Running on Real Quantum Hardware

Everything so far has run on cirq.Simulator(), an ideal, noise-free simulator. Google's real quantum processors exist too, but access works quite differently from IBM's more open model.

Google Quantum AI access

import cirq_google

# Real hardware access via Google Quantum AI requires an allowlisted Google Cloud
# project, this is not a self-serve signup the way IBM Quantum's free tier is.
engine = cirq_google.Engine(project_id='your-google-cloud-project-id')
processor = engine.get_processor('processor_id')

result = processor.run(circuit, repetitions=1000)
print(result.histogram(key='result'))

This needs an allowlisted Google Cloud project and real hardware access, so it's read-only here. Unlike IBM Quantum's public free tier (sign up and run within minutes), Google Quantum AI hardware access has historically required an approved research or partnership relationship, most learners will develop and run entirely on cirq.Simulator().

Cirq still models noise, without needing real hardware

Cirq lets you simulate with realistic noise, without needing hardware access at all, useful for understanding what real results would look like:

import cirq

q0, q1 = cirq.LineQubit.range(2)
circuit = cirq.Circuit([
    cirq.H(q0),
    cirq.CNOT(q0, q1),
    cirq.measure(q0, q1, key='result'),
])

noisy_circuit = circuit.with_noise(cirq.depolarize(p=0.02))  # 2% error rate per operation

simulator = cirq.Simulator()
result = simulator.run(noisy_circuit, repetitions=1000)
print(result.histogram(key='result'))  # small numbers of 1 and 2 (01/10) will now appear

circuit.with_noise(cirq.depolarize(p=0.02)) returns a new circuit where every operation has a small chance (p=0.02, 2%) of introducing a random error, a simplified model of the real gate errors and decoherence discussed conceptually in the Qiskit course's hardware lesson. Running this noisy circuit will occasionally produce the "impossible" 01/10 outcomes, exactly what you'd expect to see on real, imperfect hardware.