Future of computers, diamonds or optics?

[Xenocide187]

On another topic, Gibsons put a link on an article about artificial diamonds found here. After reading this, I became intrigued. I currently work in a lab which does wafer production. Currently, we are mostly working with GaAs wafers, which succeeded Si wafers. The company I work for is also conducting experiments on InP wafers, which may replace GaAs. This article suprised me. I have thought that fiber optic technology would be the next big step for technology. Now I'm not too sure. If we do start using diamonds on our semiconductors, will it still use electric signals? What are your theories on this subject?

 

[Jeffyboy]

There is digital light processing but, i'd prefer not to rely on mechanical movement of anything even if it is microscopic it can still wear out. Tough to know where to go... you'd think semiconductors would be the general answer for the next longest time. Quantum computing is a long ways off too.

http://www-03.ibm.com/chips/news/2004/1213_amd.html

http://news.com.com/Intel+to+t...00-1006_3-5327521.html

Here's a start. I agree... silicon has it's limits.
It'll be a matter of processing smarter, rather than faster. Electric Signals can only go so high in frequency before they start skinning on the surface and you get loss. Conductors can only become so small before they become insulators. Tough to say what will happen

Jeff

 

[TuxDave]

Originally posted by: Xenocide187
On another topic, Gibsons put a link on an article about artificial diamonds found here. After reading this, I became intrigued. I currently work in a lab which does wafer production. Currently, we are mostly working with GaAs wafers, which succeeded Si wafers. The company I work for is also conducting experiments on InP wafers, which may replace GaAs. This article suprised me. I have thought that fiber optic technology would be the next big step for technology. Now I'm not too sure. If we do start using diamonds on our semiconductors, will it still use electric signals? What are your theories on this subject?

What is the market for your GaAs and InP wafers? I knew those were used for very high speed specialized circuits like RF or links, but I haven't heard that they were used for microprocessors.

 

[TuxDave]

Originally posted by: Jeffyboy
There is digital light processing but, i'd prefer not to rely on mechanical movement of anything even if it is microscopic it can still wear out. Tough to know where to go... you'd think semiconductors would be the general answer for the next longest time. Quantum computing is a long ways off too.

http://www-03.ibm.com/chips/news/2004/1213_amd.html

http://news.com.com/Intel+to+t...00-1006_3-5327521.html

Here's a start. I agree... silicon has it's limits.
It'll be a matter of processing smarter, rather than faster. Electric Signals can only go so high in frequency before they start skinning on the surface and you get loss. Conductors can only become so small before they become insulators. Tough to say what will happen

Jeff

Perhaps you're right, but MEMs (microelectronic mechanical devices) are playing large roles in low power high quality circuits. I believe one of the sharpest filters can be realized by a passive FBAR device.

 

[bobsmith1492]

GaAs is used for solar cells on spaceships mostly, since it is twice as efficient as silicon (which saves space and weight) but it is incredibly expensive. (just learned this in materials class)

 

[Xenocide187]

We're not given the specifics on what we're actually making, but I also don't want to say anything that might be classified information. I can say, however, that our customers are from the government and commercial companies.

 

[TuxDave]

Originally posted by: bobsmith1492
GaAs is used for solar cells on spaceships mostly, since it is twice as efficient as silicon (which saves space and weight) but it is incredibly expensive. (just learned this in materials class)

Yeah, I agree on the expensive part. The last circuit conference I went to, GaAs and InP were all in the exotic insanely high speed circuits but most was still using silicon.

 

[Xenocide187]

Originally posted by: TuxDave

Originally posted by: bobsmith1492
GaAs is used for solar cells on spaceships mostly, since it is twice as efficient as silicon (which saves space and weight) but it is incredibly expensive. (just learned this in materials class)

Yeah, I agree on the expensive part. The last circuit conference I went to, GaAs and InP were all in the exotic insanely high speed circuits but most was still using silicon.

Hehe, you'd be suprised on how much GaAs and InP circuits are used today.

 

[Megamixman]

Well I know for a fact that Lithography on Diamond is possible. My Dad rambles on and off about random things he sees at the Universities and other Labs. I don?t remember when, but I do remember him mentioning that. So in other words we should have CPU's running over the 100C if artificial diamond production kicks into high speed. That would probably be its biggest market.

