World's First Programmable Quantum Photonic Chip

MrSeb writes "A team of engineering geniuses from the University of Bristol, England has developed the world's first re-programmable, multi-purpose quantum photonic computer chip that relies on quantum entanglement to perform calculations. With multiple waveguide channels (made from standard silicon dioxide), and eight electrodes, the silicon chip is capable of repeatedly entangling photons. Depending on how the electrodes are programmed, different quantum states can be produced. The end result is two qubits that can be used to perform quantum computing. Most importantly, though, unlike existing quantum photonic setups which require apparatus the size of a 'large dining table,' this new chip is tiny: just 70mm (2.7 inches) by 3mm."

Comments

ok

[j00r0m4nc3r]

but can you link it to the inverted phase-induced sub-space harmonic protocol analyzer to initiate a modulated tachyon pulse?

Re:ok

[masternerdguy]

Did you happen to work for the Voyager writing team?

Re:ok

[MobileTatsu-NJG]

Doubt it. He said it in the first act instead of the last.

Re:ok

[Baloroth]

Yes, but only if you reverse the polarity first.

Re:ok

[LynnwoodRooster]

Of course - there's an app for that!

Re:ok

[AlamedaStone]

Yes, but only if you reverse the polarity first.

Don't forget to reroute it all through the main deflector dish.

Re:ok

[GNious]

Always wondered what happens to the ship, while the deflector dish is being used for something else than deflecting stuff...

Re:excellent.

[Samantha Wright]

Protip: silicon is not soft. Further, it probably has, y'know, copper inputs and outputs? 40-50 thousand of those (times however many pins each chip has) poking into your supple, greasy back are also not soft.

computing power scales exponentially

[Janek Kozicki]

For those who are unaware why qubits are so powerful: the computing power provided by qubits scales exponentially if compared to bits used in ordinary computing. For example if you had 20 qubits, that would be like doing simultaneous calculations on processor with internal register size of 1048576 bits. Roughly. That's orders of magnitude more than modern CPUs, which have about dozen of 64 bit registers.

Re:computing power scales exponentially

[Janek Kozicki]

oh, and I forgot to mention - that's also the reason why quantum physics is so difficult to model using our today's computers. Monte carlo and other rough estimations are widely used. Only simplest problems (think harmonic oscilator) have analytical (and crazy complex) solutions.

Re:computing power scales exponentially

[HuguesT]

Definitely. However right now we do not have 20 qubits in a device, we have 2 qubits today. If progress in physics and electronics allows us to have 3 qubits in 18 month, 4 qubits in 36 months and so on, we have just reinvented the quantum version of Moore's law.

Re:computing power scales exponentially

[vlm]

I think he made it up. I am not making up (but could be completely wrong) that coincidentally the difficulty of preventing decoherence scales exponentially. And that is the primary limiter to # of qubits and performance, more or less correct?

In the very long run, I think quantum computing is going to be very much like DSP, in that the "hard work" is handling the analog signals to get "the problem" in and out. Inside ye olde DSP processor, a couple gnomes magically make it work, and superficially seem to be the hard part, at least partially correctly as some of the math is hideous. But the real problem is the unavoidable analog/RF work.

Kind of like how supercomputing is defined as taking a CPU bound process and making it an IO bound process, more or less.

Re:computing power scales exponentially

[alexgieg]

I know qubits can be very useful at encryption/decryption/cracking and such, but I'm curious: what else would they be useful for? I mean, is there something that a typical desktop/workstation does today that could be improved by adding some qubit-based magic behind the scenes, similar to how GPUs (and FPUs before them) resulted in improved GUIs, games, CAD/CAM etc.? Or is this the kind of thing that's most probably going to remain restricted to specific fields, with very specific needs, for the foreseeable future?

Re:computing power scales exponentially

[hweimer]

Oh, we already have a quantum version of Moore's law. However, the time constant for doubling is on the order of six years and not 18 months.

Re:computing power scales exponentially

[whovian]

I wonder if it could be used for simulating consciousness. I mean, IBM's Watson is a machine with clever brute force implementation of language parsing and data retrieval. Quantum computing seems paradigm-shifting enough to effectively implement many Watson-type machines, perhaps.

