5 ms·

Nanoimprint Lithography Aims to Take on EUV

118 points by pseudolus 2y ago

notarealllama 2y ago
This is fascinating and looks promising! I've never heard of this but expect we will more in the near future, especially if they meet that 2028 target. I wonder what the environmental impact of this is versus extreme ultraviolet. Although they mention "cost of ownership" and throughput, I wonder if this has any hidden implications.
- huijzer 2y ago
  Why should we care about the environmental impact of EUV machines? I think it's probably better to focus on things which have a real environmental impact. For example, EUV machines are estimated to 54 000 GWh per year by 2030 [1]. This number is a extremely high estimate because current usage is much lower (10 GWh per tool annually according to the same article and in 2020 ASM shipped their 100th EUV system, so current total about 1 000 GWh). This is sold as being "power hungry". Let's put these numbers in perspective. The United States alone consumes about 25 000 TWh "primary energy" pear year (includes electricy, transport, and heating) [2]. This means that in the extreme case, EUV machines consume 54 TWh / 25 000 TWh = 0.2% of total energy! In comparison, 27% of total U.S. energy consumption was used for transporting people and goods around in the US [3]. And I made the example here before that if you are considering to turn off your phone in order to save battery at the risk of taking an accidental detour, then the decision is simple. Keep the phone. Driving one kilometer extra consumes multiple orders of magnitude more energy than powering a phone for hours. I think this idea holds in many more cases. Video meetings for example can save people from traveling all over the world. This saves energy and time as well. So I would say please go full power on chip manufacturing. It's way better for the environment (and often saves people time) than deciding to stop innovation and instead keep transporting everything around physically. I'm not saying transport is bad. I'm saying that standing in the way of innovation as an argument for better "environmental impact" is nonsensical. [1]: https://www.techinsights.com/blog/euv-lithography-power-hungry-path-innovation https://www.techinsights.com/blog/euv-lithography-power-hung... [2]: https://ourworldindata.org/energy/country/united-states https://ourworldindata.org/energy/country/united-states [3]: https://www.eia.gov/kids/using-and-saving-energy/transportation.php https://www.eia.gov/kids/using-and-saving-energy/transportat...
  - saagarjha 2y ago
    Nobody has to lose for lithography to win.
  - jampekka 2y ago
    It's nice to think that global warming can be solved with some technological gimmics so people don't have to make any lifestyle changes.
    - huijzer 2y ago
      Nobody will make the lifestyle changes. Or do you? Do you sit in a cold house as you type this? Do you not shower? Do you only grow your own vegetables in your backyard? Do you never use a car? Do you not even own a car? Do you have children? It's terrible for climate to have children? I think it sounds somewhat nice in theory to make lifestyle changes, and sure it helps, but it's not a solution. It's like if you are in financial trouble. Sure you can decide to not spend any money anymore. That definitely helps. But if you sell your car to save money and then cannot make it to job interviews anymore then you saved too much. You need to focus on getting money (and maybe spending it in the process) AND saving money. Focusing only on saving money is a losing strategy. Same with climate. Focusing only on using less energy is a losing strategy. Sometimes you need to spend energy to save energy in the future.
      - atq2119 2y ago
        "Nobody", really? As a bare minimum, many people can choose to take more environmentally friendly vacations. You don't have to go on a cruise, and yet that's a booming industry. You don't have to fly across an ocean. Almost everybody who does these things has the option to go on perfectly fine, perhaps even better, less carbon intensive alternative vacations. And yes, there are people who consciously make that lifestyle change. Not enough, of course. But only a Sith deals in absolutes.
        RealityVoid 2y ago
        You guys are naive. Sure, it's possible, but most people don't do this, don't want to do this and will not do this. And of those that DO some of these things, a large subset probably does it in a ineffective way that OP so nicely illustrated in the phone/1 extra km by car example. No, the solution must be technical while people are allowed to maintain most of the comfort they are used to. Anything else and you will simply not be able to convince people to do so even if that means burning the world down.
sva_ 2y ago
> For instance, compared to an EUV system employing a 250-watt light source, Canon estimates NIL consumes just one-tenth the energy. I'm not an expert on this but feel like a 250w light is not the major driver of cost in EUV? Or am I misunderstanding this?
- szundi 2y ago
  Yes, it is 250w output but the efficiency is near zero, really
  - dylan604 2y ago
    But you can bake a cake with a mere 60w light. That 250w is also much less than what it takes to run modern GPUs. so it's all relative.
- Szpadel 2y ago
  AFAIK the main coat of EUV is cost of machines that will be obsolete in few years, so you want to produce as many chips using them as possible during that timeframe, they design a lot of around to maintain near 100% uptime of those machines. This include buffers before and after machines (so any unplanned stalls are mitigated) and technicians trained to do maintenance in F1 pitstop fashion. (source: some tour of some chipmaker I saw online, no longer remember details)
  - lnsru 2y ago
    These machines will be not obsolete for very long time. They are extremely rare and expensive. And the most of semiconductors are fabricated on mature nodes anyway.
    - rcxdude 2y ago
      Not obsolete, but the margin on fabs drops off pretty quickly once they're not at the cutting edge, as I understand it, so they need to make back their capital investment fairly quickly.
noduerme 2y ago
> So instead, Canon drew on its inkjet printing know-how to apply the resist in optimum amounts to match the circuit pattern. Hopefully for those prices it will still let you send a fax if you run out of resin.
lnsru 2y ago
We had many experiments with nanoimprint lithography at the university 20 years ago. The resolution was poor and the durability very poor. After dozens of imprints the “stamps” degraded heavily. I am curious if 20 years were enough to fix all the issues and it’s really competitive today.