 

[bobsmith1492]

What's the point of artificial diamonds? Etching circuits in them? Because last I'd heard, it cost more to manufacture an artificial diamond than to buy one, so why not just use natural ones?

 

[Xenocide187]

"This is very rare stone," he says, almost to himself, in thickly accented English. "Yellow diamonds of this color are very hard to find. It is probably worth 10, maybe 15 thousand dollars."

"I have two more exactly like it in my pocket," I tell him.

He puts the diamond down and looks at me seriously for the first time. I place the other two stones on the table. They are all the same color and size. To find three nearly identical yellow diamonds is like flipping a coin 10,000 times and never seeing tails.

"These are cubic zirconium?" Weingarten says without much hope.

"No, they're real," I tell him. "But they were made by a machine in Florida for less than a hundred dollars."

-Wired.com

 

[jagec]

Originally posted by: bobsmith1492
What's the point of artificial diamonds? Etching circuits in them? Because last I'd heard, it cost more to manufacture an artificial diamond than to buy one, so why not just use natural ones?

nope, it's SIGNIFICANTLY cheaper to manufacture them than to buy them. Or should I say, it's significantly cheaper than what De Beers will charge you after they whip their underpaid miners to death and piss all over their bodies.

I believe the artificial diamond business has an agreement with De Beers not to make gem-quality diamonds. I wish they would though, De Beers is a pretty filthy company and it would show how overrated diamonds are anyway.

 

[Xenocide187]

Originally posted by: jagec

Originally posted by: bobsmith1492
What's the point of artificial diamonds? Etching circuits in them? Because last I'd heard, it cost more to manufacture an artificial diamond than to buy one, so why not just use natural ones?

nope, it's SIGNIFICANTLY cheaper to manufacture them than to buy them. Or should I say, it's significantly cheaper than what De Beers will charge you after they whip their underpaid miners to death and piss all over their bodies.

I believe the artificial diamond business has an agreement with De Beers not to make gem-quality diamonds. I wish they would though, De Beers is a pretty filthy company and it would show how overrated diamonds are anyway.

There are a few companies that are trying to make gem-quality diamonds, a couple are mentioned on the article I posted above. Take a look at it, it's pretty interesting. And it's not really that the artificial diamond business have dealings with De Beers, it's because De Beers is doing their best to put in people's minds mined diamonds are better. Not to mention the fact that they are also investing a lot of money on trying to detect artificial diamonds.

 

[TuxDave]

Originally posted by: Xenocide187

Originally posted by: TuxDave

Originally posted by: bobsmith1492
GaAs is used for solar cells on spaceships mostly, since it is twice as efficient as silicon (which saves space and weight) but it is incredibly expensive. (just learned this in materials class)

Yeah, I agree on the expensive part. The last circuit conference I went to, GaAs and InP were all in the exotic insanely high speed circuits but most was still using silicon.

Hehe, you'd be suprised on how much GaAs and InP circuits are used today.

Too bad you can't tell us... outside of the high speed RF circuits, I would really like to know who else has been using them.

 

[Xenocide187]

I was suprised myself when I saw a comany's name written on the wafer as I was inspecting it with a microscope. Later on I found out other companies and was shocked.

 

[Eskimo]

Compound semiconductors are not used in the vast majority of semiconductor applications of today. Tthere is simply not a need or they siimply do not meet the criteria necessary.

Microprocessors: Substrates are not available in sizes necessary (200/300mm) to make them competitive with silicon. Substrate costs not competitive with bulk silicon, epi silicon, silicon on oxide, or strained silicon. Not compatible with traditional CMOS processing since they do not readily grow a usable oxide (less of a condern now that we are moving into depositied high k gate materials).

Memory (DRAM, Flash, SRAM, EEPROM, etc): Speed simply not needed (except for SRAM). All are commodity parts sold at commodity prices, cannot hope to justify cost of compound substrates.

There are a number of companies that do indeed use GaAs and other compound substrates. Primarily in the production of RF, DSP, LED, Optoelectronics, and high speed communications products. Big companies that use compound semiconductors include TI, Motorola, STMicro, Qualcomm, Jazz Semi, TriQuint, Agilent, Vitesse, Fujitsu and the list goes on.

 

[MadEye2]

Has there been any new information released lately about using organic molecules being used for electronics? I heard this would be the next big thing a few years back but haven't heard anything since.