Noisy simulations with EstimatorV2 behave as they are ideal for backend noise models

[DilhanM]

Informations

-Qiskit version: 1.3.2

• Qiskit Aer version: 0.16.0
• Python version: 3.11.5
• Operating system: macOS 15.2

What is the current behavior?

Noisy and exact simulation results for EstimatorV2 primitives are identical when noise model is built from a backend.

Steps to reproduce the problem

This is the example from docs, with the noise model built from a backend.

from time import time

from qiskit import QuantumCircuit, generate_preset_pass_manager
from qiskit_aer import AerSimulator
from qiskit_aer.noise import NoiseModel
from qiskit_ibm_runtime import QiskitRuntimeService
from qiskit_aer.primitives import EstimatorV2 as EstimatorAer

from qiskit.circuit.library import EfficientSU2
 
n_qubits = 8
circuit = EfficientSU2(n_qubits)

from qiskit.quantum_info import SparsePauliOp
from qiskit.transpiler.preset_passmanagers import generate_preset_pass_manager
from qiskit_aer import AerSimulator
from qiskit_aer.primitives import EstimatorV2 as Estimator
 
observable = SparsePauliOp("Z" * n_qubits)
params = [0.1] * circuit.num_parameters
 
exact_estimator = Estimator()
# The circuit needs to be transpiled to the AerSimulator target
pass_manager = generate_preset_pass_manager(3, AerSimulator())
isa_circuit = pass_manager.run(circuit)
pub = (isa_circuit, observable, params)
job = exact_estimator.run([pub])
result = job.result()
pub_result = result[0]
exact_value = float(pub_result.data.evs)

print("Exact value:", exact_value)

device = "ibm_brisbane"
service = QiskitRuntimeService(
    instance="YOUR INSTANCE"
)
backend = service.backend(device)
noise_model = NoiseModel.from_backend(backend)

#Works for custom noise models #
# from qiskit_aer.noise import depolarizing_error
# noise_model = NoiseModel()
# cx_depolarizing_prob = 0.02
# noise_model.add_all_qubit_quantum_error(
#     depolarizing_error(cx_depolarizing_prob, 2), ["cx"]
# )

noisy_estimator = Estimator(
    options=dict(backend_options=dict(noise_model=noise_model))
)
job = noisy_estimator.run([pub])
result = job.result()
pub_result = result[0]
noisy_value = float(pub_result.data.evs)

print("Noisy value:", noisy_value)

What is the expected behavior?

print("Noisy value:", noisy_value)` --> 0.7247404214143528

Other info

Issue is invariant to the backend used.