arXiv:2610.11026 · npj Quantum Information
Demonstration of subspace-search variational quantum eigensolver using photonic orbital angular momentum qudits
Kang-Min Hu, Myung-Hyun Sohn, Yong-Su Kim, Kwang Jo Lee, Hyun Woo Kim, Hyang-Tag Lim
Abstract
Accurate excited state energies are essential for interpreting fluorescence and photochemical processes, yet variational quantum eigensolvers (VQEs) on noisy intermediate-scale quantum (NISQ) hardware have largely been confined to ground states. Subspace-search VQE (SSVQE) offers a NISQ-friendly route to excited states without ancilla-based overlap measurements, however it has lacked an experimental demonstration since it requires the coherent preparation and manipulation of multiple orthogonal basis states under a shared variational unitary. Here we realize SSVQE on a photonic platform by encoding orthogonal qudit states in the orbital angular momentum of single photons and optimizing a shared variational circuit. We estimate excited state energies for a four-dimensional $\rm{HeH^+}$ molecule and an eight-dimensional Heisenberg chain, thereby demonstrating scalable excited state eigensolving in a high-dimensional Hilbert space. Our results establish photonic qudits as a resource-efficient pathway to excited state quantum simulation.
AI Summary
Subspace-search variational quantum eigensolving (SSVQE) targets multiple eigenstates by minimizing a weighted sum of energies from orthogonal inputs transformed by a shared parametrized unitary. This paper investigates its experimental implementation using single-photon orbital-angular-momentum (OAM) qudits, avoiding ancilla-assisted overlap measurements. The implementation does not physically apply the shared unitary to different inputs. Instead, a classical processor computes its relevant columns and converts each into a hologram that directly prepares the corresponding output state on a spatial light modulator. A second modulator implements sequential projective measurements for estimating Pauli-string expectations. Photon-counting estimates feed a COBYLA optimization loop; orthogonality is imposed in the classical parametrization. The experiments estimate all four energy levels of a four-dimensional $\mathrm{HeH}^{+}$ Hamiltonian across bond lengths, using 15 variational parameters and averaging $81.6\pm10.2$ optimization iterations. An eight-dimensional encoding of a three-spin Heisenberg chain targets its lowest three levels using 55 parameters, compared with 63 for a general eight-dimensional unitary, and averages $368\pm57$ iterations. Reported energies agree with exact benchmarks; statistical error bars use 1,000 Monte Carlo resamplings. The paper attributes residual deviations mainly to imperfect mode purity and fidelity arising from optical alignment. It explicitly notes that encoding $n$ qubits in one photon requires dimension $d=2^n$, while its measurement scheme addresses projections sequentially. Consequently, the demonstrated single-particle architecture does not establish scalable quantum computation; extensions to multiple degrees of freedom or photons would introduce additional generation, control, and detection challenges.
Central claim may not hold: The experiment classically computes and separately prepares the output states of a variational unitary, rather than physically implementing the shared variational circuit described in the abstract. The claim of scalable, resource-efficient excited-state simulation is not established by the single-photon encoding, which requires exponentially many modes in the number of encoded qubits and explicit classical synthesis of the output-state amplitudes.
Written by gpt-6.1-sol from the paper's full text. Automated: it can be wrong.
- Journal reference
- npj Quantum Information (2026)
- DOI
- 10.1038/s41534-026-01380-2
- Comments
- 8 pages, 4 figures, Supplementary Note