Comments by "Grak70" (@Grak70) on "Asianometry" channel.

  1. The idea that someone once thought combining mustard gas and anthrax was a step forward boggles the mind.
    607
  2. I lived and worked in Japan from 2010-2015 and visited Sharp sometimes for customer interaction. One of the most depressing and bloated organizations I had to interact with. And for Japan, that’s saying something.
    554
  3. Whoever marketed this as drinkable is a goddamn genius. It tastes like someone set dirty dishwater on fire.
    311
  4. They will, but Americans are not going to live in the fab for peanuts. If TSMC intends to expand further into the US, they need to get on board with the US office/fab culture. The Arizona fab is already having problems with this: they don’t want to compete hard on salaries and expect you to die in your bunny suit.
    200
  5. Fun story: the first full field functioning scanner ever made for 157 was manufactured by SVG (later acquired by ASML for their catadioptric lens designs). At the time, it was thought the entire lens train, including the condenser optics, would need to be made from ultra-pure calcium fluoride. It was later discovered that fluorine doped quartz would do the trick for all but the few final lens elements. At the time, that one scanner contained the largest concentration of lithography suitable CaF2 in existence. The projection lens was literally worth its weight in gold.
    141
  6. I’m sure the discomfort with automation that would eliminate labor was not lost on the management with a resident Party member standing behind them at all times either.
    87
  7. It’s incredible that for as many people as Motorola employed throughout its existence, I’ve never once heard a positive story about working there.
    82
  8. 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.
    74
  9. Also, half the tools I work with on a daily basis are older than I am. And I’m over 40. Compound semiconductor makers almost exclusively inherited equipment from the 1970s-early 1990s silicon world because that’s all that’s available in many cases. What’s left are often 200mm tools down-converted to 150mm. With those kind of capital restrictions, an industry that charges cents per unit for its product and can fit 5,000-15,000 chips on one 6” wafer is never going to transition to 200 or 300mm.
    67
  10.  @devondetroit2529  living in Japan is awesome. Having to work there, especially in a corporate job, sucks.
    52
  11. You can just say “saw” and “baw”. Everyone else does. Source: work in this specific industry. Also 50M parts for $1.40 is a whopper untruth. I think you confused bulk pricing per unit with order pricing. A product run capable of covering an entire generation of smart phone wouldn’t cover the cost of a single photomask for said filter if $1.40 bought 50 million units.
    51
  12. Yup. I work in GaAs. Hard to make a Cz ingot when half the atoms in your melt sublime into vapor at atmospheric pressure.
    48
  13. 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.
    34
  14.  @roc7880 did you watch the video? This was done in the interwar period, well before nuclear weapons were a thing. And it was only 4 years after Hiroshima that the Soviets got the bomb so not sure what you’re on about other than trying to stir the pot.
    32
  15.  @ryandick9649  I agree completely. No lack of talent or good ideas, just smothered by excess bureaucracy. In a lot of ways, reminds me of the problems that plagued and ultimately killed Motorola.
    30
  16.  @Braydan789  yes, by design. Almost all of it is IP protected. And the process integration matters much more than raw capital capability. Most RF fabs are using equipment that was already considered old in the early 1990’s by silicon CMOS standards.
    30
  17.  @wernerhermann9120  jfc some people will start a pissing fight about literally anything.
    28
  18.  @Taygetea  a bit different today because demand is still ludicrously strong. To their credit however, this time most semi manufacturers resisted the siren call of overbuilding capacity. And now with interest rates the way they are, there’s no way more capacity that wasn’t already under construction is going to start.
    26
  19.  @PaulFisher ere are several sources of flare in DMD. 1) the material below the mirror itself does show through a bit since it has to be exposed for a mirror to be free of its adjacent companions. While this is typically coated in a material selected to be as black as possible over the desired wavelength range, it’s not perfect. So even if 100% of light directed at the mirror is diverted, there will be some reflected light from the substrate. 2) the mirror can only tilt so far. Ideally you’d direct all the light onto a light dump orthogonal to the exit pupil, but the mirrors can’t flex 90 degrees, only about 10-12. So some light will be scattered from the mirror edges and the anchor via structure, letting a little light through even in the static full off position. 3) ideally, your light source would have total spatial coherence (or equivalently, be a point source at infinity). Real light sources are only partially coherent and have finite size, so some photons impinge on the mirror at a small angle. When these are reflected in the off state, some small amount of light will still go through the exit pupil. 4) the light dump material is chosen to be as black as possible and actively cooled from behind. However in practice, a true black body does not exist, so there will be a tiny amount of reflection which will bounce around the DMD cavity. Some of this will escape through the pupil.
    25
  20. I actually work directly for one of the former Intel guys who killed 157. 😆 it’s a fascinating story.
    25
  21. In addition to the patent issues already mentioned, at least one of the two Japanese makers (I can’t remember which) bet the farm on DPP source tech rather than LPP. That was the wrong choice and they paid the price.
    25
  22. Yeah I laughed out loud at that. Based on that price, enough SAW to cover an entire generation of iPhone wouldn’t net enough profit to cover a single photomask.
    24
  23.  @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
  24.  @wernerhermann9120  Japan-chan is not going to sleep with you. Please find a creative hobby.
    22
  25. Micromirror and microaperture arrays are fascinating. One of the NIR spectrometers on JWST has a particularly cool one. And it’s the basis for all of Texas Instruments’ IP on DLP.
