Quantum Outpost

Algorithm Zoo · Sampling

Random Circuit Sampling

Also known as: RCS, Quantum supremacy benchmark

First described: Boixo et al. (theory); Google Quantum AI (Sycamore experiment), 2019

Status: Disputed Maturity: Demonstrated Speedup: Exponential (conjectured) for the sampling task itself

The problem

Sample from the output distribution of a random quantum circuit.

Run a depth-D random Clifford+T-like circuit on n qubits, sample the output bitstring distribution. Classical simulation conjectured to require time exponential in n at fixed depth, supporting a complexity-theoretic argument for quantum advantage on this task.

Best classical

Tensor-network simulation, with several efficient variants developed since 2019.

Quantum complexity

O(1) per sample on hardware (a single circuit run).

Our verdict

The benchmark that gave us 'quantum supremacy' headlines and the years-long classical chase to invalidate them. Useful as a hardware diagnostic (XEB) and as a complexity-theory exercise; useless as an application. Every claim should be read with a sunset-by-2026 expectation.

When to use it

When not to use it

Classical baseline

Pan & Zhang (2022) simulated the 2019 Sycamore experiment in 15 hours on a Sunway supercomputer. Liu et al. (2021) Jiuzhang photonic claims similarly dented by classical tensor-network methods (Oh et al. 2023). The arms race is ongoing.

Hardware cost

n qubits, depth ~20 for the 2024 Sycamore experiment at 67 qubits. XEB fidelity drops exponentially with circuit volume — careful calibration mandatory.

Key papers

Last verified: 2026-05-24

Weekly dispatch

Quantum, for people who already code.

One serious tutorial per week, plus the industry moves that actually matter. No hype, no hand-waving.

Free. Unsubscribe anytime. We will never sell your email.