Comments by "Grak70" (@Grak70) on "Japan’s EUV Failure" video.

It’s moments like this that make me laugh when people claim laissez faire capitalism will always properly allocate resources. It may make progress in fits and starts, but it can also be obscenely wasteful.
72
ASML was and is huge in all previous UV step and repeat systems. They just have competition in that arena. There are ASML I-line steppers where I work that were built in the early 1990s with Zeiss optics stamped “made in West Germany”. ASML is famous now because they’re the only game in town for EUV and that makes them a strategic resource at a national security level. But they’ve been doing business in this space for well over 30 years.
33
I actually work directly for one of the former Intel guys who killed 157. 😆 it’s a fascinating story.
25
@ArawnOfAnnwn correct, but they also ship dozens of twinscan platforms a year at all wavelengths. Most of them are I-line or KrF. Characterizing ASML as a one trick pony with EUV is just not true.
23
Damn you are CRANKING the videos out. This is one I’ve been looking forward to.
14
EUV had already surpassed anything they hoped to achieve with proximity X-Ray and e-beam is direct write and was never going to meet throughput requirements.
10
@ArawnOfAnnwn ASML has had more market share in exposure tools than both Canon and Nikon (taken individually) since 2005. And it’s been beating Canon longer than that: the 1990s were mostly a duopoly between Nikon and ASML. You’re just making shit up at this point.
10
@akatsuki6371 you can definitely do 7nm and to some extent even 5nm nodes without EUV. The problem is your yield is crap and your cost is through the roof. What kazedcat is saying is you only use EUV on the layers you absolutely have to. Fabs still have a ton of I-line scanners because they can spit out like 250 wafers an hour with much lower unit cost and preventative maintenance cost if resolution isn’t an issue. And this is what they do; most lithography machines in a fab by count are still excimer KrF/ArF and I-line. Fab mantra is: “always use the cheapest thing that works.”
10
@ChimpyChamp if someone asked me to name a technology they thought came from aliens, it would be scanner stage positioning. You look at the requirements and can only conclude it shouldn’t work but it does. At a very simplified level, they use bands of closely spaced electromagnets to move the stage (the entire stage is levitating over a near perfectly smooth block of granite). The closed loop control uses an interferometer and powerful signal processing to feedback an electrical signal to those magnets that results in oscillation damping and acceleration control. The part of every stepper/scanner you don’t see is the rows of 2m tall control racks supporting it. They’re usually tucked away in the sub-fab, quietly and frantically sending out a flood of information to the control systems. It is truly mind boggling.
9
I think the issue with X-ray was you had to make the masks out of pure beryllium or something similarly crazy. The material science is wicked hard. And back when X-ray was being proposed, they were hoping it could do like 100nm half pitch. You can almost do that with a KrF system nowadays.
7
@LuaanTi I’m not reading all that but I’m happy for you.
6
Let’s see if they build a KrF or ArF scanner worth two farts first…
5
Imprint is cute but it can’t really overcome its throughput problem. Maybe a niche market but never gonna make consumer processors or dram with it.
4
Japan also backed the wrong source technology. DPP vs LPP was the VHS vs. Betamax of the EUV race.
4
@elchippe ok, it’s 6 years later. Where’s the production tool?
4
It’s kind of a limitation in the opposite direction. Japan is hyper protective of its cutting edge research and has been for decades. At my former job we were contacted by Selete about a product we made and they wouldn’t even let us come to their office for a sales meeting. Japan and the US may be defense partners and a macroeconomic powerhouse together, but Europe is by far easier to work with and has closer diplomatic ties that make basic research collaboration more feasible than Japan. It’s also not a coincidence that ASML had been eating the Japanese’s lunch on exposure tools since 2005 and had reached global sales parity with Nikon since the early 1990s. The DOE and US partners probably saw the writing on the wall and they were not wrong.
4
@ThePowerLover wow. Don’t see that every day on the Internet. Glad to be of help.
3
@akatsuki6371 glad to be of help!
3
It’s not like Japan and China don’t pour obscene amounts of cash into their domestic sub-industries that support semiconductors. And they’ve been doing it for decades. Especially Japan; you wanna talk about protectionism lol… At this point, sovereign financing is the only way you’re going to get anywhere in this industry.
3
@ThePowerLover ASML’s factories are in Netherlands, not Belgium. IMEC is in Belgium.
2
That’s a great sentiment provided you survive the failure and go on to succeed at something else. Plenty of people, companies, and governments don’t.
2
Comes from Greek ἀκτῖνος meaning “ray” or “beam”, so the relationship to light emission is the more proximate meaning. The actinides got their name for the majority having radioactive isotopes.
2
China can’t even make a working dry ArF scanner. 😆
2
“Necessity knows no law but to prevail.”
2
@user_375a82 shit! the West’s days are numbered!
1
@xBlackWind well, come back and let us know when this skunkworks project is for sale.
1
@xBlackWind SMEE has been talking about “soon” releasing their 28nm ArF tool for more than three years. Capability that is now OVER a decade old. My point was that the idea that some newcomer is going to leapfrog the best photo company China has is even less believable. I love the idea you think fabs are going to build an on site synchrotron. This is all pure propeller-head, nationalist rah-rah fiction.
