Can quantum parallelism be turned on and off?


The qubit is able to take on multiple states at once, unlike the bit. This property I know is superposition, but is this property perpetual (always on) or optional (controlled to be on or off only when needed by an algorithm, such that when not needed, the qubit reverts to, or takes on the disguise of, a bit)?

and how does perpetual/optional superposition play into the notion of quantum parallelism: Does the qubit always execute parallel computations and incapable of a uni-sequential workflow, or can it be controlled to execute tasks in parallel only when the algorithm deems it necessary?

Essentially, what is the difference or connection between superposition and quantum parallelism and are these properties always active in a quantum algorithm?

(optional: Circuit diagrams that demonstrate example comparisons of "superposition/parallelism on" and "superposition/parallelism off" can be included as long as each item (line, gate and symbol) is annotated as a bullet point list.)


Posted 2019-11-19T18:36:23.437

Reputation: 835

1There are multiple misconceptions in this question that cannot be reasonably addressed in a single answer. I recommend reading a proper textbook first. – Sanchayan Dutta – 2019-11-19T20:19:10.237

@SanchayanDutta could you give suggestions for develarist to start with? – C. Kang – 2019-11-19T23:45:26.693

looking back at this and just to give some perspective on why the question's loaded, the part about superposition's role in quantum parallelism is coupled with the part on whether these properties, if in fact separate, can be made optional sprung from the observation that solo classical computers have the option for parallelism once more cores or GPUs are brought in for tackling the same task simultaneously. From what i see in classical parallelism, I thought the concepts mentioned are highly intertwined that discussing them together would expose the true source of quantum parallelism – develarist – 2019-11-20T02:00:07.420

1@develarist it's pretty critical to disentangle (lol) your notion of classical HPC with quantum computing! HPC involves breaking the problem down into smaller subproblems that can be tackled with multiple cores/gpus (as you mentioned), whereas current quantum algorithms mostly use clever interference / representations of quantum computers to perform costly operations. – C. Kang – 2019-11-20T05:43:09.483



'Quantum parallelism' is a common misconception - sure, the quantum computer does compute in a 'parallel' fashion, but really the power of quantum computing emerges from constructive / destructive interference.

If you wanted, you could absolutely make a quantum computer behave in a classical format: constantly measure the qubits and correct if necessary. For example, if you had a qubit in $\frac{1}{\sqrt{2}} (|0\rangle + | 1 \rangle) $ and desired a classical bit of $|0\rangle$, you can measure, and apply $X$ if the qubit measured resolves to a $|1 \rangle$.

C. Kang

Posted 2019-11-19T18:36:23.437

Reputation: 1 411

ok so superposition can be turned on and off in a quantum algorithm, while quantum interference, where several paths that mingle with one another are brought together in the end somehow (often with high error it sounds like) in order to attain a most correct solution, might be the concept that really explains parallelism as opposed to superposition being the sole source of parallelism. – develarist – 2019-11-20T01:38:22.207

Yes - it's critical to note that superposition is mostly a vehicle for creating interference. Superposition itself would not be very useful, because we can't tell which output is tied to which superposition state! – C. Kang – 2019-11-20T05:41:38.363