Braket via Nexus

Here you can find some examples and explanations on running quantum circuits on AWS Braket via Nexus.

These will require that you have set Braket credentials via Settings->Linked Accounts on the Nexus website.

from datetime import datetime

from pytket import Circuit

import qnexus as qnx

List available Braket devices, optionally specifying a specific AWS region (by default devices for ALL regions will be returned).

qnx.devices.get_all(issuers=[qnx.devices.IssuerEnum.BRAKET]).df()

To configure Nexus to target Braket, you’ll need to provide the following arguments to the BraketConfig.

  • local: bool

    • if true then the circuit will run as a simulation in the Nexus cloud.

  • device_type: str

    • device type from device ARN (e.g. "qpu", "quantum-simulator")

  • provider: str

    • provider name from device ARN (e.g. "ionq", "rigetti", "oqc", "amazon", …)

  • device: str

    • device name from device ARN (e.g. "ionQdevice", "Aspen-8", "sv1", …)

  • s3_bucket: str

    • name of S3 bucket to store results

  • s3_folder: str

    • name of folder ("key") in S3 bucket to store results in

braket_config = qnx.BraketConfig(
    local=True,
    device_type="quantum-simulator",
    provider="amazon",
    device="sv1",
    s3_bucket="",
    s3_folder="",
)

Example of compiling and executing a circuit on AWS Braket

my_project_ref = qnx.projects.get_or_create(name="My AWS Braket Project")

circuit = Circuit(2).H(0).CX(0, 1).measure_all()

my_circuit_ref = qnx.circuits.upload(
    name=f"My AWS Braket Circuit from {datetime.now()}",
    circuit=circuit,
    project=my_project_ref,
)

compile_job = qnx.start_compile_job(
    programs=[my_circuit_ref],
    name=f"My Compile Job from {datetime.now()}",
    optimisation_level=1,
    backend_config=braket_config,
    project=my_project_ref,
)

qnx.jobs.wait_for(compile_job)

compiled_circuits = [item.get_output() for item in qnx.jobs.results(compile_job)]

execute_job_ref = qnx.start_execute_job(
    programs=compiled_circuits,
    name=f"My Execute Job from {datetime.now()}",
    n_shots=[100],
    backend_config=braket_config,
    project=my_project_ref,
)

qnx.jobs.wait_for(execute_job_ref)


# Retrieve a ExecutionResultRef for every Circuit that was executed
execute_job_result_refs = qnx.jobs.results(execute_job_ref)

# Get a pytket BackendResult for the execution
result = execute_job_result_refs[0].download_result()

result.get_counts()