Standalone, GPU-capable quantum circuit simulator in less than 500 lines.
Good for:
- Single and 2-qubit gate rotations
All you need:
- numpy (core)
- matplotlib (graph plotting)
Optional:
- cupy (replaces numpy when CUDA-enabled GPU is detected)
Functions:
- Compute final state given a sequence of gates
- Custom gate compiler
- Edit and keep track of gates (ideal/noisy) every step of the way
- Combination of time-dependent and independent noises
- Arbitrary error distribution
- Parallel execution of thousands of circuit runs as far as memory permits
- Fidelity plots
- Probability of measuring any output state
Data flow:
- Gate compilation: ideal gates ---(noise distribution)--> compiled gates (optional) --> noisy gates
- Initial states --> final states
Analysis:
- Fidelity histogram
- Probability of getting desired state
- Fidelity distribution over multiple input states/gate sets
Benchmark:
- Bell state for 10 qubits, 5000 runs
Potential applications
- Evaluate Circuit/Algorithm performance
- Characterize gate errors
Demos:
- IQFT circuit
- 6-qubit parity check
License: Apache License v2.0