Analytical Gradient Estimator for Pulses in qua_sim

[Packman2002]

Looking at the functionality of the library I would like to add another feature which is to create a function to estimate the gradients of quantum pulses. As qua is a platform for control qubits on multiple platforms I feel creating a generalized framework to estimate the gradients of pulses along with the powerful stream processing architecture of qua can help in simulating control and error mitigation techniques.

Please let me know if this new application makes sense and if you could guide me on it. @nulinspiratie @deanpoulos

[nulinspiratie]

Hi @Packman2002 this sounds very interesting and could be a highly useful feature! It would be great to hear your thoughts and brainstorm together.

Could you describe in more details what you have in mind, and if you have a specific use case you want to cover? We could also schedule a meeting to discuss, feel free to add me on LinkedIn and we can take it from there

[Packman2002]

Hi @nulinspiratie. Thanks a lot for responding. I was working on qua in a quantum lab and one of the issues we were having were implementation of optimal control algorithms on custom quantum processing units and implement variational quantum algorithms as they would take an excessively long time to optimize using chopped basis and other search space optimization techniques.

I was thinking if we could have a package which could estimate the gradients for pulse based unitaries giving user the option of which parameters in a pulse unitary he wishes to optimize and keep constant it may help to enhance the effectiveness of the above tasks on qua for custom quantum hardware. Additionally I though integrating pulse gradients along with the stream processing functionality in qua may help in gaining better insights of qubit quality during qubit spectroscopy which also I think is quite a tedious task.

These are just some of my ideas. I took inspiration from a few papers online including one by pennylane in which they do something similar which I thought may be a good idea to create implement on qua generalizing for custom hardware.

I have sent you a connection request on linkedin. Please do accept. Would love to work with
you and take this forward.

[Packman2002]

Hi @nulinspiratie if you feel this idea is worth implementing. Could we please
have a brief discussion so I would have some idea how to proceed and how to make it practical.

[nulinspiratie]

Hi @Packman2002 I'm definitely interested, could you send me a message on LinkedIn so I know it's you?

[Packman2002]

@nulinspiratie please accept my connection request. I cannot message you on Linkedin until you do that. My real name is Apoorva.