I have a quantum circuit similar to the ones used in VQE methods. One difference is that the angles of the $R_y$ rotations are limited in the range $[0, 0.2]$ (and chosen randomly). In my case, I have 30 layers of gates, where each layer contains the single-qubit gates $R_y$ and the $CNOT$ gates, so 30 layers of single-qubit gates and 30 layers of 2-qubits gates.

This is an example of this kind of circuit with 6 qubits (but with only 7 layers of gates for visualization): enter image description here

My question is: is there any classical simulation technique that doesn't scale exponentially with the number of qubits that can calculate the expectation values of the $X$ and $Z$ observables for each qubit? It is ok even if there is a small approximation.

I tried the MPS method implemented in Qiskit, and if I ask for a coefficient of determination $R^2\ge 0.99$ for 100 random circuits (that is a sufficient approximation in my case), the bond dimension seems to scale exponentially with the number of qubits, but I can try this only for $N\le 14$, so it is hard to say.

Thank you.



Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Browse other questions tagged or ask your own question.