Investigating the Exchange of Ising Chains on a Digital Quantum Computer

TL;DR

This paper simulates a braiding-like exchange operation in Ising chains on a quantum computer, analyzing fidelity and errors, and discusses the challenges of implementing such operations on current NISQ hardware.

Contribution

It demonstrates a method to simulate Ising chain exchanges on a quantum computer and analyzes the impact of various errors on fidelity, providing insights for future quantum hardware improvements.

Findings

Achieved >99% fidelity for up to 11-site chains.

Quantum gate errors dominate fidelity loss over Trotter and adiabatic errors.

Optimal single braiding operation fidelity requires circuit depth ~1000 with gate errors <10^{-6}.

Abstract

The ferromagnetic state of an Ising chain can represent a two-fold degenerate subspace or equivalently a logical qubit which is protected from excitations by an energy gap. We study a a braiding-like exchange operation through the movement of the state in the qubit subspace which resembles that of the localized edge modes in a Kitaev chain. The system consists of two Ising chains in a 1D geometry where the operation is simulated through the adiabatic time evolution of the ground state. The time evolution is implemented via the Suzuki-Trotter expansion on basic single- and two-qubit quantum gates using IBM's Aer QASM simulator. The fidelity of the system is investigated as a function of the evolution and system parameters to obtain optimum efficiency and accuracy for different system sizes. Various aspects of the implementation including the circuit depth, Trotterization error, and quantum gate errors pertaining to the Noisy Intermediate-Scale Quantum (NISQ) hardware are discussed as well. We show that the quantum gate errors, i.e. bit-flip, phase errors, are the dominating factor in determining the fidelity of the system as the Trotter error and the adiabatic condition are less restrictive even for modest values of Trotter time steps. We reach an optimum fidelity $>99\%$ on systems of up to $11$ sites per Ising chain and find that the most efficient implementation of a single braiding-like operation for a fidelity above $90\%$ requires a circuit depth of the order of $\sim 10^{3}$ restricting the individual gate errors to be less than $\sim 10^{-6}$ which is prohibited in current NISQ hardware.