Paint-on Laser Brings Optical Computing Closer

holy_calamity writes "New Scientist has a story about a laser made by painting a solution of semiconductor crystals onto glass. It could be used to break the interconnect barrier by having optical interconnects, the interconnect barrier threatens Moore's law unless a faster way of connecting chips is found."

Re:Whaaah? Maxwell 101

[pla]

It is not the travelling of electrons that gets the electric signal propagating.

The travel itself, no. The wavefront of "pressure" moving along the path of the electrons, yes. The electrons themselves move at only (depending heavily on current and wire diameter) around 1-10cm per hour.

But the wave still only travels somewhere between .25 and .75c (IIRC, hams use a factor of around .7c for the speed in a whip antenna, while tiny ethernet strands only give around .33c).

Does the difference there really matter all that much? For long-distance communication, sure. But for chip interconnects? Doubtful.

Re:Whaaah?

[955301]

yes. inductance slows electrons down and electrical traces can't touch each other so they have to be drawn around each other - laser light beans can pass through one another with no interference. So the traces can be more direct and hence faster. Finally, the scale of components in a processor has gotten small enough that individual traces are interfering with one another inductively and on a quantum level - these don't happen with light.

Quantum computers complement digital ones

[Hootenanny]

The intention of quantum computing is not to replace, but rather to complement classical (i.e. digital) computing. Quantum computing can dramatically speed up certain tasks, such as cryptography and searching. Even though they cannot yet be implemented, a number of important quantum algorithms have already been discovered. Most - but not all - quantum algorithms return probabilistic answers, rather than clear-cut answers as most classical algorithms do.

Shor's algorithm for factoring numbers could be used to rapidly crack RSA encryption. http://en.wikipedia.org/wiki/Shor's_algorithm [wikipedia.org]

Grover's algorithm can be used to search an unsorted database in O(n?2) time. http://en.wikipedia.org/wiki/Grover's_algorithm [wikipedia.org]

Speed increase

[centie]

The article and summary seem to be a bit misleading and vauge about how the speed increase arrises. The great benefit of optical computing is that it allows the signals to get much much closer together than electronic circuits, and as such allow more compact circuits, which as we know generally means faster. Interestingly, electronic signals in wires and optical signals in fibers have roughly identical upper speed limits (light in free-space optical computers is faster, but also almost impossible to do anything useful with), so its the density which is the major factor.

Electrons are charged, so as you squeeze transistors closer together, the wires get thinner and closer together, and you get cross-talk and interference between them. Photons however hardly interact at all, so you can have many beams in the same space, and theres very little heat to be dissipated. Multiplw frequencies can also be used, resulting in massivly parallel computing (another GoodThing).

There are downsides with optical computing still, photons cannot be stopped and stored (easily), meaning any kind of useful computer in the near term is likely to be some sort of electro-optical hybrid, with photons carrying signals and electrons storing them

Re:Whaaah?

[fyndor]

Yes and no. It depends on what the electricity is traveling through (freespace? wire?). Electricty generates heat and noise. Now assuming it takes less electricity to power the lasers then it will generate less heat. I think noise is the main reason. Two wires (general term, wires...traces..) with current flowing through them will affect each other, hence putting noise on each others line. This is ok at low frequencies (slow speeds), but at higher frequencies it distorts the signal up to a point of unusability. So if you replace your wires with light you dont have this problem because two light pipes next to each other will not affect each other. There is a reason why they use optical switching for the internet backbone, its fast (not just because the end product has to be light going through fiber optic cables). The problem these switches right now have to be maid out of exotic materials and cannot be integrated on a chip. Using optical switches inside of a chip is the holy grail. Don't underestimate the importance on advancing optics. If you want computers to keep getting faster we will have to go to optical pretty soon . Within the next 10 years we are expected to hit a barrier with current technology. You watch. Optical will be beating at your door before you know it.

Re:Quantum computers complement digital ones

[Hootenanny]

You may also be interested in this - I made a slight error in my first posting. We *can* currently implement quantum algorithms, through simulators such as this one for Matlab: http://www.physics.uq.edu.au/people/rohde/blog/?pa ge_id=20 [uq.edu.au]

However, a famous physicist/mathematician (whose name escapes me right now) proved that to emulate a quantum computer on a digital one will always require exponential complexity. So the benefit of speed is lost, but for the sake of curiosity and development, implementations of quantum algorithms can, at present, be tested. What we need now is the hardware. 8)

Re:Whaaah?

[pclminion]

it propagates faster than electricity. The individual electrons get hung up on imperfections in the conductor they travel along making them travel considerably slower than the speed of light.

What the hell do imperfections have to do with it? Nothing with mass can move at the speed of light. You seem to be suggesting that if the conductor was perfect, the electrons could move at the speed of light. What sort of crazy talk is that?

Anyway, the electrons have a net speed on the order of just millimeters per second. However, changes in the electric field caused by the motion of the electrons can propagate through the conductor much, much faster.

Re:Quantum computers complement digital ones

[Anonymous Coward]

>a famous physicist/mathematician (whose name escapes me right now)

His name was Richard Feynman. He sorta founded the subject of quantum computing since he was interested in modelling quantum physics on a computer but found this too be computationally expensive on a classical computer. Thus he invented the notion of a qunatum computer.