SPEAKER_00 0:00

The main idea of making masks and making transistors smaller came about because this guy, I think he was working for TI, came about uh and he was looking at a microscope and he was like, wait a minute. If I can flip the microscope over and shine light from the other side, it gets smaller.

SPEAKER_01 0:23

Hello everyone, and welcome to another semi-dope podcast. I'm Austin Lines with Chipstrat, and with me is Vic Shaker from Vic's newsletter. Hey Vic, what's going on, man?

SPEAKER_00 0:32

Yeah, I don't, not that much. Other than like everybody's panicked about uh the 13F from that situational awareness hedge fund, and uh there was this sell-off in optics, and everybody freaked out. I'm like, what's what's going on? First of all, I I still don't really know what a 13F is. I guess it's like a hedge fund has to report some holdings that you know they hold and what they sell and what they bought. I think something like that.

SPEAKER_01 0:59

Yeah, yeah, yeah.

SPEAKER_00 1:00

And uh Leopold Ashenbrenner, is that how you pronounce his name? Sorry if I got that wrong. I have no idea. It's too many syllables. Um yeah, I mean, he he I guess he made a lot of money and converted like a few million into a billion or something, and then now everybody's following him for the latest stock tips. And when everybody looks at his 13F, they panic, and there's this whole sell-off. Apparently, he shot optics and short AMD Intel and he sold in. I don't know. Yeah, so it's like everything is down in the semi-market because of some some hedge fund thingy. I it's this is what's happening. It's funny.

SPEAKER_01 1:37

I mean, so if if I'm him, I wake up and look and go, wow, I can buy a few puts because on this 13F, it doesn't talk, I don't think it talked about like the size of his puts or anything. So he can just buy a few puts at some point in time, drive the price down. Could he use that to then buy the price? I mean, that's market manipulation, but on the other hand, it's like people are trading on public data and on vibes, you know. So just like he could throw people off his trail, right? Like just buy a little tiny amount of puts in like everything, and you wouldn't know what he's investing in.

SPEAKER_00 2:08

Yeah, I don't know. I think Leopold listens to this podcast. If you do, just let us know. Maybe you can come explain to us what a 13F is, because I don't know about you, man. I have no idea what all this stuff is. So I'm happy to learn from the best.

SPEAKER_01 2:23

Yes, Leopold, you're welcome anytime. All right. So today we're gonna talk lithography. So I thought it'd be really interesting to talk about the economic challenges of lithography, modern EUV lithography, especially, um, because you know, ultimately, incentives drive outcomes, and there are challenges with the increasing cost of lithography, the increasing cost of fabs. And you you start to see, you know, TSMC uh can afford the next process node, and and Intel and Samsung are trying to be stay in the race or be in the race. Um there aren't that many other competitors. And from afar, if you're like a semi-tourist, as they like to say on X, um, you might say, like, hey, TSMC is crushing it, you know, why aren't more people in this game? And of course, as you start to unpack it, you realize that um costs are a barrier to let people enter. And so then you start pulling on the thread and you ask, well, what cost? And one of those major contributors to the capex needed to participate is modern EUV. And so, um, yeah, what do you think? Should we talk EUV?

SPEAKER_00 3:35

Yeah, that's that's a great topic for today. I think it ties in tangentially to what is happening uh to the market today and why people think that we are not in an unconstrained uh trajectory upwards. And this basically stems from the recent Gavin Baker interview that I put in in one of our daily semi-doped uh daily updates on Substack. So if anybody is not subscribed to the Substack, I recommend it. It's free. You know, you can go get subscribed on that, uh, semidope.com. But the idea that Gavin said uh you know floated in that interview was we are not in a bubble because TSMC is basically holding back the entire industry by not creating enough chips for everybody, by controlling their level of capex spend on tools that require um, you know, DUV and EUV. And most famously, I think TSMC is not pro-EUV. They think the tools are too expensive and that they want to stay on DUV with multi-patterning as far as they can manage to do so. And we'll talk about what DUV with multi-patterning means in this episode. But this is where we are right now. Tools cost a lot. I think an EUV machine costs something to the tune of uh $400 million, and that's just one machine, and you need to run many of them to produce the chips at scale that we need, and that's that scale is continuously increasing. So it really comes down to is EUV the only way forward? Um, and then there is uh Hyper Nae UV, which is like extremely amazing. I don't even know if those are in production yet. I don't think so. But those are gonna cost close to a billion dollars. That's insane.

SPEAKER_01 5:33

Yes, totally. So we'll yes, we'll unpack all this for everyone when you're like, what's DUV? What's UV? Why does it cost so much? That's what this episode's all about. Lithography, masterclass, hopefully, to set the stage for hopefully even future conversations around lithography, some startups that are out there. And actually, you know, I just saw ASML shared a roadmap. Um, I think it was IMEC had a conference. It's going on like right now, the ITF World Conference. And ASML showed some of their roadmap and how they think they're gonna be able to bring the cost per exposure down. And it would be nice to unpack this, but before we get to all of those nitty-gritty details, we want to educate our listeners on some of the some of the fundamentals. So let me throw out some numbers. So uh a brief I want to set one thing.

SPEAKER_00 6:20

I want to say one thing for Austin before Austin. So most of this stuff, Austin is actually the expert because he um actually spent time working on uh this kind of lithography stuff in a clean room, which is something I can't say I have done. Uh I've measured wafers and measured transistors, uh, that kind of stuff. So I've handled completed silicon wafers, but I've never actually done lithography myself. And you have. So this is a this is a cool thing. You can tell us a little bit more rather than just what you can read on the internet, perhaps. It's it's interesting what it looks like inside a clean room. I'm excited to hear.

