Improve performance of sampler experiment

[ikkoham]

Summary

Same idea with #1840

Details and comments

Benchmark code:

import numpy as np

from qiskit.circuit.library import EfficientSU2
from qiskit_aer.primitives import Sampler

rng = np.random.default_rng(15)

qc = EfficientSU2(3)
qc.measure_all()

num_circuits = 1000
sampler = Sampler()
parameter_values = rng.random((num_circuits, qc.num_parameters))
circuits = [qc] * num_circuits

%timeit sampler.run(circuits, parameter_values).result()

0.12.2: 716 ms ± 1.3 ms per loop (mean ± std. dev. of 7 runs, 1 loop each)
This PR: 534 ms ± 4.33 ms per loop (mean ± std. dev. of 7 runs, 1 loop each)

[doichanj]

I confirm this PR will improve performance with runtime parameter binding #1901