Skip to main content
 ██████╗ ██████╗ ██╗
██╔═══██╗██╔══██╗██║
██║   ██║██████╔╝██║
██║▄▄ ██║██╔═══╝ ██║
╚██████╔╝██║     ██║
 ╚══▀▀═╝ ╚═╝     ╚═╝

QPI QPU Driver

PyPI version CI/CD Workflow GitHub Tag License

Python QPU driver for the QPI quantum computing platform. Runs on isolated hardware nodes controlling the QPU via multiple executor backends.

📚 Read the Documentation


Install

Base package (mock executor)

pip install qpi-driver

With CLI support

pip install "qpi-driver[cli]"

With Qiskit Aer simulator

pip install "qpi-driver[aer]"

With Quantify/Qblox hardware support

pip install "qpi-driver[quantify]"

Requires Python ≥ 3.12, < 3.13.


Quick Start

CLI

# Connect a mock QPU to the server
qpi-driver process \
  --qpi-addr http://localhost:8090 \
  --token <qpu-access-token> \
  --ca-fingerprint <fingerprint> \
  --name qpu_sim_01 \
  --device mock \
  -o data_dir=./data

Environment variables are also supported for the universal flags:

export QPI_ADDR=http://localhost:8090
export QPI_ACCESS_TOKEN=<token>
export QPI_CA_FINGERPRINT=<fingerprint>
export QPI_DRIVER_NAME=qpu_sim_01
export QPI_DEVICE=mock
qpi-driver process

systemd Service (Linux)

To run qpi-driver persistently on a Linux machine in the background, you can use systemd.

We have provided a standalone interactive bash installer that automates the entire process (installing uv, installing the qpi-driver tool, prompting for your tokens/addresses, and registering the systemd service):

# Run the interactive systemd installer script directly via curl
sudo bash -c "$(curl -LsSf https://raw.githubusercontent.com/sopherapps/qpi/main/qpi-driver/install-systemd.sh)"

Alternatively, you can run the installer non-interactively by specifying all environment variables:

curl -LsSf https://raw.githubusercontent.com/sopherapps/qpi/main/qpi-driver/install-systemd.sh | sudo \
  QPI_TOKEN="<your-qpi-access-token>" \
  QPI_ADDR="http://127.0.0.1:8090" \
  CA_FINGERPRINT="<fingerprint>" \
  QPU_NAME="rigetti-aspen-1" \
  OPERATION="process" \
  DEVICE="qblox" \
  bash

Manual systemd Installation

If you prefer to configure it manually, follow these steps:

  1. Install uv (a fast Python package installer):

    curl -LsSf https://astral.sh/uv/install.sh | sh
    source $HOME/.local/bin/env
    
  2. Install qpi-driver as a tool: Make sure to specify the correct extras (e.g. [cli,qblox], [cli,aer]):

    uv tool install "qpi-driver[cli,qblox]"
    
  3. Create the systemd unit file: Replace the placeholder <values> with your actual configuration.

    sudo bash -c 'cat > /etc/systemd/system/rigetti-aspen-1.qpi-driver.service <<EOF
    [Unit]
    Description=QPI Driver Service (rigetti-aspen-1)
    After=network.target
    
    [Service]
    Type=simple
    
    Environment="QPI_ACCESS_TOKEN=<your-qpi-access-token>"
    Environment="QPI_CA_FILE=/var/qpi-driver/rigetti-aspen-1/qpi.ca.pem"
    Environment=PYTHONUNBUFFERED=1
    
    ExecStart=/home/<user>/.local/bin/qpi-driver process \
            --ca-fingerprint <your-fingerprint> \
            --qpi-addr <your-qpi-server-address> \
            --name "rigetti-aspen-1" \
            --device "qblox" \
            -o data_dir=/var/qpi-driver/rigetti-aspen-1 \
            -o quantify_device_config=/var/qpi-driver/rigetti-aspen-1/quantify.device.yml \
            -o quantify_hardware_config=/var/qpi-driver/rigetti-aspen-1/quantify.hardware.json
    
    Restart=on-failure
    User=<user>
    
    StandardOutput=journal
    StandardError=journal
    SyslogIdentifier=rigetti-aspen-1.qpi-driver
    
    [Install]
    WantedBy=multi-user.target
    EOF'
    
  4. Start and enable the service:

    sudo systemctl daemon-reload
    sudo systemctl enable rigetti-aspen-1.qpi-driver.service
    sudo systemctl start rigetti-aspen-1.qpi-driver.service
    sudo systemctl status rigetti-aspen-1.qpi-driver.service
    

Python API

from qpi_driver import run_driver

run_driver(
    qpi_addr="http://localhost:8090",
    token="<qpu-access-token>",
    ca_fingerprint="<fingerprint>",
    name="qpu_sim_01",
    executor="mock",
    data_dir="./data",
)

Custom executor

from qpi_driver import Executor, run_driver
from qpi_driver.executors.mock import MockExecutor

class MyCustomExecutor(Executor):
    def execute(self, payload):
        # Your QPU-specific execution logic
        ...

