Novel Protocol Reconstructs Quantum States in Large-Scale Experiments up to 96 Qubits

Why It Matters

Scalable quantum‑state reconstruction is essential for validating and improving large‑scale quantum computers, a prerequisite for practical quantum advantage.

Key Takeaways

• Protocol reconstructs 96‑qubit states, surpassing 35‑qubit limit
• Uses matrix‑product operators to compress quantum data
• Compatible with existing randomized measurement datasets
• Requires fewer measurements than traditional tomography
• Facilitates noise benchmarking and future channel learning

Pulse Analysis

Quantum state tomography has long been a bottleneck for scaling quantum processors. Traditional methods demand exponentially many measurements, limiting practical reconstructions to a few dozen qubits. By recasting the problem in terms of matrix‑product operators—a tensor‑network format that captures correlations with modest memory—researchers can compress the full state into a tractable representation. This shift mirrors advances in classical simulation of noisy systems, where weaker entanglement permits efficient approximations. The new protocol builds on randomized measurements and the classical‑shadows technique, extracting sufficient statistical information while tolerating experimental imperfections.

The experimental demonstration on IBM’s Brisbane chip showcases the protocol’s real‑world viability. Using only two independent randomized‑measurement datasets, the team optimized an MPO model and benchmarked its fidelity, achieving reliable reconstruction of a 96‑qubit mixed state. Compared with prior state‑of‑the‑art tomography limited to 35 qubits, the measurement overhead shrinks dramatically, and the numerical routine draws on well‑established tensor‑network algorithms such as the density‑matrix renormalization group. This compatibility with existing data pipelines means laboratories can adopt the method without overhauling their measurement infrastructure.

Beyond immediate benchmarking, the approach opens pathways for more ambitious diagnostics. Embedding noise and decoherence parameters within the MPO enables systematic error mitigation and could inform the design of fault‑tolerant architectures. The authors already envision extensions to quantum channel learning and two‑dimensional connectivity, challenges that will become pressing as hardware scales to hundreds of qubits. For industry stakeholders, the protocol offers a scalable tool to verify device performance, accelerate hardware iteration, and ultimately bring quantum advantage closer to commercial reality.