The Joint Quantum Institute published a recent paper detailing a quantum computer constructed with five qubits formed from trapped ions. The novel architecture allows the computer to accept programs for multiple algorithms.
Quantum computers make use of qubits and trapped ions–ions confined with an electromagnetic field–are one way to create them. In particular, a linear radio frequency trap and laser cooling traps five ytterbium ions with a separation of about 5 microns. To entangle the qubits, the device uses 50 to 100 laser pulses on individual or pairs of ions. The pulse shape determines the actual function performed, which is how the device is programmable. The operations depend on the sequence of laser pulses that activate it.
Because of the nature of quantum physics, each pulse has about a 99% chance of doing the right thing, so when you do 100 in a row, you have between a 70 and 90% chance of success. Among the algorithms implemented by the computer, researchers ran a QFT (quantum Fourier transform) which requires two qubit gates between all possible pairs of qubits.
The strength of quantum computing is the massive parallelism that is possible. However, this particular computer is relatively slow per operation (microsecond gate times).
If you have a spare hour, you might enjoy watching the video below from [Andreas Dewes] about the practical realization of some different quantum computers, including those using ion traps and superconducting. We’ve talked about quantum physics before. If you think that entangling particles is out of reach to the average hacker, you might be surprised.
if I have 99% chance of setting a bit to what I want. That means that the success rate for setting 100 in a row correctly is 0.99^100 = 36.6% chance. Not 70 – 90% chance. Or am I missing something?
Maybe rounded down to 99%
e^(ln(0.7)/100) = 0.99643960385
e^(ln(0.9)/100) = 0.99894694969
A quantum computer wold be perfect for calculating what the correct answer is … wait.
or wood it bee that it could give the right answer only if you were extremely specific.
Quantum algorithms have an interesting property – to put it simply, sometimes the answer is wrong.
But this is OK for many important QC applications because you can check correctness without an unacceptable time complexity, and repeat if you need to
Consider integer factorization with Shor’s algorithm, for example.
Do the factorisation in polynomial time, and suppose it’s wrong say 25% of the time.
Now multiply it back together, which you can do in poly time, and if it’s wrong repeat the factorisation and check again.
You’re still in polynomial time complexity.
A minor issue with the article. The strength of a quantum computer is not in its parallelism but in creative use of interference between different paths of the calculation which are running in superposition. By Holevo’s theorem, you can’t extract the information from a naive parallel algorithm and that’s why quantum computers don’t offer speedups over classical computers except in certain cases (or at least it seems they do. BQP’s position in the polynomial hierarchy is an open question).
Infinite impossibly hard drive. You press the button and it computes all possible cat videos simultaneously and spits out the one you want to watch… :D
The more you try to compute the lower the accuracy of the result. It’s like playing yahtzee with 10,000 dice. Then applying a for statement and setting the delims to what your not looking for.
Ask yourself how many zx spectrums can fit on an i7 [ not to mention the bigger server chips out there] if you could drill holes in the die and keep it cold you’d have enough cores to do that sort of calculating.
Who knows what’s hiding in the skunk works waiting to come out a decade from now.
Beijing launches the world’s first quantum-communications satellite into orbit
Surely, the information that you presented, wasn’t good enough or were the institute deploying novel masking system of the global folks.
The video URL changed, the new one is this:
https://www.youtube.com/watch?v=aXtE0Zeszho