Anyone interested in discussing the use of light in a processor?

youcef

Beta member
Messages
4
Location
Belgium
Hi everyone

I'm a new member in this board and i hope to make some friends while i'm here.

I was wondering if there were people interested in discussing or giving advice about a project that have.

The project is about the hyphothesis of using light in a processor instead of current. An introduction has been posted on a website dedicated to this project in which I discuss the properties of light that could be used in a processing unit.

What I don't understand is why there is so little info sources about this on the web. I have a hard time finding an audience to discuss this topic which is mind boggeling since we have almost reached the limitations of traditional cpu's with electric transistors.

So if you are interested or if you know people that might want to enter this discussion feel free to share your thoughts on this thread about it.

Ps: is it okay if I post a link of that website?
 
as you haven't hosted your video on youtube, and added it as an attachment it is hard to go back and re-watch with out repeating the whole thing...
so I watched it twice, here are some notes I made...


What do you mean no new breakthroughs are made? - outright saying nothing new is happening shows a lack of understanding in the field,

processors don't opporate in serial all the time (think parallels cores)
and when they do there is a great reason (think race hazzard)



light doesn't bend in crystal liquids, it refracts when it passes through mediums of different densities.


then you post a bunch of what looks like clip art.
what you've labelled as subtraction colour mixing probably isn't applicable to light in the way that you are thinking of transmitting beams of light.

why do you include pictures of the double slit experiment.? - you know what that is and what it shows?

then you go on about if we want to use light as a carrier of information we need to understand how a transistor works...

firstly we already use light as a carrier of information. you never heard of fibre optic?
secondly, you show a picture of a transistor, and say it;s an electronic switch, that's how it functions, not how it works, i.e by having P and N type silicon junctions, looking at electron mobility etc.

then you jump from transistors to logic gates... again how does this apply to light? - had you spend a tiny bit more time instead of looking at schematic pictures of a transistor, and instead worried about HOW they do the thing that they do, you'd see that this isn't quite as simple as you wish!



your final example, showing the light gate is pretty clever as a thought experiment, but I don't think it'll work the way you are thinking,
right now we use gates to look at binary signals,

your device takes two (separate) beams of white light, filters, the beams, to complimentary colours, then combines them, so we see *IF* the resultant beam is white then both inputs are on, but *if* the beam is any other colour then one bean must have been off...

the trouble is that you have two inputs, that you combine to a single input, then you have to split those two combined (hopefully white) inputs into three outputs to check the RGB values to check that the light is white.

you take two easily checkable things, that you could just use a photo transistor to check if the light source is on, you then filter, combine and transmit using expensive options, then at the far side that you omitted, you'd need more optics, more filters and photo transistors to electronically check the result...

meanwhile it's about $2 in parts to check with standard silicon parts.

it's kind of a cool way to transmit multiple parts of information, but then so is multimode fibre optics, which is kind of the same thing, passing different information on different wavelengths... the difference being that multimode fibre works!



I really hope that this isn't a final year thesis project.
 
light doesn't bend in crystal liquids, it refracts when it passes through mediums of different densities.


Refraction is the bending of light (it also happens with sound, water and other waves) as it passes from one transparent substance into another. ... Even our eyes depend upon this bending of light. Without refraction, we wouldn't be able to focus light onto our retina.

Didn't read any further...

Gangstalking = constantly attacking someone

Your post is nothing but cheap provocation: no science in it at all or good intention.
 
The video works fine and you can go to specific sections on my browser (Firefox), don't worry about it.

It's an interesting concept, and one I'm sure that's being looked into.

I'm no scientist, but I guess compared to silicon you have to consider:
- Transmission speed of a single bit - Which light wins by a country mile
- Capability of transmitting multiple bits - Can light pulsate quickly enough? - Again, I'm sure it can.

Inside a processor, I can only surmise that this could be difficult to utilize as not only would the manufacturing process be more intricate, the transmission pipes themselves would have to be hollow which I'm sure would cause a gigantic headache on a chip-scale production.

On a large internet level, I can see how the speed of transmission of a single bit is massively beneficial due to the speed at which light travels, but on the scale of a single processor, I struggle to see the benefit across such a small distance in comparison to current chips.

I think they're primarily looking into materials that can replace silicon by withstanding much greater temperatures, allowing us to effectively shove more power into a chip of the same size with less concern about how the heat is affecting the chip.

The proof of concept already exists for you with fibre optics in networking. I suspect next steps would be to consider the viability of production on such a microscopic scale?

Again, disclaimer, I'm no scientist, just basing my hypothesis purely on what I know about light and current tech!
 
light doesn't bend in crystal liquids, it refracts when it passes through mediums of different densities.


