Researchers Test Binary Quadrature Encoding on a 56-Qubit Chip

Introduction

Recent experiments by researchers at the Delft Institute of Applied Mathematics, QuTech, and Quantinuum demonstrate that binary quadrature encoding can simulate boson‑matter interactions on a 56‑qubit trapped‑ion processor (Quantinuum H2‑2). By leveraging the quantum Fourier transform, this method balances qubit count and circuit depth, providing a practical alternative to earlier approaches that required thousands of qubits.

Comparing Encoding Strategies

Three encoding schemes were evaluated for representing continuous bosonic modes on discrete qubits:

1. Unary (Fock) encoding – maps each boson number to a distinct qubit, leading to simple logic but rapidly escalating qubit requirements.
2. Dicke state encoding – a newly explored collective‑state representation that yields very shallow circuits but demands a substantially larger qubit register.
3. Binary quadrature encoding – encodes bosonic amplitudes using binary digits and the quantum Fourier transform, achieving moderate circuit depth with a modest qubit budget.

Classical simulations confirmed that binary quadrature encoding requires fewer gate operations than unary encoding for comparable boson numbers, while Dicke encoding, despite its shallow depth, consumes significantly more qubits, making it less scalable on current hardware.

Experimental Results on the H2‑2 Device

Using Quantinuum’s 56‑qubit H2‑2 trapped‑ion system, the team implemented displacement and Dicke‑model simulations across the three encodings. Key observations include:

• Resource efficiency: Binary quadrature encoding succeeded with the full 56‑qubit register, whereas unary encoding would have needed well beyond the available qubits for the same boson count.
• Circuit depth: The binary approach produced circuits of intermediate depth—deeper than Dicke but far shallower than unary—aligning with the hardware’s coherence times.
• Noise sensitivity: Early‑stage simulations at low boson numbers already exhibited deviations from ideal behavior, indicating that noise remains a limiting factor even for the more compact binary encoding.

These findings suggest that while binary quadrature encoding offers a promising path toward scalable bosonic simulations, robust error‑correction or mitigation techniques will be essential for larger or longer computations.

Implications and Future Directions

The work establishes a practical benchmark for boson‑simulation methods on near‑term quantum hardware. Binary quadrature encoding’s favorable trade‑off between qubit usage and circuit complexity makes it a strong candidate for early applications such as quantum chemistry and lattice‑gauge models, where bosonic degrees of freedom play a central role. However, the observed noise‑induced errors underscore the need for:

• Improved error mitigation: Techniques like zero‑noise extrapolation or dynamical decoupling could extend usable simulation depth.
• Hardware advances: Enhancements in ion‑trap stability and gate fidelity will directly benefit the accuracy of binary quadrature simulations.
• Algorithmic refinements: Hybrid approaches that combine the shallow depth of Dicke encoding with the qubit efficiency of binary quadrature may further optimize performance.

Conclusion

Binary quadrature encoding on a 56‑qubit trapped‑ion platform demonstrates that boson‑matter interactions can be simulated with far fewer qubits than previously required, offering a balanced solution between space and time complexity. While noise remains a challenge, the method provides a clear roadmap for scaling bosonic simulations as quantum hardware and error‑correction strategies continue to improve.