- actionfromafar 2y ago
  Wasn't this sort of how Intel got started in the 1970s? Some kind of contact printed ICs?
  - formerly_proven 2y ago
    You're probably thinking of contact lithography, where a 1:1 mask is placed directly on the wafer and illuminated. This would've been used for the earliest IC processes, where you'd still be able to see the structures with the naked eye or a loupe.
    - f1shy 2y ago
      That is basically like home grown PCB but at very tiny.
- hinkley 2y ago
  I would think you’d want to make them out of something crazy like diamond or titanium carbide. What did you make yours out of?
- swordsmith 2y ago
  Came to the comments to see this. Stamping has all sorts of problems with alignment, stamp resolution, contact quality, etc, it's not clear whether this so called "simplicity" still holds after scaling up the resolution even more.
P_I_Staker 2y ago
[flagged]
- gliptic 2y ago
  Apparently you don't believe in reading titles either.
- smokel 2y ago
  An EUV is not a second-hand EV. The term refers to Extreme UltraViolet lithography: https://en.m.wikipedia.org/wiki/Extreme_ultraviolet_lithography https://en.m.wikipedia.org/wiki/Extreme_ultraviolet_lithogra...
usrnm 2y ago
Anything that disrupts the monopoly ASML currently has is a good thing
- ClumsyPilot 2y ago
  Why do you say that? Does ASML hurt the consumer or the market in some way? Has it been caught using dodgy practices against its workers, like Amazon has, against its customers, like the Wells Fargo has, or something else?
  - asdasdsddd 2y ago
    Because its bad to have a single point of failure in our supply chain
- hinkley 2y ago
  How did that happen? I’ve never been clear on how we got here. Just some vague stuff about how others tried and failed.
  - cma 2y ago
    Intel thought the tech wasn't mature enough and decided to wait longer. They and the US were originally part of inventing it in the 90s (EUV LLC public private partnership), but sold it off with terms that allowed us to have export controls over ASML's part of it.
    - hinkley 2y ago
      Man, US tech and the 90’s. Al Gore pushed for The Partnership for a Next Generation Vehicle. They say the only vehicle the PNGV produced was the Prius, and Toyota was not eligible to participate. But they freaked out because the Americans were taking it seriously and created a dream team by recruiting their best people into a new program. Only it was all for show. So TM ended up with a serious car and everyone else with demonstrators.
spuz 2y ago
I wonder how big the wafers can be in the NIL system. It definitely sounds like the larger the wafer, the more problems you will have with deformation, alignment etc. if they have to reduce the wafer size in then that would also affect their ability to compete with EUV.
- ov_ov_ov 2y ago
  Think is on roll-to-roll flexible substrate, no wafers.
abdullahkhalids 2y ago
OP states that this can go down to 14nm. What I am interested in is whether older and larger processes (say ~50nm) can be done at a much cheaper cost than traditional methods. A lot of stuff simply does not require the most advanced chips.
- nabla9 2y ago
  Canon has been selling nanoprint litography machines for a long time. FPA-1200NZ2C came out 2015-2016. Press release from a sale 2017 https://global.canon/en/news/2017/20170720.html https://global.canon/en/news/2017/20170720.html
- atq2119 2y ago
  Note that 14nm processes (which are quite old by now) are not the same as 14nm feature sizes. I'm not sure what these machines are capable of, since some details may well be lost in translation in this kind of publication. And I'm only an interested enthusiast, I don't work in the field directly. But towards the end of the article they talk of targeting 8nm line width in 2028, which is impressive. Maybe this time around NIL actually becomes real for high-end processes?
  - pama 2y ago
    These numbers are all mindboggling. I understand that the modern specs for EUV dont mean wire width, but if with this future NIL we truly get down to 8nm wide wires, perhaps we should start counting the number of atoms across the width of the wire (around 30).
  - phonon 2y ago
    These are feature sizes. (Comparable to the 13.5 nm of EUV)
- shaism 2y ago
  The answer almost certainly is no. While lithography is one of the largest single contributor to manufacturing costs, the contribution to overall cost is still far below 10%. And one cannot simply substitute an optical lithography with a nano imprint machine without redesigning some part of the process (etch, metrology etc.). Investing R&D resources for a (best case) 10% reduction in costs while still having a decent probability of failure in a big but declining node is not worth it.
y04nn 2y ago
For everyone interested on technical details of the TSMC EUV process I would highly recommend this CCC talk [1] (From Silicon to Sovereignty: How Advanced Chips are Redefining Global Dominance). [1] https://news.ycombinator.com/item?id=42546231 https://news.ycombinator.com/item?id=42546231
- sschueller 2y ago
  I knew the process was complex especially with the light source but I didn't realize that diffraction was something they also use which is absolutely insane.
- modeless 2y ago
  Cool! I uploaded the video to YouTube here: https://www.youtube.com/watch?v=sB-y-tDlOSA https://www.youtube.com/watch?v=sB-y-tDlOSA (It's licensed CC-BY so this should be allowed, and I like having videos like this on YouTube where I can easily watch them from anywhere and add them to my playlists.)
  - mjrpes 2y ago
    The "transistors shipped" in the history of computing was an interesting number. In 2024 it is now over 10^24. That's a massive number, more than estimate number of stars in the universe. But, in another sense, still quite small. It finally surpassed Avogadro's number, or 6*10^23 particles. This is the equivalent of a small shot glass filled with water (molecules).
westurner 2y ago
Nanoimprint lithography (NIL) https://en.wikipedia.org/wiki/Nanoimprint_lithography https://en.wikipedia.org/wiki/Nanoimprint_lithography
hinkley 2y ago
I feel like this tech would be better suited to flexible circuitry, because flexible can be continuous feed, and why try to limit or size your stamp to the surface area of a wafer when you could just size it to the width of a spool? Also flexible circuits tend to be at a much larger feature size and so it’s okay if they’re a couple generations behind, but this is still far ahead of printed circuitry.

