and TVs, requires the exact 00:15:46.460 --> 00:15:48.860 same foundational technologies as making chips. 00:15:49.059 --> 00:15:51.679 Really? Yes. They use clean room environments. 00:15:51.980 --> 00:15:54.500 They rely on photolithography to pattern the 00:15:54.500 --> 00:15:57.620 transistors onto huge glass substrates. And crucially, 00:15:57.779 --> 00:16:00.120 they use chemical and physical vapor deposition 00:16:00.120 --> 00:16:02.980 to lay down the thin films. So it all comes back 00:16:02.980 --> 00:16:05.080 to their core expertise in deposition. It all 00:16:05.080 --> 00:16:08.279 comes back to deposition. AMAT. simply leverages 00:16:08.279 --> 00:16:10.580 its world -leading expertise in high -precision 00:16:10.580 --> 00:16:13.220 deposition and coating systems and adapts them 00:16:13.220 --> 00:16:16.720 from 300 millimeter wafers to massive glass panels 00:16:16.720 --> 00:16:19.120 that can be several square meters in size. It's 00:16:19.120 --> 00:16:21.299 a perfect illustration of how deeply rooted that 00:16:21.299 --> 00:16:24.120 core technology is across different fields. That 00:16:24.120 --> 00:16:26.340 gives us a fantastic summary of their operational 00:16:26.340 --> 00:16:29.070 structure. Now let's rewind the clock and look 00:16:29.070 --> 00:16:30.629 at the narrative of how they built this structure, 00:16:30.730 --> 00:16:33.429 because their history is really marked by chaos, 00:16:33.570 --> 00:16:36.250 some very sharp strategic shifts, and as we've 00:16:36.250 --> 00:16:38.809 hinted, some major regulatory roadblocks. Here's 00:16:38.809 --> 00:16:41.879 where the story gets really interesting. Applied 00:16:41.879 --> 00:16:44.519 Materials was founded way back on November 10th, 00:16:44.519 --> 00:16:48.340 1967, by Michael A. McNeely and others. They 00:16:48.340 --> 00:16:51.299 were part of that first big Silicon Valley boom. 00:16:51.519 --> 00:16:53.700 And they went public pretty quickly. They did. 00:16:53.860 --> 00:16:56.639 They went public on the then -new Nasdaq exchange 00:16:56.639 --> 00:17:01.259 in 1972. But the early years were not a smooth 00:17:01.259 --> 00:17:03.480 ride to the top. There were some real challenges. 00:17:03.720 --> 00:17:06.279 What was the main problem? Diversification, or 00:17:06.279 --> 00:17:09.279 really a lack of focus. The initial companies 00:17:09.279 --> 00:17:11.519 sort of drifted into various other ventures that 00:17:11.519 --> 00:17:13.619 were outside of their core equipment business. 00:17:13.880 --> 00:17:16.319 They spread themselves too thin. Exactly. And 00:17:16.319 --> 00:17:19.259 the critical turning point came in 1976 when 00:17:19.259 --> 00:17:22.180 James C. Morgan took over as CEO. He recognized 00:17:22.180 --> 00:17:24.680 this immediately. He saw that the company needed 00:17:24.680 --> 00:17:27.000 to commit entirely to the one thing they were 00:17:27.000 --> 00:17:29.759 best at, semiconductor manufacturing equipment. 00:17:30.059 --> 00:17:32.680 And that single decision to narrow the focus 00:17:32.680 --> 00:17:35.500 seems to have defined the company's entire trajectory 00:17:35.500 --> 00:17:38.190 for the next 40 years. It did. The financial 00:17:38.190 --> 00:17:40.769 results were immediate and they completely validated 00:17:40.769 --> 00:17:44.690 his pivot. By 1978, just two years after Morgan's 00:17:44.690 --> 00:17:47.109 decision, sales had increased by a substantial 00:17:47.109 --> 00:17:51.210 17%. That one act of strategic clarity basically 00:17:51.210 --> 00:17:53.529 rescued the company and set it on the path to 00:17:53.529 --> 00:17:55.950 becoming a global leader. That focus eventually 00:17:55.950 --> 00:17:58.990 allowed them to innovate. And the high watermark 00:17:58.990 --> 00:18:02.029 of their early innovation Has to be the Precision 00:18:02.029 --> 00:18:05.470 5000 in 1987. Why should we care about a piece 00:18:05.470 --> 00:18:07.950 of CVD equipment from the late 80s? Because the 00:18:07.950 --> 00:18:10.309 Precision 5000 didn't just perform a process 00:18:10.309 --> 00:18:12.869 better, it revolutionized the entire factory 00:18:12.869 --> 00:18:15.750 floor. It changed the whole concept of manufacturing. 00:18:16.289 --> 00:18:18.490 How so? What was it like before the P5000? Before 00:18:18.490 --> 00:18:21.089 the P5000, different process steps required moving 00:18:21.089 --> 00:18:23.910 the wafer between highly specialized, isolated, 00:18:24.109 --> 00:18:26.490 single chamber machines. You do one step, take 00:18:26.490 --> 00:18:28.230 the wafer out, carry it across the clean room 00:18:28.230 --> 00:18:30.150 in a cassette, and put it in the next machine. 