State of the Semiconductor Business – 2022

Well, I'll tell you, I I see the demand increasing in the semiconductor market moving forward. You know, mainly due to the fact that we're seeing semiconductors used in so many places. You know, in the past, you were thinking about semiconductors in the world of personal computers, enablement smartphones came along, but but but now, I mean, we're seeing them everywhere from your sensors for the Internet of things, to, of course, automobiles, high end gaming, personal computers, so they're everywhere. So Just in general, the overall demand for semiconductors is increasing, just for our daily consumption for daily use. And so, to be honest, it's truly exciting from where we sit to be, to be part of that. Yeah. I would hazard a guess that The demand has accelerated faster than expected because of COVID because everybody was remote. Everybody wanted connectivity Tevony, everybody wanted bandwidth, everybody wanted the widget. And so COVID's basically steepened the steepened the ramp. You know, it people wanted it now, you know, it was maybe five years down the road. So there's this big rush to do it now. And so COVID's driven this. And then with that, because there's more devices, there's more data, so there's more data to be crunched. And so it's a cycle that's not gonna end. We can solve that information, edge computing, how you collect that data, how you do something meaningful with it, then monetize it. So Yeah. I agree with Kelvin, but I think COVID has accelerated this whole new industrial revolution, if you will, for for electronic Yeah. I kinda concur. I mean, I don't know. Angus, as you said, a pandemic is truly, like, kind of changed everything. Right? The semiconductor in my opinion is is, you know, semiconductor industry has benefited greatly. From folks meeting to work from home. I was reading this article the other day that said, you know, since twenty twenty. Each year, a laptop, you know, computer demand has risen by fifty percent. So and as you said, Kelvin, I mean, you know, lucky for us, semiconductors are becoming the fabric of people's everyday lives. So I agree, I think it's gonna continue to to move in a positive way. Yeah. Just to put some some numbers to it, if I recall at two thousand nineteen, the industry at a top level was actually a a down year for us, and we were you know, expecting twenty twenty to, you know, be kinda flat or, you know, maybe even down a little bit. But, you know, to Angus's point when the pandemic hit and the work from home and you needed cheaters, you needed you know, the kids needed the gaming systems, and all of that it really it really escalated it. And we saw an increase in the industry. And then last year, in twenty twenty one, there was another increase in, you know, twenty five percent year over year growth. So You know, there there's it's just continuing and to see that growth. Or even this year, there's an anticipation of about ten percent growth. At the top level for the industry. And so I think all these factors combined is is just leading to to what we're seeing with an increase in demand for for chips. I think the funny thing is that the automotive companies, you know, that's the big headline. What happened with your car? I think they saw the pandemic and they backed off on orders because they thought sales were gonna go down, and it went the other way. And so they canceled orders, and non automotive companies use that capacity. So they couldn't go back and say, oh, no, I want it back now. And so they they got caught short, and they had a not as a sophisticated supply chain for semiconductors as other companies do, and I think what happened was They backed off. Somebody else bought the capacity, and when they come back, the capacity wasn't there. So, you know, General Motors and these companies, they're in a crunch, Danler's in a crunch. A Volkswagen. And so, you know, while that's only about six percent of the market, as a growing market, it kinda caught the automotive industry flat footed. So they'll be proactive today, and and and they're working on solutions to to make sure that doesn't happen again. Alright. What do you guys think about, you know, this this you know, the automotive industry is is is lots of articles out there about them? Taking production and semiconductor chip production inside our internal. You know, Angus, in particular, I know you've got an experience in the in the fabs, and it's not as simple as just plugging it in and pushing a button. Right? There's still a lot of workmanship there and I have a feeling you know, there's a lot of they don't know what they don't know yet. And I don't know how how do you feel about that? In general meeting? Well, you know, General Motors specifically announced and and I think maybe it's been misconstrued they're now collaborating with foundries. They're designing with a TSMC, a chipset for their automotive content. In the past, they had maybe three hundred