Qiskit Aer worker¶

A Tierkreis worker that compiles and runs circuits with Qiskit Aer.

The Aer worker largely wraps the functionality from pytket-qiskit. In addition to the elementary tasks exposed, there are also prepackaged graphs to make using the worker more convenient.

Installation¶

pip install tkr-aer-worker

will install an executable Python script tkr_aer_worker into your virtual environment.

Elementary tasks¶

The Aer worker exposes the following elementary tasks to the user.

get_compiled_circuit. A wrapper around AerBackend.get_compiled_circuit, which is intended to be parallelised using a Tierkreis map.
run_circuit. A wrapper around AerBackend.run_circuit, which is intended to be parallelised using a Tierkreis map.
run_circuits. A wrapper around AerBackend.run_circuits, which performs multiple simulations according to the logic in the process_circuits method in pytket-qiskit.
to_qasm3_str. Converts a pytket Circuit to QASM3 using the Qiskit qasm3 module.

Prepackaged graphs¶

The Tierkreis Python package provides a few prepackaged graphs to make it easier to compile and run circuits with Aer.

tierkreis.graphs.simulate.compile_simulate.compile_simulate is intended for the common use case of compiling a list of circuits in parallel and then running them in parallel. It can be included within a custom graph using GraphBuilder.eval or run as a standalone graph.

An example use is in examples/simulate_parallel.py in the Tierkreis repo, which looks like:

simulator_name = "aer"
circuits = ...your circuits here...

g = compile_simulate()
storage = FileStorage(UUID(int=107), do_cleanup=True)
executor = UvExecutor(PACKAGE_PATH / ".." / "tierkreis_workers", storage.logs_path)

run_graph(
    storage,
    executor,
    g,
    {
        "circuits": circuits,
        "n_shots": [30] * len(circuits),
        "config": config,
        "compilation_optimisation_level": 2,
    },
    polling_interval_seconds=0.1,
)
res = read_outputs(g, storage)
print(len(res))

The subgraph tierkreis.graphs.simulate.compile_simulate.compile_simulate_single can also be used if the user wants to do additional customisations.