Running with Threadpool and DASK

Qiskit Aer runs simulation jobs on a single-worker Python multiprocessing ThreadPool executor so that all parallelization is handled by low-level OpenMP and CUDA code. However to customize job-level parallel execution of multiple circuits a user can specify a custom multiprocessing executor and control the splitting of circuits using the executor and max_job_size backend options. For large scale job parallelization on HPC clusters Qiskit Aer executors support the distributed Clients from the DASK.

Installation of DASK packages with Aer

If you want to install dask client at the same time as Qiskit Aer, please add the dask extra as follows. This option installs Aer, dask, and distributed packages.

pip install .[dask]

Usage of executor

To use Threadpool or DASK as an executor, you need to set executor and max_job_size by set_options function. If both executor (default None) and max_job_size (default None) are set, Aer splits the multiple circuits to some chunk of circuits and submits them to the executor. max_job_size can control the number of splitting circuits. When max_job_size is set to 1, multiple circuits are split into one circuit and distributed to the executor. If a user executes 60 circuits with the executor and max_job_size=1, Aer splits it as 60 jobs each of 1 circuit. If there are 60 circuits and max_job_size=2, Aer splits it as 30 jobs, each with 2 circuits.

Example: Threadpool execution

import qiskit
from concurrent.futures import ThreadPoolExecutor
from qiskit_aer import AerSimulator
from math import pi

# Generate circuit
circ = qiskit.QuantumCircuit(15, 15)
circ.h(0)
circ.cx(0, 1)
circ.cx(1, 2)
circ.p(pi/2, 2)
circ.measure([0, 1, 2], [0, 1 ,2])

circ2 = qiskit.QuantumCircuit(15, 15)
circ2.h(0)
circ2.cx(0, 1)
circ2.cx(1, 2)
circ2.p(pi/2, 2)
circ2.measure([0, 1, 2], [0, 1 ,2])
circ_list = [circ, circ2]

qbackend = AerSimulator()
# Set executor and max_job_size
exc = ThreadPoolExecutor(max_workers=2)
qbackend.set_options(executor=exc)
qbackend.set_options(max_job_size=1)
result = qbackend.run(circ_list).result()

Example: Dask execution

The Dask client uses multiprocessing so you need to guard it by an if __name__ == "__main__": block.

import qiskit
from qiskit_aer import AerSimulator
from dask.distributed import LocalCluster, Client
from math import pi
def q_exec():
    # Generate circuits
    circ = qiskit.QuantumCircuit(15, 15)
    circ.h(0)
    circ.cx(0, 1)
    circ.cx(1, 2)
    circ.p(pi/2, 2)
    circ.measure([0, 1, 2], [0, 1 ,2])

    circ2 = qiskit.QuantumCircuit(15, 15)
    circ2.h(0)
    circ2.cx(0, 1)
    circ2.cx(1, 2)
    circ2.p(pi/2, 2)
    circ2.measure([0, 1, 2], [0, 1 ,2])

    circ_list = [circ, circ2]

    exc = Client(address=LocalCluster(n_workers=2, processes=True))
    # Set executor and max_job_size
    qbackend = AerSimulator()
    qbackend.set_options(executor=exc)
    qbackend.set_options(max_job_size=1)
    result = qbackend.run(circ_list).result()


if __name__ == '__main__':
    q_exec()