'Hybrid' Logic Gate For Quantum Computers Demonstrated

hypnosec writes: Researchers at Oxford University have demonstrated a 'hybrid' logic gate using two isotopes of calcium, the abundant isotope calcium-40 and the rare isotope calcium-43. One of the leading technologies for building a quantum computer is trapped atomic ions, and researchers at Oxford's Networked Quantum Information Technologies (NQIT – pronounced 'N-kit') Hub are working to develop the constituent elements of a quantum computer based on these ions. As explained by researchers in the study published in the journal Nature, each of the trapped ions is used to represent one 'quantum bit' of information. The quantum states of the ions are controlled with laser pulses of precise frequency and duration. Two different species of ion are needed in the computer: one to store information, a 'memory qubit', and one to link different parts of the computer together via photons, an 'interface qubit'."

Hybrid logic gate

[U2xhc2hkb3QgU3Vja3M]

What do you mean? Like a Toyota Prius or Chevrolet Volt?

Re: Hybrid logic gate

[Anonymous Coward]

Don't be silly, it's just what it says, in a quantum mechanical manner. That is, they've invented a gate that is a superposition of a NAND gate and a NOR gate, which collapses to the appropriate gate when you observe the circuit. That is, the circuit computes every possible function until you observe it, at which point it gives you the answer you want.

Re:

[Anonymous Coward]

Oh how I love it when the whoosher is in fact the whooshee :)

Re: Hybrid logic gate

[Aighearach]

Nonsense, just because you don't know what gate it is doesn't mean that it is both. It would have to have already collapsed to one or the other in order to compute anything.

Until it interacts with something (what you describe as being "observed," though that is specific to inputs and this has outputs too) it doesn't have inputs or outputs and you not only can't prove that it is calculating anything, you can't even make a basic argument that there is a reason to believe that it is capable of calculation. Just as, holding a rock in your hand that has a high silicon content, with a little copper mixed in, doesn't mean you have a computer that is calculating everything that could be calculated with the devices you could make from those ingredients. Or put another way, it isn't that you don't know where something is until you interact with it, it is that it doesn't even have a specific location until something interacts with it. It is spread out over the probabilities. NAND and NOR gates don't operate on probabilities; they operate on achieved electromagnetic properties. Nothing at all can be happening until it is one gate or the other.

Not that that is what the thing in TFA is. This is more about a type of logic gate that can connect to what they're using as quantum logic gates, and also what they're using as pointers, which you could think of as being like a quantum system bus for communication between different parts of the machine. Also note that the part they just demonstrated is entirely deterministic. It is sortof analogous to Harvard architecture microcontrollers, which have separate data areas for program code and data. Quantum computers as being researched by academia use different types of logic structures for storing memory and doing calculations, so they have this separation inherently. This is a special logic gate that can connect to both. So presumably this will end up being part of some kind of programmable register with branching capabilities.

While your comment was clearly in jest, what it might actually be like is that you set up a complex calculation, expose that system of gates to this gate, the whole thing collapses at the average speed of the photons exchanged, and the branching based on the result happened that fast; literally as the calculation is "being" made, the branch associated with the answer is being activated. As close to instantly as humans can detect, minus whatever considerable lag and blanking periods are required before triggering these events, of course. Quantum computers might not have or need branch prediction.

Once again...too soon to care.

[Myria]

Wake me when I can factor 2048-bit composite numbers. The original Xbox's public key and I have some unfinished business from a decade ago.

Consequences For Software Engineers

[KermodeBear]

So, as a typical corporate software engineer, should I care at all about quantum computing?

Re:Consequences For Software Engineers

[JoshuaZ]

It depends how long term your outlook is. Nothing here will be practical for at least a decade and very likely more. But once quantum computers become a practical thing, there are going to be a lot of uses (e.g. factoring large numbers using Shor's algorithm https://en.wikipedia.org/wiki/Shor's_algorithm [wikipedia.org], and detailed molecular simulation). Note that one thing we don't expect to see is extreme speedup of generic NP-complete problems, although the popular press often says otherwise. Also, as with any computer technology, once the technology becomes more common, people will have more of an incentive to find uses for it. GPUs are a fun example of that: made for a very narrow purpose, but once they were made, people quickly started realizing they could do other things with them. Quantum computers will likely be in a similar boat.