Quantum Noise

Quantum Wednesday

Sept 6th 2023

A. System-environment perspective:

- Once the excitation leaves the system (e.g., qubit), it's lost forever (Markovian system-environment interaction)

- Once the excitation leaves the system (e.g., qubit), it can come back (Non-Markovian system-environment interaction)

​

Environment

System

​

A bit more detailed description of noise:

- T1 de-excitation of each qubit (e.g., amplitude damping)

- T2 dephasing of each qubit (e.g., depolarizing noise)

- Crosstalk among qubits

- Classical and quantum noise on the control lines

- Leakage into dark states or out of the qubit subspace (can be T1 but also non-Markovian)

​

​

D. Quantum process / channel perspective

- noise as additional, unwanted unitary quantum gates in a quantum circuit (e.g., random X, Y, Z gates)

- noise as a superoperator in a bigger space than that of the qubits (effectively: can be reset gates or other non-unitary gates)

​

​

E. Gate level / Pulse level

- noise is described at the gate level

- noise is described at the pulse level; this can include more easily time-correlated noise of various form

• https://arxiv.org/pdf/1204.2137.pdf

Questions

• Nate: How accurate it is that noise model is the same for inverse of gates? (As in unitary folding)
• KIK as an example of noise that is not same for the inverse
• Brian: How consistent is the noise, over multiple operations? How do you account for fluctuations for computations?
• Some data from IBM on characterizing this dependence
• Sycamore data showed quite a variability
• https://www.nature.com/articles/s41534-019-0168-5
• Aryan: QCs have been proposed as detectors for cosmic radiations
• SQMS at Fermilab is using SC qubits for particle detection, also for dark matter candidates (axions)
• Experiments have been done to evaluate the impact of cosmic rays for SC qubits; for example at INFN Gran Sasso, the chips were characterized in that shielded environment and compared to non-shielded
• Andrea: Cosmic rays could be sources of correlated noise as they disrupt multiple qubits
• Victor: How do you use choose the best qubits, e.g. using IBM devices?
• https://github.com/Qiskit-Partners/mapomatic (it seems like it’s used by the Qiskit Runtime transpiler)
• Nate: Correlated noise: what are more advanced ways to
• Misty: stim simulates noise faster thanks to symmetries, https://github.com/quantumlib/Stim

Questions

• Nate: meaning of the positive value lambda > 1 in the quantum depolarizing channel?
• Discussion on dimensionality

​

• Nathan: Can we visualize error mitigation techniques vs. noise type…?
• ZNE:
• PEC:
• DDD:
• REM:
• CDR:
• Can we simulate this with Mitiq? With ancillas maybe?

Some Resources Mentioned

• https://arxiv.org/pdf/2303.05001.pdf, “Adaptive quantum error mitigation using pulse-based inverse evolutions”
• https://arxiv.org/pdf/1204.2137.pdf “Hybrid quantum circuits: Superconducting circuits interacting with other quantum systems”
• https://arxiv.org/abs/1809.02573 “Tackling the Qubit Mapping Problem for NISQ-Era Quantum Devices”
• https://arxiv.org/pdf/2305.02939.pdf “Tackling the Qubit Mapping Problem with Permutation-Aware Synthesis” (LBL)
• https://en.wikipedia.org/wiki/Quantum_depolarizing_channel
• ​

​