00:18:30.599 --> 00:18:33.740 Which sounds inefficient and risky. Extremely. 00:18:33.740 --> 00:18:36.299 Every time you handle the wafer, you risk contamination 00:18:36.299 --> 00:18:39.859 from dust particles, which can kill a chip. The 00:18:39.859 --> 00:18:43.740 P5000 was a massive conceptual leap. It incorporated 00:18:43.740 --> 00:18:46.500 diverse processes, like multiple different processing 00:18:46.500 --> 00:18:49.900 chambers, into a single integrated system with 00:18:49.900 --> 00:18:53.130 a central robotic handler. So instead of transporting 00:18:53.130 --> 00:18:55.950 the wafer across the factory floor, they could 00:18:55.950 --> 00:18:58.690 complete complex sequences of steps all inside 00:18:58.690 --> 00:19:02.099 one machine under vacuum. Exactly. This drastically 00:19:02.099 --> 00:19:04.980 reduced handling, which cut contamination, increased 00:19:04.980 --> 00:19:07.359 precision, and significantly boosted the yield. 00:19:07.500 --> 00:19:10.039 It was the birth of the integrated process platform 00:19:10.039 --> 00:19:12.599 concept, which is completely standard today. 00:19:12.779 --> 00:19:14.880 And it was a big deal at the time. A huge deal. 00:19:14.980 --> 00:19:18.200 It was so influential that in 1993, the Precision 00:19:18.200 --> 00:19:21.079 5000 was inducted into the Smithsonian Institution's 00:19:21.079 --> 00:19:23.500 permanent collection of information age technology. 00:19:23.880 --> 00:19:26.359 Wow. That's a testament to its impact. It really 00:19:26.359 --> 00:19:28.359 is. At the same time this was happening, AMAT 00:19:28.359 --> 00:19:31.119 was also showing incredible foresight by aggressively 00:19:31.119 --> 00:19:33.720 pursuing global market expansion long before 00:19:33.720 --> 00:19:36.180 many of its American competitors. They were true 00:19:36.180 --> 00:19:39.079 pioneers in that respect. They recognized the 00:19:39.079 --> 00:19:41.339 rise of Japan as a manufacturing powerhouse. 00:19:41.559 --> 00:19:45.220 So AMAP became the very first U .S. semiconductor 00:19:45.220 --> 00:19:48.299 equipment manufacturer to open its own technology 00:19:48.299 --> 00:19:51.359 center there back in 1984. And they were early 00:19:51.359 --> 00:19:54.460 in China, too. Very early. Crucially, they were 00:19:54.460 --> 00:19:56.640 also the first semiconductor equipment company 00:19:56.640 --> 00:20:00.339 to operate a service center in China. This early, 00:20:00.420 --> 00:20:03.519 deliberate global vision positioned them perfectly 00:20:03.519 --> 00:20:05.750 for the massive scale scale of manufacturing 00:20:05.750 --> 00:20:09.009 that would eventually migrate to Asia. They saw 00:20:09.009 --> 00:20:11.569 where the world was going. Before we delve into 00:20:11.569 --> 00:20:13.490 their big strategy of acquisitions, we should 00:20:13.490 --> 00:20:16.650 probably pause for a moment on that bizarre and 00:20:16.650 --> 00:20:20.059 unique detail the sources included. The 1992 00:20:20.059 --> 00:20:23.900 lawsuit involving Applied Scholastics and Scientology. 00:20:24.039 --> 00:20:25.839 It's a strange corporate footnote, for sure, 00:20:25.980 --> 00:20:28.019 but it shows some of the management challenges 00:20:28.019 --> 00:20:30.140 that can pop up during periods of hypergrowth. 00:20:30.440 --> 00:20:33.079 So what happened? AMAT ended up settling a lawsuit 00:20:33.079 --> 00:20:36.559 for an estimated $600 ,000 with a group of former 00:20:36.559 --> 00:20:39.019 employees. They complained that they were essentially 00:20:39.019 --> 00:20:41.400 driven out of the company. Their issue was with 00:20:41.400 --> 00:20:44.079 some mandatory workplace training courses that 00:20:44.079 --> 00:20:47.519 were being taught by a group called Applied Scholastics, 00:20:47.740 --> 00:20:50.490 an organization. that has associations with the 00:20:50.490 --> 00:20:52.650 Church of Scientology, which had been hired by 00:20:52.650 --> 00:20:56.049 AMAT. It's a culture clash. A big one. It's just 00:20:56.049 --> 00:20:58.009 a moment of corporate turbulence that reminds 00:20:58.009 --> 00:20:59.910 us that even these highly focused technology 00:20:59.910 --> 00:21:02.650 companies can sometimes struggle to manage their 00:21:02.650 --> 00:21:05.490 internal culture, especially during periods of 