devices that they would buy from multiple places. Now they're trying to streamline that at and get custom devices, they're not going to go back. GM used to have Delphi in Kokomo, Indiana making sensors. Think it out of the business. It's difficult for an automotive company to own away for fab because you need so much volume. So what we know and as an example, GM just signed a deal with Qualcomm who is fabulous, but they've got their new snapdragon that's a three chipset for autonomous driving. So They're just being they're being proactive like I said before. They're thinking a little bit further ahead. They're trying to manage the content in their vehicle. You know, right now, forty percent of the cost to build a car is in electronics. By two thousand and thirty, forty five percent So these guys have to manage that because cost is, you know, on cost and profitability. So they're going to get smart about it. They're going to build instead of having A big circuit board with a hundred chips on it will build me a device that's multifunctional and does this, and I can program it. And so that's where that's going to go. Yeah. You know, I guess it makes sense. I mean, they're trying to control what they can control, you know, when there is a shortage and and they're looking to their suppliers to get you know, chips that are needed for their cars, you know, there's there's not much they can do at the time and weren't able to even though the fabs and the founders were adding more capacity, they weren't able to cut in line and get what they want. So now I think that makes sense where they're trying to control what they can control and design specifically with the TSMC. And ensure that they have the capacity needed so that it can do what they wanna do. So it makes sense. And you know, probably to your point, I mean, is is that, you know, they may not have the skill set yet, but I'm sure they'll, you know, try and get the right people and do what they need to do. But we'll just kinda see how it works as we move forward. Yeah. You know, there is a wealth of automotive suppliers. There's TI, there's On Sammy, there's ST, there's Bosch. You know, a lot of that's done in in Europe, And so you're right. They're just gonna have to be smarter about it, and they are being proactive. You know, GM announced seven billion dollars for a new factory in the US, there's probably all going to be electric vehicles, a lot more content. So they're going to customize their offering from a semiconductor standpoint. They can more functional across platforms forms. So instead of buying, I need a hundred of these. I need two of these. And and and it's it's managing their supply chain. So, you know, it's it probably have a higher value, but there's a way to do this that I think makes sense for the auto manufacturers to be more involved with their customers with the founders so that they get something that basically is customized for them. And so, you know, what's interesting, you talk about, you know, the shortage, the automotive shortage, and obviously, the increased demand in the pandemic. And one thing we're seeing is so much new capacity coming on board or at least being announced that's coming on board, you know, across the world. You know, a lot of it right now is in is in Asia. Most of the manufacturing and and expansion is in is in Asia. But we're seeing so much here in the States as far as new announcements for fabs, for, you know, Intel, DSMC, Samsung, TI, Micron, all the guys, global foundries, all the guys in the United States are are talking about increasing their their their their manufacturing. So really exciting. But I guess I'll ask, you know, Angus, Paul, You you have any thoughts on that as far as just what's going on here in the United States? You know, I I'm I'm super excited about that. Personally. You know, I've been working in manufacturing, you know, sales capacity for a long time. And typically, what used to happen is you develop prototype and get it you know, get your your r and d technology incorporated, but it was always shipped off seats to overseas, to manufacture First of West Mexico, necessarily for semiconductors. First was Mexico and then it transitioned to a Southeast Asia and I don't know. If if you look back at when America really prospered is when we were manufacturing, you know, we were very strong. And, you know, with Intel announcing, you know, a twenty five billion dollar investment in Arizona for a new facility and another facility in Ohio, I think, like, twenty billion dollars in TSMC coming here, and that just means a lot of a lot of our next generation is gonna be studying more I think we're gonna be producing more engineers and just manufacturing more, and and I I just feel like it's a really good time for the economy in general. Yeah. I think to some degree, it's driven also by some geopolitical pressure, you know, from the Far East. Also, if you look at the US, you know, so Samsung's already here, they're