Magnon-Based Quantum Computing

Updated 2 December 2025

Magnon-based quantum computing is a quantum approach using spin-wave excitations in magnetic materials for robust, long-lived quantum information processing.
It integrates hybrid architectures, combining superconducting qubits, microwave cavities, and optical components to implement high-fidelity quantum gates.
The platform supports scalable processors through tunable gate primitives, error mitigation techniques, and on-chip integration under cryogenic conditions.

Magnon-based quantum computing leverages collective spin-wave excitations—magnons—in magnetically ordered materials as quantum information carriers. Magnons provide a unique combination of long coherence times, strong collective coupling to microwave photons and superconducting qubits, intrinsic bosonic character, and versatile connectivity to photonic, mechanical, and spintronic degrees of freedom. By integrating magnons into hybrid quantum architectures, researchers have demonstrated quantum logic operations, nonclassical state preparation, and the prospect of scalable quantum processors and networks.

1. Physical Realizations and Model Hamiltonians

The fundamental element is a magnon mode, typically the uniform Kittel mode in a ferromagnetic insulator such as yttrium iron garnet (YIG). The minimal system Hamiltonian for a hybrid device comprising a superconducting qubit (frequency $\omega_q$ ), a magnon mode ( $\omega_m$ ), and a microwave cavity ( $\omega_c$ ) reads (rotating-wave approximation):

$H = \omega_c\,a^\dagger a + \omega_q\,\sigma_+\sigma_- + \omega_m\,b^\dagger b + g_q(a^\dagger \sigma_- + a\,\sigma_+) + g_m(a^\dagger b + a\,b^\dagger),$

where $a$ , $b$ , and $\sigma_-$ are cavity, magnon, and qubit operators, and $g_q$ , $g_m$ are coupling strengths. In the strong dispersive regime ( $|\omega_c-\omega_{q,m}| \gg g_{q,m}$ ), the cavity can be eliminated to yield an effective Jaynes–Cummings–type qubit–magnon coupling $g_{qm} = g_q g_m/(\omega_c-\omega_q)$ (Tabuchi et al., 2014, Tabuchi et al., 2015).

Advances are not limited to ferromagnets. Cavity-induced strong magnon–magnon coupling has been predicted for altermagnetic thin films, where cavity photons mediate beam-splitter–type interactions between chirality-split magnon branches, leading to direction-dependent strong coupling and tunable multimode gates (Jin et al., 2023).

2. Gate Primitives, Quantum Operations, and Protocols

Single-Qubit Operations: The two lowest Fock levels of a magnon mode, $|0\rangle_m$ , $|1\rangle_m$ , encode a qubit. Arbitrary single-qubit rotations are implemented by direct microwave driving of the magnon port at resonance. Calibrated pulses allow preparation of Rabi oscillations and fast $\pi$ -rotations in $\sim 25$ ns (1906.08103). In spin-chain quantum circuits, plane-wave single-magnon states can be initialized efficiently on small NISQ processors (Ranu et al., 2022).

Two-Qubit Gates: At resonance, the effective Hamiltonian $H_\text{eff} = g_{qm}(\sigma_+ b + \sigma_- b^\dagger)$ mediates iSWAP-type gates between qubit and magnon, with gate time $\tau_\text{swap} = \pi/(2g_{qm})$ . Parametrically oscillating the coupling enables fast on/off gating. Controlled-phase (CZ) gates are achieved in the dispersive regime, where virtual magnon–qubit exchange results in a state-dependent frequency shift $\chi \simeq g_{qm}^2/\Delta$ ; waiting for time $\tau_\text{CP} = \pi/(2\chi)$ realizes a CZ (Tabuchi et al., 2014).

Multi-Qubit and Entangling Gates: In ensembles where multiple qubits couple to a common magnon mode, synchronized multi-qubit flipping and GHZ-state generation are possible, especially in platforms with squeezed magnons providing enhanced coupling and access to the deep-strong regime (Skogvoll et al., 2021). Cavity-mediated interactions in altermagnets allow partial- or full-SWAP gates between distant chiral magnon modes, supporting spatially distributed entanglement and quantum information transfer (Jin et al., 2023).

