Speakers
Description
This work introduces a hardware-realistic simulation framework for modeling quantum circuits under physically motivated stochastic gate noise, addressing a key limitation in existing simulators that decouple decoherence from gate execution and ignore non-trace-preserving dynamics. Built on the quantum-gates library and extended with CERN's Quantum Technology Initiative, the framework adds several capabilities for multi-cycle circuit simulation: mid-circuit measurements with Born-rule sampling that preserves trajectory weight across measurements, noisy qubit reset and readout, and support for native two-qubit gates on non-adjacent qubits without requiring SWAP decomposition. The noise model integrates decoherence continuously during gate execution via the Lindblad formalism, producing stochastic trajectories whose ensemble average recovers physical Lindblad evolution, with hardware-calibrated parameters sourced from IBM Quantum backends.
The framework is demonstrated on planar and rotated surface-code QEC circuits. While idealized, untranspiled distance-3 codes show clear error suppression, transpilation to IBM's Eagle backend introduces substantial additional circuit depth and gate count, distributing errors in patterns that fall outside decoder correction capability. This highlights the necessity of hardware-topology-aware noise simulation for accurately predicting real-device performance. Future work will extend the model to correlated multi-qubit noise channels for better crosstalk modeling, with applications extending to broader quantum program benchmarking.
| Topic | Quantum algorithms for machine learning applications |
|---|---|
| Additional Topic | Theoretical foundations of quantum learning |