SPEAKER_01 6:57

Yes, yes. So, well, thank you for mentioning that. So, yes, when I was in grad school, uh, I worked at the University of Illinois or Bana Champaign in doing research as a research assistant, making graphene-based transistors. So I was in a lab with Professor Eric Popp, he's at Stanford now, and he studies 2D materials. So we were studying our people on our team were studying carbon nanotubes, and then I got to work in graphene. And so, yes, I got to be in a clean room with you know silicon wafers, um, etching, you know, depositing, doing lithography. And in fact, I even got to do e-beam lithography because we were trying to make you know very precise, kind of one-off little uh nanoelectronic systems. And it didn't matter that the throughput of e-beam is really low, which we're not gonna talk about e-beam lithography much here, but it's just a cool thing that I got to experience. And in fact, when I first started, we were mechanically exfoliating graphene, which means we were taking um basically tape, and mechanically we get graphite, like layers of pencil lead essentially, and then we take like literally like clear tape, um, and then stick it on top of the graphite and then pull it off, and then take it under a microscope and look, and you can kind of like figure out how many layers of graphene you have there. And so I'd just scan around, you know, I'm I'm like doing some of this stuff in a bunny suit and then going back into the lab and moving this thing around, you know, and just thinking like, oh, what am I doing with my life? I'm trying to find single layer graphene, that's what I'm doing. But of course, um, once you find single layer graphene, then we would make trans, we'd pattern the transistors right there on it and then take measurements and publish it because it was a hard thing to do. And so we kind of had an advantage that we knew how to do it. Um, I won't I won't talk too much more. We eventually did chemical chemical vapor deposition um to create graphene using like uh copper foil essentially, and then we could grow graphene, and it wasn't as pristine and pure, but it was more scalable. And anyway, uh good memories. I spent a lot of time.

SPEAKER_00 9:02

Yeah, it's amazing stuff. I I bet it was a great experience doing all this stuff. At the time, it's probably frustrating as all hell as all research is, but then looking back at it, it's like, ah, I don't mind doing that again now. At least like going into a clean room and like pottering around would be nice. Totally, man.

SPEAKER_01 9:18

It was like it was challenging in that you had to, you couldn't just do the research, you had to do a lot of exploration to figure out how to even make the thing before then you could actually measure it and do the research. Um, and so yes, it was frustrating at the time because you'd spend all this time a whole day and then you get to the end and be like, oh, that didn't work, you know. But yes, looking, looking back now, it's like, oh, how fun was that? How much intellectual freedom there to just be like, you know, your research professor says, Hey, go make this thing. And you're like, Well, I guess I gotta read and experiment and try to figure out how to make it and then measure it, and then see if our hypothesis was correct, you know?

SPEAKER_00 9:53

Yes, yes, yeah. It's nice. Okay. Now we're good, now we're good. Now, you know, we hit carbon nanotubes already, like research that's never made it out of the lab. You spend time doing that. Awesome. Now let's talk about stuff we can actually make.

SPEAKER_01 10:06

Yes. Okay, so let's talk about. Okay, so um, have you heard of Rocks Law, by the way?

SPEAKER_00 10:12

I have not.

SPEAKER_01 10:13

Okay, so so uh this is a you know, a quote unquote law, just like Moore's law. It's it's really an observation, um, named after Arthur Rock, and he was uh an early investor in Intel, maybe um kind of one of the founders of the venture capital industry, if you will. But um, the the observation was that the cost of a semiconductor fab doubles every four years. So, you know, it gets more and more expensive. So now this is interesting. This is related to Moore's law, where Moore's law said that the number of transistors in an integrated circuit doubles every two years. And so if the cost of a fab doubles every four years, that's slower than the number of transistors doubling every two years. And so what that means is that Moore's law is really like an economic statement that says for roughly the same price, we can get more transistors per dye area, per chip, per unit area over time. And that's good, that's really what drove the industry was that economic realization that for the same dollars, you could get more compute over time. We are getting to the point where the cost of fabs is getting so high that that, you know, we've seen decades ago really, Moore's law, the number of transistors doubling, shrinking for physical reasons, but also from an economic perspective, actually the cost of transistors flatlining and potentially even coming up the cost per transistor. And again, the question is well, what's driving that fundamental change to some of these important economic scaling laws that we've seen for decades? And lithography is a big contributor. Like you said, low NA EUV tools used to cost around 250 million. Um, now high NA is coming out around 400 million. And give or take, you know, these numbers change over time, and it probably depends per customer. Um, but you can back some of this stuff up out of uh ASML's reports because they only sell uh you know they don't sell many of these per year. But anyway, yes, the rumored hyper NA, which would be a future tool, could cost anywhere from 600 to 800 million. Potentially, if the Strait of Hormuz stays closed, maybe a billion dollars because that's driving everything up. Yep. Um, so which means the lithography, all of these tools, like you said, and we'll get into it. Um, you know, you might need 15 tools to open up a new fab. So I think I'd seen some CNBC coverage where it uh it said that Intel's 18A fab in Arizona, Fab 52, needed 15 EU V machines. So imagine that, you know, half a billion dollars a pop and you need 15 of them. That is no joke. And that's why uh a brand new fab costs on the order of 20, 30 billion dollars. So I just wanted to quick paint the picture of the enormous CapEx cost to just build one new fab to try to stay in this race. And of course, that ultimately means the cost per wafer will have to go up too, because you only can do so many wafer starts per month, say tens of thousands of wafer starts per month or a hundred thousand, and you need to amortize the cost of all those tools over that fixed number of wafers.

SPEAKER_00 13:26

Okay, cool. So let me back up a little bit and you know quickly frame it uh in a way that you know I usually process things. So one of the biggest problems of modern lithography is cost, and that primarily stems from having to make smaller and smaller transistors. And we have gone from um deep ultraviolet, which is what we'll refer to as DUV, uh, to eventually extreme ultraviolet, which is what EUV stands for. And then we have within EUV, we have uh several levels of numerical aperture. I think we should define what that is, you know, going forward next, so that we actually have all these basic terms in place. So we have low numerical aperture or low Nae UV and high Na uh EUV and then hyper Nae UV. And the higher you go from low to high to hyper, you can make smaller and smaller transistors, right? Yes. So basically higher the numerical aperture, the smaller the transistors, the smaller the feature sizes you can make. And now what you need is regardless of whether you choose a deep ultraviolet or an extreme ultraviolet machine, you're gonna have to need like at least 10 or 20 of them in a fab. And if each of them costs like, I don't know, half a billion dollars, and you want to put in 10 of them, you're spending five billion dollars in just these EOV machines. And then not only that, you have to put a whole lot of other infrastructure like cooling, uh like clean rooms are difficult because you would have to have the vacuum the HVAC systems that pulls out all of the dust particles from the air. And so, based on the how many dust particles you can find per unit volume, these clean rooms have different classifications, right? So you've got class one, class ten, class hundred. Um, and I think the higher the number, you've got more particles per cubic volume of air. So all of this stuff takes like an enormous amount of money and time to build. And if you actually, this is like another point, if you look at like the TSMC's construction of their uh fabs, it's actually critical machines are suspended on pistons, like entire factory floors are suspended on pistons so that it's immune to like earthquakes and stuff. So it is very expensive to build a fab and takes uh years, it takes years. I it could take three to five years, uh, you know. So this is why we can't simply add chip capacity willy-nilly, right? Totally, totally.

