DevOps for Quantum Apps: Hybrid Pipelines.
Hybrid DevOps pipelines execute tests and deployments on quantum circuits and codes. Quantum DevOps targets NISQ with simulators like Qiskit Aer, cloud providers like IBM Quantum and Amazon Braket, and hybrid gate approaches. GitHub Actions execute Pytket transpilation and Cirq circuit execution. ArgoCD manages fault-tolerant queue-based deployments. Noise profiling drives automated error mitigation. Around half of the QML workload is automated by 2026.
Hybrid Pipeline Stages
- Code: Hybrid codes use Q#, Qiskit, and PennyLane programming languages.
- Test: Tests use noisy simulators to execute unit tests and integration tests.
- Build: It includes transpilation to target topology with noise awareness included.
- Deploy: It includes submitting to queues and polling results.
- Monitor: It tracks fidelity metrics and monitors any drifts.
- Orchestration: Node.js is used to manage workflows with Django-based dashboards to provide results.
Benefits
- Reproducible: Pre-built noise models provide reproducibility.
- Scalability: Parallel simulation on multiple cloud providers is scalable.
- Speed: Hybrid DevOps pipelines provide rapid execution with simulators under 5 minutes. Real cloud execution takes longer compared to simulators.
- Performance: A 10 times velocity increase is achievable with hybrid DevOps pipelines.
Challenges
- Qubit reservation with API-based approaches.
- Hybrid debugging with visualizers.
Toolkit
- GitLab Quantum
- QDevOps runners
Conclusion
Quantum DevOps is a pipeline of classical and quantum software development. It is implemented with a React.js-based visualization of quantum circuits, Node.js-based job queuing, Python-based and Django-based noise analysis, Laravel-based hybrid tooling, and Java Spring Boot-based workflow engines. With this, it becomes seamless to move from classical to quantum computing.