00:21:05.490 --> 00:21:09.329 rapid expansion. Okay, that detour aside, let's 00:21:09.329 --> 00:21:11.490 get back to the core strategy that defined their 00:21:11.490 --> 00:21:15.490 growth. By the mid 1990s, the company was using 00:21:15.490 --> 00:21:18.369 its capital to strategically fill in any capability 00:21:18.369 --> 00:21:21.150 gaps they had. Yes, specifically around that 00:21:21.150 --> 00:21:23.130 idea of yield enhancement, which we discussed 00:21:23.130 --> 00:21:26.150 earlier. That 1996 acquisition of the two Israeli 00:21:26.150 --> 00:21:28.890 companies, Opel and Orbot, seems like a textbook 00:21:28.890 --> 00:21:31.829 case. It is. They weren't just buying more deposition 00:21:31.829 --> 00:21:33.990 tools, they were buying control over the whole 00:21:33.990 --> 00:21:36.930 process. Oprah provided high -speed metrology 00:21:36.930 --> 00:21:39.789 systems, and Orbot provided inspection tools 00:21:39.789 --> 00:21:42.390 for patterned wafers. These buys ensured that 00:21:42.390 --> 00:21:44.390 AMET could offer the full lifecycle solution, 00:21:44.710 --> 00:21:46.710 the tools to build the chip, and the tools to 00:21:46.710 --> 00:21:49.309 verify its quality. And that pattern just continued, 00:21:49.450 --> 00:21:51.609 right? It did. We see them purchase E -Tech systems 00:21:51.609 --> 00:21:54.970 in 2000, then Oramir semiconductor equipment 00:21:54.970 --> 00:21:58.829 in 2001 for about $21 million, which brought 00:21:58.829 --> 00:22:01.690 them new laser cleaning technologies. And then 00:22:01.690 --> 00:22:04.390 the really big one we mentioned earlier, Varian 00:22:04.390 --> 00:22:07.309 Semiconductor in 2011. And the Varian deal was 00:22:07.309 --> 00:22:09.750 transformative. Completely. Because it gave them 00:22:09.750 --> 00:22:13.309 that key ion implantation technology, it meant 00:22:13.309 --> 00:22:15.910 AMAC could now provide the tools for deposition, 00:22:16.170 --> 00:22:18.990 etch, polishing, inspection, and the fundamental 00:22:18.990 --> 00:22:22.130 electrical doping process. It solidified their 00:22:22.130 --> 00:22:24.289 status as a single -source solution provider 00:22:24.289 --> 00:22:27.529 for many of the core fab steps. Their motto became 00:22:27.529 --> 00:22:30.309 pretty clear. If we don't own the best tool in 00:22:30.309 --> 00:22:32.150 a critical segment, we'll just buy the company 00:22:32.150 --> 00:22:34.509 that does. Precisely. That relentless pursuit 00:22:34.509 --> 00:22:36.650 of capability brings us to the really dramatic 00:22:36.650 --> 00:22:38.730 failed mergers, which I think reveal the true 00:22:38.730 --> 00:22:41.869 regulatory boundaries of this business. The first 00:22:41.869 --> 00:22:44.410 major attempt was the planned merger with Japan's 00:22:44.410 --> 00:22:46.990 Tokyo Electron. This was a colossal attempt at 00:22:46.990 --> 00:22:48.849 market consolidation. It was announced in 2013. 00:22:49.210 --> 00:22:51.390 The combined company, which they were tentatively 00:22:51.390 --> 00:22:53.309 going to call Eteris, would have been valued 00:22:53.309 --> 00:22:56.890 at $29 billion. It would have created an undisputed 00:22:56.890 --> 00:22:59.829 number one global dominant supplier. Without 00:22:59.829 --> 00:23:02.140 a doubt. They were trying to become the Microsoft 00:23:02.140 --> 00:23:05.079 of the semiconductor equipment world. So why 00:23:05.079 --> 00:23:07.319 did the birth of the largest equipment supplier 00:23:07.319 --> 00:23:10.720 in history fail? Antitrust concerns, pure and 00:23:10.720 --> 00:23:13.279 simple. After nearly two years of regulatory 00:23:13.279 --> 00:23:15.920 reviews, the deal was officially scrapped in 00:23:15.920 --> 00:23:19.339 April 2015. What was the specific problem? Regulators 00:23:19.339 --> 00:23:22.279 across multiple jurisdictions, the U .S., Europe 00:23:22.279 --> 00:23:25.960 and Asia, all feared the insane level of dominance 00:23:25.960 --> 00:23:28.440 the merged entity would have in critical sectors 00:23:28.440 --> 00:23:31.680 like deposition and etch. They basically concluded 00:23:31.680 --> 00:23:33.980 that this level of concentration was unhealthy 00:23:33.980 --> 00:23:36.599 for competition and for innovation in what is 00:23:36.599 --> 00:23:39.079 a foundational global industry. So they learned 00:23:39.079 --> 00:23:40.720 their lesson, but they tried again a few years 00:23:40.720 --> 00:23:43.079 later with the more targeted purchase of Kokusai 00:23:43.079 --> 00:23:45.599 Electric. Right. The intention, which they announced 00:23:45.599 --> 00:23:48.819 in 2019, was to buy Kokusai Electric. a former 00:23:48.819 --> 00:23:52.980 Hitachi group member, for $2 .2 billion. Still 00:23:52.980 --> 00:23:55.640 a big deal, but not as massive as Tokyo Electron. 