    22
  26. This is exactly correct. More cowboy engineering in action. The engineers who invented plasma etch used 13.56MHz because it combined RF generators that were commercially available off the shelf with license-free use.
    18
  27. This idea that you can run a company with hyperrationalism and no office politics is a pipe dream of folks who have never managed anything more complex than their next white paper.
    16
  28.  @ad2094  are you being pedantic or just not understanding the difference? I liked being there except for the work environment. Vacations were great. Weekends were always fun. If you’re independently wealthy, it’s a playground. If not, you spend a third of your life doing something soul crushing.
    15
  29.  @snorman1911  yeah and as someone else pointed out he says “more than what the bucket costs” then said he felt bad for the manufacturer’s profit margins. It’s just a dumpster fire series of sentences. lol
    15
  30.  @souvikrc4499  actually no. The power to control who ASML sells to through the EUV license is proving, at least so far, to benefit US national security interests, which was the plan all along. There wasn’t and still isn’t any company in the US with anything like the skills or capital to build out EUV infrastructure and produce tools.
    14
  31. Damn you are CRANKING the videos out. This is one I’ve been looking forward to.
    14
  32. An enormous amount of III-V production has never scaled beyond 0.5um. There’s simply no need in many cases, and the tools and processes used to create them aren’t capable of doing much better.
    13
  33. If a supply of raw materials is more expensive to sort for authenticity than to cosmetically reprocess and there is no downside to counterfeiting, the bad actor has no incentive to ensure quality. The same mindset that cuts baby formula with poison sure as hell isn’t going to care if you get a bad Atari sound chip.
    13
  34.  @Adamrc98  people have been smoking for thousands of years and that’s nasty as shit. Gross Chinese mud booze is gross.
    13
  35.  @TamagoHead  SRAM is definitely still in use; its applications are just different to DRAM. SRAM cells are significantly larger, complex (they use latching to store data instead of simple charge like DRAM) and must be continuously energized to retain data. But it’s stupid fast, so it’s usually only used for onboard cache.
    12
  36.  @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.”
    12
  37.  @dxelson  yet somehow Intel manages to make billions a year? The fab I work for (in the US, not Intel) is decades old and still bonkers profitable. The idea that a cutting edge fab (or even one built in the 80s) can’t make money in the US because our engineers are paid significantly more than an engineer in Taiwan is laughably wrong.
    11
  38. It’s quite difficult to build a late 90s/early 2k PC because so many motherboards surviving from that era have Capacitor Plague.
    11
  39.  @curtlundgren6867  sounds about right. Off state black levels are a major reason DLP became obsolete as LC black levels improved and then LED off-current improved. Then came active LED, OLED, and then quantum dots with ever better black levels. There was a lot of room at the bottom for solid state light solutions, but DLP could never quite keep up. MEMS in general are much more finicky and lower yielding to fab, not to mention they face very different challenges to CMOS. TI repurposed a lot of silicon equipment to make DMDs and I think the ROI of using apples to make orange juice was difficult to justify in the long term, especially with such stiff competition from Japan, Korea and later China. Once China started pumping out decent quality LED displays that kicked even the best DLP’s ass, the margins in display evaporated.
    11
  40. You need bigger problems.
    11
  41.  @connormauk3210  I literally don’t give a shit what you think. I’ve been forced to drink this slop at multiple dinners in Taiwan and Shanghai. Moutai is fucking gross.
    11
  42. 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
  43. I assure you, stepper optics makers knew about immersion as a concept from the beginning. The problems were 1) why solve a bunch of novel engineering issues before you absolutely have to and 2) once you have solved them, how to convince fabs that having liquid in contact with their wafers during exposure wouldn’t cause killer defects.
    10
  44.  @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
  45.  @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.
    10
  46. The fact that photo rework works at all is almost as much of a miracle as lithography itself.
    9
  47.  @rrai1999  “more than what the bucket costs” then saying “I feel bad for the saw manufacturers” then makes ZERO sense.
    9
  48. You can brute force a lot of things. We still don’t know what their yield or cost is. I expect it’s pretty poor. Not particularly problematic for SMIC, they can just dip into public money to make up for it until those things improve.
    9
  49.  @Daz912  I ran the entire division. But thanks for playing.
    8
  50. What are you even talking about. “Micron” is standard diction in the entire semiconductor and electronics industry.
    7
  51.  @GalvayraPHX  pretty poorly I’d say, considering you’re just puking up whatever cack-handed scribblings you read five minutes ago on Breitbart.
    7
  52. @Formosan I retract what I’ve said about the pay. On a cost of living basis, TSMC pay is very generous (although less senior engineers live like paupers on their salary waiting for that bonus). Median engineering pay seems to be about $120k/year after bonuses which is pretty generous. The only thing I stand firm on here is the complete lack of work-life balance. I spent years going to Fab12 and Fab14 and having meetings with photo engineers and even after years working there, they basically have no life. I never had a meeting where anyone who wasn’t a big shot didn’t have a hairnet line in their forehead. They only come out of the fab to take a dump.
    7
  53. It’s always etch’s fault. Even when it’s photo’s fault. 💕
    7
  54.  @Daz912  if you like that environment you’re the exception. And if you’re a foreigner, you’re being given special treatment and leeway anyway, so nobody should take your experience seriously.
    7
  55.  @Daz912  more than some token gaijin who loves his strings even though all the Japanese corpos around him privately think he’s a talking monkey. Assuming anything you’ve said is remotely true that is.