Re:computing power scales exponentially

[mikael]

The general rule for qubits seems to be anything that requires a unique solution but has to consider every possible combination of boolean states. Since they are Boolean zero or one values, that leads to cryptography because a relatively few number of bits would be required; 256,512,1024.

GPU's do floating-point calculations in parallel, which is really good for those problems which have to apply the same algorithm to different data points, like CFD, physics, AI, image and signal processing.

To represent floating-point data would require at least 16 qubits for half-floats, 32-bits for IEEE 754 standard floats, and 64-bits for doubles. But to do anything useful like CFD, would require storage of the entire state of the system which would require gigabits of data.

Unless someone could shrink the problem of CFD modelling down to atomic scales using phantom atoms, and overlapping qubits onto the same logic, GPU's won't have any competition.

Re:computing power scales exponentially

[Anonymous Coward]

No, that's false. Quantum computing offers polynomial time algorithms for a very small set of problems for which classically only exponential time algorithms are known, particularly, instances of the hidden subgroup problem (including integer factorization in the form of Shor's algorithm). More generally, Grover's algorithm gives some speedup to general NP problems, but not exponential. As I understand it, you would need a rather large quantum computer before it would actually faster than existing classical computers.

Re:computing power scales exponentially

[TheRealMindChild]

To simulate conciousness, the thinking entity needs to be able to genuinely ask a question... not just look for different types of metadata

Re:computing power scales exponentially

[mug funky]

that's an arse-backward definition of "simulation" that you have.

it needs to be able to make meaningful predictions as well.

for example, a VFX explosion in a space-opera versus a simulation of a nuke explosion on a supercomputer at Los Alamos

Re:computing power scales exponentially

[Surt]

http://en.wikipedia.org/wiki/Quantum_computer#Potential

Re:computing power scales exponentially

[rbmyers]

I think it was five computers, and what no one ever mentions is that Thomas Watson, who made the comment, wanted the world to continue using punch cards, which were the backbone of IBM's profitability. That is to say, all you IBM'ers out there who think IBM practically invented computing, IBM's first take on the "opportunity" was to try to kill it.

Re:computing power scales exponentially

[Anonymous Coward]

It's Mersenne primes. And use powers of 2, not 10.

Re:computing power scales exponentially

[fsckmnky]

what else would they be useful for?

https://secure.wikimedia.org/wikipedia/en/wiki/Grover's_algorithm

Re:computing power scales exponentially

[Prune]

He did make it up. The speed-up is polynomial, not exponential. You still cannot effectively solve large NP-hard problems. Most algorithms that are susceptible to quantum-computing speedup end up running in O(sqrt(original_time)).

Re:excellent.

[Denihil]

i know! one time i slept on about 60 cat-5 cables on a 200+ person man LAN, it wasn't too bad. maybe if i laid in bed, and then had the chips on top of me? thanks for the input though :) i slept next to my frigid bitch of a ex girlfriend, so if i can handle near absolute zero temperatures, this should be a walk in the park.

Bristol group

[vlm]

I assume this is the same group (how many quantum computer groups are there likely to be in Bristol?) that did the whole "lets run Shors algorithm on a silicon etched chip" a couple years ago. So the new news right now is ... Or is this a re-reporting of that historical event, or another paper about that historical work? I'm just trying to figure out the whole timeline thing here.

Hey /. editors, the recent interviews have been very interesting and all that, I'm just thinking interviewing the quantum group in Bristol would be even more interesting...

Entangling photons is a bad idea.

[ross.w]

It takes hours to sort them out afterwards.

Re:Entangling photons is a bad idea.

[The Askylist]

The trick is to keep Schroedinger's cat away from your photon strings.

Imagine...

[jeremiahstanley]

a beowulf cluster of these...

Had to go there, this is /. afterall...

The Downside

[MyHair]

Unfortunately, after you program it you no longer know where it is.

Result=99% of CPUengineers will be without job ?

[Anonymous Coward]

Help me out.

This would have happened sooner or later, and I am assuming it happened. Quantum computer are here to stay.
Question is, what this means to general community of engineers and software developers ?

I am perfectly aware that we don't have Hardware that is capable of supporting the work of this chip (RAM and HD don't make sense). Maybe in another 15 years.
Does it mean complete shift of computing paradigm ?
Instead of 100 servers, we have just small black-box in a backroom ?
What will happened to all the engineers who created silicon chips ?
What all this means for programmers ?

