The number of the Grover's diffuser iterations is proportional to square root of the amount of input data, but how is this realized when building a real quantum circuit? Is it simply by concatenating the needed number of diffusers one after another (in which case I assume the circuit designer is required to know the needed number of iterations beforehand)?


A quantum logic circuit isn't a literal physical object, it's a plan for what operations to perform. It's like sheet music. The piano maker doesn't have to know how long the song is when making the piano, they instead provide reusable keys that the player will just keep hitting until the song is done.

In a superconducting qubit device, the "keys" are microwave pulses bounced off the superconducting circuit. Each pulse performs some fixed operation. You would run Grover's algorithm by decomposing its operations into those pulses and playing them until the song is done.

  • $\begingroup$ Following your piano metaphor, a person who builds a quantum circuit can then be said to be a music composer. If there is a number of iterations in a specific melody, he/she must know exactly how many times that melody repeats in order to complete the composition. Similarly, we as a circuit builder must know how many times the Grover diffuser will be run to configure the circuit. In other words, the number to iterate cannot be assigned by output from previous operations in the circuit. Have I got it right? $\endgroup$ Feb 18 at 2:25
  • 1
    $\begingroup$ @Alternative7 No, you can have adaptive rules like "keep doing X until Y happens". $\endgroup$ Feb 18 at 8:37
  • $\begingroup$ Ah, then I can let the algorithm/circuit set the number of repeat? That should make things a lot more convenient. Thank you! $\endgroup$ Feb 18 at 12:56

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.