SPEAKER_01 16:03

And not only that, if if you haven't seen a fab, I'd encourage people to figure out how to get on a fab tour if they can somehow. I know it's very difficult, but I've never been in one. Vic Vic wants to go. Um, I I actually had the good fortune recently of getting to go to Intel's uh Fab 52 and tour it. But a fab, you also need a so not only do you not want the machines to move around even a tiny bit from earthquakes, but even from like passing traffic and stuff, because of course we're making transistors on like the atomic scale, nanometer scale, and so you just you don't want all that any sort of mechanical wiggling and movement. But you also uh these EUV machines, they have like big power sources, and those uh to the in the light source, like a lot of that actually goes in the um subfab floor. So there's a floor beneath, like the main floor where the tool sits, and they have all this equipment underneath it. And then, of course, you know, there's like a floor above it, and you've got you've got to flow all that air through. And so I just also wanted to illustrate that it's not just like um like data centers where you're just like, oh, find some real estate, slap a building up, throw some racks in, and you're good. But yes, it takes uh even from a construction and HVAC perspective, building a fab is is no joke.

SPEAKER_00 17:22

And it goes beyond that because if you look at Intel fabs and the way they've built it in the past, they had this copy exactly method, which means they copy exactly. This is not something that they mess around with. They use like similar plumbing. I've heard that they even use the same brand of paint because they do not want anything to go wrong. Because if small things happen and change the way a fab functions, you can't get the yield up. And if you can't get the yield up after spending $20 billion, you can't make enough wafers, which means you can't sell them and make a profit. So Intel decided to copy exactly, and that actually slowed a lot of stuff down for them. That's a different story, but really, that's how difficult it is to build a fab. Really, I mean that's that's insane.

SPEAKER_01 18:09

Totally, totally. So, all right, going back, you mentioned DUV and EUV, and so let's tell listeners a little bit. So, back in the DUV days, the light source that was used um had uh a wavelength that eventually made its way to 193 nanometers. And I guess even zooming out even further, so lithography at the end of the day, for those who don't know, I think everyone probably does these days because ASML is an awesome company and everyone wants to invest or has invested, and so at a high level understands what lithography is. But we're talking about ultimately being able to expose light to sort of, and I put in quotes, uh draw, you know, the shape of transistors or the shape of areas that you want to etch away that you will leave parts of the transistor, but etch away other parts of the surface of the chip. And so ultimately, to make transistors smaller and smaller, um, you can either make the wavelength of ultimately you need to make the wavelength of light smaller and smaller. Um, but there's also something which we can get into, which is the numerical aperture, the mirrors that you talked about making on changing the numerical aperture, but just focusing on making the wavelength of the light smaller. You know, the canonical example here is like writing with a Sharpie, you're gonna draw like fat lines. And if you can write with a fine tip marker, a fine tip pen, you can ultimately make a lot thinner lines and you could draw smaller precision features. So that's what the industry was trying to go to from deep DUV, um, deep ultraviolet lithography, to EUV, which uses 13.5 nanometer light. So ultimately you're going in order of magnitude smaller from the fat marker down to the fine-tip pen.

SPEAKER_00 19:57

Yeah. So this whole relationship that you main you mentioned here, where you want a smaller wavelength of light, but you want a higher numerical aperture. This is governed by what uh is known as the Raleigh criterion, which means that the smallest dimension you can make on a wafer is literally proportional to the you know wavelength, but inversely proportional to the numerical aperture. There is also a constant factor here that's often called K1, which we won't get into here, but think of it as another knob which you can use by designing the masks that you know selectively allow light or don't allow light in regions. You know, they do all kinds of tricks on those masks to improve this proportionality factor K1. We won't get into it, but these are the factors. So there are some bunch of tricks, then there's the wavelength, and then there's a numerical aperture. So the smaller the wavelength you go, the better. And to Austin's point here, deep ultraviolet lithography was most uh famously you know ended at what is called uh argon fluoride lithography, is that right? At 193 nanometers. And then there was a like a quantum leap down to 13.5 nanometers with ultra EUV with extreme ultraviolet. So that's a big change. That's like more than a 10x change, right? And going down another 10x hasn't happened yet, but we will get to how that can happen at the end of this episode. But yes, so keep going. Let's let's go with it.

SPEAKER_01 21:35

Okay, yeah, let's let's let's talk about DUV for a second. Um, so do you want to explain like where the 193 nanometer light comes from with argon fluoride?

SPEAKER_00 21:47

Yeah, there was a whole uh there was a whole evolution to that as well. It's not like we we just landed up there right, you know, when we started lithography. Like in the 1980s, it was mostly like Like what was called eyeline lithography, which had a wavelength of like 365 nanometers. Then over the years, people realized like, wait, we've got to make this better. Um, and then they came up with um you know krypton fluoride uh lithography, KRF lithography that went to you know 248 nanometers. So just by changing the kind of uh light source that you're shining through and the wavelengths of the light source, you could get better features. So this was like going through the 90s, you could have like 248 nanometers. That evolved to like uh argon fluoride lithography, where they went to you know 193 nanometer, and that was pretty cool, but then uh that lasted all the way through the 2000s, let's say. And kind of they kind of ran out of light sources. They did try some other light sources uh along the way, but they didn't really like work out for various reasons. Um, and then they were kind of stuck with like argon fluoride for a while, but then like they thought about it and were like, how do we improve numerical aperture? Somehow we have to improve it. And the the answer was extreme, actually. It's amazing. If you come to think of the history of lithography, it's insane. Some smart guy came up with the idea and said, How about we put water on the wafer? Like, we let's just put water on it. Like, what do you mean you're gonna put water? So, yeah, that's literally what they did. They put extremely pure water on top of the mask and then put the light through the water onto the mask. And that came to be called like immersion lithography. So that actually helped scale transistors further by just like putting water on the wafer. It's like insane, right? So the the history of lithography is amazing. I wanted to tell you the way it all started, I don't know if you know this, but the way it all started was in the early days, uh I I forget the name of this guy, but you know, uh look out on semi-doped. We'll we'll have a poll post on this thing. The main idea of making masks and making transistors smaller came about because this guy, I think he was working for TI, came about uh and he was looking at a microscope, and he was like, wait a minute, if I can flip the microscope over and shine light from the other side, it gets smaller, right? Everybody knows you look the wrong way at these things, like stuff gets smaller. I'm like, this is it. Like I'm gonna turn the microscope upside down and shine light through the wrong end, and everything gets smaller. And that that's how all of lithography came about by turning the optics in lithography came about because this one guy had the idea to turn the microscope upside down. So that's how it all started, right? And then we've been continuously going down the path of these various laser materials, down to putting immersion lithography with water, and ultimately coming down to like extreme ultraviolet lithography, which is an engineering feat that is uh an achievement for like humankind. That's how big it is. We'll talk about it too, but yeah.

