What question did this study set out to answer?

This investigation aims to understand how fidelity of GHZ states varies with qubit number and to identify the noise that affects this fidelity.

June 10, 2026Open Access

Collapse of Multipartite Entanglement: A First-Person Benchmark of GHZ State Fidelity

Key Points

This investigation aims to understand how fidelity of GHZ states varies with qubit number and to identify the noise that affects this fidelity.
Conducted systematic measurements on a 127-qubit superconducting quantum processor from 4 to 80 qubits.
Analyzed computational-basis populations and global Pauli-X parity oscillations to assess state fidelity.
Identified noise channels including CNOT gate errors and dephasing that impact entanglement.
4-qubit GHZ state achieved fidelity F = 0.9307, but fidelity dropped to 0.0963 by 32 qubits.
At 80 qubits, fidelity fell to F = 0.0453, indicating the state was nearly indistinguishable from a random state.
Adaptive compilation and dynamical decoupling methods could recover fidelity above the genuine multipartite entanglement threshold at 120 qubits.

Abstract

I report a hands-on, systematic investigation of Greenberger–Horne–Zeilinger (GHZ) state generation on a 127-qubit supercon-ducting quantum processor, covering system sizes from 4 to 80 qubits. I measured both computational-basis populations andglobal Pauli-X parity oscillations to decompose the state fidelity into its population purity and phase-coherence components. Mydata reveal an exponential collapse of fidelity with qubit number: a 4-qubit GHZ state achieves a fidelity of F = 0.9307, wellabove the threshold for entanglement-based quantum key distribution, yet by 32 qubits the fidelity has plummeted to 0.0963 andthe parity signal is extinguished. At 80 qubits, the state is indistinguishable from a maximally mixed state (F = 0.0453). I isolatethe dominant noise channels—CNOT gate errors, dephasing, readout crosstalk—and map the current usability frontier for rawmultipartite entanglement. I then contrast my results with mitigation-enhanced benchmarks from other platforms, showing thatadaptive compilation and dynamical decoupling can recover fidelity above the genuine multipartite entanglement (GME) thresholdeven at 120 qubits. This work provides a deeply personal, quantitative perspective on the challenges and near-term solutions forscaling entanglement in quantum communication, sensing, and error-mitigated computation.

Collapse of Multipartite Entanglement: A First-Person Benchmark of GHZ State Fidelity

Key Points

Abstract

Cite This Study