Example quantum tasks on Amazon Braket

This section walks through the stages of running an example quantum task, from selecting the device to viewing the result. As a best practice for Amazon Braket, we recommend that you begin by running the circuit on a simulator, such as SV1.

Specify the device

First, select and specify the device for your quantum task. This example shows how to choose the simulator, SV1.

# choose the on-demand simulator to run the circuit
from braket.aws import AwsDevice
device = AwsDevice("arn:aws:braket:::device/quantum-simulator/amazon/sv1")

You can view some of the properties of this device as follows:

print (device.name)
for iter in device.properties.action['braket.ir.jaqcd.program']:
    print(iter)

SV1
('version', ['1.0', '1.1'])
('actionType', <DeviceActionType.JAQCD: 'braket.ir.jaqcd.program'>)
('supportedOperations', ['ccnot', 'cnot', 'cphaseshift', 'cphaseshift00', 'cphaseshift01', 'cphaseshift10', 'cswap', 'cy', 'cz', 'h', 'i', 'iswap', 'pswap', 'phaseshift', 'rx', 'ry', 'rz', 's', 'si', 'swap', 't', 'ti', 'unitary', 'v', 'vi', 'x', 'xx', 'xy', 'y', 'yy', 'z', 'zz'])
('supportedResultTypes', [ResultType(name='Sample', observables=['x', 'y', 'z', 'h', 'i', 'hermitian'], minShots=1, maxShots=100000), ResultType(name='Expectation', observables=['x', 'y', 'z', 'h', 'i', 'hermitian'], minShots=0, maxShots=100000), ResultType(name='Variance', observables=['x', 'y', 'z', 'h', 'i', 'hermitian'], minShots=0, maxShots=100000), ResultType(name='Probability', observables=None, minShots=1, maxShots=100000), ResultType(name='Amplitude', observables=None, minShots=0, maxShots=0)])

Submit an example quantum task

Submit an example quantum task to run on the on-demand simulator.

# create a circuit with a result type
circ = Circuit().rx(0, 1).ry(1, 0.2).cnot(0,2).variance(observable=Observable.Z(), target=0)
# add another result type
circ.probability(target=[0, 2])

# set up S3 bucket (where results are stored)
my_bucket = "amzn-s3-demo-bucket" # the name of the bucket
my_prefix = "your-folder-name" # the name of the folder in the bucket
s3_location = (my_bucket, my_prefix)

# submit the quantum task to run
my_task = device.run(circ, s3_location, shots=1000, poll_timeout_seconds = 100, poll_interval_seconds = 10)
# the positional argument for the S3 bucket is optional if you want to specify a bucket other than the default

# get results of the quantum task
result = my_task.result()

The device.run() command creates a quantum task through the CreateQuantumTask API. After a short initialization time, the quantum task is queued until capacity exists to run the quantum task on a device. In this case, the device is SV1. After the device completes the computation, Amazon Braket writes the results to the Amazon S3 location specified in the call. The positional argument s3_location is required for all devices except the local simulator.

Submit a parametrized task

Amazon Braket on-demand and local simulators and QPUs also support specifying values of free parameters at task submission. You can do this by using the inputs argument to device.run(), as shown in the following example. The inputs must be a dictionary of string-float pairs, where the keys are the parameter names.

Parametric compilation can improve the performance of executing parametric circuits on certain QPUs. When submitting a parametric circuit as a quantum task to a supported QPU, Braket will compile the circuit once, and cache the result. There is no recompilation for subsequent parameter updates to the same circuit, resulting in faster runtimes for tasks that use the same circuit. Braket automatically uses the updated calibration data from the hardware provider when compiling your circuit to ensure the highest quality results.

from braket.circuits import Circuit, FreeParameter, Observable

# create the free parameters
alpha = FreeParameter('alpha')
beta = FreeParameter('beta')

# create a circuit with a result type
circ = Circuit().rx(0, alpha).ry(1, alpha).cnot(0,2).xx(0, 2, beta)
circ.variance(observable=Observable.Z(), target=0)
# add another result type
circ.probability(target=[0, 2])
# submit the quantum task to run
my_task = device.run(circ, inputs={'alpha': 0.1, 'beta':0.2})

Specify shots

The shots argument refers to the number of desired measurement shots. Simulators such as SV1 support two simulation modes.

For information about the maximum number of shots per quantum task, please refer to Braket Quotas.

Poll for results

When executing my_task.result(), the SDK begins polling for a result with the parameters you define upon quantum task creation:

This asynchronous execution facilitates the interaction with QPU devices that are not always available. For example, a device could be unavailable during a regular maintenance window.

The returned result contains a range of metadata associated with the quantum task. You can check the measurement result with the following commands:

print('Measurement results:\n',result.measurements)
print('Counts for collapsed states:\n',result.measurement_counts)
print('Probabilities for collapsed states:\n',result.measurement_probabilities)

Measurement results:
 [[1 0 1]
 [0 0 0]
 [1 0 1]
 ...
 [0 0 0]
 [0 0 0]
 [0 0 0]]
Counts for collapsed states:
 Counter({'000': 761, '101': 226, '010': 10, '111': 3})
Probabilities for collapsed states:
 {'101': 0.226, '000': 0.761, '111': 0.003, '010': 0.01}

View the example results

Because you’ve also specified the ResultType, you can view the returned results. The result types appear in the order in which they were added to the circuit.

print('Result types include:\n', result.result_types)
print('Variance=',result.values[0])
print('Probability=',result.values[1])

# you can plot the result and do some analysis
import matplotlib.pyplot as plt
plt.bar(result.measurement_counts.keys(), result.measurement_counts.values());
plt.xlabel('bitstrings');
plt.ylabel('counts');

Result types include:
 [ResultTypeValue(type={'observable': ['z'], 'targets': [0], 'type': 'variance'}, value=0.7062359999999999), ResultTypeValue(type={'targets': [0, 2], 'type': 'probability'}, value=array([0.771, 0.   , 0.   , 0.229]))]
Variance= 0.7062359999999999
Probability= [0.771 0.    0.    0.229]