gonna build a new factory. TSMs, coming to town. Intel stumbled a few years ago, they had a wafer fab, five forty two. This is all let's hold off, and then they ran out of capacity. So opened up Fab forty two. But if you look at the US as a nation, and for critical infrastructure, for telecom, For communications, for defense, we don't have a strong technical partner, and I think Intel is gonna fill that void. They're gonna build capacity the two factories they're building in Arizona are gonna be foundries. They're gonna take, you know, they were a secure partner for for the US government and for the big, you know, military industrial complex. So I think there's a desperate need for that in this country because of the geopolitical situation. But they're getting into the foundry game, and they're gonna be able to offer an integrated solution, a microprocessor, a memory, and some logic with their new packaging technologies. I think they're gonna be a viable player against TSMC and Samsung. They're not gonna be the same size because these guys are way ahead of the curve. But I think Intel has some tricks up that sleeve from a technology standpoint. They're just announced that they're buying some staffers from ASML, and they're looking at two nanometer, which is quite a leap for them. But I think that It's kind of like a sleeping giant, if you will. I think we might be surprised, and I hope they're less successful because it's It's exciting just like Paul said. You know, it's good to see some of them on our home turf. Yeah. And then, you know, the advanced node manufacturing, you know, that's always been in Asia, you know, to bring, I think TSMCs bring a five nanometer when they come, you know, although which would probably be a couple of years when they can actually manufacture and like you said, intelligent looking at two nanometer already. So so just having the advanced nanometer advanced technology here as well. Where, you know, in the past it's been more, you know, two hundred millimeter at some lower level. So that's really exciting. I think like, you know, Paul mentioned earlier about just the talent and what's what we're gonna see here. Based upon that. That that's really exciting as well. So I I think it's really, you know, exciting time not only for the industry, but specifically even here, you know, in the United States with what's going on. With the expansions. So, you know, I think one of the biggest things is, you know, no, people are saying let's go, let's build. Historically TSMC, if you use them sort of like the gold standard because they've built a lot of capacity. They can go for mud, what meaning they've got permits that got the land, they're signed off to silicon in about thirty months. So they can start breaking ground, get to silicon in thirty months. But today, There's not enough steel to do frames. The supply chain from semiconductor capital equipment is running as fast as possible. So While Intel and TSMC are building in Arizona, you know, and they're pushing as hard as they can, that thirty month from mud to silicon, as I would call it, is going to be tough, because the supply chain, in general, not just for semiconductor, for construction, for insulation, for concrete, is pretty tight right now. So that's gonna be another interesting dynamic to see how that plays out. And, yeah, that's interesting as far as how long it takes for a fab kinda to come up to speed, especially with all the announcements that that that we're hearing. So, you know, we know there's a shortage. Automotive has talked about, you know, when can we get over this? I've looked at a ton of articles that talk about Is it gonna be the second half of twenty two? Is it gonna be the beginning of twenty three? And it's really hard to know, you know, when we're kinda kinda get out of this you know, what we consider a a shortage situation. But I do think with the amount of fives that are coming on board, here in United States being announced and also in in Asia that, you know, we are bringing on a lot of capacity into the market. So I think that's a good thing where we should be able to, you know, maybe kinda get over the hurdle, you know, in the next year or so. At least that's kinda what I'm what I'm hearing. But as far as, you know, moving forward, it's also there to support the, you know, number of devices that we talked about, you know, with semiconductor being everywhere. So I think the capacity that we're seeing being added into the market is just vital for us to kinda get over that that next hump to be able to support what we're seeing with the growth. Yeah. I don't I agree with you. I don't think we're building and going to have an overcapacity, you know, like a flattening of the ramp of the demand, because, yeah, we've got this. Today, we've got the automotive problem, and that's for internal combustion engines. But just to say, GM's building a new EV factory. The content goes up exponentially. So it's just not control systems, it's