    7
  56. Nano imprint will never replace EUV. It’s too slow, too defect prone, and they’ll never solve the overlay issue. All the single digit overlay data you see in Canon’s propaganda material is massaged to hell.
    7
  57. While it’s obsolete for silicon technologies, resist undercut is extensively used as a desirable feature of liftoff patterning where gold is used to carry current (gold has no volatile halides, so it cannot be dry etched). This is still very common in GaAs power amplifier technologies.
    7
  58. At the time, it was seen as the only way ASML was going to force Intel’s hand to buy from them. I think if Intel had understood how unstable and hollowed out SVG was, they would have switched earlier.
    7
  59.  @TamagoHead  all good. It’s a ridiculously specialized field. To the point I probably can’t get a job doing anything else at this point. And yeah, Motorola was truly a king of industry in semiconductors for a long time. In the end, I think bad management, bad culture, and bureaucratic bloat got them.
    7
  60. No, he is correct. Memory capacity IS reported in bits. Because the cell size to store one bit is the critical scaling factor, not the word line size.
    7
  61. 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
  62.  @epicsuper6775  so economically unviable, Samsung has one of the biggest fab complexes in the world in Austin, TX making one of the lowest margin commodities in semiconductors? And has for two decades?
    6
  63.  @epicsuper6775  then it’s pretty weird how Samsung did it like a dozen times and still makes money on razor thin margin products.
    6
  64. Yeah that was either bad writing or a misunderstanding. He was describing bulk pricing but misinterpreted it as an insanely low unit cost.
    6
  65.  @neilkurzman4907  I wanted to make a post about this exact thing but you said it better. For decades semiconductor companies have been in a feast to famine cycle of greedily chasing capacity. Much like the oil and gas refining industry this time around, when customers and governments pissed and moaned about building extra capacity, knowing the business cycle has busted every time and they’ve been left holding the bag, they said no. It feels like maybe we finally learned our lesson about chasing our own tail to overproduce in good times.
    6
  66. It’s a PAS5500 body DUV scanner. The Illuminator optics train in the back is behind the projection optics, in the direction of the excimer laser unit peeking out behind it.
    6
  67.  @LuaanTi  I’m not reading all that but I’m happy for you.
    6
  68. Let’s see if they build a KrF or ArF scanner worth two farts first…
    5
  69.  @ganganbam  workshop that joke. Don’t give up, you’ll be funny someday.
    5
  70.  @sooocheesy  very true. It’s much more difficult to make a large piece of CaF2 with uniform optical properties than fused silica/quartz. As such, high NA optics were always going to be a problem.
    5
  71.  @tonysu8860  that would strongly depend on whether you stock it with new tools or used ones; preferably used ones you already own. If the latter is the case and the equipment is already depreciated, it might make financial sense to build a fab shell and start it up. But then why weren’t you doing it already? Semiconductors just don’t allow for this kind of “go back and capture old nodes/tech” business model. A good example of one case where this worked was when TI built RFab in Richardson, TX. The only reasons they were able to start up a 300mm analog line was 1) they had already built the fab shell when the subprime mortgage crisis shit the bed and 2) Qimonda had just gone under hard and sold a massive amount of 300mm tool assets for cents on the dollar. That sort of confluence of chance almost never happens.
    5
  72. He’s pretty much just trolling at this point to avoid admitting he made a mistake early on. Don’t waste your breath.
    5
  73. In practice you always trade selectivity for directionality. A heavy non-reactive ion like argon will etch straight down, but it won’t distinguish between layers and the etch products will sputter rather than diffuse away from the surface. Purely chemical reactive species with low ionization will be very selective, but won’t interact with the plasma electric field strongly, so they will etch isotropically like a wet etch. In practice, you either alternate etch chemistry to form a passivation layer that prevents isotropic etching or strike a balance between etch species to achieve the profile you want.
    5
  74. I had always thought one of the main reasons ASML bought SVG is because SVG held the patents for catadioptric lens design. Of course nothing came of that, but it wouldn’t be the first time a M&A happened for IP that never got used.
    5
  75. I should clarify: while indeed consumers are marketed to with DIMM capacities in bytes, manufacturers and semi industry in general spec memory chips in bits.
    5
  76. 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
  77.  @4473021  fab engineering is not IT. What are YOU on about?
    4
  78.  @azuaraikrezeul1677  all well and good until Winnie the Pooh steals your assets.
    4
  79. Surprising because it’s laughably untrue. Think about it for even ten seconds and you’ll see why.
    4
  80.  @elchippe  ok, it’s 6 years later. Where’s the production tool?
    4
  81.  @PaulFisher  no problem! By the way, notice that cause 1 and cause 2 are in competition! If you tilt the mirror more, you can divert more light, but then you are exposing the substrate even more. So there is a complex trade-off between the torsion material flexibility and lifetime, the substrate reflectivity, and the view factor of the mirror all interacting with a non-ideal source. Still pretty cool it works at all though!
    4
  82.  @Hallettjs7957  Intel did it, yes. The largest most cash flush semiconductor company on the planet did it. Once. Analog and discrete makers aren’t working with that kind of capital. And if they’re III-V substrate forget it; the wafers don’t even exist.
    4
  83.  @dieSpinnt  with all due respect you have no idea what you’re talking about. Semiconductor fabs are not like Chinese sweat shops spitting out wax lips and Tupperware.
    4
  84.  @p-jbroodbakker1303  yup. That 157 tool was complete insanity. And a few months after it was completed they realized only the final few elements needed to be pure CaF2 and the rest of the optics train could be fluorine-doped fused silica. A real head slapper.
