Noise resilience of Bayesian quantum phase estimation tested on a Si quantum photonic chip

Summary form only given. Quantum Phase Estimation (PE) is a fundamental building block in the framework of Quantum Computing. Accurate estimation of the true eigenphase φ
₀
of a known eigenstate |φ〉 is fundamental for the implementation of many promising quantum algorithms. The interest in PE is also due to the modest quantum hardware requirements of the Iterative Phase Estimation Algorithm (IPEA) implementation [1], successfully implemented on small-scale devices [2]. However, IPEA makes hard decisions at each step of the algorithm, relying on majority voting schemes for its robustness against noise. It has been observed how non-error-corrected machines may enter a regime where error-rates for IPEA diverge quickly, making this approach impractical [3]. New Bayesian approaches, such as the Rejection Filtering Phase Estimation (RFPE), have been recently proposed to overcome these limitations [4]. Here, we report the implementation of RFPE in a photonic device embedding state-of-art integrated elements. Spontaneous Four Wave Mixing sources generate photon pairs. State preparation and evolution are accomplished via integrated Mach-Zehnder interferometers and thermo-optical phase shifters operating in the hundreds of kHz regime. Fast reconfigurability is key for RFPE, as the algorithm adaptively chooses the parameters for each experiment. Arbitrary CU(2) controlled operations are made possible by an entanglement-based scheme, integrated on this chip for the first time.