SPEAKER_01 24:58

Yes, yes. So, yeah, Vic, you make some interesting points here, which optical lithography, it's all about light sources and about the optics, about the mirrors, about how do you bend the light. And so when you talk about uh yeah, the guy having the insight of like, oh, when I look at a microscope, it makes small things seem bigger. So if I flip that lens, I could make big things seem smaller. What an amazing way to take a big mask and make it smaller to be patterned. And then ultimately, uh, you know, when you're talking about, you know, moving through um various materials uh and getting unlocking smaller wavelengths, we're talking about lasers. These are light sources to shine through the optics that we've been talking about. And ultimately, the industry was just playing with like, what are different materials that can lase at lower and shorter and shorter wavelengths? Um, and this actually leads me to, you know, we got to argon fluoride, 193 nanometer, um, and the industry sort of stuck for a while, waiting to figure out what's that next way of unlocking even lower, even shorter wavelengths, um, ultimately EUV as we know it today, which we'll get into. Um, but in the meantime, the industry came up with this nice trick called multi-patterning. And I thought I'd explain it really quickly because there's also economic trade-offs to multi-patterning. So, multi-patterning is ultimately about like the question is how do you draw smaller than the single wavelength features? How would, for example, um, here, let me let's let's come up with an analogy. Let's say you're drawing the lines on a football field, like an American football field, you know, the end zone, zero-yard line, 10 yard line, 20-yard line, so on and so forth. And maybe you have a machine that is like, I don't know, really fat and it can only draw a line every 10 yards, you know, the 10-yard line, 20 yard line. Well, then maybe the coach comes to you and says, Oh, hey, we also need uh markers at the five-yard line and the 15-yard line and the 25-yard line. And at first you're like, well, wait, my machine, it can only print them every 10 yards. Like, how am I gonna possibly do that? And then some clever person comes up and says, Oh, well, just draw 10, 20, 30, 40, and then go back to start and scooch it over five yards and draw five, 15, 25, 35. And ultimately, when you're it takes twice as many steps, but instead of having to get a new machine that can now print every five yards, zero, five, ten, fifteen, twenty, you just draw them every 10 yard space, and then you offset by five yards, and then you draw 15, 20, 20. So when you zoom out and you're done, you're like, wait a minute, now I've drawn lines every five yards, even though I didn't have to get a new machine. And that's like a very crude analogy for what's going on in multi-patterning, which is drawing features in step one that are only spaced at the distance that you can comfortably make, and then coming back in with another step and drawing a second set of features and just offsetting it. And so the amazing thing is with a trick like multi-patterning, you can unlock shorter dimensions between the drawn features, but of course, the uh economic cost to this is it takes twice as many steps or it decreases your throughput by half.

SPEAKER_00 28:21

Awesome. So I love the analogy, by the way. That's like a super cool way to understand it. Um so what you're saying is basically you can do uh a coarse etch, scooch it over, do a coarse etch again, and what you're left with is like a fine etch. Because you can now by scooching over somewhere in between the last two etches, you can get a you know finer spot, you know. And uh if I remember the right, this terminology is called litho h, litho etch. So you'll see this as L-E-L-E, right? Is this the same thing I'm talking about? Yes, exactly. You you nailed it. Okay, cool, cool. Now, I think that people have taken this to more than two levels of litho h, right? They've gone to like triple patterning and even quad patterning, which is all cool and all because now we you're stuck with two problems. One, it becomes increasingly difficult to even align masks between the yard lines. Like, okay, like when you had to align the mask at like the 15-yard line, it was okay, okay, whatever, it was between 10 and 20. But now you want to align it at, you know, 12, 14, 16, 18. And you're like, okay, that's the problem. The second problem is you're going to run through four different quad patterning steps, which each one takes the same time, so it's kind of scales linearly. And now it takes four times as much time to make that one lithography step. Um, and I'm not sure like how many levels this can be applied to quad patterning, but you know, making a transistor isn't like one etching step or one lithography step. There are many of them. And if you have to quad pattern on multiple steps, it adds up a whole lot of time. And the throughput decreases, which means uh the cost per transistor goes up, or you don't get enough amortization of your original 20 billion investment. And now we are at like a crossroads here.

SPEAKER_01 30:13

Yes, yes. And um, case in point, I know SMIC had to, which is the fab in China, um, they're not allowed to get EUV, and so they were able to take DUV and use tricks like quad patterning to get to you know seven nanometer class and then five nanometer class transistors, um, which I wanted to point out, by the way, because it's related to lithography. Um nowadays, when we're talking about making transistors, it's no longer just like two-dimensional transistors, but it's really three-dimensional transistors with thin fets that have these fins. We should find some pictures and you know, people go Google it, and ribbon fets. Um, and so now you've got these three-dimensional shapes. So it's also making a transistor actually takes on the order of like 60 or 70 or 80 steps because you have to pattern and etch and deposit material um kind of over and over and over to build up this 3D-shaped transistor. Um, so it's not only which, but but there's a kind of a marketing thing that um, like you know, the semiconductor tourists, for lack of a better word, which just means you're new to semis, it's no shade. We have a I was a semi-tourist at one point. You're welcome. Everybody's welcome into semiland. We love it. That's this podcast exists for you. Um very inclusive. Yeah, exactly. When a fab says we make two nanometer transistors or 1.8 nanometer transistors, it's not the smallest dimension, this you know, critical dimension, like we talked about before, the distance between any two really close lines is not two nanometers. It used to be, you know, like back when they were 90 nanometers uh and 180 nanometers and 45 nanometers, that was a lot closer, but it became a marketing term. And so actually, something that's called two nanometers, the smallest dimension may still be on the order of like 30 nanometers.

