Experimental CPU with integrated optical logic promises significant power reductions

[Toddwjp]

on a much bigger scale it is definitely possible, the problem is shrinking it down... Like I said...

 

which is something we will get too like i said, discussing future potential not what we can do now, which is why i dont get why you were disagreeing so much 

[ZetZet]

which is something we will get too like i said, discussing future potential not what we can do now, which is why i dont get why you were disagreeing so much 

Because I don't think you understand how these two technologies work. I am trying to explain to you that in the universe we currently live in, we can't possibly reduce light emitting and receiving devices to be smaller than a simple switch.

[Toddwjp]

Because I don't think you understand how these two technologies work. I am trying to explain to you that in the universe we currently live in, we can't possibly reduce light emitting and receiving devices to be smaller than a simple switch.

 

that is not what you said and still not right, you said its against the laws of physics and again you are saying the universe we currently live in. 

 

Yes right now we cannot build it but the issue is not one of physics being wrong but material and engineering constraints on what we are able to build, there are mathematical models and proof of concepts out there for exceedingly small optical transistors (the main part of any cpu) while not as quite as small as current transistors they are close, which i why i was talking about bigger cpus.

 

Also the light emitting device and light receiving device does not need to be on the CPU just like they are not on there right now either. I think you need to look into optical computing a little more and realise that it is theoretically possible but not currently manufacturable. there is some doubt in how close it will get to standard semiconductors now but there is still room for breakthrough and as we are discussing the future you are allowed to assume a few problems to be fixed here and there otherwise its impossible to decide on whether its worthwhile attempting to solve them.

[ZetZet]

that is not what you said and still not right, you said its against the laws of physics and again you are saying the universe we currently live in. 

 

Yes right now we cannot build it but the issue is not one of physics being wrong but material and engineering constraints on what we are able to build, there are mathematical models and proof of concepts out there for exceedingly small optical transistors (the main part of any cpu) while not as quite as small as current transistors they are close, which i why i was talking about bigger cpus.

 

Also the light emitting device and light receiving device does not need to be on the CPU just like they are not on there right now either. I think you need to look into optical computing a little more and realise that it is theoretically possible but not currently manufacturable. there is some doubt in how close it will get to standard semiconductors now but there is still room for breakthrough and as we are discussing the future you are allowed to assume a few problems to be fixed here and there otherwise its impossible to decide on whether its worthwhile attempting to solve them.

do you not see the faults in your own reasoning. There is no way we get optical on the same level as semiconductor and we still keep shrinking our process. We don't need a bigger cpu, it's not efficient, it never will be. Optical cpu has no useful applications whatsoever.

[Toddwjp]

do you not see the faults in your own reasoning. There is no way we get optical on the same level as semiconductor and we still keep shrinking our process. We don't need a bigger cpu, it's not efficient, it never will be. Optical cpu has no useful applications whatsoever.

 

i was discussing that an optical cpu will be bigger but may use the same or less power as a semi conductor equivalent and thus potentially be easier to make in the future and cost less if it uses less power. saying it has no useful application when we haven't even got a viable test bed yet to conclude if it will be worthwhile is just being a close minded

[elfensky]

TL;DR

 

New experimental chip uses fiberoptics & light instead of the normal "physical" connections(cables/motherboard circuits) that use electrons to communicate. Something along these lines, if I understood it correctly.