MicroQuantum¶
A lightweight, NumPy-only Python quantum computing SDK foundation. Build a circuit. Execute it locally. Understand the result. Formulate a problem, run an algorithm, run an experiment and analyze the outcome — with a pluggable backend architecture that is ready for future hardware providers.
Get Started
Concepts
Algorithms
Execution & Backends
Analysis
Examples & Tutorials
API Reference
Developer Guide
Quality
Project
- Release 1.1.0
- General Availability (v1.0.0)
- Developer Preview
- Changelog
- Compatibility
- Platform Support
- Contributing
- FAQ
- What is MicroQuantum?
- Who is this for?
- Why NumPy-only?
- Do I need a GPU or a quantum computer?
- Which qubit ordering does MicroQuantum use?
- How do I bind circuit parameters?
- How reproducible are results?
- Can I add my own backend / algorithm?
- Are provider/enterprise features in this SDK?
- How do I report a bug or ask a question?
- Troubleshooting
pip install microquantumfails to find a version- A circuit fails with “unbound parameters”
validate()returns diagnostics instead of raising- QAOA/VQE give a degenerate “energy” for disconnected problems
- Phase estimation reports a non-unitary Hamiltonian
Sphinxbuild fails with warnings-as-errors- Results look non-deterministic
- Performance is slow for many qubits
Overview¶
MicroQuantum is an independent quantum computing SDK: every circuit, operator, simulator and algorithm is implemented from scratch with NumPy as the only hard dependency. It is not a wrapper around Qiskit, Cirq or OpenQASM.
What the SDK provides today:
Core engine — quantum circuits, operators, Pauli algebra,
StateVector/DensityMatrix, measurement, gradients, registers, serialization, a transpiler and aPassManagerpipeline.Problems — plain, JSON-safe descriptions of computational tasks (
SamplingProblem,OptimizationProblem,HamiltonianProblem,EigenvalueProblem,SearchProblem).Algorithms — VQE, QAOA, Grover search, phase estimation, QFT, HHL, Hamiltonian simulation and more, through a uniform
validate(problem) -> solve(problem, runtime)lifecycle.Execution runtime — declarative
ExecutionPlans, batch/sweep orchestration and anExecutionRuntimethat routes work to backends.Backends & providers — statevector, density-matrix, MPS and tree-tensor network simulators, a deterministic
MockBackend, a capability-drivenBackendcontract and aProviderdiscovery boundary for future hardware.Experiments —
ExecutionRecords,ParameterSweeps andExperiments with reproducibility fingerprints and JSON-safe serialization.Analysis — sampling, expectation, state and result-aggregation analysis on top of the existing result contracts.