3
$\begingroup$

I'm looking to write an algorithm (potentially) using Qiskit, and am trying to find out what the maximum number of Qubits is that Qiskit can support with various simulation methods.

From reading other answers on here, and using Qiskit a bit, I have found that the Qasm simulator has a maximum of 30(?), but that Qiskit supports also using a GPU.

I assume the number of qubits a GPU can simulate depends on the GPU in question, but I am wondering if anyone knows how to find out what that number would be for a given GPU?

In addition, are there any other methods using Qiskit or otherwise wherein you can simulate a quantum computer running a complex algorithm, for example, one which requires several real numbers (including decimals) as variables? Not looking for specifics necessarily - any pointers will do, but specifics are very welcome!

Thanks in advance, and apologies if this is a silly question or I've missed something.

$\endgroup$
1
  • $\begingroup$ You can use backend.configuration().n_qubits. This should return the number of qubits whether the backend is local simulator, cloud simulator, or quantum processor. $\endgroup$ Jun 30, 2021 at 18:13

1 Answer 1

4
$\begingroup$

I think this page from the IBM Quantum services could answer your question about how to simulate differently and gain in the number of qubits : https://quantum-computing.ibm.com/services/docs/services/manage/simulator/#simulators-overview

You have a list of all simulators available in the cloud, and as you can see depending on the method you can go from 30 cubits up to 100 with MPs method, even 5000 for stabilizer, although be aware that this will have its limits, for example you cannot model noise with the MPS simulator.

On the link you put on your question, you can see as well all the possible methods that are available via the Aer simulators.

Hope this will be useful to you, if you have other questions coming from this feel free to ask :)

$\endgroup$
1
  • 1
    $\begingroup$ Thanks! That is helpful. A follow up question would be whether (as far as you are aware) any of the options available allow for simulating an algorithm which requires using multiple (sometimes fairly large (or really small)) decimal/real numbers as variables, and if there are any examples anywhere of this? $\endgroup$
    – Jk9898
    Jun 30, 2021 at 13:31

Your Answer

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

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