Diamond-based CPU's: Advantages & Applications in Physics

[albert281]

So I was looking online at a company called Apollo diamond...they have a method for making diamond called chemical vapor deposition. Essentially, it is a method where a seed diamond is placed in a carbon gas and layer upon layer of carbon atoms are deposited on the seed and the diamond grows accordingly.

Currently the semiconductor industry uses silicon as its medium of choice...but diamond is a better semiconductor. Given the high melting temperature of diamond, as opposed to silicon, clearly CPU's would no longer be constrained at running around 4Ghz...and with multiple cores, we are talking about computing power increases that are almost unimaginable.

Does anyone know of any applications in physics that such a semiconductor would be advantageous over current designs? I am thinking high temp environments...but surely there are other applications?

 

[Vagn]

So I was looking online at a company called Apollo diamond...they have a method for making diamond called chemical vapor deposition. Essentially, it is a method where a seed diamond is placed in a carbon gas and layer upon layer of carbon atoms are deposited on the seed and the diamond grows accordingly.

Currently the semiconductor industry uses silicon as its medium of choice...but diamond is a better semiconductor. Given the high melting temperature of diamond, as opposed to silicon, clearly CPU's would no longer be constrained at running around 4Ghz...and with multiple cores, we are talking about computing power increases that are almost unimaginable.

Does anyone know of any applications in physics that such a semiconductor would be advantageous over current designs? I am thinking high temp environments...but surely there are other applications?

At a guess, I'd say that the main factor would be the price, I'd say the process to make a semiconductor from silicon is probably cheaper than the process for diamonds.

 

[albert281]

At a guess, I'd say that the main factor would be the price, I'd say the process to make a semiconductor from silicon is probably cheaper than the process for diamonds.

I disagree that price in manufacturing would be a problem...as it is, prices for chips are mostly for the R&D into the design, then of course operating profit and anything else they can get from the consumer. The same would hold true for diamonds. Think about it...a small seed diamond can be wafer sized merely by depositing carbon from a gas. The cost for a one carat white diamond of exceptional clarity should drop to a very low value. De'Beers may have something to say about this in that the guys at apollo may have to hire bodyguards (if they already haven't), but cost shouldn't be an issue when manufacturing at this scale. Incidentally...diamonds are only as expensive as they are because of hoarding by De'Beers.

 

[Integral]

You are missing a major factor. In a few hours they pull a large Si ingot from a furnace full of molten Si. This ingot it then sawed into hundreds of SI wafers.

In that same few hours a SINGLE diamond wafer would be created by deposition in a vacuum chamber. So not only is the process itself more expensive for diamond then Si it is also much slower. Further it is likely the deposition process is in need of a significant amount of development in order to produce a diamond wafer.

I would not count on diamonds stealing the wafer market from Si just yet.

 

[albert281]

You are missing a major factor. In a few hours they pull a large Si ingot from a furnace full of molten Si. This ingot it then sawed into hundreds of SI wafers.

In that same few hours a SINGLE diamond wafer would be created by deposition in a vacuum chamber. So not only is the process itself more expensive for diamond then Si it is also much slower. Further it is likely the deposition process is in need of a significant amount of development in order to produce a diamond wafer.

I would not count on diamonds stealing the wafer market from Si just yet.

No doubt there is some work left on manufacturing silicon wafers...but theoretically, what would stop them from laying out a football field of seed diamonds and building a giant sheet of wafers? I don't think there is a need for a significant amount of development for building a wafer...they are already selling diamonds using this process for jewelry. Technically, there should be no reason, aside from engineering, that this method couldn't be made economically feasible.

 

[albert281]

Let's get this thread back on the original question please...

 

[NobodySpecial]

No doubt there is some work left on manufacturing silicon wafers...but theoretically, what would stop them from laying out a football field of seed diamonds and building a giant sheet of wafers?.

CVD is used to grow Sapphire (AlOx) for some image sensors.
Diamond has some advantageous as a base for ICs - because of it's high thermal conductivity - but I haven't heard of any big advantages as an actual semiconductor.

 

[Kmenex]

"Does anyone know of any applications in physics that such a semiconductor would be advantageous over current designs? I am thinking high temp environments...but surely there are other applications?"

This is the original topic of the post. Indeed it is a good one. There is a growing body of work concerning the issues of diamonds for signal transduction. I think another good property is that since they are made of carbon there is an extra advantage to any type of organic coupling. However i am not knowledgeable about the physics too much.

 

[albert281]

CVD is used to grow Sapphire (AlOx) for some image sensors.
Diamond has some advantageous as a base for ICs - because of it's high thermal conductivity - but I haven't heard of any big advantages as an actual semiconductor.

The main advantage in the semiconductor industry, I believe, would be the thermal characteristics. Running a CPU at 81Ghz is very alluring.

 

[Gordianus]

As far as I know, the problem with diamond is about the impurities needed to turn it into P or N type. The impurities that match the lattice constant provide energy levels too far away from the conduction or the valence bands. Thus, at room temperature there is very little ionization and the conductivity is too small to manufacture a practical doped semiconductor.

 

[albert281]

As far as I know, the problem with diamond is about the impurities needed to turn it into P or N type. The impurities that match the lattice constant provide energy levels too far away from the conduction or the valence bands. Thus, at room temperature there is very little ionization and the conductivity is too small to manufacture a practical doped semiconductor.

I don't know anything about this...but I do know that a transistor has been made from diamond doped with boron. The latest news I found regarding diamond transistors:

http://www.physorg.com/news158946763.html
http://www.theinquirer.net/inquirer/news/1044396/-diamond-transistor-clocks-120ghz

 

[albert281]

"Does anyone know of any applications in physics that such a semiconductor would be advantageous over current designs? I am thinking high temp environments...but surely there are other applications?"

This is the original topic of the post. Indeed it is a good one. There is a growing body of work concerning the issues of diamonds for signal transduction. I think another good property is that since they are made of carbon there is an extra advantage to any type of organic coupling. However i am not knowledgeable about the physics too much.

This, if it is true, is very exciting! Do you have any more information on this?