Dwarkesh Patel: uncertainty

Transcript context

…That’s completely compatible, if you think about it. TSMC and the entire ecosystem have something like 250 to 300 EUV tools already. Then you stack on 70 this year, 80 next year, growing to 100 by 2030. You’re at 700 EUV tools by the end of the decade. 700 EUV tools, at three and a half tools per gigawatt—assuming it’s all allocated to AI, which it’s not—gets you to 200 gigawatts worth of AI chips for the data centers to deploy. Sam wants 52 gigawatts a year. He’s only taking 25% share then. Obviously, there’s some share given to mobile and PC, assuming we’re even allowed to have consumer goods still and we don’t get priced out of them. But roughly, he’s saying 25% market share of the total chips fabbed. That’s very reasonable given that this year alone, I think he’s going to have access to 25% of the Blackwell GPUs that are deployed. It’s not that crazy. When did ASML start shipping EUV tools, when 7 nm started? I don’t know when that was exactly. You’re saying in 2030, they’re going to be using machines that initially were shipped in 2020. So for ten years, you’re using the same most important machine in this most technologically advanced industry in the world? I find that surprising. ASML’s been shipping EUV tools now for roughly a decade, but it only entered mass volume production around 2020. The tool’s not the same. Back then, the tools were even lower throughput. There are various specifications around them called overlay. I was mentioning you’re stacking layers on top of each other. You’ll do some EUV, you’ll do a bunch of different process steps—depositing stuff, etching stuff, cleaning the wafer—dozens of those steps before you do another EUV layer. There’s a spec called overlay, which is: you did all this work, you drew these lines on the wafer, now I want to draw these dots. Let’s say I want to draw these dots to connect these lines of metal to holes, and then the next layer up is another set of lines going perpendicular, so now you’re connecting wires going perpendicular to each other. You have to be able to land them on top of each other. It’s called overlay. Overlay is a spec that’s been improved rapidly by ASML. Wafer throughput has been improved rapidly by ASML. The price of the tool has gone up, but not as much as the capabilities of the tool. Initially, the EUV tools were $150 million. Over time, they’re now $400 million as I look out to 2028. But the capabilities of the tools have more than doubled as well, especially on throughput and overlay accuracy, which is the ability to accurately align the subsequent passes on top of each other even though you do tons of steps between. ASML is improving super rapidly. It’s also noteworthy to say that ASML is maybe one of the most generous companies in the world. They have this linchpin thing. No one has anything competitive. Maybe China will have some EUV by the end of the decade, but no one else has anything even close to EUV, and yet they haven’t taken price and margins up like crazy. You go ask some other folks that we talk to all the time, like Leopold, and they’re like, “Let’s have the price go up.” Because they can. The margin is there. You can take the margin. Nvidia takes the margin. Memory players are taking the margin. But ASML has never raised the price more than they’ve increased the capability of the tool. In a sense, they’ve always provided net benefit to their customers. It’s not that the tool is stagnant, it’s just that these tools are old. Yes, you can upgrade them some, and the new tools are coming. For simplicity’s sake, we’re ignoring the advances in overlay or throughput per tool for this podcast.…

Stored transcript either side of the excerpt. The highlighted words are the published quote; the surrounding text is unedited source, never generated.