1
@xBlackWind let’s resume this conversation when your Chinese CERN complex is built and cranking out chips. I suspect we won’t be talking again any time soon.
1
X-Ray mask technology didn’t ever really work and the resolution of x-ray was quickly overrun by traditional UV scanners in the early 2000s so it was always going to be behind the curve.
1
@gyzq in other words they will be 15 years behind forever, assuming they can even build a contemporary machine by then. High-NA EUV is a whole other ball of wax that you need low-NA experience to even attempt. Honestly the only way China is ever going to catch up in this space is if they invade Taiwan and steal one from a TSMC fab.
1
@thisiskevin1000 I will believe the SMEE DUV scanner works when I hear they’ve sold a couple dozen to SMIC.
1
@gyzq I’ll believe it when I see it. Right now, China can’t even put out a working dry ArF system. They are light years behind in EUV and the only way they’re going to catch up is the usual way: stealing IP from others who’ve already spent the money to learn the hard way.
1
Unrelated, other than how it limited capital allocation.
1
@xBlackWind what’s its overlay specification? What’s its throughput? SMEE “had” 90nm ArF 10 years ago with its 600/10 model, but it was trash and sold no units. Now, having sold few or no units of the “improved” 600/20 model, they are confidently going to release an immersion ArF tool to dominate the Chinese market…and yet their optics supplier estimated it wouldn’t even finish the factory to supply lens glass to SMEE before this year. Which means SMEE is hand crafting or Frankenstein-ing every tool they build, and likely rejecting numerous test articles in the process just to get one passing lens. This kind of development hell waste will play out in multiple system elements just to get a single tool, meaning SMEE is entirely reliant on sucking the government tit to survive. SMEE is a CCP and local government fund dumpster, even in more traditional lithography spaces. The idea that they’re on the verge of releasing a production EUV tool is laughable. The idea that EUV will net them a profit in China’s domestic market, even more so. China has consistently been about 10 years behind in lithography for decades. I see no reason for that to change. And given the unique challenges EUV poses, I think you will see the gap widen even further.
1
@didyoumissedmegobareatersk2204 you can have parity with other countries in one technology and still suck at something else, so no, that’s not relevant.
1
@leechjim8023 it’s ironic you used an EUV video to plop down this soap box because the basic research behind EUV was developed through a government agency and the IP is licensed to ASML through a US holding company. That’s why the US government can stop ASML from selling to China: they’d revoke their IP licenses.
1
@RandomPerson-vf3ld gee thanks Neil Degrasse Autism. If only you’d been around in 1929 all those people wouldn’t have thrown themselves out of windows.
1
Second this in general. KLA’s story just as a company is long and interesting.
1
@_Chad_ThunderCock “possible” is not necessarily “manufacturable”. If you work in a fab, you know this. I do.
1
Nanoimprint is a cute idea but it’s a dead end. Nobody’s going to buy hundreds of nanoimprint tools just to keep up with their backend scanners than can crank out 200+ wafers per hour.
1
@xBlackWind I’ll believe it when I see it. EUV is not something you can just rapidly catch up on. The high NA work being done now can only be built on a working knowledge of low NA technology. And even so, the same problems are going to plague any Chinese EUV scanner than plague their excimer systems: their overlay technology sucks. That’s leaving aside all the blank production, blank inspection, and pellicle technology ecosystem that had to evolve in parallel. I don’t believe this is an area where China can just play catch-up by stealing, reverse engineering, or scraping data.
1
@causewaykayak you’re welcome!
1
@DwAboutItManFr I really don’t think so. The jump in resolution is just not that much, even if you assumed x-Ray was fully developed already (if they need it in 15 years, they should have started five years ago and they haven’t, so smarter people than me probably think they don’t need to). It’s a very similar situation to what happened with immersion ArF and F2. Why go through the enormous pain in the ass of F2 excimer patterning when it’s really only a 1-2 node solution? And when you can develop 1 additional system component (the immersion apparatus) but keep the rest of the ArF infrastructure the same? Why go back and do massive amounts of basic science on synchrotrons, beryllium masks, and unknown-to-science x-Ray transparent materials when you can just slowly improve EUV? It’s less sexy for sure, but most progress in engineering is slow, incremental, and (to most of us anyway) boring. Making big investments in new physics is something companies want to avoid unless they absolutely must. That was the case with EUV: thankfully it worked. But so far there’s no “drop dead” problem clearly visible for EUV: just engineering challenges. Personally I think EUV is the end of the road for patterning tech: it will continue to get better, but never be supplanted by a next gen tech. Increasingly clever designs and integration are going to have to take up the slack from here. This also means that the Moore’s Law component of ever decreasing costs will die. Performance will come at an outrageous cost, so the era of consumer electronics rapid obsolescence and the cutthroat capitalism of fast, continuous improvement will end. Institutional bodies and governments will ironically once again be the only customers for truly state of the art computing, if indeed they need it at all. We left the era of single chip supercomputers behind long ago; networking of smaller compute units and AI optimization will rule from now on.
1
@DwAboutItManFr I specifically said as much.
1