00:23:55.839 --> 00:23:58.200 Exactly. It was a critical acquisition aimed 00:23:58.200 --> 00:23:59.900 at strengthening their deposition portfolio, 00:24:00.140 --> 00:24:03.279 particularly in batch bonuses. But they ran into 00:24:03.279 --> 00:24:06.240 the exact same wall, only this time the blocker 00:24:06.240 --> 00:24:08.839 was explicitly geopolitical. And the sources 00:24:08.839 --> 00:24:11.400 are very clear on this. The deal was terminated 00:24:11.400 --> 00:24:15.630 in March 2021. specifically due to delays in 00:24:15.630 --> 00:24:17.730 receiving approval from China's regulator. And 00:24:17.730 --> 00:24:19.990 that is a huge signal about the shifting landscape, 00:24:20.190 --> 00:24:22.710 a massive signal. Well, think about the difference. 00:24:22.950 --> 00:24:26.569 In 2015, the Tokyo Electron deal failed because 00:24:26.569 --> 00:24:29.150 of traditional Western antitrust concerns about 00:24:29.150 --> 00:24:32.410 market competition. By 2021, the Kogusai deal 00:24:32.410 --> 00:24:34.789 failed because China, which had by then grown 00:24:34.789 --> 00:24:36.869 into the largest market for many of these companies, 00:24:37.069 --> 00:24:39.769 asserted its own regulatory power. They just 00:24:39.769 --> 00:24:41.829 delayed the approval until the transaction deadline 00:24:41.829 --> 00:24:44.289 expired and the deal collapse. So the foundational 00:24:44.289 --> 00:24:46.769 technology space had officially become a political 00:24:46.769 --> 00:24:50.069 football. It had. And that failure to execute 00:24:50.069 --> 00:24:53.029 their M &A strategy is a perfect segue into one 00:24:53.029 --> 00:24:55.470 of AMAT's most ambitious and ultimately most 00:24:55.470 --> 00:24:59.069 costly strategic diversifications, their foray 00:24:59.069 --> 00:25:01.509 into the solar industry. Right. The Energy and 00:25:01.509 --> 00:25:04.849 Environmental Solutions, or EES, sector. The 00:25:04.849 --> 00:25:07.450 solar chapter. This is a fascinating case study 00:25:07.450 --> 00:25:09.869 in trying to apply your core competence to a 00:25:09.869 --> 00:25:13.299 new market and just failing dramatically. Applied 00:25:13.299 --> 00:25:15.779 Materials truly believed that their world -leading 00:25:15.779 --> 00:25:18.480 expertise in high precision deposition and coating 00:25:18.480 --> 00:25:21.460 systems could be pivoted to solve the complexity 00:25:21.460 --> 00:25:24.700 of large -scale solar panel manufacturing. They 00:25:24.700 --> 00:25:27.339 entered the solar market around 2006 after acquiring 00:25:27.339 --> 00:25:30.019 Applied Films, and they quickly ramped up their 00:25:30.019 --> 00:25:32.059 commitment. Organizing everything under this 00:25:32.059 --> 00:25:35.259 new EES sector. Their entry was incredibly aggressive 00:25:35.259 --> 00:25:38.420 and very expensive. They spent heavily on acquisitions. 00:25:38.559 --> 00:25:41.740 Like what? Like the 2007 purchase of HCT shaping 00:25:41.740 --> 00:25:45.339 systems for $475 million, which brought them 00:25:45.339 --> 00:25:48.140 wafer sawing tools. Then in 2008, they bought 00:25:48.140 --> 00:25:51.539 Italy's Buccini for $330 million, which specialized 00:25:51.539 --> 00:25:54.339 in the metalization steps of making solar cells. 00:25:54.500 --> 00:25:56.519 You know, printing the conductive silver lines 00:25:56.519 --> 00:25:58.480 onto the cells. They were trying to build a full 00:25:58.480 --> 00:26:00.990 stack solar manufacturing offering. And the centerpiece 00:26:00.990 --> 00:26:02.809 of this whole ambition was something called the 00:26:02.809 --> 00:26:05.740 Applied SunFab line. The SunFab was a visionary 00:26:05.740 --> 00:26:09.099 concept announced in 2007. It aimed to apply 00:26:09.099 --> 00:26:13.019 silicon thin film layers to huge glass substrates, 00:26:13.200 --> 00:26:16.440 effectively creating photovoltaic modules using 00:26:16.440 --> 00:26:18.660 the same deposition techniques they use for displays. 