Performance of Uncoded Implementation of Grover's Algorithm on Today's Quantum Processors

TL;DR

This paper evaluates the performance of uncoded Grover's algorithm on current IBM quantum hardware for small search spaces, highlighting the impact of noise and the necessity for error correction.

Contribution

It introduces two new Grover's circuits for N=16, compares simulation and real device results, and emphasizes the need for error correction for larger search spaces.

Findings

Real devices show fluctuations due to noise

New circuits for N=16 are proposed

Error correction is necessary for N≥32

Abstract

This work tests the performance of Grover's search circuits on some IBM superconducting quantum devices in case of the size of search space $N=2^4$ and $N=2^5$. Ideally, we expect to get an outcome probability distribution that is clearly peaked at the goal (marked) state. However, the quantum circuit executed on real devices is vulnerable to noise which leads to fluctuations in the results. The contributions of the paper are therefore the following: a) it presents two new Grover's search circuits for $N=16$ which were not yet reported in the state of the art; b) it shows performance difference between simulation results and results obtained on real devices; c) it shows the need of adding error-correction on the circuit for $N\ge 2^5$.