Circuit execution

Qibolab can be used as a qibo backend for executing executions. The purpose of this section is to show how to do it, without entering into the details of circuits definition that we leave to the Qibo documentation.

import qibo
from qibo import Circuit, gates

# create a single qubit circuit
circuit = Circuit(1)

# attach Hadamard gate and a measurement
circuit.add(gates.H(0))
circuit.add(gates.M(0))

# execute on quantum hardware
qibo.set_backend("qibolab", "tii1q_b1")
hardware_result = circuit(nshots=5000)

# execute with classical quantum simulation
qibo.set_backend("numpy")
simulation_result = circuit(nshots=5000)

# retrieve measured probabilities
hardware = hardware_result.probabilities(qubits=(0,))
simulation = simulation_result.probabilities(qubits=(0,))

In this snippet, we first define a single-qubit circuit containing a single Hadamard gate and a measurement. We then proceed to define the qibo backend as qibolab using the tii1q_b1 platform. Finally, we change the backend to numpy, a simulation one, to compare the results with ideality. After executing the script we can print our results that will appear more or less as:

print(f"Qibolab: P(0) = {hardware[0]:.2f}\tP(1) = {hardware[1]:.2f}")
print(f"Numpy:   P(0) = {simulation[0]:.2f}\tP(1) = {simulation[1]:.2f}")

Returns:

> Qibolab: P(0) = 0.54 P(1) = 0.46
> Numpy:   P(0) = 0.50 P(1) = 0.40

Clearly, we do not expect the results to be exactly equal due to the non ideality of current NISQ devices.

Note

Qibo circuits and gates are backend agnostic. The same circuit can be executed on multiple backends, including simulation and quantum platforms.

A slightly more complex circuit, a variable rotation, will produce similar results:

import numpy as np
from qibo import Circuit, gates


def execute_rotation():
    # create single qubit circuit
    circuit = Circuit(1)

    # attach Rotation on X-Pauli with angle = 0
    circuit.add(gates.RX(0, theta=0))
    circuit.add(gates.M(0))

    # define range of angles from [0, 2pi]
    exp_angles = np.arange(0, 2 * np.pi, np.pi / 16)

    res = []
    for angle in exp_angles:
        # update circuit's rotation angle
        circuit.set_parameters([angle])

        # execute circuit
        state = circuit.execute(nshots=4000)
        p0, p1 = state.probabilities(qubits=(0,))

        # store probability in state |1>
        res.append(p1)

    return res


# execute on quantum hardware
qibo.set_backend("qibolab", "tii1q_b1")
hardware = execute_rotation()

# execute with classical quantum simulation
qibo.set_backend("numpy")
simulation = execute_rotation()

# plot results
exp_angles = np.arange(0, 2 * np.pi, np.pi / 16)
plt.plot(exp_angles, hardware, label="qibolab hardware")
plt.plot(exp_angles, simulation, label="numpy")

plt.legend()
plt.ylabel("P(1)")
plt.xlabel("Rotation [rad]")
plt.show()

Returns the following plot:

Note

Executing circuits using the Qibolab backend results to automatic application of the transpilation and compilation pipelines (Transpiler and Compiler) which convert the circuit to a pulse sequence that is executed by the given platform. It is possible to modify these pipelines following the instructions in the How to modify the transpiler? example.