Refraction is the bending of light (it also happens with sound, water and other waves) as it passes from one transparent substance into another. ... Even our eyes depend upon this bending of light. Without refraction, we wouldn't be able to focus light onto our retina.

Didn't read any further...

.
perhaps you should?
maybe I'm wrong... but...

what I mean to say is that when light enters your liquid it will refract in one way, and then when it enters a (more dense) crystal suspended in the liquid it refracts more, then it exists the dense crystal medium back to a liquid and refracts less. You've given no indication of how you plan to align the crystals in the medium, so really your ideal of pass light through some sort of liquid crystal is nothing more than you saying "I know that electronically controllable liquid immersed crystals exist, surely they could be used in some other way..." - and that is not science!
actually you said crystal liquids, and as far as I know, that is not really a thing anyway...

I understand how light is refracted and focused in the eye, critically it is via a process where a lens is controlled, the density of the lens does not change the width of the lens changes, in a VERY controlled way, when that control is lost you pretty much loose sight.
You haven't proposed any control.

Gangstalking = constantly attacking someone
1, That's not what that means,
2, you made one post, I made one reply, I'd hardly say that was "constantly attacking" but, whatever makes you happy.

But I get it, if someone comes and takes a crap on your ideas it's much easier to call them names than defend your ideas.

Your post is nothing but cheap provocation: no science in it at all or good intention
I really think that my post had slightly more science in it than your clip art extravaganza.
For example I actually described how a transistor worked, and explained why your idea does not work.



Clearly we've gotten off to a bad start. as a serious question, is this a project that you hope to explore through academia? like is this a final year project that you want to explore.

Honestly, if this is something that you are hoping to "citizen science" a new technology via random post on forums to create new processors that use light rather than electricity, replacing logic gates based on known quantities with devices that rely on light, then go nuts, I'll happily stay out of the discussion.


if this is a final year project for university making something like three or four modules of marks ~25% of your final year mark... then can you post the video on youtube and provide a link. then I'll make my post a little more polite, and include time codes for parts that I'm asking questions on.

Then we can get into times that you say things like light bends, but if you're writing a paper, you should really use terms like refracts.
and lot at your light gate setup in more detail (since that's the only thing that currently you have proposed that goes towards making an actual processor...)
 
The video works fine and you can go to specific sections on my browser (Firefox), don't worry about it.

It's an interesting concept, and one I'm sure that's being looked into.

I'm no scientist, but I guess compared to silicon you have to consider:
- Transmission speed of a single bit - Which light wins by a country mile
- Capability of transmitting multiple bits - Can light pulsate quickly enough? - Again, I'm sure it can.

Inside a processor, I can only surmise that this could be difficult to utilize as not only would the manufacturing process be more intricate, the transmission pipes themselves would have to be hollow which I'm sure would cause a gigantic headache on a chip-scale production.

On a large internet level, I can see how the speed of transmission of a single bit is massively beneficial due to the speed at which light travels, but on the scale of a single processor, I struggle to see the benefit across such a small distance in comparison to current chips.

I think they're primarily looking into materials that can replace silicon by withstanding much greater temperatures, allowing us to effectively shove more power into a chip of the same size with less concern about how the heat is affecting the chip.

The proof of concept already exists for you with fibre optics in networking. I suspect next steps would be to consider the viability of production on such a microscopic scale?

Again, disclaimer, I'm no scientist, just basing my hypothesis purely on what I know about light and current tech!

Indeed what they are currently researching is hybrid processors in which they only want to solve the heat issue. This would be the ideal solution since they don't have to change anything about the architecture itself.

However heat build up is not the main issue it's the serial mode of a CPU and the use of the ascci code which is limited to 0 and 1 states: that truly limits computational power.

There is a quasi infinite amount of color contrasts which each represent a proper wavelenght each of them could be it's own unique input signal which would result in millions if not billions of different bits of input rather than just 0 and 1...

The main advantage of light is not necesarily it's speed but the fact that it can be used in a paralel mode effectively achieving quantum computing.

The only difficulty would be to create a frequency.
 
If we can manipulate light the way we manipulate electricity that will be a big advancement in micro technology.

The biggest advantage of this is heat control and speed.
 
But they can (and do) already transmit multiple frequencies at once sending different information on different wavelength carriers. - look up multi mode fibre.

It's quite hard to break out of binary thinking, especially as most number systems rely on it in computing. (Using schemes like twos compliment.)

I guess the exciting think about this is that a RGB colour scheme allows what it currently three binary states of data to be transmitted in a single (whatever system this is...) bit.
 
Back
Top Bottom