Go-No go criteria for performing quantum chemistry calculations on quantum computers

Quantum chemistry is envisioned as an early and disruptive application where quantum computers would provide a genuine advantage with respect to purely classical approaches. In this work, we propose two criteria for evaluating the potential of the two leading quantum approaches for this class of problems. The first criterion applies to the Variational Quantum Eigensolver (VQE) algorithm and sets an upper bound to the level of noise that can be tolerated in quantum hardware as a function of the target precision and problem size. We find a crippling effect of noise with an overall scaling of the precision that is generically less favourable than in the corresponding classical algorithms. This is due to the studied molecule being unrelated to the hardware dynamics, hence its noise; conversely the hardware noise populates states of arbitrary energy of the studied molecule. The second criterion applies to the Quantum Phase Estimation (QPE) algorithm that is often presented as the go-to replacement of VQE upon availability of (noiseless) fault-tolerant quantum computers. QPE suffers from the phenomenon known as the orthogonality catastrophe that generically leads to an exponentially small success probability when the size of the problem grows. Our criterion allows one to estimate quantitatively the importance of this phenomenon from the knowledge of the variance of the energy of the input state used in the calculation.