Can Substrate disrupt ASML using particle acceleration?(nytimes.com)
nytimes.com
Can Substrate disrupt ASML using particle acceleration?
https://www.nytimes.com/2025/10/28/business/can-a-start-up-make-computer-chips-cheaper-than-the-industrys-giants.html
12 comments
I hope this is not the case but a charismatic, politically connected college dropout in the past promised to revolutionize a multi billion dollar industry and ended up going to jail for fraud (Elizabeth Holmes).
If anything, the most complicated problem in lithography is not EUV at all. The hardest problem is the overlay mismatch. ASML had spent decades minimizing overlay mismatch. A new player in semiconductor may claim to master EUV, or X-Ray or contact printing, but everyone must solve the overlay mismatch problem. That is no more easier than the EUV part.
For those who don't know, ASML didn't invent EUV lithography. It was invented in the US, paid for largely by the US government, and licensed to ASML for commercialization.
https://en.wikipedia.org/w/index.php?title=ASML_Holding§...
https://en.wikipedia.org/w/index.php?title=ASML_Holding§...
I don’t have a times subscription but this looks to be Xray lithography using a synchrotron to produce the X-rays? Can anyone confirm?
https://archive.ph/xKqnk
There is also this from out of China: https://pubs.aip.org/aip/app/article/10/8/086105/3357991/Ele...
Via: https://phys.org/news/2025-10-electrohydrodynamics-machine-e...
Not to forget https://global.canon/en/technology/nil-2023.html , https://www.usa.canon.com/shop/business/semiconductor/lithog... which recently went online in https://www.tomshardware.com/tech-industry/canon-delivers-fi... , https://www.eenewseurope.com/en/canon-delivers-5nm-capable-n... , https://global.canon/en/news/2024/20240926.html , https://spectrum.ieee.org/nanoimprint-lithography
OTOH: https://newsletter.semianalysis.com/p/nanoimprint-lithograph...
And whatever https://www.searchforthenext.com/ & https://semefab.com claim for their stuff
There is also this from out of China: https://pubs.aip.org/aip/app/article/10/8/086105/3357991/Ele...
Via: https://phys.org/news/2025-10-electrohydrodynamics-machine-e...
Not to forget https://global.canon/en/technology/nil-2023.html , https://www.usa.canon.com/shop/business/semiconductor/lithog... which recently went online in https://www.tomshardware.com/tech-industry/canon-delivers-fi... , https://www.eenewseurope.com/en/canon-delivers-5nm-capable-n... , https://global.canon/en/news/2024/20240926.html , https://spectrum.ieee.org/nanoimprint-lithography
OTOH: https://newsletter.semianalysis.com/p/nanoimprint-lithograph...
And whatever https://www.searchforthenext.com/ & https://semefab.com claim for their stuff
Thanks for the link. What I'm wondering is how this guy aims to succeed where previous X-ray lithography efforts failed. For reference X-ray litho was competing with EUV. What was it that held the approach back? My guess would be that that the industry simply coalesced around EUV and X-rays were abandoned. But that just makes me wonder why the industry coalesced there. Perhaps synchrotons are not cheap / easy? But don't university labs build these all the time?
Prof. Hank Smith of MIT spent decades trying to push X-ray lithography but yes, the industry coalesced around DUV and then EUV.
https://dspace.mit.edu/bitstream/handle/1721.1/16801/5050370...
https://dspace.mit.edu/bitstream/handle/1721.1/16801/5050370...
As far as I understand it real X-Rays are too hard to handle in general, and though more precise because of shorter wavelengths etch too much away from the substrates. EUV is just soft enough, to not do that(as much). That seems to be the main reason all the effort went into EUV of 'just the right wavelength and energy.'
That searchforthenext link shows an illustration captioned "Conventional CMOS unipolar NOR gate."
It's actually a NAND gate. Seems like a pretty big red flag if they want to revolutionize transistors but they can't even get super basic stuff right.
It's actually a NAND gate. Seems like a pretty big red flag if they want to revolutionize transistors but they can't even get super basic stuff right.
It claims that they can print 12nm features with their particle accelerator. This looks weird.
We need to see their detailed implementation strategy to understand whether this can ever be viable. If it is, you can be sure that ASML is already working on it.
We know it can't be used in a fab because the mask in EUV scanners is in a hydrogen ambient to protect the optics and masks whereas the synchrotron is in a perfect vacuum. Now EUV has tons of problems with stochastics and as the wavelength decreases these errors increase. Second 13.5nm has optics with highest reflectivity which lower or higher wavelengths do not. It was tried by IBM but it failed and many in IBM and AMD's lithography team ended up at ASML.