search

qiboteam / boostvqe

Using DBI top boost VQE optimization
3 stars 1 forks source link

Speeding up `rotate_h_with_vqe` #40

Closed andrea-pasquale closed 3 months ago

andrea-pasquale commented 3 months ago

rotate_h_with_vqe was slow when dealing with large number of qubits. The bottleneck seems to be related to circuit.unitary() which computes the unitary transformation implemented by the circuit. To speed up the execution I opted for fusing the circuit before asking for the unitary transformation. I need to test it to check if everything is working correctly but the speed up is already visible.

marekgluza commented 3 months ago

Thanks!

Another thing is that we need to often use the vqe unitary cf #34 notebooks for the manipulations that we consider for testing things on the states i.e. exposing the vqe unitary out of rotate_h_vqe might be needed down the line?