Computing Library › Quantum Foundations
Quantum Foundations

Two-Qubit Gates

Two-qubit gates are 4x4 unitaries that can entangle qubits; at least one entangling gate is required for universal quantum computation.

Why two qubits are special

A two-qubit gate is a 4x4 unitary acting on the joint space of two qubits. Unlike single-qubit gates, a two-qubit gate can create entanglement, which is why at least one is indispensable for universal computation.

Common examples

Kronos motion — thermal gate

Entangling versus non-entangling

Not every two-qubit gate entangles. SWAP, for instance, only relabels qubits and creates no entanglement from a product input. A gate is entangling if it can turn some product state into an entangled one; CNOT and CZ do, SWAP does not. Universality requires an entangling gate specifically.

Hardware-native gates

Each platform has a natural two-qubit interaction determined by its physics. Superconducting chips often implement CZ or iSWAP; trapped ions use the Molmer-Sorensen gate. Compilers translate abstract CNOTs into whatever the hardware provides, so the logical circuit and the physical pulses can differ substantially.

The cost of entangling gates

Two-qubit gates are typically slower and noisier than single-qubit gates, so they dominate the error budget of a circuit. Reported two-qubit fidelities lag single-qubit ones, and reducing two-qubit gate count is a central goal of circuit optimisation. The threshold for error correction is set largely by two-qubit gate quality.