Descripción del proyecto
Quantum computers bring the promise of solving efficiently problems that would take a considerable amount of time on the best classical computers. This is called the quantum speedup. However, an experimental proof of a quantum speedup on a problem useful for society remains to be shown. One of the main issues preventing it to occur is the noise: the hardware is too noisy which makes the algorithms outputs unreliable. These algorithms are usually designed under the assumption that the qubits are perfect, and the errors are corrected or mitigated afterwards. However, mitigation techniques are typically not scalable (they stop working if the algorithm is too large). Quantum error correction is scalable but very challenging to implement as it requires very high quality qubits. This project aims to design quantum algorithms that, by construction, would efficiently work in highly noisy regimes, because of a noise-aware design without the need for error correction or error mitigation techniques. Our first objective will be to see if it is possible to carefully design quantum algorithms (by looking at how errors propagate in the circuits) for which the consequences of the noise will be that the experimentalist has only to re-run the algorithm a polynomial number of times, naturally preserving any exponential speedup. Our second objective will be to study the possibility to design algorithms operating on mixed-states that do not introduce any entanglement but nonetheless provide an exponential speedup. Because such algorithms would work with mixed-states without entanglement, they could also lead to interesting noise-resilience properties. For both of our objectives, we will aim to find useful applications for our algorithms, beyond the mere proof of concepts. Overall, our project could lead to a new class of quantum algorithms, noise-resilient from their very design, and would provide fundamental insights on the necessary ingredients allowing a speedup to occur.