I am running a 3-qubit simulation using ionq.simulator in azure-quantum. When I send the task, I get the result from one shot + probability distribution that includes all the possible 8 bitstrings. I am not interested in getting a probability distribution, just the result from 1 shot. This is a big problem when I try to do 29-qubit simulation, because I cannot even retrieve the results from Azure; the results are 12GB!
Here is the 3 qubit example:
from azure.quantum.qiskit import AzureQuantumProvider
from azure.quantum import Workspace
from qiskit import QuantumCircuit
workspace = Workspace(resource_id = "", location = "")
provider = AzureQuantumProvider(workspace)
circuit = QuantumCircuit(3, 3)
circuit.name = "Qiskit Sample - 3-qubit GHZ circuit"
circuit.h(0)
circuit.cx(0, 1)
circuit.cx(1, 2)
circuit.measure([0, 1, 2], [0, 1, 2])
device = provider.get_backend("ionq.simulator")
job = device.run(circuit, shots=1)
res = job.result()
The results are:
Result(backend_name='ionq.simulator', backend_version='1', qobj_id='Qiskit Sample - 3-qubit GHZ circuit', job_id='...', success=True, results=[ExperimentResult(shots=1, success=True, meas_level=2, data=ExperimentResultData(counts={'000': 1}, probabilities=defaultdict(<class 'int'>, {'000': 0.5, '111': 0.5})), header=QobjExperimentHeader(qiskit='True', name='Qiskit Sample - 3-qubit GHZ circuit', num_qubits='3', metadata={}, meas_map='[0, 1, 2]'))], date=None, status=None, header=None, error_data=None)
The troublesome element for me is probabilities=defaultdict(<class 'int'>, {'000': 0.5, '111': 0.5}) since it grows exponentially with system size for an arbitrary circuit.
Then, my question is, how do I avoid Azure computing the probability distribution? Which includes $2^n$ elements, with $n$ being the number of qubits.