118 points by pseudolus 2y ago

This is fascinating and looks promising! I've never heard of this but expect we will more in the near future, especially if they meet that 2028 target. I wonder what the environmental impact of this is versus extreme ultraviolet. Although they mention "cost of ownership" and throughput, I wonder if this has any hidden implications.

huijzer 2y ago
Why should we care about the environmental impact of EUV machines? I think it's probably better to focus on things which have a real environmental impact. For example, EUV machines are estimated to 54 000 GWh per year by 2030 [1]. This number is a extremely high estimate because current usage is much lower (10 GWh per tool annually according to the same article and in 2020 ASM shipped their 100th EUV system, so current total about 1 000 GWh). This is sold as being "power hungry". Let's put these numbers in perspective. The United States alone consumes about 25 000 TWh "primary energy" pear year (includes electricy, transport, and heating) [2]. This means that in the extreme case, EUV machines consume 54 TWh / 25 000 TWh = 0.2% of total energy! In comparison, 27% of total U.S. energy consumption was used for transporting people and goods around in the US [3]. And I made the example here before that if you are considering to turn off your phone in order to save battery at the risk of taking an accidental detour, then the decision is simple. Keep the phone. Driving one kilometer extra consumes multiple orders of magnitude more energy than powering a phone for hours. I think this idea holds in many more cases. Video meetings for example can save people from traveling all over the world. This saves energy and time as well. So I would say please go full power on chip manufacturing. It's way better for the environment (and often saves people time) than deciding to stop innovation and instead keep transporting everything around physically. I'm not saying transport is bad. I'm saying that standing in the way of innovation as an argument for better "environmental impact" is nonsensical. [1]: https://www.techinsights.com/blog/euv-lithography-power-hungry-path-innovation https://www.techinsights.com/blog/euv-lithography-power-hung... [2]: https://ourworldindata.org/energy/country/united-states https://ourworldindata.org/energy/country/united-states [3]: https://www.eia.gov/kids/using-and-saving-energy/transportation.php https://www.eia.gov/kids/using-and-saving-energy/transportat...
- saagarjha 2y ago
  Nobody has to lose for lithography to win.
- jampekka 2y ago
  It's nice to think that global warming can be solved with some technological gimmics so people don't have to make any lifestyle changes.
  - huijzer 2y ago
    Nobody will make the lifestyle changes. Or do you? Do you sit in a cold house as you type this? Do you not shower? Do you only grow your own vegetables in your backyard? Do you never use a car? Do you not even own a car? Do you have children? It's terrible for climate to have children? I think it sounds somewhat nice in theory to make lifestyle changes, and sure it helps, but it's not a solution. It's like if you are in financial trouble. Sure you can decide to not spend any money anymore. That definitely helps. But if you sell your car to save money and then cannot make it to job interviews anymore then you saved too much. You need to focus on getting money (and maybe spending it in the process) AND saving money. Focusing only on saving money is a losing strategy. Same with climate. Focusing only on using less energy is a losing strategy. Sometimes you need to spend energy to save energy in the future.
    - atq2119 2y ago
      "Nobody", really? As a bare minimum, many people can choose to take more environmentally friendly vacations. You don't have to go on a cruise, and yet that's a booming industry. You don't have to fly across an ocean. Almost everybody who does these things has the option to go on perfectly fine, perhaps even better, less carbon intensive alternative vacations. And yes, there are people who consciously make that lifestyle change. Not enough, of course. But only a Sith deals in absolutes.
      - RealityVoid 2y ago
        You guys are naive. Sure, it's possible, but most people don't do this, don't want to do this and will not do this. And of those that DO some of these things, a large subset probably does it in a ineffective way that OP so nicely illustrated in the phone/1 extra km by car example. No, the solution must be technical while people are allowed to maintain most of the comfort they are used to. Anything else and you will simply not be able to convince people to do so even if that means burning the world down.