00:26:19.019 --> 00:26:20.680 And it got some recognition, right? The technology 00:26:20.680 --> 00:26:23.559 was seen as legitimate. It was. The SunFab line 00:26:23.559 --> 00:26:26.160 was certified by the International Electrotechnical 00:26:26.160 --> 00:26:30.380 Commission, the IEC, in 2009. For a moment, they 00:26:30.380 --> 00:26:32.940 looked like industry leaders. They even topped 00:26:32.940 --> 00:26:35.740 VLSI. research's list of PV equipment suppliers 00:26:35.740 --> 00:26:39.380 in 2008, with sales of almost $800 million in 00:26:39.380 --> 00:26:41.559 that segment. They also went all in on the R 00:26:41.559 --> 00:26:44.440 &D, opening that massive solar technology center 00:26:44.440 --> 00:26:47.539 in Xi 'an, China in 2009. It was noted at the 00:26:47.539 --> 00:26:49.480 time as the world's largest commercial solar 00:26:49.480 --> 00:26:52.059 energy R &D facility. They were not messing around. 00:26:52.299 --> 00:26:55.460 So if the technology was recognized and the investment 00:26:55.460 --> 00:26:58.420 was massive, Why did the whole thin film bet 00:26:58.420 --> 00:27:01.019 collapse so spectacularly? The market just shifted 00:27:01.019 --> 00:27:03.740 way faster than AMAC could adapt. Their entire 00:27:03.740 --> 00:27:06.240 EES strategy was predicated on the belief that 00:27:06.240 --> 00:27:08.420 their thin film technology would become the low 00:27:08.420 --> 00:27:11.660 cost standard. But it didn't. No. Intense global 00:27:11.660 --> 00:27:14.440 competition, especially from Chinese manufacturers, 00:27:14.779 --> 00:27:17.420 drove down the cost of the competing technology 00:27:17.420 --> 00:27:20.779 crystalline silicon far faster than anyone expected. 00:27:21.099 --> 00:27:23.680 The economics for the SunFab line just fell apart. 00:27:23.940 --> 00:27:26.539 The cost to manufacture their thin film panels 00:27:26.539 --> 00:27:29.180 became completely unsustainable compared to the 00:27:29.180 --> 00:27:31.599 cheap crystalline silicon flooding the market. 00:27:32.109 --> 00:27:34.509 And that forced a rapid and, I imagine, a very 00:27:34.509 --> 00:27:38.170 painful strategic retreat. It was. By 2010, just 00:27:38.170 --> 00:27:41.369 three years after the SunFab's big launch, AMET 00:27:41.369 --> 00:27:43.069 announced they were abandoning the thin film 00:27:43.069 --> 00:27:45.369 market entirely and closing the whole division 00:27:45.369 --> 00:27:47.650 down. And they kept cutting from there. For years. 00:27:47.829 --> 00:27:50.430 In 2015, they left the solar wafer sawing and 00:27:50.430 --> 00:27:53.089 ion implantation businesses. The process was 00:27:53.089 --> 00:27:56.210 finally formalized in July 2016 when the EES 00:27:56.210 --> 00:27:58.490 sector ceased to be reported separately in their 00:27:58.490 --> 00:28:01.009 financials. It just got folded into corporate 00:28:01.009 --> 00:28:03.000 and... others. So what's the big lesson here? 00:28:03.079 --> 00:28:05.359 What that tells us is that while AMAB possesses 00:28:05.359 --> 00:28:07.720 this unparalleled expertise and deposition tools, 00:28:08.039 --> 00:28:10.119 they can't just pivot that success into a new, 00:28:10.220 --> 00:28:12.359 fundamentally different market without major 00:28:12.359 --> 00:28:14.880 consequences. It's a massive lesson in staying 00:28:14.880 --> 00:28:17.440 within your core competency, even when you see 00:28:17.440 --> 00:28:19.579 what look like shiny, adjacent opportunities. 00:28:20.359 --> 00:28:22.940 They were punished financially for straying outside 00:28:22.940 --> 00:28:25.559 the high margin, hyper complex world of integrated 00:28:25.559 --> 00:28:28.180 circuits and high end displays. That strategic 00:28:28.180 --> 00:28:30.859 focus that James C. Morgan demonstrated back 00:28:30.859 --> 00:28:34.339 in 1976, really is paramount to their success. 00:28:34.660 --> 00:28:36.880 It is. But when we look back at that core business, 00:28:37.160 --> 00:28:39.940 the resilience and growth are just undeniable. 00:28:40.349 --> 00:28:42.829 And that brings us to the financial context of 00:28:42.829 --> 00:28:45.170 their current geopolitical battles. The financial 00:28:45.170 --> 00:28:47.930 trajectory is staggering. It just reflects the 00:28:47.930 --> 00:28:50.309 ever -increasing global demand for digital infrastructure. 