SPEAKER_00 32:07

Yeah, yeah. So it's not exactly two, but that's how we now call it because it's somehow the equivalent of two.

SPEAKER_01 32:16

Correct, correct, yeah, correct, right? It's like the equivalent of like when you think about like transistor density and whatnot. But I'll say it's important because you know, naturally, when we say 13.5 nanometer um EUV wavelength, someone might go, oh well, that's still way too big to draw two nanometer lines. But it's it's not exactly. So you might think then, oh, if we went from big fat marker DUV to fine-tip sharpie EUV, we must not have to multi-pattern anymore, right? And actually, your intuition is correct, but from a resolution perspective, we don't have to, but actually from a yield perspective, the industry um can still need to rely on some multi-patterning. Um, and there's a really nice graphic from Fred Chen's Substack. He wrote a nice article on it. We'll link to it in the show notes. But ultimately, we are getting so small that when you're shining very short wavelength light um at a certain dose, there's only really so many photons that are that are hitting there, and you can only control them so precisely, and you've got like resist chemist chemistry going on, and there might be some ideally there's not there might be some impurities or even dopants in the way, and so you end up getting like this stochastic nature. When you draw with the Sharpie, you don't actually get a very fine line, but if you zoom in, there's some little dots around the edges and stuff. Um, think of I don't know, maybe like spraying with a spray paint can or something. It's like not a perfect line, you know.

SPEAKER_00 33:49

I'm I'm looking at the picture and I was thinking of spray paint exactly. So if you didn't you always nail these analogies, and I was like, I'm gonna nail the spray paint analogy.

SPEAKER_01 33:56

I'm I stole it from you. I'm I'm sorry. Yeah, so ultimately, yeah, that's what they do is basically you might draw with the spray paint twice to get a better defined line, especially as you're starting to go in three dimensions. Um, so I just wanted to throw that in there to mention that yes, we now we've jumped up to these, you know, $300 million, $400 million EUV tools, but it the throughput isn't just immediately solved because there's still some multi-pattering that may have to happen. Um and there's other things about the power of the light source and the dose, but we won't get into those now because we're really starting to get into the weeds. Um but okay, what do you say we jump in? Should we talk about high NA next? Um, or do you have anything else to add here that's useful at a high level?

SPEAKER_00 34:45

I think that we should conclude before we talk about NA, we should talk about how we can generate light at 13.5 nanometers in UV. Because we mentioned that these were like laser light sources based on argon fluoride lasers. But um it's quite different when it comes down to 13.5 nanometer UV, and that is where the hardest uh innovation actually was um holding back the industry from going to this for a very long time. And fundamentally, what in a simple way, it's far more complex than I'm explaining it, but in a simplest way, it is basically tin droplets uh that you know fall through a chamber, and you hit it with um laser light and it gets activated, and then you hit it again with a laser light. Remember, you have to hit a falling tin droplet that's about 50 micron in size twice as it falls through this you know chamber, and the second time it gives you an explosion of 13.5 nanometer light. And that keeps happening precisely. Uh, ASML has an awesome video on their website where you can see these tin droplets falling. It's it's an animation, you can't really see this thing. But then these droplets are falling, and these like laser sources are like continuously hitting the droplets, and you see these explosions of EUV light. That is then it goes through like a mirroring, it goes through like 13 different mirrors because it has to be focused ultimately onto the wafer. And then ultimately it lands up on the wafer where it hits a mask and then it selectively exposes or doesn't expose stuff. But this whole power that this you went, you the one of the big problems is that you went through all this trouble to get extreme ultraviolet light by you know shooting thin lasers, but then you reflect it through so many mirrors, and at each reflection you lose some power. Like less, like a single digit percentage of the actual generated EUV power actually gets to the wafer. It's a big loss because of these mirrors. There's literally no way around it, or so we think. But yeah, that's what I wanted to talk about because now that we've finished talking about how lasers entirely work and how light sources work, numerical aperture is a good transition to get into right now.

SPEAKER_01 36:59

Yes, no, this is good. You I tried, I almost skipped over EUV entirely, uh, at least low N at EUV. So it's a good introduction, which is we were stuck at DUV, we tried multi-patterning. In the meantime, the industry was trying to work on EUV. And as Vic talked about, you know, ultimately we're trying to find a light source that has a much shorter wavelength. And, you know, there work had been done that show that showed with tin you could um basically induce a plasma, like that's why you hit it twice ultimately, and that plasma would generate 13.5 nanometer wavelength light. But there was a lot of engineering challenges and optics challenges around, okay, great. Yes, when we're under vacuum, we can generate a plasma and we and it will emit this really low wavelength light, short wavelength light. But how do we ultimately harvest all that light? How do we reflect it back and then use, you know, with mirrors, like aim it? Ultimately, you need to like gather this light because it's just gonna shoot in any direction, presumably from the tin droplets. And you need to gather it all, and then you need to like get it to where it needs to be, to where the mask is ultimately. And in in while you're doing that, you're trying to focus all the light. And like Vic said, there's a lot of losses. Every time light hits a mirror, it's not gonna all bounce perfectly exactly in the direction that you want. There's gonna be some scattering and some loss. And so then ultimately you end up losing so much light in the process that you don't have enough to like uh expose the photoresist. And so then the question is that the industry was working on for a long time is not only how do we just make all this work and repeatedly, but also how do we increase the light source so that we ultimately, by the time we harvest all this light, get it exactly where we need to, get it focused all the way down, we still have enough to actually expose the photoresist and draw the transistor. So that, of course, that's why we're stuck at DUV for a while, because this is an amazing engineering feat. And of course, it takes, read the book Focus by Martin something, I don't remember his last name, but it's about ASML. And what's really interesting is it talks about the entire supply chain and all the co-innovation needed. For example, famously from Zeiss with their mirrors. Um it's so it's it's no joke to even build the laser uh produced plasma light source, but then you have all the optics, and of course, there's something called a scanner. We won't talk about it a ton, but ultimately the mask, you're patterning, you don't want to just pattern one, you don't pattern like one die or one chip, you pattern a die on the chip, and then you like we talked about before, it's like a checkerboard pattern on a big dinner plate. You need to draw these transistors for every checkerboard square. So you need some like um mechanical, you know, mechatronics that ultimately move the everything around so that you can repeatedly print all of this. So there's a ton of engineering to make this even possible.

