Light Processors

[SilentRavens]

Just out of curiosity; just how possible is a light processor. As in small lasers would fire white light down fiber optic channels. This light would then operate light XOR gates with the minute pressure of photons, and of course this would require a light computer to take advantage of, but if a light processor could be done the rest would be relative easy.

Just wondering...

 

[Shalmanese]

The biggest problem with light processors is having Finding a way to get "light memory". We can construct relatively simple gates and stuff but we are still relying on ordinary DRAM for storage and cache which slows the thing down.

 

[Mday]

this is what they are doing now. they have nano tech mirrors in them. they are still in the labs. btw.

 

[User1001]

are the mirrors easy to break through shaking?

 

[AbsolutDealage]

Originally posted by: User1001
are the mirrors easy to break through shaking?

In short, No. There is plenty of info about this all over the web. One of the more interesting applications of this technology, IMO, is here.

Google on MEMs (Microelectromechanical devices), nanotechnology mirrors, and DLP for more info.

 

[Cogman]

Light would be awsome to say the least. I think that we could remove a lot of the heat issues and be able to run a computer with minimal wattage. be like Shalmanese said, the jump to "Light Speed" would be pointless in the CPU only as any non light device would severily hurt the preformance of the system. be what about quantum computing? which would be superior? I fear that in which ever goes public first will become the new standerard for home computers and the other will die out. (out of light and quantum computing)

 

[SilentRavens]

About light memory

I read somewhere that use can use holographs to hold data, this would have do been done in or on some sort of reflective surface. I guess this memory would have to be manipulated by colored lasers, and then read using low intensity white light. The problem I see with this is that the physical size of these memory devices would go back to when computers were still using vacuum tubes. Of course "memory" back then was in the form of paper-punch cards, but the reference still stands. How this could possibly be integrated into something as small a processor cache is beyond me.

An interesting thought about holographic memory though, is that in holographic form you can hold more than two positions ? on/off. If you change the structure so that in stead of being a magnet with north and south; you have north, south, east, and west, thus four positions as opposed to two. It gets more complicated if you then combine the directions getting you eight possible positions. The next step would be then to be able to place the north/south magnet in any place or position, and as a result creating a holographic sphere that can hold an (in theory) infinite number of positions.

 

[jswjimmy]

its a cool idea but its to sast for the ram

 

[User1001]

Intel is developing lithography based proccesors. I read somewhere that the production would start in 2005.

 

[Sohcan]

Originally posted by: User1001
Intel is developing lithography based proccesors. I read somewhere that the production would start in 2005.

You might want to check your sources...we started producing lithography-based microprocessors in 1971.

 

[User1001]

Originally posted by: Sohcan

Originally posted by: User1001
Intel is developing lithography based proccesors. I read somewhere that the production would start in 2005.

You might want to check your sources...we started producing lithography-based microprocessors in 1971.

I'm sorry. I meant EUV Lithography.

 

[OddTSi]

Originally posted by: User1001
I'm sorry. I meant EUV Lithography.

That's just a different/better method of creating current processors. Has nothing to do with optical processors.

 

[User1001]

Extreme Ultraviolet lithography is truly a technology breakthrough. The 13.4 nm wavelength of this light is more than 10x shorter than optical alternatives and will allow the patterning of lines below 50 nm dimensions. Light at this wavelength does not travel through the atmosphere and is absorbed by glass. Therefore, much of Intel's research on EUV is directed toward producing efficient light sources, all-reflective optics, and novel reflective masks to support this technology. Intel is leading a consortium of six semiconductor companies, called the EUV LLC (Limited Liability Corporation), to develop this technology. The LLC announced completion of the first full-scale prototype machine for making computer chips using EUV lithography on April 11, 2001. Processors built using EUV technology are expected in 2005.

it uses low wavelength light

 

[Falloutboy]

that refers to the light doing the etching not how the processor works

 

[RadioactiveHamzter]

Anyone remember the research done on 'stopping' light? From what I remember they fired a burst of photons carrying information through an odd gas, that was opaque normally, but turned transparent when a second laser was fired at it. Anyway, the photons travelled into the gas by shining the laser at it, then stopping the second laser so the photons were 'frozen' in the gas. Switching the laser on again allowed the photons to continue on their path, still carrying the information.

Could this be somehow implimented as RAM for these light processors? It'd obviously be pretty big and difficult to use, but would it be possible, or even at all practical?