00:28:50.670 --> 00:28:53.190 If you look at their overall revenue from 2005, 00:28:53.609 --> 00:28:56.349 it was around $7 billion. It vaulted to over 00:28:56.349 --> 00:29:00.049 $27 billion by 2024. And that growth isn't just 00:29:00.049 --> 00:29:02.710 volume, it's complexity. Each new generation 00:29:02.710 --> 00:29:04.990 of chips requires more specialized, more expensive 00:29:04.990 --> 00:29:07.250 equipment from AMET, which just boosts their 00:29:07.250 --> 00:29:09.980 top and bottom lines. In fiscal year 2020, was 00:29:09.980 --> 00:29:12.619 a banner year. It really illustrates that explosive 00:29:12.619 --> 00:29:15.799 post -pandemic demand for chips. A huge year. 00:29:16.079 --> 00:29:19.599 Revenue hit $23 billion, which was a massive 00:29:19.599 --> 00:29:23.240 34 % increase over the previous year. And earnings 00:29:23.240 --> 00:29:27.000 soared to almost $6 billion. This kind of consistent 00:29:27.000 --> 00:29:29.279 financial performance is why they're always highly 00:29:29.279 --> 00:29:31.839 ranked, like being named one of Fortune's world's 00:29:31.839 --> 00:29:34.259 most admired companies in 2018. So they were 00:29:34.259 --> 00:29:36.200 riding the absolute crest of the wave. They were. 00:29:36.339 --> 00:29:38.880 High demand combined with their irreplaceable 00:29:54.160 --> 00:29:56.960 Okay, this is it. This is where the story of 00:29:56.960 --> 00:29:59.279 the quiet equipment maker slams directly into 00:29:59.279 --> 00:30:01.680 the headlines of national security. Let's start 00:30:01.680 --> 00:30:05.279 with the market reaction in August of 2025. That 00:30:05.279 --> 00:30:07.579 event was a really stark reminder that even the 00:30:07.579 --> 00:30:09.779 most indispensable companies are not immune to 00:30:09.779 --> 00:30:12.920 political risk. In August 2025, applied material 00:30:12.920 --> 00:30:16.500 shares tumbled 12 % in a single day. It was after 00:30:16.500 --> 00:30:19.200 the company issued unexpectedly weak sales and 00:30:19.200 --> 00:30:21.799 profit forecasts, and they didn't hide the reason. 00:30:22.000 --> 00:30:25.299 They explicitly cited U .S.-China trade tensions 00:30:25.299 --> 00:30:28.180 as weighing heavily on both demand and critically 00:30:28.180 --> 00:30:31.980 on their visibility. Visibility is the key operational 00:30:31.980 --> 00:30:34.559 term here, isn't it? It's not just that sales 00:30:34.559 --> 00:30:37.079 are slowing down. It's the uncertainty. Exactly. 00:30:37.180 --> 00:30:39.900 If you are selling a multimillion dollar piece 00:30:39.900 --> 00:30:42.299 of equipment that is subject to immediate export 00:30:42.299 --> 00:30:45.720 control changes, your customer, a fa - in China 00:30:45.720 --> 00:30:48.799 or South Korea or Taiwan cannot confidently place 00:30:48.799 --> 00:30:50.819 an order. Because a license could be revoked 00:30:50.819 --> 00:30:53.759 overnight. Or a whole category of equipment could 00:30:53.759 --> 00:30:55.920 suddenly be sanctioned. Right. And your forecasting 00:30:55.920 --> 00:30:58.519 becomes impossible. The entire capital expenditure 00:30:58.519 --> 00:31:01.779 cycle for a fab operates on five to 10 year horizons. 00:31:02.099 --> 00:31:04.599 Geopolitical risk completely scrambles that planning. 00:31:04.720 --> 00:31:07.059 If the U .S. government restricts the export 00:31:07.059 --> 00:31:10.619 of your advanced ALD or CMP tools, the client 00:31:10.619 --> 00:31:12.619 might just cancel the purchase entirely rather 00:31:12.619 --> 00:31:14.980 than risk buying a multimillion dollar paperweight. 00:31:15.160 --> 00:31:17.400 And that uncertainty is amplified by the ongoing 00:31:17.400 --> 00:31:20.640 and incredibly serious legal issue that surfaced 00:31:20.640 --> 00:31:27.380 in 2020. Yes. This is no longer just about trade 00:31:27.380 --> 00:31:31.099 tariffs or weak forecasts. This is about alleged 00:31:31.099 --> 00:31:34.240 criminal circumvention of U .S. sanctions. This 00:31:34.240 --> 00:31:36.359 investigation raises the stakes dramatically. 00:31:37.680 --> 00:31:39.759 AMAC was reported to be under criminal investigation 00:31:39.759 --> 00:31:42.319 for allegedly routing specialized semiconductor 00:31:42.319 --> 00:31:45.480 equipment to a very specific Chinese company. 00:31:45.880 --> 00:31:48.119 Semiconductor Manufacturing International Corporation, 00:31:48.440 --> 00:31:51.859 or SMIC. SMICAC. They allegedly did this via 00:31:51.859 --> 00:31:54.619 intermediaries in South Korea to get around U 00:31:54.619 --> 00:31:57.039 .S. export controls. So for our listeners, we 00:31:57.039 --> 00:31:59.000 need to really understand the significance of 00:31:59.000 --> 00:32:01.980 SMIC. They aren't just any company. No. AMAC 00:32:01.980 --> 00:32:04.700 is China's largest and most technologically advanced 00:32:04.700 --> 00:32:07.180 state -backed foundry. It's a national champion. 