SPEAKER_00 39:53

Yeah, that's insane. That's what 39.5 nanometer EUV make. It's an incredible feat of engineering. And um we are here today because ASML took 20 years to develop this. And uh the the whole question of how come ASML landed up with this uh is another interesting question because uh this technology was actually developed in the United States. And at some point it was sold to ASML, and at that time the United States government didn't actually come in and say, no, this is critical technology, we want to hold it. You know, the US government has blocked many such things before, like including like protecting 5G technology. They've done all of this stuff. Even now, they have like there's so much export control. This was like before the day of export control. So we, as you know, from the United States have handed over the uh keys to the kingdom to ASML like a few decades ago. And to kudos to them, they spent like 20 years developing it. And there's an enormous supply chain that goes into ASML's uh machines that all need to come together to make this work. So it's it's built on a massive amount of effort. But I just wanted to point out that this was actually US technology at one point.

SPEAKER_01 41:14

Totally. Yeah, that's a good, a great history lesson there. Of course, we should write more about that history sometime. Um, okay, so we're running long, but let me blow through. So, okay, wow, it's an engineering marvel to get 13.5 nanometer light, but we want to make transistors smaller. What do we do? Okay, like we talked about with the Raleigh criterion, um, you ultimately have two big knobs that you can turn. One is the wavelength of light, but if you're like, dude, we spent so long to get here, we're not just gonna turn that all of a sudden. Just 13.5 was hard enough. The other um knob is the numerical aperture, which ultimately has to do with like the size of mirrors. Um, and so that's where we get into high NA and extreme, like extreme NA or whatever it was called, hyper NA. Um, but but but maybe really quick, the industry's trying to move from 0.33 numerical aperture in low NA to 0.55 in high NA, which makes features on the order of like one and a half, 1.7 times smaller possible. But there's a catch. There's always a catch in engineering, there's always trade-offs. You need bigger mirrors. When you have bigger mirrors, you've got these steeper light angles ultimately as they bounce in, and you have something called anamorphic optics that come into play. Um, and I won't get way in into how that works and what that means, other than to say you ultimately end up can you can only pattern an area that's like half the size of what you could with low NA. They call this the half field. So basically, now instead of your $250 million machine printing an area, you've got you know a $400 million machine printing half the area. Of course, that sounds horrible. Now you need, you're telling me I need, okay, Mr. Salesman, I just bought a $250 million machine from you, and now you say I need not only your $400 million machine, but I need two of them, right? That's crazy. So what um ASML has done a ton of amazing engineering where they've said, yes, we can only do a smaller size, but what if we speed up like the scanner and the mechatronics to go even faster to make up for it? So it's like, sure, we we keep this the area is gonna be smaller, but we're just gonna move that thing around the wafer even faster. And again, um ASML has all these amazing videos on YouTube where they show like how fast they're accelerating and moving this stuff, and it's crazy. It's like fighter jet style acceleration, but with nanometer precision, moving things perfectly around, stopping and reversing, like it's crazy that it all works. But again, things are expensive, there's more trade offs, there's a lot more innovation that needed to happen. And ultimately, even with the proposed hyperNA, even bigger mirrors, there's even more trade offs, um, even with stuff like photo resist. Um, so I'll probably just leave it at that and we won't do dive on high end. Or hyperNA, but just trying to illustrate that like not only are there economic challenges, but there's also just like engineering challenges, probably presumably reliability challenges. So then we'll leave you with this. The question is: well, instead of the mirrors, could we make the wavelength smaller? How could we make the wavelength smaller?

SPEAKER_00 44:20

Yeah, I want to add one more thing about the mirrors that's like an engineering challenge, but then we're going to go and talk about how to go even smaller wavelengths, right? These mirrors are not simple. You just, it seems like, what's the big deal going from low NA? You just have to make a bigger mirror. Make a bigger mirror. What's the problem? These are not ordinary mirrors because they are actually made up of multiple layers of 40, there's like 40 or 50 layers of alternating layers of very thin molybdenum and silicon layers. They are layered like this, and it is insanely smooth. And I read this book, Chip War by Chris Miller. It's a good book, I recommend it. It talks through a lot of history, and a lot of what I've said here is from that book. And I have a quote here from that book. It says, if the mirrors in the EUV system were scaled to the size of Germany, their biggest irregularities would be a tenth of a millimeter. Think about that. Think about how flat those mirrors are. Yeah. And we're going to put up a picture here, and you'll see like, you know, how how smooth it is. It is very difficult to even hold this thing. And I feel like I wouldn't even want to breathe on it, like at this level. I don't know. They probably have protection, protective gear. But making bigger mirrors isn't easy. It is an incredible engineering feat to make irregularities a tenth of a millimeter when the mirror scale is the size of Germany. That's that's really flat, right? So that's very smooth surface. So it's not simple that we can go from a 0.55 Na, which is like hyper Na, to like 0.75 like next year. You know, if we are used to like the incredible pace of AI. Everybody's like, oh, what's the big deal? Like we can go to like 3.2T, 6.4T, 12.8T networking, right? No problem. Like, when are we gonna get there? Like, you know, two years, three years, what's the time frame? No, no, this stuff is difficult. You cannot make a mirror that simply, that easily. So that's where we are right now. And now the question is, what's next? Like a machine costs a billion dollars, and now you tell me like this is only half field, and now I need two billion dollar machines. It's just like the economics is exploding. Something is going wrong. And so this is where we have new ideas to go where no human has gone before.