Blockade and Single-Magnon Sources: Magnon blockade exploits Jaynes–Cummings nonlinearity and quantum interference to suppress double excitation, yielding pure single-magnon emission, as quantified by $g^{(2)}(0) \ll 1$ . The blockade effect arises under $g_{qm} \gg \kappa_{m,q}$ and specific detuning and drive regimes; $g^{(2)}(0) \sim 10^{-7}$ has been predicted in realistic systems, exceeding photon blockade performance (Liu et al., 2019, Jin et al., 2023, Jin et al., 16 Dec 2024).

Dissipative and Continuous-Variable Gates: Recent proposals use photon–phonon–magnon three-body couplings, enabling programmable dissipative Ising interactions between magnon "spins." Tuning the pump phase selects ferromagnetic vs antiferromagnetic coupling, supporting reconfigurable Ising machines and reservoir-computing architectures (Dou et al., 27 Nov 2025).

3. Quantum State Engineering and Measurement

Nonclassical Fock and Cat States: The strong dispersive regime permits resolution of individual magnon number states via qubit spectroscopy; the qubit resonance splits into $n$ -dependent lines for $|n\rangle$ magnons, enabling direct preparation and tomography of Fock states and number-resolved superpositions (Lachance-Quirion et al., 2016). Protocols from circuit QED, such as number-selective SNAP gates, facilitate mapping arbitrary qubit superpositions to magnon cat or binomial states.

Squeezed and Entangled States: Bosonic squeezing arises intrinsically in anisotropic ferromagnets (via Bogoliubov transformation) or can be engineered dynamically (e.g., via cavity-enhanced Raman processes in antiferromagnets). The squeezing parameter $r$ directly determines continuous-variable entanglement, with two-mode squeezed states serving as resource for EPR correlation and continuous-variable quantum computing (Mousolou et al., 2021, Parvini et al., 16 Sep 2024). Squeezed Perelomov coherent states of magnon pairs can be created and stabilized by optical cavity back-action, supporting resource-efficient cluster states (Parvini et al., 16 Sep 2024).

Measurement: QND (projective) measurement of magnon number can be realized by dispersive qubit readout, where the qubit frequency shift per magnon exceeds both linewidths. Homodyne detection on the cavity output and optomagnonic measurement enable continuous-variable quadrature readout (Lachance-Quirion et al., 2016, Mousolou et al., 2021).

4. Hybrid Architectures and Integration Strategies

Magnons function both as memory and as quantum interconnects in hybrid systems:

Quantum Memories: Long-lived magnon modes in YIG spheres ( $T_2 \sim 1$ – $10\,\mu$ s) serve as quantum memories for superconducting qubits. Storage and retrieval are performed via SWAP or dispersive gates (Tabuchi et al., 2015, Lachance-Quirion et al., 2019).
Quantum Buses and Networks: Cavity-embedded magnon modes provide a natural quantum bus linking microwave, optical, spin, and mechanical subsystems. Circuit QED integration enables multi-qubit processors and arrays of YIG microresonators, with on-chip routing via microwave waveguides (Lachance-Quirion et al., 2019, Jiang et al., 2023).
Quantum Transducers: Coupling magnons to optical (optomagnonics) and mechanical (magnomechanics) modes enables coherent microwave–optical and microwave–phonon conversion, foundational for quantum networks (Lachance-Quirion et al., 2019, Yuan et al., 2021).
Dissipative Machines: Tripartite photon–phonon–magnon units now allow dynamically controllable, dissipatively engineered Ising couplings for analog quantum processing and combinatorial optimization (Dou et al., 27 Nov 2025).

5. Coherence, Error Sources, and Mitigation

Decoherence: The dominant loss mechanisms are magnon damping ( $\gamma_m/2\pi \sim 1$ –$2$ MHz in YIG spheres at millikelvin $T$ ), qubit relaxation, and cavity photon loss. Intrinsic nonlinearity in the qubit is critical for nonclassical state generation and blockade effects (Tabuchi et al., 2014, Lachance-Quirion et al., 2019).