00:32:07.579 --> 00:32:10.160 For national security purposes, the U .S. government 00:32:10.160 --> 00:32:12.380 has placed significant restrictions on the transfer 00:32:12.380 --> 00:32:14.839 of any equipment that would allow SMIC to advance 00:32:14.839 --> 00:32:17.000 to more cutting -edge nodes. We're talking about 00:32:17.000 --> 00:32:19.819 roughly 14 nanometers and below. Yes, the kinds 00:32:19.819 --> 00:32:21.740 of nodes that are crucial for advanced military 00:32:21.740 --> 00:32:23.839 applications and high -performance computing. 00:32:24.079 --> 00:32:26.839 So the U .S. policy is designed to create a technological 00:32:26.839 --> 00:32:29.859 ceiling, preventing SMIC from getting the most... 00:32:30.079 --> 00:32:32.940 advanced tools, particularly the deposition and 00:32:32.940 --> 00:32:35.839 etch tools where AMAT dominates. That's the policy. 00:32:36.059 --> 00:32:38.680 So if AMAT allegedly routed equipment through 00:32:38.680 --> 00:32:41.460 South Korea, how would that even work? Well, 00:32:41.519 --> 00:32:43.160 it generally suggests a pretty sophisticated 00:32:43.160 --> 00:32:46.240 attempt to get around the export controls. This 00:32:46.240 --> 00:32:48.920 could involve things like mislabeling the ultimate 00:32:48.920 --> 00:32:52.200 end user on shipping documents, routing the equipment 00:32:52.200 --> 00:32:54.579 through a subsidiary or even a shell company 00:32:54.579 --> 00:32:57.119 in South Korea that acts as a front. Or maybe 00:32:57.119 --> 00:32:59.819 selling older generation tools to a non -sanctioned 00:32:59.819 --> 00:33:02.599 company, knowing those tools would then be illegally 00:33:02.599 --> 00:33:05.359 transferred or maybe even upgraded to perform 00:33:05.359 --> 00:33:07.700 advanced functions once they reached CIMIC. All 00:33:07.700 --> 00:33:10.200 of those are possibilities. The core of the allegation 00:33:10.200 --> 00:33:12.519 is that AMAP may have knowingly facilitated the 00:33:12.519 --> 00:33:14.940 transfer of sensitive control technology to a 00:33:14.940 --> 00:33:17.380 designated company. entity on the U .S. entity 00:33:17.380 --> 00:33:20.019 list. The gravity of this just cannot be overstated. 00:33:20.099 --> 00:33:22.079 We're talking about potentially violating U .S. 00:33:22.079 --> 00:33:24.700 national security sanctions. What are the potential 00:33:24.700 --> 00:33:27.319 consequences if these allegations are proven 00:33:27.319 --> 00:33:30.079 to be true? The implications are massive. I mean, 00:33:30.079 --> 00:33:32.900 beyond potentially facing enormous criminal fines 00:33:32.900 --> 00:33:35.420 that would dwarf any past settlements, AMINT 00:33:35.420 --> 00:33:37.599 could face crippling new restrictions on its 00:33:37.599 --> 00:33:40.180 export licenses. Which could impact their ability 00:33:40.180 --> 00:33:42.259 to sell even non -sanctioned equipment globally. 00:33:42.839 --> 00:33:45.440 Right. Or, in the most extreme case, you could 00:33:45.440 --> 00:33:48.920 even see executives facing prosecution. This 00:33:48.920 --> 00:33:51.259 investigation confirms that this equipment manufacturer 00:33:51.259 --> 00:33:53.920 is no longer just a commercial tech company. 00:33:54.119 --> 00:33:56.859 It is operating directly on the front lines of 00:33:56.859 --> 00:33:59.779 technology and national security policy. And 00:33:59.779 --> 00:34:02.759 those two major failures, the failed Kokusai 00:34:02.759 --> 00:34:05.180 merger in 2021 because of Chinese regulatory 00:34:05.180 --> 00:34:08.820 obstruction, and now the 2023 DOJ investigation 00:34:08.820 --> 00:34:12.239 into circumventing U .S. controls, they seem... 00:34:12.460 --> 00:34:15.679 inextricably linked. They're caught in this geopolitical 00:34:15.679 --> 00:34:17.519 crossfire. They are caught in what you could 00:34:17.519 --> 00:34:20.099 call the technology squeeze. They are simultaneously 00:34:20.099 --> 00:34:22.820 too globally essential for either side to ignore, 00:34:22.980 --> 00:34:25.719 yet too politically sensitive to operate freely. 00:34:25.980 --> 00:34:28.300 China blocked their acquisition attempts, flexing 00:34:28.300 --> 00:34:30.480 its power as a massive market. And the U .S. 00:34:30.480 --> 00:34:32.380 government is investigating their sales practices, 00:34:32.659 --> 00:34:34.980 flexing its power as the regulator of foundational 00:34:34.980 --> 00:34:37.519 American technology. So their management team 00:34:37.900 --> 00:34:40.440 led by Gary Dickerson as the CEO and Bryce Hill 00:34:40.440 --> 00:34:43.940 as the CFO. They are navigating. what is arguably 00:34:43.940 --> 00:34:46.239 the most complex and risk -laden environment 00:34:46.239 --> 00:34:49.159 this specialized equipment sector has ever faced. 