read "The Diamond Age" by Neal Stephenson for continuation of this story, and because it's an awesome book.

Sweet, now I can finally manufacture my "Feeze Ray" like Arnold did in Batman and Robin!

You suck. I was looking for this clip right after it aired. Unfortunately it took a while for them to post it. Anyway, it's an awesome idea and I'm glad you found it.

...now about that rash?

Um, except all modern electronics are based on semi-conductors.. which diamonds are not. Can you dope diamond wafers with III-V? Can you grow Si on diamond? Otherwise the heat dissipation is useless. Plus the whole point is growing one 12" diameter silicon slug...

The only possible application would be to take s fully processed silicon wafer and then thin the back of the wafer to insanely thin, then CVD the synthetic diamond to act as the substrate then package it in something that takes heat out better.

They have an inherently superior material that can take away the overheating problem engineers face today at every aspect of miniaturization and you're presenting a technical difficulty to the manufacturing process...

From wikipedia:
Natural blue diamonds containing boron and synthetic diamonds doped with boron are p-type semiconductors. N-type diamond films are reproducibly synthesized by phosphorus doping during chemical vapor deposition. Diode p-n junctions and UV light emitting diodes (LEDs, at 235 nm) has been produced by sequential deposition of p-type (boron-doped) and n-type (phosphorus-doped) layers.

http://en.wikipedia.org/wiki/Material_properties_of_diamond#Electrical_properties

Theres your answer joe.
You can dope diamonds and make them semi-conductors.

You are also right, this would make a fantastic substrate if they could grow Si. on them in the mean time.

The point is that feature size in silicon is getting to be as small as it possible can. (when a gate is only a few atoms wide, it cannot withstand manufacturing variability, and so the process is unreliable).

So when you can't shrink any smaller, the performance of the transistors will reach a maximum limit, which is determined by the material properties of the substrate and doping elements. (which is why SiGe is faster than Si).

Artificial diamond substrates (replacing silicon) promise the next "leap" in integrated circuit performance. It's just a matter of making the manufacturing process cheap and the product reliable. When I say cheap, I mean on the same order of magnitude of a 32nm process fab ($3-$7 billion). It might not happen anytime soon (20+ years from now).