Dead kernel when running noisy simulation on Apple silicon

[Dran-Z]

Environment

• Qiskit Terra version:
• 0.22.2
• Python version:
• 3.9.6
• Operating system: mac OS Ventura 13.2.1 (MBP 2021 Apple M1 Pro)

What is happening?

I am writing code to do noisy simulations using the Aer simulator. When I execute my code in my local jupyter notebook, I will get the error.

The Kernel crashed while executing code in the the current cell or a previous cell. Please review the code in the cell(s) to identify a possible cause of the failure. Click [here](https://aka.ms/vscodeJupyterKernelCrash) for more info. View Jupyter [log](command:jupyter.viewOutput) for further details. Canceled future for execute_request message before replies were done However, if I put my code on the online IBM-Quantum-Lab, I won't have this issue and my code can be executed normally.

My noise model contains depolarizing_noise and thermal_relaxation_error. I tried to remove the thermal_relaxation_error out from my noise model, and my code can be executed locally without any problem.

How can we reproduce the issue?

from qiskit_aer.noise import (NoiseModel, thermal_relaxation_error, depolarizing_error)
from qiskit import QuantumCircuit
from qiskit_aer import QasmSimulator

def get_noise_model(p_gate, t1, t2, time):
    single_gate_error = depolarizing_error(p_gate/10, 1)
    double_gate_error = depolarizing_error(p_gate, 2)
    meas_error = thermal_relaxation_error(t1=t1, t2=t2, time=time)
    noise_model = NoiseModel(basis_gates=['crx','cx', 'id', 'rz', 'sx', 'u1', 'u2', 'u3'])
    noise_model.add_all_qubit_quantum_error(thermal_relaxation_error(t1=t1,t2=t2,time=time), ['id'])
    noise_model.add_all_qubit_quantum_error(double_gate_error,['crz','cx'])
    noise_model.add_all_qubit_quantum_error(single_gate_error,['rz', 'sx', 'u1', 'u2', 'u3'])
    return noise_model

qc = QuantumCircuit(1)
noise_model = get_noise_model(p_gate=1e-3, t1 = 90, t2 = 120, time = 0.1)
backend = QasmSimulator(method='statevector', noise_model=noise_model)
res = backend.run(qc,optimization_level=0, shots=1000).result()

What should happen?

The code example above should be executed.

Any suggestions?

I found that if I change `method='statevector'' to 'method='density_matrix'', the above code can be executed. However, the execution time will be far longer.

[hhorii]

I reproduced this issue on my Macbook. This is from installed BLAS (heevx) problem because this is not happened in Linux environment though their inputs are the same. Anyway, I would like to find workaround.

[ajavadia]

I have the same issue on my mac.