Two-Qubit Gate Compilation

Two-Qubit Gate Compilation

Compiling an arbitrary two-qubit gate into the limited native gate set of real hardware is a fundamental problem in quantum circuit compilation, since most devices provide only one fixed entangling interaction plus arbitrary single-qubit rotations. This Cirq example demonstrates an algorithm that solves it using the GateTabulation class, which approximates any desired two-qubit unitary as a short product of the form k3 A k2 A k1 A k0, where A is a fixed base two-qubit gate (here a square-root-of-iSWAP combined with a controlled phase) and the k_j are arbitrary single-qubit gates. The approach precomputes a tabulation, a dense sampling of the two-qubit gates reachable with a small number of applications of the base gate, so that compiling a new target gate reduces to looking up a nearby tabulated point and solving for the surrounding single-qubit rotations. The script builds the GateTabulation for the chosen base gate, generates a collection of one thousand random two-qubit gates, and compiles each of them, reporting how accurately the tabulation-based decomposition reproduces the targets. It is a practical demonstration of how a hardware-native entangling gate, interleaved with single-qubit rotations, can synthesize the full continuum of two-qubit operations, a core capability for any quantum compiler.

Run it

pip install -r requirements.txt
python two_qubit_gate_compilation.py

Source and license

Imported from examples/two_qubit_gate_compilation.py in quantumlib/Cirq at v1.6.1, under the Apache License 2.0. Original authors: The Cirq Developers. The upstream LICENSE is included alongside this example.