    4
  85. The whole "I can't take feedback about a mistake I made so I'm going to play it off as trolling" attitude is dumb. Taking feedback from people who actually work in this industry with a little grace improves your credibility.
    4
  86. Japan also backed the wrong source technology. DPP vs LPP was the VHS vs. Betamax of the EUV race.
    4
  87. It’s all fun and games until that MESFET discrete your fancypants state-of-the-art module needs is backordered!
    4
  88. Lol why would they give it to some rando for free?
    4
  89. Not to mention you cannot make field term corrections if previous layers were exposed by stepper/scanner.
    4
  90. 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
  91. Awesome! I asked for this topic last year! I’m so glad to see you explored this cool tech.
    4
  92. A lot of the reason 450mm died was the equipment folks NOT wanting to get screwed like they did with 300mm. They still got screwed on 450, but not as bad as if they’d gone all in.
    3
  93.  @rkan2  ah…no? Lol
    3
  94. It depends how quickly you can turn a mask layer. If the process is complicated and has to go back through lithography more than a couple times, even not factoring in yield loss from errors this introduces, you can’t justify buying and depreciating the tool for the output of that process. When your EUV scanner costs $150M-$200M, how fast you can pump money out of it is basically all that matters.
    3
  95. 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
  96. The bigger problem is that 157 is not simpler and wouldn’t really push very far beyond what can be done with KrF and phase shift masks. Even before 157 was de facto cancelled by Intel dropping its support, researchers were struggling to find resists transparent enough for decent patterning. And the search for high RI liquids transparent to 157 that would work for immersion was even less fruitful. Combined with CaF2 birefringence issues even the experts at the time struggled with, and I can’t see this ever being worthwhile or fruitful.
    3
  97.  @imeakdo7  a lot of papers were written about 157 for a solid decade and it’s still dead. I worked in one of the chemistry labs whose focus was resist formulation when 157 resist screening was hot. It’s not coming back. Inevitably any fluid used for immersion is going to be a fluorinated hydrocarbon and they release fluorine and HF when irradiated. The last thing you want in your quartz optics train is HF. And that’s just one problem with 157 nobody ever figured out a way around. China is not going to solve the myriad problems with 157 when it can’t even produce a 248 scanner that works.
    3
  98. It has interesting properties, but in general not scalable to replace silicon. GaN wafers are formed by MBE on either sapphire, SiC, or silicon. Wafer sizes above 4” on sapphire or SiC and 6” on silicon just aren’t practical, therefore they can’t benefit from the economies of scale already present at 300mm silicon.
    3
  99.  @J_X999  not sure what that entails if they’re already embargoed. The West isn’t going to stop them developing their own native tech ecosystem. Or at least trying to.
    3
  100. Basically irrelevant. These are different worlds.
    3
  101.  @brodriguez11000  I agree. The point is East Germany, as a Soviet satellite state, fell into a lot of traps of that culture that doomed them to failure. Some bureaucrat denying you static protection shoes, for example, is going to doom your efforts no matter what. Playing the “you pretend it’s fine, I pretend you didn’t fuck up” game is Soviet as hell. Technology difference aside (and that is like comparing cave paintings to AI art), China at least tends to listen to the science and put its money where its mouth is. If you’re going to be authoritarian, it’s far better for your industry to be ruled by technocrats than ideologues and apparatchiks.
    3
  102.  @Daz912  did you read what I wrote? I loved living in Japan. Working there is what blows.
    3
  103.  @ZxZ239  I never said any such thing. Stop being so defensive.
    3
  104.  @catsspat  some people just can't take public criticism or correction by people who know what they're talking about.
    3
  105.  @dieSpinnt  maybe it’s a second language issue but your tone is incredibly condescending for someone with no professional experience in this field.
    3
  106.  @dieSpinnt  it’s not an assumption because if you worked in this industry you wouldn’t be coming in here saying something as foolish and uninformed as that a 150mm or 200mm fab must inevitably be upgraded eventually. It’s exceedingly rare to do that because of the opportunity cost of lost production time. Those fabs are still producing chips and revenue streams while they’re online. Taking them completely offline to make a massive capital investment is ground-floor stupid. If you want the capacity of a 300mm fab and you have a 200mm one, you build a new 300mm fab and keep the 200mm one running in the meantime. It’s likely completely depreciated and running better than when you built it due to small improvements and engineering tweaks over time and it’s still making you money.
    3
  107.  @ugarit5  I continue to be perplexed by people who think every fuck up on YouTube is a joke or a troll.
    3
  108.  @tykjpelk  well you certainly wouldn’t do it for silicon CMOS in any case. Gold contamination forms trap states in the silicon bandgap. You can’t have gold anywhere near a silicon fab.
    3
  109. At the operator and tech level I 100% agree. Grunt labor is in short supply at most fabs right now. You can make the same hourly wage working at a grocery store in most semi-heavy US states. Why bust your ass in a fab for the same amount and get fired for dropping one FOUP worth $2M?
    3
  110.  @ThePowerLover  wow. Don’t see that every day on the Internet. Glad to be of help.
    3
  111.  @Gameboygenius  correct. 5500 refers to the body type. /xxxx is the model #. So for example a /100 is an i-line stepper with fixed pupil illumination and lower lamp power. A /275 is the same but with variable illuminator/pupil, better lens, and higher lamp power. Etc. A 5500/700 would be a single wafer stage body (limited to 200mm wafers max) containing a mid-range KrF exposure system.