run_driver(
    qpi_addr="http://localhost:8090",
    token="<token>",
    ca_fingerprint="<fingerprint>",
    name="my_qpu",
    executor="custom",
    custom_executor=MyCustomExecutor(),
)

Executor Backends

Backend Description Extra
mock Qiskit BasicSimulator (default)
qiskit_aer Qiskit Aer simulator [aer]
quantify Quantify-scheduler + Qblox instruments [quantify]
qblox Qblox scheduler (legacy) [qblox]

Architecture

The driver uses Python's multiprocessing library to isolate responsibilities:

  • Main Process: NNG PULL listener, receives commands from server
  • Worker Process: Executes quantum circuits via the configured executor
  • Result Sender Process: NNG PUSH, sends results back to server
┌─────────────┐     NNG PUSH      ┌─────────────────┐
│ Server│ ────────────────> │  Main Process   │
│  Dispatcher │                   │  (PULL listener)│
└─────────────┘                   └────────┬────────┘
                                           │
                              multiprocessing.Queue
                                           │
                                           ▼
                                   ┌───────────────┐
                                   │ Worker Process│
                                   │  (Executor)   │
                                   └───────┬───────┘
                                           │
                              multiprocessing.Queue
                                           │
                                           ▼
                                   ┌───────────────┐
                                   │Result Sender  │
                                   │  (PUSH)       │
                                   └───────┬───────┘
                                           │ NNG PUSH
                                           ▼
                                   ┌───────────────┐
                                   │  Server │
                                   │   Listener    │
                                   └───────────────┘

CLI Reference

A driver is run by its operation subcommand — process (a QPU) or monitor (e.g. a cryostat) — on a specific --device. Both share the same universal options; each device's own settings are passed as repeatable -o key=value.

qpi-driver process|monitor [OPTIONS]

Universal options:
  -a, --qpi-addr TEXT     QPI server URL [env: QPI_ADDR]
  -t, --token TEXT        Access token identifying the driver [env: QPI_ACCESS_TOKEN]
  -n, --name TEXT         Human-readable driver name [env: QPI_DRIVER_NAME]
  -d, --device TEXT       Backend within the operation, e.g. mock, qblox, bluefors_gen1 [env: QPI_DEVICE]
  -o, --option KEY=VALUE  Operation-specific config, repeatable
  --ca-file PATH          Path to the CA root certificate [env: QPI_CA_FILE]
  --ca-fingerprint TEXT   Fingerprint pinning the CA root certificate [env: QPI_CA_FINGERPRINT]
  --help                  Show this message and exit.

process -o options: data_dir, is_dummy, job_timeout, quantify_hardware_config,
                     quantify_device_config, use_sdk
monitor -o options (bluefors_gen1): channels (required), base_url, api_key,
                     poll_interval, timeout

Documentation


License

MIT — see the main repository for details.

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

qpi_driver-0.1.2.tar.gz (62.9 kB view details)

Uploaded Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

qpi_driver-0.1.2-py3-none-any.whl (62.0 kB view details)

Uploaded Python 3

File details

Details for the file qpi_driver-0.1.2.tar.gz.

File metadata

  • Download URL: qpi_driver-0.1.2.tar.gz
  • Upload date:
  • Size: 62.9 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? Yes
  • Uploaded via: twine/6.1.0 CPython/3.13.14

File hashes

Hashes for qpi_driver-0.1.2.tar.gz
Algorithm Hash digest
SHA256 e4f19b91bcc3fa4d1fdb2997c215d6628ae42c2c4038902e4ea32846e893567a
MD5 ebabc59f84609d6fbcb75da106b491cc
BLAKE2b-256 ba33a881f80e09b20526d6f9a2984434dc8054dfda2b3bed872a8a40ec6c3414

See more details on using hashes here.

Provenance

The following attestation bundles were made for qpi_driver-0.1.2.tar.gz:

Publisher: ci.yml on sopherapps/qpi

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

File details

Details for the file qpi_driver-0.1.2-py3-none-any.whl.

File metadata

  • Download URL: qpi_driver-0.1.2-py3-none-any.whl
  • Upload date:
  • Size: 62.0 kB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? Yes
  • Uploaded via: twine/6.1.0 CPython/3.13.14

File hashes

Hashes for qpi_driver-0.1.2-py3-none-any.whl
Algorithm Hash digest
SHA256 ffacce55efff6467866446257fdbefdb33eb7216af0f387db399fcb05c3a7af9
MD5 54a69fcc2ee5d73c11129ef4274d5bb7
BLAKE2b-256 416651ff707195c9ba8767565f96697a7733064540c136e778978bad3b01a1f4

See more details on using hashes here.

Provenance

The following attestation bundles were made for qpi_driver-0.1.2-py3-none-any.whl:

Publisher: ci.yml on sopherapps/qpi

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

Supported by

AWS Cloud computing and Security Sponsor Datadog Monitoring Depot Continuous Integration Fastly CDN Google Download Analytics Pingdom Monitoring Sentry Error logging StatusPage Status page