Dynamic T-decomposition for classical simulation of quantum circuits

Abstract

It is known that a quantum circuit may be simulated with classical hardware via stabilizer state (T-)decomposition in $O(2^{αt})$ time, given $t$ non-Clifford gates and a decomposition efficiency $α$. The past years have seen a number of papers presenting new decompositions of lower $α$ to reduce this runtime and enable simulation of ever larger circuits. More recently, it has been demonstrated that well placed applications of apparently weaker (higher $α$) decompositions can in fact result in better overall efficiency when paired with the circuit simplification strategies of ZX-calculus. In this work, we take the most generalized T-decomposition (namely vertex cutting), which achieves a poor efficiency of $α=1$, and identify common structures to which applying this can, after simplification via ZX-calculus rewriting, yield very strong effective efficiencies $α_{\text{eff}}\ll1$. By taking into account this broader scope of the ZX-diagram and incorporating the simplification facilitated by the well-motivated cuts, we derive a handful of efficient T-decompositions whose applicabilities are relatively frequent. In benchmarking these new 'dynamic' decompositions against the existing alternatives, we observe a significant reduction in overall $α$ and hence overall runtime for classical simulation, particularly for certain common circuit classes.