    3
  112. Because that part of the radio spectrum is allocated to exactly these kinds of devices. 13.56+/-0.007Mhz is one of the so called “ISM” bands reserved for “industry, science, and medical” use. As such, there are no licensing requirements to use it. Although it’s possible to broadcast and receive in that band, machines that do so typically don’t exist since your data would be corrupted every time someone within 100ft microwaved a burrito.
    3
  113. Are you a fellow Perun fan?
    3
  114. There isn’t anything after lithography. There’s only different ways of doing lithography. At the moment the only two technologies that have a chance of equaling high-NA EUV are nanoinprint and DSA. And both of those have unresolved problems with defectivity that have never quite been solved.
    3
  115.  @inomo  oh well in that case I’ll tell Nvidia to make their intellectual property available to internet randos who “work for ASML” immediately. What are you smoking?
    3
  116.  @inomo  Somehow I think it’s much more likely that you’re an NPC who has no reason to have access to the design kit. Just because YOU can’t get access doesn’t make it fake. And honestly, Nvidia has very little reason to give you access: this likely competes with ASML’s in-house work on OPC algorithms and computational lithography. So not only do they not have a reason to give it to you, they have a reason to NOT give it to you. 🧂
    3
  117.  @inomo  not sure why you’d want that. You haven’t said anything of value yet.
    3
  118.  @tobiwonkanogy2975  different industry (beam or vapor phase epitaxy), but cool application. Not sure I want arsenic that close to my face, but I assume it will be encapsulated somehow.
    3
  119.  @akatsuki6371  glad to be of help!
    3
  120. “Micron” is standard industry jargon. Which you would know if you worked in this industry.
    3
  121. Even though you can look through the template, overlay is not a problem that’s going to go away because rather than adjust the image of the mask (which can scale linearly), changing the shape of the imprint template is limited by the modulus response of the template: a decidedly non-linear process. This means when you apply a correction to say, X-mag, different areas of the template will respond completely differently. Nano imprint is cute for some niche applications but it will never be used to print anything at the bleeding edge where its resolution capabilities shine. And on top of that, it’s slower than molasses in January.
    3
  122.  @JigilJigil  I’ve long suspected this was payback for Japan keeping the US out of Selete and other Japanese nationally funded semiconductor programs, not to mention sour memories of the 1980s. “Two can play at the protectionist game.” And two did.
    3
  123.  @MrDanisve  I work in a fab. Micron and nanometer are the trade terms in use the world over in this industry. This guy is either trolling or he’s a 14 year old pedant.
    2
  124.  @Gameboygenius  yes. It’s practically a pastime in Japan to come up with creative ways to avoid the NHK guy who *comes to your front door to ask for money*. For the longest time “I don’t have a TV” was a convenient lie that worked. Then NHK got wise and claimed any device that could play NHK content, including a laptop, meant that you have to pay the fee. But they have no powers of enforcement so most people just don’t answer the door or say the owner’s not home. For all its quirky creativity, Japan can be ass-backwards. Prolific and mandatory use of fax machines, the NHK guy, and the paper hanko system are just a few examples.
    2
  125.  @NightRogue77  indeed. Nobody loves a good story more than a frightened, ignorant human.
    2
  126. Because putting all your eggs into one basket is stupid, especially when GPU prices are being driven by an unpredictable crypto frenzy, not stable retail market pressure. That’s to say nothing of the fact they have existing contracts to produce for other customers that can’t be breached without penalties. It would also rapidly deplete stock of chips used to make things like smartphones.
    2
  127.  Olaf Willocx  I suspect they will be persuaded with the chips money, in which case, yes: they want to.
    2
  128.  @Luxcium  sure, maybe if your bucket cost $10M
    2
  129. I felt the same way about the newer Sony Xperia offerings, specifically for their cameras. The stuff you can do is light years beyond any other smart phone camera. Build quality is great too. But it’s almost impossible to buy one outside Japan and even counting domestic sales they’re an unmitigated failure. Yet somehow Sony keeps releasing a new one every year. It’s baffling.
    2
  130.  @fouzaialaa7962  200mm equipment is in high demand, but it’s not going to make a dent in your purchase of new 300mm equipment. There is very little used 300mm equipment on the market because all of it is still in use. The only time fabs get a chance to buy significant amounts of used 300mm capital is when a major company croaks and has to liquidate their assets to pay creditors. That’s rare. The last big one I know of is when TI outfitted R-Fab in Richardson, TX as an analog chip facility with 300mm equipment they bought from the dissolution of Qimonda. It was also a golden opportunity for TI as they had already built the R-fab shell, having previously intended for it to be a follow up to DMOS6. Bottom line is if you’re going to build a 300mm fab, your capital orders needed to be in a couple years ago because it’s all likely to be new build tools.
    2
  131. Assuming everyone here who hates the taste of Moutai is American pretty much pegs you as American yourself.
    2
  132.  @michaelkeudel8770  they may be on the same site, but fab 10 and fab 24 are separate entities. Fab 10 was shut a while ago.
    2
  133.  @michaelkeudel8770  well the point being it’s still extremely rare to do that for all the reasons laid out in this thread. Having another 300mm fab on the same campus is probably a huge advantage. As well as just being Intel in the first place.
    2
  134.  @nvignesh  the thing is on this channel: some of us ARE those people.
    2
  135. The proportion of Western progressives who actually want a full-on communist state is quite small, and suggestions to the contrary are mostly right-wing scare propaganda. Nordic socialism for example still relies on capital markets and functions just fine.
