Quantum Phase Estimation

Quantum Phase Estimation (QPE) solves a deceptively simple problem: given a unitary operator U and one of its eigenstates |ψ⟩ satisfying U|ψ⟩ = e^(2πiθ)|ψ⟩, estimate the phase θ. The construction uses two registers. A counting register of n qubits is put into equal superposition with Hadamard gates, then a sequence of controlled-U^(2^k) operations is applied for k = 0 to n-1. Through phase kickback, each controlled operation writes a phase onto the control qubit rather than disturbing the eigenstate, encoding θ in the Fourier basis across the counting register. Applying the inverse QFT converts that encoding into a binary number, and measuring the counting register reads it out — n counting qubits yield n bits of precision on θ. QPE is the engine inside Shor's algorithm, where order finding on modular exponentiation is exactly a phase estimation problem, and inside fault-tolerant quantum chemistry, where the phase of a time-evolution operator gives a molecular energy eigenvalue. The catch is depth: the controlled-U^(2^k) chain demands long coherent circuits, making QPE generally impractical on today's NISQ hardware — which is precisely why variational methods like VQE exist as a near-term alternative.