sva_ 2y ago

> For instance, compared to an EUV system employing a 250-watt light source, Canon estimates NIL consumes just one-tenth the energy. I'm not an expert on this but feel like a 250w light is not the major driver of cost in EUV? Or am I misunderstanding this?

szundi 2y ago
Yes, it is 250w output but the efficiency is near zero, really
- dylan604 2y ago
  But you can bake a cake with a mere 60w light. That 250w is also much less than what it takes to run modern GPUs. so it's all relative.
Szpadel 2y ago
AFAIK the main coat of EUV is cost of machines that will be obsolete in few years, so you want to produce as many chips using them as possible during that timeframe, they design a lot of around to maintain near 100% uptime of those machines. This include buffers before and after machines (so any unplanned stalls are mitigated) and technicians trained to do maintenance in F1 pitstop fashion. (source: some tour of some chipmaker I saw online, no longer remember details)
- lnsru 2y ago
  These machines will be not obsolete for very long time. They are extremely rare and expensive. And the most of semiconductors are fabricated on mature nodes anyway.
  - rcxdude 2y ago
    Not obsolete, but the margin on fabs drops off pretty quickly once they're not at the cutting edge, as I understand it, so they need to make back their capital investment fairly quickly.

noduerme 2y ago

> So instead, Canon drew on its inkjet printing know-how to apply the resist in optimum amounts to match the circuit pattern. Hopefully for those prices it will still let you send a fax if you run out of resin.

lnsru 2y ago

We had many experiments with nanoimprint lithography at the university 20 years ago. The resolution was poor and the durability very poor. After dozens of imprints the “stamps” degraded heavily. I am curious if 20 years were enough to fix all the issues and it’s really competitive today.

actionfromafar 2y ago
Wasn't this sort of how Intel got started in the 1970s? Some kind of contact printed ICs?
- formerly_proven 2y ago
  You're probably thinking of contact lithography, where a 1:1 mask is placed directly on the wafer and illuminated. This would've been used for the earliest IC processes, where you'd still be able to see the structures with the naked eye or a loupe.
  - f1shy 2y ago
    That is basically like home grown PCB but at very tiny.
hinkley 2y ago
I would think you’d want to make them out of something crazy like diamond or titanium carbide. What did you make yours out of?
swordsmith 2y ago
Came to the comments to see this. Stamping has all sorts of problems with alignment, stamp resolution, contact quality, etc, it's not clear whether this so called "simplicity" still holds after scaling up the resolution even more.