Thermal Effects: Thermal magnon occupation $n_\mathrm{th}$ is suppressed below $10^{-9}$ at $T\simeq 10$ mK for $\omega_m/2\pi \sim 8$ GHz; plateauing decoherence and preserving blockade effects at higher $T$ require mitigation strategies including improved filtering and infrared shielding (Tabuchi et al., 2014, Jin et al., 2023).

Material Optimization: High-purity, strain-free YIG spheres or microdisks, along with chemical polishing, reduce inhomogeneous linewidth. Purcell and TLS-related losses are limited by cavity design and substrate annealing (Lachance-Quirion et al., 2019, Jiang et al., 2023).

Dynamical Control: Dynamical decoupling on qubits, echo sequences for magnons, and pulse-shaping to avoid crosstalk support high-fidelity operation. Parametric drives enable fast and programmable on/off coupling necessary for scalable logic (Tabuchi et al., 2014, 1906.08103).

6. Scalability and Prospects for Large-Scale Systems

Architectural Extensions: Multi-mode magnon registers, either by exploiting higher magnetostatic modes or tiled YIG spheres in a shared cavity, can store multiple qubits or serve as high-density bosonic registers (Tabuchi et al., 2014, Tabuchi et al., 2015). Hybrid systems with dispersive cavity-mediated magnon–magnon or qubit–qubit couplings enable the construction of quantum networks and distributed quantum memory.

Programmable Couplers: Recent demonstrations of dynamically switchable dissipative Ising couplings via photon–phonon–magnon three-body interactions enable universal, reconfigurable coupling topologies for coherent Ising machines and quantum simulators (Dou et al., 27 Nov 2025).

Continuous-Variable Processing: Magnon-pair squeezing in antiferromagnets and altermagnets, accessible via cavity or optical parametric control, opens a route to cluster state–based continuous-variable quantum computing entirely in the magnonic platform (Parvini et al., 16 Sep 2024).

Integration: On-chip co-fabrication of YIG microstructures, superconducting circuits, and low-loss waveguides provides pathways toward monolithic, scalable devices. Integration of NV centers or optomechanical elements with magnonics supports local readout, error correction, and modularity (Jiang et al., 2023).

Error Correction and Fault Tolerance: Encoding logical information in bosonic cat or binomial codes, enabled by the large Hilbert space of collective magnon modes, offers robustness to single-excitation loss and supports active error correction architectures (Liu et al., 2019, Jin et al., 2023).

Current experimental gate fidelities for magnon-mediated iSWAP and CZ gates approach or exceed 99%, with gate times in the $10$–$30$ ns regime for optimized parameters (Dols et al., 21 Jun 2024). Further improvements in material quality, cryogenic engineering, and microwave/optical integration are critical for extending coherence times and scaling to many-qubit logical processors.

Table 1: Key Figures of Merit in Representative Magnon-Based Devices

Parameter	Typical Value	Reference
Qubit–magnon coupling $g_{qm}/2\pi$	$7$–$12$ MHz	(Tabuchi et al., 2014, Tabuchi et al., 2015)
Magnon linewidth $\gamma_m/2\pi$	$1$–$2$ MHz	(Tabuchi et al., 2014, Tabuchi et al., 2015)
Swap gate time $\tau_{\rm swap}$	$25$–$35$ ns	(Tabuchi et al., 2014)
$g^{(2)}(0)$ in optimal blockade regime	$\sim 10^{-7}$	(Jin et al., 2023)
Coherence time $T_2$ (YIG sphere at 10 mK)	$100$–$1000$ ns	(Tabuchi et al., 2014)
iSWAP/CZ gate fidelity	$\gtrsim 99\%$	(Dols et al., 21 Jun 2024)
Typical temperature for $n_\mathrm{th} < 10^{-9}$	$T < 20$ mK	(Liu et al., 2019)

Magnon-based quantum computing thus constitutes a comprehensive and technically mature subfield at the intersection of quantum information and spintronics. The unique physical properties of magnons—coherence, bosonic structure, collective enhancement, and hybrid connectivity—enable both discrete-variable gate-based architectures and continuous-variable quantum information processors. Ongoing developments in material synthesis, on-chip integration, and hybrid control schemes continue to advance the scalability, fidelity, and versatility of magnonic quantum technologies.