SPEAKER_01 46:45

Totally. So, okay, transitioning here, people will say you could never compete with ASML. It took the industry so long to figure out this 13.5 nanometer light, and they have a supply chain like they have a relationship with Zeiss, the only person in the world who can make these perfect mirrors. Why would Zeiss sell their mirrors to you, dumb startup? Of course they're not, because they don't want to um make ASML mad, right? And so now you're gonna have to go get another person to be the next AS or the next Zeiss, you're gonna be the next ASML. It's never gonna happen, right? And so some startups are saying, okay, hold the phone. Let's just like forget all that. Let's just think simple from um first principles. Could we get a smaller wavelength light? Um, how do we tackle the optics? How do we tackle the integration, the mechatronics, all that stuff? So one startup, X Lite, out of um California, and I think uh Pat Gelsinger is on their board maybe now, maybe he's the chairman of the board or something. Um what they're trying to do is they're saying, hey, what if we use free electron lasers as the light source? So we'll replace LPP, laser-produced plasma, that's the tin droplet, shooting it with the laser machine gun and all the magic that happens. Um, but what if we start by using this different light source that can ultimately, which by the way, a free electron laser, um, it essentially like, think of it as like accelerating electrons to like near light speed. You've got these undulators that like wiggle them, you can get this coherent light. It can ultimately scale down to like one nanometer, sub-1 nanometer. But what if we start by using this new technology but still producing 13.5 nanometer light so that it can plug in to existing ASML scanners and ASML optics? So, what if we could generate light in a new fashion? You know, by the way, FEL has a much higher total power source. So you can what you could ultimately do is have higher dose, which is better for yield. But actually, what X Lite's trying to do is say, what if we use um one free electron laser and we can actually split the beam and feed many EUV scanners? So they're ultimately trying to decouple the light source from the scanner. So what if you could buy um, you know, 10 scanners and feed it with one light source? Or obviously maybe they would have to have two light sources, one as a backup in case one doesn't work. Um, but you get the gist. And so that's the approach they're trying to take is say, hey, what if we build one massive free electron laser next to the fab and pipe the light into all of your ASML scanners? You can amortize the cost of your FEL across all those scanners. Um, and then ultimately it there will be some integration, but we're not gonna ask everyone to change, not only the optics, but the photo resists, and we're not gonna ask anyone else in the industry to change. We are just going to decouple the light source.

SPEAKER_00 49:42

That's fancy. Yeah. I haven't looked into X Lite, so I'm actually learning on the fly right now. That's amazing. One of the things that you can do with a laser source that has a higher output power is that tell me if I'm wrong, if you can get more light onto a wafer, the throughput actually increases, doesn't it? Not only yield, but the throughput grows up.

SPEAKER_01 50:03

Correct, correct, correct. The throughput, exactly, exactly. You it um, you know, whatever. If you only need like a small flashlight to shine on something and now you got really powerful light, you could get the same amount of light by actually taking your really powerful light and shining it for less long, exposing it for less long, right? You just like it.

SPEAKER_00 50:22

How many photons get in?

SPEAKER_01 50:23

Yeah. Exactly, exactly, exactly. There, so so therefore you can increase the throughput, but you could say, okay, well, hey, the yield maybe isn't that great at this the way the industry is doing it now. So we'll shine it for just a little bit longer than we need to, and you'll get even more extra photons, right? So you can, you know, have a higher dose, um, but ultimately have a so both better yield and better throughput.

SPEAKER_00 50:48

So who would be the end customer of uh X Lite? Would it be ESML?

SPEAKER_01 50:53

No, it would be the Fab. So the Fab would be buying, and then the crazy thing, and I wrote about it on Chipstrat, you can go check it out, is the business model here is ultimately selling light sort of like a utility, like photons as a service. So um you might ask, like, okay, well, if why would TSMC go build an FEL from some startup and then they'd have to go like re-jigger, you know, and work with ASML to say, we don't want your um LPP like light sources, we just want your scanner part. And like that seems like a lot of risk and a lot of effort for TSMC. But what if, and a lot of CapEx, by the way, what if um X Lite came in and they said, we will pay to build this utility right next to your fab, just like you get electricity delivered, just like you get water delivered, and even um just like you buy gas. So like uh these fabs, they will buy inputs like gas in sort of this consumption-based way. Let us build uh the FEL, the light source, and then um we will just charge you for what you consume. So it's on our books, we take the CapEx hit, and then we'll just charge you. So if you want to just spin up um three scanners, fine. We'll feed you three scanners. Now, of course, um X Lite wants to ultimately have you spin up as many scanners as possible, but X Lite, there's a way that X Lite can take a lot of the risk and uh do a lot of the upfront investment, and then they will just sell light to TSMC over time. And by the way, then once they build that relationship, not only could they sell you 13.5 nanometer light, but maybe for a premium later, once you and the industry are ready, they could sell you one nanometer light. So it's a very interesting business model.

SPEAKER_00 52:39

So the optics and stuff still comes from ASML, but then you've got this uh free electron laser sitting on premises in TSMC, just like supplying light. So exactly. You they count the number of photons you use and charge you for it? Is that the whole business model?

SPEAKER_01 52:55

Yep, presumably. How they how they do that, how they you know track how much light you're consuming, it would be also very interesting to know. But that's that's exactly it's like your electricity bill at the end of the month is gonna be your light bill, your light for lithography.

SPEAKER_00 53:10

Amazing. So, what what other ways are there to make uh one nanometer wavelength of light?

SPEAKER_01 53:15

All right, so one more that we'll hit on today. Substrate is another startup. They're also in California, in San Francisco, and um they are throwing out the playbook and also taking a different approach. And instead of FELs, they're saying, hey, what if we use X lay X-ray lithography? Historically, um X X-rays were generated by big synchrotrons, um, football stadium size, you know, particle accelerators essentially. Um, but there's actually precedent. And in those, again, you you speed up these particles, they get super high energy. Super high energy means really short wavelength, and you can ultimately control and use it as a light source. Um, there's actually the industry has actually explored using X-rays as a light source. And again, if you Google chipstrat substrate, you'll find this. I wrote about the history, but IBM did a ton of work. So, again, a lot of this early research happening in the United States, IBM did a ton of work here to see could this be a path forward for the industry? And they actually made um uh synchrotron or an X-ray light source um that fit on a truck. So it's a bit of a myth that it has to be massive. They figured out a way to make it a lot smaller. And this is the approach that Substrate's taking, which says, hey, ultimately I'd I'd phrase it this way. Hey, uh IBM and a bunch of other people back in the 80s and 90s, they explored X-ray lithography, and it was a working prototype. Um, it wasn't economical yet, but a lot has changed in 30 years. Uh uh, not only about light sources, but um with photoresists and with optics and everything else that it takes to build a light source and a scanner and do lithography. What if we went back and revisited from first principles and we took a stab at X-ray again and said, hey, given all that we've learned in the last 30 years, could it now be economically possible to do lithography using um X X-ray, you know, particle accelerator-based X-ray lithography?

