0
$\begingroup$

Nowadays quantum learning is hiring. And we can see mainly two different area. One of them is variational algorithms part. And the other one is classical learning for quantum systems like NISQ. (Some of scientists call both areas as quantum learning wheras some scientists call the second one as a classical learning for quantum system) Regarding variational part, there are many libraries like tensorflow quantum, qiskit ... But for the second part, I would like to learn what is good. For instance we can simulate photonic chips with Simphony or gds helpers. But my aim is not to simulate photonic chips. My aim is to train integrated photonic chips for noises. So which libraries are good for that. For instance is cirq enough for that? or should I use another libraries? Or should I start to simulate my chip and after that I should use some additional libraries like tensorflow quantum or qiskit or cirq?

$\endgroup$
3
$\begingroup$

You will have very limited support simulating photonic circuits in Cirq (and therefore Tensorflow Quantum). Those libraries mainly deal with evolving systems of qubits where a state $|\psi\rangle$ is expressed over a discrete basis like $|\psi\rangle = \sum_i c_i |i\rangle$, as opposed to photonic systems containing continous variables (such as a wavefunction $\psi(x)$ expressed in the basis of a continous operator $\hat{x}$).

One of the main uses for Tensorflow Quantum is when you have an algorithm that involves both a quantum simulation part and a classical processing part that interact. For example a parameterized quantum circuit and some kind of neural network layer that are both trained using gradient-based optimizers, then the library is able to efficiently simulate both the circuit and its derivatives which are then propagated through both of the layers to differentiate a given cost function. This does not sound like a good fit for your application since you would need to outsource the circuit simulation to a different program.

Maybe check out Strawberry Fields since this was designed for photonic simulation and interfaces with Xanadu's Pennylane and so should have similar capabilities as I described above, but for a photonic qc simulator. I'm not very familiar with the library so I don't know what kind of support it has for device noise.

$\endgroup$
2
$\begingroup$

Here's some examples:

  • IBM: Aqua
  • Google: Tensor-Quantum
  • Xanadu: PennyLane
$\endgroup$
1
  • 1
    $\begingroup$ I believe you meant "Tensorflow Quantum" $\endgroup$ – forky40 Mar 2 at 19:06

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.