Source-verifiedLevel 4

Quantum Phase Estimation

Quantum phase estimation (QPE) estimates the eigenvalue phase of a unitary operator: given U and an eigenstate |u> with U|u> = e^(2*pi*i*phi)|u>, it extracts phi to a chosen precision using controlled-U operations and an inverse quantum Fourier transform.

What it means

Phase estimation (Kitaev, 1995) uses two registers: t ancilla qubits initialized in uniform superposition via Hadamard gates, and a target register holding an eigenstate |u> of the unitary U.Controlled-U^(2^j) operations write the phase phi into the relative phases of the ancilla register (phase kickback).Applying the inverse quantum Fourier transform to the ancillas and measuring yields the best t-bit approximation of phi with high probability.Precision improves with more ancilla qubits: t qubits give roughly t bits of phi.QPE is the workhorse subroutine behind Shor's factoring algorithm (period finding), quantum chemistry energy estimation (eigenvalues of a Hamiltonian's evolution operator), and quantum counting.If the target register is not an exact eigenstate, QPE returns each eigenphase with probability given by the overlap with the corresponding eigenstate.

Everyday analogy

It is like finding out how fast a merry-go-round spins by peeking at it after 1 turn, 2 turns, then 4 turns -- each peek tells you one more piece of the answer, and putting the pieces together reveals the speed.
Think of the inverse QFT as a decoder ring: the controlled-U operations scramble the hidden phase into many qubits, and the decoder ring reads the digits back out one binary digit at a time.

Common misconceptions

  • QPE does NOT read out the phase in a single direct measurement -- the phase is first encoded across an ancilla register via phase kickback, then decoded by the inverse QFT.
  • QPE gives FINITE precision set by the number of ancilla qubits (t qubits for about t bits of phi), not an exact real number.
  • The input does not have to be a perfect eigenstate, but then the output is probabilistic: each eigenphase appears with probability equal to the squared overlap with that eigenstate.

Key takeaways

  • Estimates phi in U|u> = e^(2*pi*i*phi)|u> using controlled-U powers plus inverse QFT.
  • Precision scales with the number of ancilla qubits: t ancillas give about t bits of phi.
  • Core subroutine of Shor's algorithm and quantum chemistry energy estimation.

Check your understanding

Which transform is applied to the ancilla register just before measurement in quantum phase estimation?

  1. A.Hadamard on every qubit
  2. B.Inverse quantum Fourier transform
  3. C.Grover diffusion operator
  4. D.CNOT ladder
Show the answer

Answer: B. Inverse quantum Fourier transform

Why: After controlled-U^(2^j) operations encode the phase into the ancilla register, the inverse quantum Fourier transform converts that phase information into a measurable bit string approximating phi.

Builds on

QubitSuperpositionHadamard GateQuantum Fourier TransformQuantum Measurement

Primary source: Kitaev, Quantum measurements and the Abelian Stabilizer Problem, arXiv:quant-ph/9511026 (1995)

Graded 2026-07-10 (human sign-off): established per Kitaev arXiv:quant-ph/9511026 (1995) and Nielsen & Chuang (2010).

Learn it hands-on

This concept is part of a 46-level curriculum with an interactive simulator and Lumen, a tutor whose answers are verified before you see them. Levels 1–5 are free.

Start freeSee the curriculum