How hyped should consumers be for the full adoption of this technology?
Yields? Costs?
Quality? Longevity?
Leaps to better more efficient chips?
Extreme Ultraviolet Lithography
Jeremiah Price
Other urls found in this thread:
semiwiki.com
twitter.com
Leo Harris
noone on g knows about this shit
if they did theyd be making 100k a month and definitely wouldnt post here with us faggots
Benjamin Reyes
But it's just consumer enthusiast placebobullshit. Surely some nerds are vaguely excited and know what benefits it will bring.
James Baker
exactly
you'll get consumer placeboshit here
nothing actually useful or real
William Sullivan
>How hyped should consumers be for the full adoption of this technology?
Not at all hyped.
The average consumer has no idea what EUVL is, and likely never will.
Nor should they.
Daniel Cox
Oh well
Ayden Martin
That "light source" is drops of molten tin being vaporised by a CO2 laser to form a tin plasma that emits EVU light.
Asher Fisher
so it's sputtering?
Caleb Gomez
>How hyped should consumers be for the full adoption of this technology?
What would the average consumers do with this? Or do you mean for hobbyist?
Luis Murphy
>>Yields?
low
>>Costs?
high
Personally I'm holding out for EXTREME EXTREME EXTREME EXTREME ultraviolet lithography and by that I mean gamma ray lithography. We already use x-ray lithography, next logical step is to use gamma rays.
Charles Morales
No. It's not sputtering, that's how they make the EUV light. Consider how fucking crazy that is for making light.
Jordan Campbell
No one should be hyped for EUV until 5nm, 7nm EUV processes have very limited EUV steps.
5nm will be a big inflection point in the market again. The nodes should have far better yields than current 7nm lines. Even as wafer costs increase further there should be a net decrease in total cost to clients by eliminating heavy use of multipatterning.
For TSMC at least they expect a 45% reduction in area, potentially up to an 80% improvement in xtor density. Power at ISO performance improvements of 20%, possibly higher, and it should clock at least 10% better.
Of course actual performance metrics depend on the design of the chip being run on the process, but this is an incredible uplift all around versus current processes.
semiwiki.com
Luis Campbell
what good does increasing frequency do for lithography?
Aaron Sanchez
>increasing frequency
decreasing wavelength
Aiden Wright
Those gains are insane. All hail human engineering.
Decreasing size? (layman)
Chase Edwards
>Longevity?
I'm interested in this, do 7nm chips blow up sooner than a 22nm?
Jordan Edwards
Electromigration increases with the electric field strength and increasing temperature. Voltage isn't decreasing much after the end of Dennard scaling and the distances are smaller in 7nm, the increased resistance in smaller conductors lead to local heating.
TL;DR yes.
John Long
>Decreasing size? (layman)
Yes. EUV have a basic problem with "fuzziness" so the decreased wavelength from DUV isn't helping as much as it could appear. Still much better than quadruple patterning.
Jack Jones
Frequency is essentially the same thing as wavelength. And wavelength is literally the physical size of the light thats being used in lithography. You should read into diffraction gratings and physical spectrum separation. Its pretty interesting.
And basically the size of the light being used is their minimum pitch they can etch the resists off. Its the smallest they can make any individual detail.
Jacob Thompson
>Yes. EUV have a basic problem with "fuzziness" so the decreased wavelength from DUV isn't helping as much as it could appear. Still much better than quadruple patterning.
Its because the EUV spectrum created by the tin vaporization isn't as tight as what one would get from a laser.
I'm not sure if a free electron laser could even generate an EUV beam in the intensities required for photolithography.
Chase Lewis
Everyone should be exceptionally hyped. We're guaranteed to hit the absolute physical limits of integrated circuits in the next 10 years at this rate.
Benjamin Reyes
Get ready for quantum junctions
Christian Cox
Yeah maybe if magic gets invented and they get that shit to work without cryogenics
Grayson Kelly
Quantum junctions have nothing to do with Qbits. It does not require near absolute zero temps to operate.