> > > >
> >>John Lark> >>
> >>[...]
> >>
> >>> What stuns me is that the various parts of a modern wafer scanner
> >>> (formerly called a stepper) like mask, lens, and wafer, are in
> >>> continuous motion during exposure. To sub-nm precision.
> >>>
> >>> And euv is not even coherent light, just an emission line from blasted
> >>> tin droplets.
> >>
> >>Thanks. I often wondered how they could start and stop a massive moving
> >>weight, then wait for the vibrations to cease, expose and start moving
> >>again. Repeat every 5nm. That is not going to work.
> >>
> >>The solution is don't stop. Why didn't I think of that.
> >>
> >>Now you need femtosecond accuracy in the clocks. That brings it to
> >>Time-Nuts domain:
> >>
> >>formatting link
> >
> > The lens of an EUV system looks like a diesel engine. Nothing refracts
> > 13 nm light so the lens is a mess of shaped grazing-incidence metal
> > things with movers and benders. I think Zeiss makes it.
> Apparently the beam is focused and directed by mirrors
>
> How Carl Zeiss Crafts Optics for a $150 Million EUV Machine
>formatting link
>
> The next level is 1nm. I wonder if they use the same method.
>
> 1nm chip manufacturing factory! TSMC has once again set a world first!
>formatting link
Someone commented: "This is a totally fabricated story. TSMC has never announced anything about the 1 nm node."
I agree. 3nm is more likely.