Shot noise (or sampling noise) arises because quantum measurement is probabilistic. A quantum circuit must be run many times (called "shots") to estimate the probability distribution of outcomes. With a finite number of shots N, the statistical uncertainty in an estimated probability p scales as √(p(1−p)/N). To halve the statistical error, you need four times as many shots. Typical circuit runs use 1,000–10,000 shots for reasonable precision. Shot noise is separate from hardware gate errors — it exists even for perfect simulators. The number of shots needed depends on the precision required and the type of measurement: expectation values (VQE) may need thousands of shots; simple bitstring sampling may need fewer. HLQuantum lets you set shots=N in hlq.run().
Related Terms
Measurement
FundamentalsThe act of observing a qubit's state, which collapses the superposition to a definite 0 or 1.
Fidelity
MetricsA measure (0 to 1) of how close an actual quantum operation or state is to the ideal target.
VQE
AlgorithmsVariational Quantum Eigensolver — a hybrid quantum-classical algorithm for finding ground state energies.