Intel is has 65nm cores on the market now and AMD will be following suit soon. Both companies have or nearly have the capability to begin taping out 45nm cores and Intel hopes to have 32nm in production by 2009. IBM mentioned a month or two ago that they have begun working on lithography in the <20nm range. How long until we reach picometer lithography? Would engineering on such a small scale be possible with current SI-based substrates?
![[H]ard|Forum](/styles/hardforum/xenforo/logo_dark.png){kind=logo}
Picometer lithography
- Thread starter Order
- Start date
Eva_Unit_0
[H]ard|Gawd
- Joined
- Jun 1, 2005
- Messages
- 1,991
Order said:
you can't. Physics only allows you to go so far. The wavelength of the light itself becomes an insurmoutable obstacle. In order to go much below the 20nm mark you need to start going below the shallow UV down into X-ray land, but unfortunately there is no known way of making a photoresist that would work with an X-ray source.
Now of course other cool techniques come into play such as immersion lithography which let you use a particular wavelength farther than you should be able to, but they only help so much. I honestly don't think we'll ever really be able to go below 20nm, at least not on silicon using optical lithography.
EDIT: however, ever since the dawn of semiconductors engineers have been screaming "omg physics won't let us go below xxxum or xxxnm. and every time we've found a way. So who knows. However, I don't think we'll see picometers. You're talking sub-angstrom lengths, which obviously can't work. The atomic radius of a silicon atom is 110 picometers, so somehow I suspect that on silicon that would be as small as you could possibly go.
Great response, Eva. Thank you very much for it .
My guess is that you're right. By the time we get near that scale, anyway, we should be getting around to quantum computing which, from what I've understood to be the case, operates in a completely different way than our SI-based technology. Once that has come about, I figure there will be another scale on which we can quantify technological progress.
.My guess is that you're right. By the time we get near that scale, anyway, we should be getting around to quantum computing which, from what I've understood to be the case, operates in a completely different way than our SI-based technology. Once that has come about, I figure there will be another scale on which we can quantify technological progress.