P_I_Staker 2y ago

[flagged]

gliptic 2y ago
Apparently you don't believe in reading titles either.
smokel 2y ago
An EUV is not a second-hand EV. The term refers to Extreme UltraViolet lithography: https://en.m.wikipedia.org/wiki/Extreme_ultraviolet_lithography https://en.m.wikipedia.org/wiki/Extreme_ultraviolet_lithogra...

usrnm 2y ago

Anything that disrupts the monopoly ASML currently has is a good thing

ClumsyPilot 2y ago
Why do you say that? Does ASML hurt the consumer or the market in some way? Has it been caught using dodgy practices against its workers, like Amazon has, against its customers, like the Wells Fargo has, or something else?
- asdasdsddd 2y ago
  Because its bad to have a single point of failure in our supply chain
hinkley 2y ago
How did that happen? I’ve never been clear on how we got here. Just some vague stuff about how others tried and failed.
- cma 2y ago
  Intel thought the tech wasn't mature enough and decided to wait longer. They and the US were originally part of inventing it in the 90s (EUV LLC public private partnership), but sold it off with terms that allowed us to have export controls over ASML's part of it.
  - hinkley 2y ago
    Man, US tech and the 90’s. Al Gore pushed for The Partnership for a Next Generation Vehicle. They say the only vehicle the PNGV produced was the Prius, and Toyota was not eligible to participate. But they freaked out because the Americans were taking it seriously and created a dream team by recruiting their best people into a new program. Only it was all for show. So TM ended up with a serious car and everyone else with demonstrators.

spuz 2y ago

I wonder how big the wafers can be in the NIL system. It definitely sounds like the larger the wafer, the more problems you will have with deformation, alignment etc. if they have to reduce the wafer size in then that would also affect their ability to compete with EUV.

ov_ov_ov 2y ago
Think is on roll-to-roll flexible substrate, no wafers.

abdullahkhalids 2y ago

OP states that this can go down to 14nm. What I am interested in is whether older and larger processes (say ~50nm) can be done at a much cheaper cost than traditional methods. A lot of stuff simply does not require the most advanced chips.

nabla9 2y ago
Canon has been selling nanoprint litography machines for a long time. FPA-1200NZ2C came out 2015-2016. Press release from a sale 2017 https://global.canon/en/news/2017/20170720.html https://global.canon/en/news/2017/20170720.html
atq2119 2y ago
Note that 14nm processes (which are quite old by now) are not the same as 14nm feature sizes. I'm not sure what these machines are capable of, since some details may well be lost in translation in this kind of publication. And I'm only an interested enthusiast, I don't work in the field directly. But towards the end of the article they talk of targeting 8nm line width in 2028, which is impressive. Maybe this time around NIL actually becomes real for high-end processes?
- pama 2y ago
  These numbers are all mindboggling. I understand that the modern specs for EUV dont mean wire width, but if with this future NIL we truly get down to 8nm wide wires, perhaps we should start counting the number of atoms across the width of the wire (around 30).
- phonon 2y ago
  These are feature sizes. (Comparable to the 13.5 nm of EUV)
shaism 2y ago
The answer almost certainly is no. While lithography is one of the largest single contributor to manufacturing costs, the contribution to overall cost is still far below 10%. And one cannot simply substitute an optical lithography with a nano imprint machine without redesigning some part of the process (etch, metrology etc.). Investing R&D resources for a (best case) 10% reduction in costs while still having a decent probability of failure in a big but declining node is not worth it.

y04nn 2y ago

For everyone interested on technical details of the TSMC EUV process I would highly recommend this CCC talk [1] (From Silicon to Sovereignty: How Advanced Chips are Redefining Global Dominance). [1] https://news.ycombinator.com/item?id=42546231 https://news.ycombinator.com/item?id=42546231

sschueller 2y ago
I knew the process was complex especially with the light source but I didn't realize that diffraction was something they also use which is absolutely insane.
modeless 2y ago
Cool! I uploaded the video to YouTube here: https://www.youtube.com/watch?v=sB-y-tDlOSA https://www.youtube.com/watch?v=sB-y-tDlOSA (It's licensed CC-BY so this should be allowed, and I like having videos like this on YouTube where I can easily watch them from anywhere and add them to my playlists.)
- mjrpes 2y ago
  The "transistors shipped" in the history of computing was an interesting number. In 2024 it is now over 10^24. That's a massive number, more than estimate number of stars in the universe. But, in another sense, still quite small. It finally surpassed Avogadro's number, or 6*10^23 particles. This is the equivalent of a small shot glass filled with water (molecules).

westurner 2y ago

Nanoimprint lithography (NIL) https://en.wikipedia.org/wiki/Nanoimprint_lithography https://en.wikipedia.org/wiki/Nanoimprint_lithography

hinkley 2y ago

I feel like this tech would be better suited to flexible circuitry, because flexible can be continuous feed, and why try to limit or size your stamp to the surface area of a wafer when you could just size it to the width of a spool? Also flexible circuits tend to be at a much larger feature size and so it’s okay if they’re a couple generations behind, but this is still far ahead of printed circuitry.