    2
  136. 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.
    2
  137. Because it’s a red ocean market with high risk and volatile prices. If their margins are 20x higher making GPUs and snapdragon mobile processors, making DRAM would be a poor allocation of capital.
    2
  138. Fun fact: Perkin Elmer was also the company that fucked up the Hubble primary mirror. I remember a coworker of mine recalling he would bring it up constantly to his Perkin-Elmer FSE at Motorola just to get a rise out of the guy. It’s not often the US government has to launch a space shuttle just to fix someone’s incompetence. 😆
    2
  139.  @tykjpelk  sure, if you don’t mind sputtering gold all over everything and failing reliability testing. What works in a university lab or a white paper to get a pretty SEM image rarely works in the field.
    2
  140.  @ThePowerLover  ASML’s factories are in Netherlands, not Belgium. IMEC is in Belgium.
    2
  141.  @tykjpelk  totally gonna start calling mask files “GD slayers” starting this coming Monday.
    2
  142. Their AOI tools are pretty good, especially if you can’t or don’t need to pay KLA prices.
    2
  143. Other than TEL, Japan maintained dominance in one lithography area where it has never been surpassed: photoresist and multilayer spin-on materials. I worked for an American competitor to these companies (comprised mainly of Shin Etsu, TOK, and JSR, as well as Nissan Chemical's semiconductor materials division) and it was laughable how far behind we were. I eventually left because continuing to be employed there while standing in front of customers with these entrenched suppliers was a personal embarrassment. Competing with someone like Shin Etsu whose entire history was built on organosilicon materials and polymers is a fool's errand.
    2
  144. BTW my memories of working on a pair of Micrascan III+ tools at my first job are a constant source of long-term trauma. Everyone loved the reticle cassette design y’all came up with where if you picked it up wrong the mask would fall out. 😆
    2
  145. Lower reticle demagnification reduces active write area on the mask, making defects easier to tolerate. 5x masks still exist at all because legacy equipment requires them. Canon for example used 5x reduction in its steppers, but left the advanced lithography market after dry ArF.
    2
  146. Animation at 3:20 technically shows a scanner, not a stepper. When the shutter is open on a stepper, there is no stage motion. Minor quibble though; I love this channel.
    2
  147. 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
  148. Precision diamond saws have long been the standard. IR and UV laser saws are more common these days. A lot of it depends on the substrate, what sort of overcoat protection is needed, and the street dimensions.
    2
  149. 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
  150.  @thisiskevin1000  Indeed. ASML bought SVG for the patents on their catadioptric lens designs, which they needed to design and deploy advanced 193 and 157 scanners a few years later.
    2
  151.  @skwalka6372  global warming is real, but it’s not a Venus scenario by any stretch of the imagination. CO2 levels have been far in excess of projected 2100 levels in the past and life was just fine. The issue is with sustaining 8B people in those conditions, not survival of life itself.
    2
  152. China can’t even make a working dry ArF scanner. 😆
    2
  153. “Necessity knows no law but to prevail.”
    2
  154. This tech is cute but it will never replace projection lithography. Too slow, takes too much fab floor space, and the overlay issues inherent to a rigid template are completely intractable.
    2
  155. I’d never heard of split manufacturing as a counterfeit countermeasure. That is kind of brilliant.
    2
  156. No matter what you think of either, deglobalization is not imperialism. If this seemed like a bad deal to TSMC, they wouldn’t have taken the CHIPS money. And in terms of technical achievement, where was Uncle Sam going to go for tech not even Intel can build? Nobody twisted TSMC’s dick off to build the AZ fabs.
    1
  157.  @punditgi  my point is regardless of the “official” word, micron is in extremely wide use. ESPECIALLY in the semiconductor industry. There’s also IUPAC names for complex organic molecules but chemists don’t use them. In this industry, “micron” is the correct choice of word. Saying “micrometer” in this field to a colleague makes you sound like a pretentious asshole.
    1
  158.  @Gameboygenius  I knew someone would come in hot with the whataboutism. It never fails. Yes, every country is fucked up in myriad ways. Right now, we’re talking about Japan and its stereotype as some kind of high tech wonderland when in reality a door to door salesman comes to your door to ask for tax money and you need to do physical paperwork for practically everything. Lots of places don’t even take credit cards: a technology invented over 50 years ago (although the past ten years they’ve gotten way better about this). And yes, some places in the US technically still take personal cheques. But this practice is rapidly fading. Pulling out your checkbook at a grocery store is bound to garner eye rolls from the people in line behind you, if not the cashier. It’s quintessential boomer behavior and it’s dying off about as fast as they are. I don’t know anyone under 40 who has a checkbook and uses it.
    1
  159.  @causewaykayak  you’re welcome!
    1
  160.  @WhiteOwlOnFire_XXX  so glad a figurative turn of phrase got some weird white supremacist with a keyboard to show up and tell us how brave he is. 😆 ISTFG, get more than 100 likes on a comment on an Asianometry video and 100% the weirdos and trolls will come out in droves. This channel’s fans have become the absolute worst thing about it.
    1
  161.  @lu544  I am pretty sure the government told them to come up with some heinous shit to kill people. No need to bring your screeching about the left into the discussion.
    1
  162.  @Samṣtrainss  I’d hardly call Elon Musk a scientist. 😆
    1
  163. Also, try tripling your chip prices to Nvidia, one of your biggest customers, and telling them to just pass on the cost to their retail customers, see what they say. 😂
    1
  164.  @alexanderphilip1809  I lived in Texas for 30 years and its people are no more “intelligent” than any other large state. They’re just very pro-business, which is fine for encouraging fab construction. But A, this fab won’t be built in Texas and B, TSMC work-life imbalance won’t fly in Texas either.
