Qiskit is an open-source framework for working with noisy quantum computers at the level of pulses, circuits, and algorithms.
This project contains a provider that allows access to Quantinuum quantum devices.
You can install the provider using pip:
pip3 install qiskit-quantinuum-providerpip will handle installing all the python dependencies automatically and you
will always install the latest version.
Once the package is installed, you can access the provider from Qiskit via the following import:
from qiskit_quantinuum import QuantinuumYou will need credentials for the Quantinuum Quantum Service. Credentials are tied to an e-mail address that can be stored on disk with:
Quantinuum.save_account('[email protected]')After the initial saving of your account information, you will be prompted to enter your password which will be used to acquire a token that will enable continuous interaction until it expires. Your password will not be saved to disk and will be required infrequently to update the credentials stored on disk or when a new machine must be authenticated.
The credentials will then be loaded automatically on calls that return Backends, or can be manually loaded with:
Quantinuum.load_account()This will load the most recently saved credentials from disk so that they can be provided for each interaction with Quantinuum's devices.
Storing a new account will not invalidate your other stored credentials. You may have an arbitrary number of credentials saved. To delete credentials you can use:
Quantinuum.delete_credentials()Which will delete the current accounts credentials from the credential store. Please keep in mind this only deletes the current accounts credentials, and not all credentials stored.
With credentials loaded you can access the backends from the provider:
backends = Quantinuum.backends()
backend = Quantinuum.get_backend(device)You can then use that backend like you would use any other qiskit backend. For example, running a bell state circuit:
from qiskit import *
qc = QuantumCircuit(2, 2)
qc.h(0)
qc.cx(0, 1)
qc.measure([0,1], [0,1])
result = execute(qc, backend).result()
print(result.get_counts(qc))