Crosstalk
Crosstalk is a control signal intended for one qubit causing unwanted evolution on another — an operation nobody asked for.
What it means
Crosstalk is a control signal intended for one qubit causing unwanted evolution on another.Qubits on a chip are not perfectly isolated: a microwave pulse aimed at Q1 partially 'leaks' — through stray electromagnetic coupling, shared wiring, or residual qubit-qubit coupling — and nudges Q2, executing an operation nobody asked for.The problem grows with qubit count and density.Each qubit needs its own transition frequency, but the usable band (roughly 4–8 GHz for transmons) is finite, so larger chips suffer frequency crowding: neighboring qubit frequencies sit ever closer together, and drive pulses increasingly address the wrong neighbor.As of 2026, research shows that hardware design alone cannot eliminate XY crosstalk at scale — pulse-level countermeasures such as frequency modulation combined with dynamical decoupling are required, making crosstalk suppression an active research frontier.Crosstalk is also one of the three reasons calibration must be continuous, alongside fabrication inhomogeneity and environmental drift: it is a property of the whole system, not a defect of any single qubit.Everyday analogy
Common misconceptions
- Crosstalk is NOT eliminated by chip design alone — as chips scale, frequency crowding makes some XY crosstalk unavoidable at the hardware level, requiring pulse-level countermeasures such as frequency modulation combined with dynamical decoupling (per 2026 research).
- Crosstalk is one of the three reasons calibration must be continuous (together with fabrication inhomogeneity and environmental drift) — it is a system property, not a single qubit's defect.
Key takeaways
- Crosstalk: a control signal intended for one qubit causes unwanted evolution on another.
- It grows with qubit count and density — frequency crowding packs neighboring qubit frequencies ever closer as chips scale.
- Pulse-level suppression (e.g., frequency modulation plus dynamical decoupling) is an active research frontier as of 2026.
- Crosstalk is one of calibration's three enemies, alongside fabrication inhomogeneity and environmental drift.
Check your understanding
Why can't chip design alone eliminate XY crosstalk on large superconducting processors?
- A.Because qubits must be kept below 15 mK
- B.Because of frequency crowding: as qubit count grows, neighboring transition frequencies pack ever closer within the usable band
- C.Because microwave pulses cannot be shaped
- D.Because T1 times are too short
Show the answer
Answer: B. Because of frequency crowding: as qubit count grows, neighboring transition frequencies pack ever closer within the usable band
Why: The usable frequency band for transmons is finite (roughly 4–8 GHz), so larger chips force neighboring qubit frequencies closer together — some crosstalk then becomes unavoidable in hardware and must be cancelled at the pulse level.
Crosstalk is one of the three reasons a QPU must be calibrated continuously. What are the other two?
- A.Fabrication inhomogeneity and environmental drift
- B.Leakage and decoherence
- C.Readout error and SPAM error
- D.Transpilation and routing
Show the answer
Answer: A. Fabrication inhomogeneity and environmental drift
Why: Calibration's three enemies are fabrication inhomogeneity (every qubit is a snowflake), environmental/temporal drift (properties wander over time), and crosstalk (signals leak between qubits).
Builds on
Primary source: Sarovar et al., Detecting crosstalk errors in quantum information processors, Quantum 4, 321 (2020)
Crosstalk taxonomy/definition per Sarovar 2020; pulse-level suppression frontier claims are phrased as-of-2026 in text (arXiv:2601.05231 existence-verified 2026-07-03).
Learn it hands-on
This concept is part of a 46-level curriculum with an interactive simulator and Lumen, a tutor whose answers are verified before you see them. Levels 1–5 are free.