Simulation of noisy Clifford circuits without fault propagation

TL;DR

This paper introduces an efficient simulation method for noisy Clifford circuits that avoids fault propagation by leveraging the mathematical structure of the spacetime code, significantly reducing computational effort.

Contribution

The authors develop the adjoint-based code (ABC) simulation algorithm, enabling fault simulation in Clifford circuits without propagating faults, unlike traditional methods.

Findings

Reduces simulation complexity for Clifford circuits

Eliminates the need for trillions of fault propagations

Applicable to large-scale fault-tolerant quantum architectures

Abstract

The design and optimization of a large-scale fault-tolerant quantum computer architecture relies extensively on numerical simulations to assess the performance of each component of the architecture. The simulation of fault-tolerant gadgets, which are typically implemented by Clifford circuits, is done by sampling circuit faults and propagating them through the circuit to check that they do not corrupt the logical data. One may have to repeat this fault propagation trillions of times to extract an accurate estimate of the performance of a fault-tolerant gadget. For some specific circuits, such as the standard syndrome extraction circuit for surface codes, we can exploit the natural graph structure of the set of faults to perform a simulation without fault propagation. We propose a simulation algorithm for all Clifford circuits that does not require fault propagation and instead exploits the mathematical structure of the spacetime code of the circuit. Our algorithm, which we name adjoint-based code (ABC) simulation, relies on the fact that propagation forward is the adjoint of propagation backward in the sense of Proposition 3 from [14]. We use this result to replace the propagation of trillions of fault-configurations by the backward propagation of a small number of Pauli operators which can be precomputed once and for all.