    1
  165. The number of these Japanese "pivot to success" stories where they moved in on a related but tenuously connected industry seems endless. Nintendo for example used to make playing cards.
    1
  166. Brace yourself. Terrible pronunciations are coming.
    1
  167. sto-KA-stick
    1
  168. Gate-all-around (a finFET with the gate completely wrapped around the channel) is already planned by Samsung and TSMC. High-NA EUV is a beast so difficult it’s really more of a successor technology to today’s EUV. Beyond that, there’s not really any clear solution for further pattern scaling. EUV itself was a pipe dream for decades until it became a necessity. We don’t really even have a serious contender for pipe dreams at this point…
    1
  169.  @Adamrc98  literally nobody here is saying they’re a professional. Most people are saying they’ve tried it and it sucks. That’s not incurious. And honestly, it makes you sound like exactly what you’re criticizing.
    1
  170. Always heard the company name pronounced “Air Li’keed” in industry. It is French after all.
    1
  171. That advert at 15:00 is fucking hilarious.
    1
  172.  @neilkurzman4907  don’t bother, it’s just some FCA Chinese troll copy-pasting their comments.
    1
  173.  @imeakdo7  this is an important caveat to anything China does to play catch up. Even their home grown stepper/scanner electromechanics aren’t up to snuff. China is still missing a great deal of basic know-how to even build a KrF scanner worth two farts, let alone try to solve 157’s problems.
    1
  174.  @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
  175. Aside from not knowing the yield, not knowing the cost is just as important. MP lithography for layers that should use EUV is stupidly expensive.
    1
  176. Anyone can fart and file a patent on it. China can’t even make a competitive DUV tool and that’s 30 year old tech. Believe it when you see it.
    1
  177.  @Stephen_Ito  lol what. China has one of the biggest public capital markets in the world.
    1
  178.  @Stephen_Ito  the Chinese government is not going to tank its Crown Jewels by collapsing their stock price. Investors still matter hugely if those companies want to maintain market capitalization and credit worthiness. Modern China is state capitalist, not communist.
    1
  179. Sharp: [craters into the dirt] YouTube weebs: “another charpter on the anti-Japan war”
    1
  180. Also it's pretty obvious that Trudeau, while phrasing his statement like a moron, clearly meant that China's one-party rule allows it to dictate policies and pivot quickly to implementing them. In a political system that inherently must compromise to enact policy, that IS a challenge Western democracies have to contend with.
    1
  181. Nikon tried, but they backed the wrong horse on EUV source technology (DPP vs LPP). That coupled with ASML’s patents on dual stage technology to augment the throughput requirements at the low wafer plane powers EUV started at and ASML’s strong partnership with Zeiss who were way ahead on the reflective optics pretty much doomed Nikon’s efforts. Canon, the only other major exposure tool maker, largely dropped off the map after dry ArF and never recovered.
    1
  182.  @agsystems8220  I’ll believe China will catch up on EUV after they’ve produced, say, a working KrF scanner.
    1
  183.  @agsystems8220  I’m not saying they can’t do it. They just have a hell of a long row to hoe first.
    1
  184.  @chrismofer  somewhere between $0.05 and $0.15 per unit seems about right. They are usually not sold directly though. The filter is part of an amplifier or signal package chip that sells to the end user for a couple bucks.
    1
  185.  @_Chad_ThunderCock  no, it won’t. It’s too slow. I worked for the research group that invented step and flash 20 years ago. It has the same problems it did back then despite all the advancement. In the meantime, EUV went from impossible to production. Most hurdles can be overcome, but if your final solution isn’t cost effective, it’s dead. NIL may be useful for some applications, but you will never see one in production for advanced node logic or memory. I’d bet my house on it.
    1
  186.  @_Chad_ThunderCock  companies spend tons of money on strategic failures all the time. Look how much Meta has wasted on its moronic business VR platform. I have worked in this industry for 25 years. NIL has applications, but as far as challenging ASML for mass production of semiconductors, it’s a turd that simply will not polish.
    1
  187. Second this in general. KLA’s story just as a company is long and interesting.
    1
  188. They also wear our gearboxes faster because when the rotor is in vertical alignment, one blade experiences a drop in lift due to the wind shadow of the support structure, putting a backward directed torque perpendicular to the rotation on the rotor shaft.
    1
  189. I mean, good luck inventing immersion and EUV from scratch. China has been dumping money into government tech programs and domestic foundries for 20 years and still can’t build a stepper worth a shit.
    1
  190. There is no stepper infrastructure for reticles other than 4x and 5x. You can do whole wafer masks aligners, but those have resolution limits and defect issues that can’t even break 3-5um in most cases.
    1
  191. They reflect the light from a source. In the on state, the pixel is bright. When the mirror tilts, it turns the pixel off.
    1
  192.  @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
  193.  @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
  194. In this case it was a mistake. The original quote was clearly about bulk pricing on a unit basis, not the total cost of 50 million filters.
    1
  195. The part that doesn’t get mentioned very often is how much of a pain in the ass it is to send objects into orbits or intercept trajectories closer to the sun than earth. Think about how easy it is to throw something out from a carousel versus in, relative to the ground frame of reference.