power of control, so Industry environment, global warming's causing changes that aren't going to go away, and so we're having to adjust those automotive go to adjust. The demand is going to continue to go up with high performance computing, cloud computing, data centers, the metaverse from, you know, what you know, it's it's it's not a sort of Facebook only thing, but AI and AR consume massive amounts of data bandwidth and computing power. So that's not going to go away because this working remotely also comes with advancements in and how people can improve manufacturing. So it's all gonna tie together, and I don't think it's got to slow down. It's like like when Henry Forbes started the car, look where we are today, I think we're right back there. We're at a massive ramp And it's just going I'm gonna say it's exponential, but I don't think I don't see this slowing down in the next ten years at all. Now I I concur with both, you know, what you guys said. And, you know, in general, companies, you know, will tend to be pretty conservative in terms of investing right, to expand capacity for instance. But I mean, the last two years for the semiconductor industry has been, you know, just crazy in terms of growth. And with recorded sales, instead of forecast, I think that really does give companies the confidence to actually make those investments quickly before they're left out. Right? So there's a race to support this growth, overall industry growth, and I think that's really gonna facilitate it hopefully shorten, you know, when these when these fabs come online and and that like you said, Angus and and Kelvin, Seems like there's always something coming in behind, you know, last two years, laptop was driving it. Five g is just starting to come on. So there's always something behind it to to facilitate the growth. So, yeah, I think I think the market's in a good place. Yeah. I mean, the industry, if I, you know, will get one of the I think it's a WSTS article world semiconductor trade statistics, mentioned that You know, the the expectations get to a six hundred billion dollar market this year at twenty twenty two. I wanna say the market was in the, you know, four hundred billion level, you know, maybe about, you know, six eight years ago. So, you know, serious serious growth. And then looking at getting to a trillion dollars by twenty thirty. And that's just the expectation when you look at just a, you know, you know, fairy conservative CAGR, you get to a trillion dollars by by twenty thirty. So again, just speaks of, you know, not only the importance, but the demand that's out there for for the chips. And I just know, you know, we've all been in this industry for, you know, decades and You know, we've seen in the past the cyclical nature of it. And as you see these types of gross we've seen recently, you're you're really expecting, you know, kind of a downturn on the other end. All of a sudden the forecasts are showing a little bit of a depth. Maybe it's a slowing growth Nothing like we've seen before. And I kinda think the cyclicality of it, you know, we're not gonna see like we used to. And there was one one article I saw, and they mentioned that a lot of cyclicality you've seen over the last ten years is really due to, you know, memory and the the changing in ASP is just due to supply and demand. So it's really not a unit issue as much as it's a change in in the dollar value of of memory that has caused some some cyclicality on on the revenue win. So I I do think that just the industry itself is so much more healthier than it used to be based upon a lot of the drivers that we we talked about. No. I've agreed and you know, from a semiconductor capital equipment side, if you look at the four or five companies that buy ninety percent of the CapEx, now you know, it used to be a hundred it's like the automotive industry. There was a hundred car companies now there's four or five. And that's what's happened to semiconductor. So But TSAMC orders, they know what they're ordering. It's not like they're ordering and competing against somebody else. So the market's condensed, four or five big guys, it's much smaller. And so they're forecasting is much more accurate. They're not competing against another guy, winning business, so Oh, I could've win that, let's cancel the order. So it's become much more stable. And and the capital equipment manufacturers are much more sophisticated because they're they're at the tail when it when that changes when it when it changes and they cancel orders, it's it's it's it's deadly for for companies like Applied to Lem. So They're very careful how they do it now and how they place an order. So that cyclicality has damped them significantly from OEM and an IDM standpoint to where, you know, it's pretty stable. I think it's it's much more predictable. And, yeah, I like help, and I'm feeling good about it. Yep. Definitely agree. I think The peaks and valleys are are gonna be a lot more friendly.