Optical Interference as Logic gates

[neptune]

Here this patent talks about using optical interference as basic logic gates (AND,OR,NOT,XOR)
I think that's cool. its very simple to understand and make.
https://www.google.com.ar/patents/US20130094070
full PDF -
https://www.google.co.in/url?sa=t&rct=j&q=&esrc=s&source=web&cd=2&ved=0CCIQFjAB&url=https://patentimages.storage.googleapis.com/pdfs/US20130094070.pdf&ei=mBB3VJbJJ9iTuATHgYKwDA&usg=AFQjCNF_WjFZjmw0W5JLHEQeU0JqzWTN-w&sig2=KGmN1wLOZdx_8zWmQMwInA
if its working then why didn't we see optical computers yet.
this patent was filed in 1992, 20 years old.

 

[alec_t]

if its working then why didn't we see optical computers yet.

Probably because it's a lot cheaper/easier to use solid state electronics.

 

[neptune]

I think its because miniaturization is not possible, the interference requires a great distance to see the perfect pattern. what do you say ?

 

[steveB]

One of the things people forget about when they talk about optical computing is that what people have done with integrating circuits in silicon is truly amazing. A transistor does not make a computer and a gate does not make a computer. It takes many of them integrated in a controlable way at small scale to make a computer with impressive computational power.

You can make gates in many ways with almost any technology, but the issue is always whether you can integrate the approach at small scale and with high yield.

 

[neptune]

hey what is high yield ?

 

[steveB]

Yield is the percentage of manufactured devices that actually work.

 

[neptune]

one more question. what will be the minimum criterion that optical computer would need to full fill if it needs to come into the eyes of companies, research labs, universities etc. what be the program that it needs to run so that world takes it seriously.
I am thinking of the possibility of miniaturization, the power consumption, the heat dissipation (but I don't know by how much), modularity and interface with current semiconductor chips, on the software side I am betting on sorting algorithms, what do you say ?

 

[alec_t]

what will be the minimum criterion that optical computer would need to full fill if it needs to come into the eyes of companies, research labs, universities etc.

Compared to existing electronic computers it would have to be cheaper, faster, more energy-efficient, easier to manufacture, less noise susceptible, operable over a wider temperature range, more radiation-resistant, and/or offer some other advantage.
Why do you think sorting algorithms would be performed better optically?

 

[neptune]

I read that big companies have big databases, which is ever growing, they always need to sort the information in some order but the traditional computers takes much time.
here they have sorting competition also look
https://www.jacobsschool.ucsd.edu/news/news_releases/release.sfe?id=967
apart from this I couldn't think of any program that would measure speed

 

[MrAl]

Hi,

They (Japan) are currently working on using silica glass made using a new process for various applications including precision optics for telescopes and they think they can use it for integrated circuits too. Before that the cost of making optics with silica glass was extremely high but the new technique will bring the cost way way down, which may make this kind of thing possible.

As an interesting side note, one of the companies was working on a new fuel cell driven car for some 20 years and now finally this December (which actually is this very month) a new line of fuel cell driven cars will be rolling off the assembly line. Probably still upwards of 50 grand or more but it's an interesting thing to see finally happening.