00:34:49.300 --> 00:34:52.059 It requires them to constantly manage the intense 00:34:52.059 --> 00:34:54.780 technological demands of their clients while 00:34:54.780 --> 00:34:57.179 ensuring compliance with a constantly shifting 00:34:57.179 --> 00:35:00.500 matrix of export policies. And these policies 00:35:00.500 --> 00:35:02.500 are designed not just for commercial fairness, 00:35:02.599 --> 00:35:06.000 but for strategic competitive advantage on a 00:35:06.000 --> 00:35:08.400 global stage. It requires not just technical 00:35:08.400 --> 00:35:11.079 prowess, but incredible political agility. And 00:35:11.079 --> 00:35:13.260 it forces them to make really tough decisions. 00:35:13.519 --> 00:35:15.980 Where do you invest your R &D dollars? Who can 00:35:15.980 --> 00:35:18.960 you legally and safely sell to? And what level 00:35:18.960 --> 00:35:20.880 of risk are you willing to accept as these two 00:35:20.880 --> 00:35:23.039 global powers continue to fracture the global 00:35:23.039 --> 00:35:25.599 supply chain? So let's try to synthesize everything 00:35:25.599 --> 00:35:28.260 we've uncovered about applied materials. We've 00:35:28.260 --> 00:35:30.840 traced their path from a young company that was 00:35:30.840 --> 00:35:32.960 saved by James C. Morgan's strategic refocus 00:35:32.960 --> 00:35:35.860 back in 1976 through the technological triumph 00:35:35.860 --> 00:35:38.440 of the Precision 5000 and their strategic attempts 00:35:38.440 --> 00:35:40.219 to dominate the market through acquisitions. 00:35:40.300 --> 00:35:42.980 And we define their technical dominance, that 00:35:42.980 --> 00:35:45.440 semiconductor product sector, which provides 00:35:45.440 --> 00:35:49.440 the irreplaceable tools for ALD, CMP, and ion 00:35:49.440 --> 00:35:52.179 implantation that enable the 3D structures of 00:35:52.179 --> 00:35:54.780 every modern chip. We also examine the stability 00:35:54.780 --> 00:35:57.159 provided by their Applied Global Services division 00:35:57.159 --> 00:35:59.760 and the strategic failure of their ambitious, 00:35:59.960 --> 00:36:02.500 but ultimately misplaced, IVIT into the solar 00:36:02.500 --> 00:36:04.800 energy market. The essential takeaway here is 00:36:04.800 --> 00:36:07.360 that AMAT remains absolutely foundational to 00:36:07.360 --> 00:36:10.239 the entire electronics ecosystem. They are the 00:36:10.239 --> 00:36:12.719 toolmaker that defines the very limits of what 00:36:12.719 --> 00:36:14.980 can be fabricated in modern chips and displays. 00:36:15.340 --> 00:36:18.260 But that foundational role is precisely what 00:36:18.260 --> 00:36:21.010 makes them a central player. and a target in 00:36:21.010 --> 00:36:23.269 today's high -stakes geopolitical competition. 00:36:23.570 --> 00:36:25.630 And this raises a really important question for 00:36:25.630 --> 00:36:27.829 you, the listener, to think about further. We 00:36:27.829 --> 00:36:29.570 know that specialized equipment manufacturers 00:36:29.570 --> 00:36:32.329 like Applied Materials are indispensable to global 00:36:32.329 --> 00:36:34.769 chip manufacturing, yet they are increasingly 00:36:34.769 --> 00:36:37.250 becoming tools of statecraft. We see it with 00:36:37.250 --> 00:36:40.630 the failed Cocoside deal and the 2023 DOJ investigation. 00:36:41.130 --> 00:36:42.849 So what does this concentration of foundational 00:36:42.849 --> 00:36:45.389 manufacturing technology really mean for the 00:36:45.389 --> 00:36:48.030 future? Right. If true technological progress 00:36:48.030 --> 00:36:50.619 is held hostage, by a handful of machine suppliers 00:36:50.619 --> 00:36:52.940 who have to operate under the increasing pressure 00:36:52.940 --> 00:36:55.760 of national security restrictions can global 00:36:55.760 --> 00:36:58.019 technological independence truly be achieved 00:36:58.019 --> 00:37:01.079 or are we just entering a new era where all advanced 00:37:01.079 --> 00:37:03.780 technology is subject to the geopolitical whims 00:37:03.780 --> 00:37:06.260 of the flu who make the machines it's a profound 00:37:06.260 --> 00:37:07.840 question about the fundamental vulnerability 00:37:07.840 --> 00:37:09.519 of our entire digital age