    1
  196.  @calvinhobbes1617  I work with these tools daily. No ASML tool in existence uses both refractive and reflective optics in the actinic path.
    1
  197.  @calvinhobbes1617  well your information is about 20 years out of date then.
    1
  198.  @transcrobesproject3625  it’s stupid.
    1
  199. ASML did produce a model 1100A and 1100B ArF scanner, but as I recall not many were produced. It was the first commercial Twinscan Atlas body ArF tool ASML made. I worked on one in my first photo engineering job. It really felt like ASML was using us to test their software. Constant bugs, software updates, and crashes. It sucked.
    1
  200.  @dexlab7794 in other words, communism can’t succeed unless the whole world is communist to start with. Way to blame the failures of communism on literally anything but itself.
    1
  201. I’d be interested to read that if you have a link handy.
    1
  202. ​ @filippalexandrov1554 a single dispense for one wafer is about 0.5-1.5ml typically for a front end process. So a gallon bottle can process about 2500-7500 wafer passes. In practice it's lower due to run in waste and periodic dummy dispenses to keep the nozzles clean. But that's the ballpark.
    1
  203. ​ @lucasrem EMD is also not Dupont.
    1
  204. A big problem with this idea is getting enough equipment to operate at large enough scale to disrupt the market. In many ways it’s more difficult to tool up an old node fab because cheap, vintage used equipment isn’t available or gets snapped up as soon as it’s on market.
    1
  205. @卩乇尺丂ㄖ几 tell me you’ve never worked in semiconductors without telling me you’ve never worked in semiconductors. Chips are not priced as market commodities and customer pressure to lower prices as each generation of product ages is unavoidable. The fab I work at is at capacity. We would love to start more wafers and sell them. People are beating down our door. What we’re not going to do though? Bend our existing customers over a barrel and fuck them when they’re hamstrung by supply chains. That’s how you get customers to leave for your competitors when things get better.
    1
  206.  @justinhealey-htcohio3798  General Oligarkov salutes you!
    1
  207. Thanks for all your great work and research. Take a break if you need it, we’ll still be here.
    1
  208. Some of these pronunciations continue to give me brain damage. I’m starting to roll with it though.
    1
  209.  @GordonFreemayne or maybe all the people standing up are being exploited by Big Chair.
    1
  210. It is in a lot of places, but the US or its western allies never did, so the accomplishment is downplayed in the history of space flight.
    1
  211.  @yagoa  that has nothing to do with why it isn’t well known. The V-2 was a Nazi terror weapon and we all know about it. The truth is it was a scientific accomplishment the west never achieved, so it’s not mentioned very much in western space flight history. This can be understood WHILE acknowledging that the Soviet Union was an abusive, genocidal authoritarian state.
    1
  212.  @skwalka6372  TF are you talking about? XD
    1
  213.  @skwalka6372  I’m objectively correct. CO2 levels in excess of 600-800ppm were common until about 34M years ago. Levels dropping below 600ppm are thought to have triggered Antarctic glaciation. NONE of this is the same as saying we should be pumping more CO2 into the atmosphere, because that is already altering the climate and at the brink of triggering mass extinction. But it’s also a far cry from a runaway greenhouse effect turning Earth into Venus. Get some damn perspective in your life.
    1
  214.  @user_375a82  shit! the West’s days are numbered!
    1
  215.  @didyoumissedmegobareatersk2204  you can have parity with other countries in one technology and still suck at something else, so no, that’s not relevant.
    1
  216.  @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
  217.  @xBlackWind  well, come back and let us know when this skunkworks project is for sale.
    1
  218.  @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
  219.  @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
  220. 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
  221.  @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
  222.  @thisiskevin1000  I will believe the SMEE DUV scanner works when I hear they’ve sold a couple dozen to SMIC.
    1
  223.  @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
  224. “Apple was an 800lb gorilla dick to their suppliers and forced them to innovate in exchange for their business.” Ever was it thus.
    1
  225. Purchasing Power Parity matters.
    1
  226. Unrelated, other than how it limited capital allocation.
    1
  227.  @punditgi  I am familiar with metrology. My job is fab metrology. Policing the language of people who actually work in this industry just makes you look like a pedantic dick.
    1
  228. You joke but this WAS a topic that garnered some serious consideration. The problem, as the EUV folks discovered, is almost nothing worth building reflects light well below 100nm wavelength. X-rays and gamma rays would also cause soft errors in the underlying SRAM, if not permanent midband traps. Even making a DMD that can reflect DUV light is difficult from a materials perspective. I always thought it would be cool to replace the reticle in a lithography scanner with a strip of DMD mirrors and then use that as your pattern master. But UV DMDs like to get stuck in one state or the other after getting beat on with ultraviolet light, and so tend to get an unacceptable accumulation of dead pixels over time.
    1
  229.  @_Chad_ThunderCock  “possible” is not necessarily “manufacturable”. If you work in a fab, you know this. I do.
    1
  230. State of the art production is TSMC 3nm and 5nm node and whatever Samsung is calling their equivalent node, I forget.
    1
  231. Isn’t there a simpler explanation? The people in charge of raising wages are the ones being overpaid and benefit from not raising wages, while the labor isn’t organized enough to push back on them.
    1
  232. Totally agree. That taxi/car ride from HSR to TSMC sucks so bad lol
    1
  233.  @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
  234.  @DwAboutItManFr  I specifically said as much.
    1
  235.  @zhongxingjin846  if I need a drinking pedigree to even tolerate your next, more disgusting swill, you’re bad at making alcoholic drinks.
    1