Re:excellent.

[mikael]

That would be awesome to see - a hammock made of woven cat-5 cables.

Once saw the interconnect of a supercomputer/rack server "styled" into ocean waves, rather than just some snake-pit of cables.

Re:Bristol group: uncertainty

[petes_PoV]

how many quantum computer groups are there likely to be in Bristol?

You can either know where they are, or how many there are - but not both.

Bad news for crypto

[gillbates]

If what you say is true, this is truly bad news for cryptography. Algorithms like AES owe their security largely to the fact that brute-forcing all of the keys is generally impractical; with a 256 qubit machine, AES 256 would be cracked in *a single clock cycle*.

If they can do this with two qubits, why not 4? Why not 8, or 128, or 512?

In the same way the WWII cipher designers probably had a hard time imagining that in 40 years there would exist a machine which could crack their ciphers in real time, the designers of block ciphers like DES and AES probably had a difficult time imagining that their ciphers would be insecure in mere decades. DES took 30 years before brute force became practical; will AES survive even 20?

It was just 20 years from the invention of the transistor to the first 32 bit computer. How long will it be before a machine with more computing power than in all of recorded history can be built on something the size of a postage stamp, for a few dollars?

Re:Bad news for crypto

[Prune]

This is plain wrong. Under a quantum-computer attack, AES256 is as strong as AES128. Thus, you simply need to double your key size. Most symmetric ciphers are safe. Most public-key, on the other hand, is indeed broken by quantum-computation. People often forget that for most things, quantum algorithms can only provide a quadratic speedup--not an exponential one!

Not exactly exponentially

[marcosdumay]

The computing part does indeed act on every combination your register can have at the same time. An exponential speedup here, that part is right. What is missing on your post is that reading the result is kind of hard. We only know how to get usefull data from a few kinds of calculation, and we don't know if it is possible to get anything usefull from the general case.

The good news is that if we ever discover a way to read the result of a general computation (if it is possible), we'd have discovered a nondeterministic computer. And forget about P ?= NP.

Re:CPUengineers will be without job ?

[gweihir]

Quantum computers are not "here" in any meaningful sense. Nobody ever has demonstrated a meaningful larger-numbers quantum computation (say, with numbers > 1000). At the moment, the there is no proof these will even work. It is still entirely feasible that the theory is wrong and large quantum computers are not possible or not useful. Even some tiny deviations from the current theory could cause that. Remember the results have to be physically measured and the input has to be physically put in. Both operations with huge, huge errors when compares to the precision classical computers achieve.

Then, even if meaningful sizes can be built (which is entirely unclear at this time) they are not effective or efficient for most problems.

Example: For breaking ciphers like AES, you get a square root on the key size, i.e. breaking AES-256 becomes as difficult as breaking AES-128 (both by brute force). Breaking AES-128 by brute force without quantum computers is quite infeasible in this universe. Breaking AES-256 by brute-force with quantum computers is quite infeasible in this universe as well.

Forget about any large data-set problems as well. Unlike classical computers, you cannot break problems down for quantum computers. You always have to solve the whole thing in one go, or you lose the advantages.

Bottom line: This is not a revolution, even if it turns out not to be bogus in the first place.

Re:CPUengineers will be without job ?

[neonsignal]

log2(sqrt(x)) = log2(x)/2

(by key size, he meant the magnitude of the key, not the number of bits in the key)

Re:Result=99% of CPUengineers will be without job

[Surt]

Nothing happens to engineers. They just design quantum chips instead, at worst (but most likely: a mix of quantum and conventional computers is still required). Most likely it will still be decades yet before most even need to care.

Nothing happens to programmers. A handful of library designers will work out the interesting bits. The rest will continue building applications on top of the libraries as usual.

Quantum Co-Processor?

[Reteo Varala]

Before mathematical capability was baked into the main processor cores, motherboards used to have mathematical co-processors, which could handle the advanced math in a computer. Even if a quantum chip cannot compare with a classical chip's calculation performance at this time, just how feasible would it be to include a quantum co-processor on a classical motherboard for quantum calculations? Would the two combined provide any benefit that either by itself could not?