Practical Subarchitectures for Optimal Quantum Layout Synthesis

Abstract: Quantum Layout Synthesis (QLS) maps a logical quantum circuit to a physical quantum platform. Optimal QLS minimizes circuit size and depth, which is essential to reduce the noise on current quantum platforms. Optimal QLS is an NP-hard problem, so in practice, one maps a quantum circuit to a subset of the complete quantum platform. However, to guarantee optimality, one still has to consider exponentially many subarchitectures. We introduce an effective method to enumerate relevant subarchitectures. This reduces the number of considered subarchitectures, as well as the number of expensive subgraph isomorphism checks, thus boosting Optimal QLS with subarchitectures. To do so, we assume a fixed number of ancilla qubits that can be used in the mapping. We guarantee optimality of the quantum layout, for the selected ancilla bound. We evaluate our technique on a number of benchmarks and compare it with state-of-the-art Optimal QLS tools with and without using subarchitectures.