SPEAKER_00 55:24

Yeah, I wanted to just step back one minute and just quickly explain what a synchrotron is. The idea of a synchrotron is that you accelerate a charged particle in a ring, right? In a circle or an ellipse or something like that. And as the charged particle that is continuously being accelerated turns around and changes angle, it spits out X-rays as it uh turns around. That's that's basically how a synchrotron works. And typically, in the past, you know, like you mentioned, they these are really big uh installations, particle accelerators tend to be really big, depending on what energy you have to accelerate them to. And uh I think they the invention for making tabletop uh synchrotrons has been around already like 20, 30 years. So it's not something that you really need uh a whole lot of like space to do. So that's one thing. So that's very important because people shouldn't be like, oh, what do you mean? You need a football field, we don't have that kind of space, so we can't do X-ray lithography. No, no. I think it I think it's it can be done in a smaller way. But the one thing that I uh learned when I wrote about this, it's on my sub stack too, about uh substrate and x-ray lithography, is that it's very difficult to like actually focus X-rays. So, you know, we spoke about the mirrors for EUV lithography, but you can't do that for like um X-rays because they go through things, you can't reflect them. That's a big problem. So the optics for X-rays is uh a challenge, it really is a challenge. So one of the ways that you can do is uh do this is that you have to do what is called proximity printing. Because, you know, like we mentioned uh earlier, the the the inverted microscope approach means that you could scale down a mask, you could put the mask on the big end of the microscope, and then the other end you know scales down, let's say five times. Uh that is called um you know uh reductive printing or something. I forgot the exact term. Basically, you can reduce the magnification factor by a factor of five because you've got this inverted uh microscope approach. But uh that by the way, it came to me, the person who did that was Jay Lethra. Um and so he's the guy who came up with this idea. It came to me later. But uh yeah, so you can't do that with X-rays because you there is no optics that works for this. So you have to do proximity printing, which means that you have got to make masks the same critical dimension as the stuff that you are patterning. So the masks are actually very fine. And so for this purpose, the mask making is significantly harder when you're using X-ray lithography because you don't have the optics for them. So there are a whole lot of challenges that require to be solved. So it's not just like, oh, we've got X-rays now that go to one nanometer. So just let's just just swap out uh, you know, the the LPP 13.5 nanometer source for a one nanometer X-ray, and then voila, you can like print, you know, 0.1 nanometer transistors gate all around or whatever it is. It doesn't work that way. So once you change the wavelength, everything changes. And that's where we are now. And there's this startup called Substrate that's working on this. They made quite a splash sometime back because they feel that not only can they make smaller transistors and continue to scale Moore's law, but X-ray lithography can be significantly cheaper, and that you don't need to spend uh one billion dollars for an EOV machine anymore. Which means that in most people now with far lesser capital investment, going back to the whole economics angle that we started with in this podcast, can make more fabs. And then if this technology is held within US soil this time and not given away, maybe all of manufacturing will come back to US soil if we can make X-rays work. And now we can own all of the supply chain uh you know required to make uh I don't think we can own the supply chain, but if we can at least make wafers on US soil and have so many fabs that we don't rely on anybody else, that would really propel the chipmaking industry like we have never seen before. So that is the case for making X-ray lithography on US soil.

SPEAKER_01 59:54

Totally, man. There's so many implications. We have to have a full episode on this. Hopefully, we'll talk to them. Because, first of all, it's good that you point out that there's lots of engineering that has to happen, not only with the light source, but with the optics. There's implications for the mask. How do you draw a mask at such small dimensions? Maybe it's e-beam, there's gonna be a cost to that, right? So there's lots of technical questions to get answered. But to your point, the implications are very profound. If, in fact, it can reduce, ultimately reduce the cost. Um, hey, could global foundries make two nanometer chips? Could Texas instruments, why not? So, what are the implications that I think it's super interesting of these legacy fabs, trailing edge fabs, now being able to make even smaller transistors at the cost of maybe their trailing edge nodes? Tons of implications. What does that mean for fabless design companies? Where, you know, you're like, well, yeah, maybe we'd make our own chip, but that's pretty expensive. And I don't know, probably we can't amortize $100,000 per wafer across what how many weights? We only need five wafers or something, but what if all of a sudden, again, it was you know the cost of a 90 nanometer chip that you can now buy wafers for $10,000 instead of $100,000, but get two nanometer you know transistors. Yeah, crazy implications. And then to your point, the geopolitical implications are fascinating too. So that's why I get excited.

SPEAKER_00 1:01:20

Ultimately, you know what will happen ultimately what Jerry Sanders said. Real men will have fabs again.

SPEAKER_01 1:01:25

Totally, totally. So that even that, like, why did every company at the start of semiconductors have a fab? Well, because ultimately, if you're vertically integrated, you're gonna get a better product. If you can co-design across the fabrication, across the design and the fabrication, if you can design, but also design for manufacturability all in the same house, you're just gonna go faster, you're gonna build a better product. But ultimately the cost, because of Rock's Law, got so big that people had to drop out because they couldn't afford a billion dollars for the next fab, $2 billion, $4 billion, $8 billion. Everyone has to drop out because, like the Global Foundries or the TIs, like they just can't, they don't have enough volume or high enough ASPs to amortize that cost. So it's just dropping off. But yes, in an ideal world, some of these players would still love to build, design, and build their own chips. And then, of course, from a wafer allocation perspective, you own your own destiny. Like, there's just so many amazing implications. So I know everyone gets super hung up on like the technology is impossible. Who dares think that they can take on ASML and Zeiss and all that crap? But I'm more excited about all the positive implications that will happen that will benefit all of us.

SPEAKER_00 1:02:41

If you've been watching this on YouTube, you'll notice that I've been drinking from this lens cup. So now that it's over, I guess our episode is two. We've spoken a lot about lithography. So let's let's get on with it.

SPEAKER_01 1:02:54

Totally. Okay, that's it for today. Everyone, thanks for listening. Thanks for hanging with us. Uh, we hope you're enjoying Semi Dopes. Please tell your friends about it, pass it along. If they want to learn about lithography, send this to them. Send us questions, comments on the YouTube, subscribe at semidope.com to our Substack that we've started. And thanks, as always, uh for joining us in this journey.