Following programming an Ising model, we have qs storing the target system to simulate. We numerically simulate it via SimuQ’s QuTiP provider, which employs a translation from SimuQ to QuTiP and solve the Schrodinger equation.
Install dependencies
To enable QuTiP providers, we need to install the related dependencies by
pip install simuq[qutip]
Create a QuTiP provider
A provider is a user interface for convenient manipulations of functionalities of SimuQ. We use QuTiP provider as a basic example on how to use providers to deploy quantum simulation problems on devices and obtain results.
We can create a QuTiP provider via the following code
from simuq.qutip import QuTiPProvider
qpp = QuTiPProvider()
Compilation in provider
To simulate a quantum system qs programmed in HML via SimuQ, we need three major steps of a provider: compile, run, results.
We call the compile function of the provider to process the system into a runnable executable. For QuTiP provider, we can execute
qpp.compile(qs)
QuTiP provider processes the quantum system qs and translate it into a Hamiltonian in QuTiP. When succeeds, a message Compiled. will show up.
For other providers, compile command may specify the backend device, AAIS, and compiler specifications.
When compilation succeeds, the job will be recorded in the provider.
Run and obtain results from providers
Running a job will send the compilation results to backend devices to execute. For QuTiP provider, we execute
qpp.run()
When successfully executed, the command will show a message Submitted, meaning that the executables are sent to devices to run.
For other providers, you can choose whether to run on classical simulators, and how many shots to run on the real device or simulators. Note that these classical simulators are mostly from hardware providers and executed on the cloud.
To retrieve the results, we can execute
qpp.results()
If the task is still in the queue on the device, an exception will be raised. When the retrieval succeeds, a dictionary is returned, which contains the frequencies of obtaining a measurement array (encoded as a 0/1 string). A bit in the string corresponds to a site of the quantum system. We can call qpp.print_sites() to show the order of the sites in the measurement output.
The resulting dictionary of the 3-site Ising model is
{'000': 0.6972053600815633,
'001': 0.0,
'010': 0.0,
'011': 0.05973358916427098,
'100': 0.0,
'101': 0.18332746158989466,
'110': 0.05973358916427098,
'111': 0.0}