An exact branch-and-bound solver for generalized qubit mapping shows that the standard layering constraint raises optimal SWAP counts and circuit depth, most strongly on sparsely connected hardware graphs.
Depth-Optimal Quantum Circuit Placement for Arbitrary Topologies
1 Pith paper cite this work. Polarity classification is still indexing.
abstract
A significant hurdle towards realization of practical and scalable quantum computing is to protect the quantum states from inherent noises during the computation. In physical implementation of quantum circuits, a long-distance interaction between two qubits is undesirable since, it can be interpreted as a noise. Therefore, multiple quantum technologies and quantum error correcting codes strongly require the interacting qubits to be arranged in a nearest neighbor (NN) fashion. The current literature on converting a given quantum circuit to an NN-arranged one mainly considered chained qubit topologies or Linear Nearest Neighbor (LNN) topology. However, practical quantum circuit realizations, such as Nuclear Magnetic Resonance (NMR), may not have an LNN topology. To address this gap, we consider an arbitrary qubit topology. We present an Integer Linear Programming (ILP) formulation for achieving minimal logical depth while guaranteeing the nearest neighbor arrangement between the interacting qubits. We substantiate our claim with studies on diverse network topologies and prominent quantum circuit benchmarks.
citation-role summary
citation-polarity summary
fields
quant-ph 1years
2025 1verdicts
CONDITIONAL 1roles
background 1polarities
unclear 1representative citing papers
citing papers explorer
-
An Exact Branch and Bound Algorithm for the generalized Qubit Mapping Problem
An exact branch-and-bound solver for generalized qubit mapping shows that the standard layering constraint raises optimal SWAP counts and circuit depth, most strongly on sparsely connected hardware graphs.