Physicists Demonstrate Quantum Router

Diggester tips news that physicists from Tsinghua University in China have published "the first proof-of-principle demonstration of a genuine quantum router." The group's paper (PDF) is available at the arXiv. MIT's Technology Review describes it thus: "In this new device, the information is encoded in the polarization of photons, either horizontal or vertical. The Chinese group begin by creating a single photon that is in a superposition of both horizontal and vertical polarization states. They then convert this single photon into a pair of lower energy photons that are entangled, a process called parametric down conversion. Both of these photons are also in a superposition of polarization states. The router works by using the polarization of one of these photons as the control signal to determine the route of the other, the data signal. The device is simple, little more than a collection of half mirrors for guiding photons and waveplates for rotating their polarization."

Comments

Does This mean...

[sycodon]

...I'll start getting at least half of the advertised speed from my AT&T DSL connection?

Re:Does This mean...

[zlives]

no you need the higgs field router for that

Re:Does This mean...

[dkleinsc]

No, that's just the spin they're putting on it.

Re:Does This mean...

[bjoast]

He will have to make do with the standard model.

Comment removed

[account_deleted]

Comment removed based on user account deletion

Re:v6?

[RaceProUK]

Only if you measure it in the right way.

So...

[Dins]

How long until we have a working Ansible?

Re:So...

[bluefoxlucid]

Actually I was thinking more of something like that, but with store/drain, depending on how far entanglement goes and how long you can maintain it.

Think about 1 meter entanglement range. You entangle two particles, shift one a meter away. It takes time to move that particle, of course; but once it's entangled, flipping one flips the other instantly.

Bandwidth like this is infinite. If you have 1Gbit/s real bandwidth and can store 3000Gbit of entangled particles stably. and they're stable for 3000 seconds on average, you can eventually build up 3000Gbit of entanglement between two nodes.

Now space these nodes a meter apart down the wire, 1500Gbit on each end relaying, 2km. While there's less than 1Gbit/s in use, store entangled particles. At a point, you have 1500Gbit of bandwidth charge ... and then you can dump 1500Gbit of data straight on the link (if your computer is fast enough to send that to the hardware!) and your burst data rate is 1500Gbit INSTANTANEOUS. It's 1500Gbit in however many cycles it takes to move 1500Gbit from memory onto the output buffer--if it takes 10 cycles to get a bit from RAM onto the "wire" (that's 320 cycles per 32-bit word on a 32-bit bus, which is realistic--random RAM read can be 200-400 cycles to precharge, set RAS, CAS, etc, but sequential can be 10-20 cycles, plus overhead messing with the actual transmit hardware), at 1.5GHz CPU that's 150MHz or 10 seconds to transmit 1500Gbit. 3GHz that's 5 seconds, and with 3GHz quad core and a distributed architecture that's a little over a second--you've achieved nearly 1.5Tbit/s burst with near-zero latency over 2km!

By the time the technology is actually affordable for such use, the range will probably be in the km. That means repeaters entangled across miles and miles. You think we need 1500Gbit/s burst? Saturation will happen often enough to warrant bursting. It'd take 1500 seconds over a 1Gbit link running idle, like 2.5 hours. Probably 2-3 times capacity (2-3 Gbit) entangled store being used to artificially raise peak capacity (when off-peak, i.e. 50% utilization of 1Gbit/s over a few seconds means you can constantly burst, store up 3Gbit and then 3Gbit comes and you transmit it instantly instead of over 3 seconds). and decrease off-peak latency (50% saturation? Store for 50%, use for the other 50%, zero ping!).

All that untapped capacity of a 30% or 70% utilized network will ACTUALLY MAKE IT FASTER. Imagine!

Re:So...

[ByteSlicer]

Please, someone correct me if im wrong, but my understanding is its like having a red, and blue ball in seperate bags. You throw one ball (in its bag) across the room, then open the other bag. The open bag is blue, you now know the red one is across the room.

It's more like having two balls, both half-red half-blue, in separate bags. The balls contain a magnet that causes them to align: when one is red-up, the other is blue-up.

Then you shake the bags, open one, and pick out the ball. If it's red-up, you'll know the other ball will be blue-up, and vice versa.

At first glance this may look the same as your model with the full red or blue ball (a so called hidden variable model), but the statistics differ from real entanglement (which includes superposition of states).

In the first (classical) model FTL information transfer is impossible because nothing is actually transferred. The ball was blue or red to begin with, and opening the bag doesn't change that.

In the second (quantum) model FTL information transfer actually does occur (with entanglement, the magnet alignment in the model occurs slower than light obviously). But the observer can't choose what this information will be: picking a ball out of a bag will result in a random color. And the other bag will then contain a ball in the opposite but equally random color. You can't pick the color, so you can't choose the message/information.

Statistics from experiments have shown that the second model is the correct one, and that entanglement and the FTL state transfer are real.

Re:Professor Frink Reference

[mooingyak]

But to comply with the OP's request, it would need to be a bunny cake.

Re:Me chinese. Me play joke.

[FatdogHaiku]

But what about the quantum backdoors implanted for the Chinese government?

Quick, somebody lock down quantumbackdoors.com for use as a porn site.
When the Chinese government tries to use their secret backdoor they will encounter a black hole!

Re:Very Nice

[rts008]

Is it going to be light on the pocket book too when it comes out?

No.

It will create a black hole in your wallet that won't stop until you have lost your ass.

It's just simple consumerism, but you were mistaken that you are the consumer, instead of the consumed....

Re:Professor Frink Reference

[Jane Q. Public]

Certainly.

Take two stuffed bunnies. Lay one on its side, and stand the other on its head. These are two "polarizations".

But with QUANTUM stuffed bunnies, it is possible to create a pair of stuffed bunnies that are actually in both positions at the same time... "superposition"... and "entangled", meaning the the positions of both are linked. (This has also led to the creation of a new illustrated book, the Quanta Sutra... but that's another story.)

So, you send these superposed stuffed bunnies down different pipes, and when they get to the end, one of them is knocked over the head with a laser, which causes them BOTH to resolve into one or the other position. Then by determining the position of that one you can determine the position of the other, even at a distance.

I am thinking about naming this the "stuff your bunny sideways" principle.