Quantum computing encodes information in qubits — two-state systems, analogous to classical bits. Every operation on a quantum computer manipulates these two states or their superpositions. A photon's polarization, for instance, can be horizontal or vertical, giving exactly two basis states. The restriction to two dimensions is not a physical law but a convenience: simpler to build, simpler to control, simpler to analyze. But it means that encoding a single classical digit (0 through 9) requires four qubits, most of which are overhead.
Reader note (2026-11-19): Checked against the arXiv full text of Liu, Ren, Wan, Zhu, Cheng, Wang, Shi, Xi, Huber, Friis, Gao, Wang and Wang, 'Heralded High-Dimensional Photon-Photon Quantum Gate' (2407.16356, version 1 of July 2024); I did not compare it with the published Nature Photonics text. The main description matches: a heralded controlled phase-flip gate for two four-dimensional qudits encoded in orbital angular momentum, with authors at Nanjing University and TU Wien. The figures the essay leaves out: the abstract gives a process fidelity 'within a range of [0.64 +/- 0.01, 0.82 +/- 0.01]' and says the four-dimensional gate's decomposition 'would require at least 13 two-qubit entangling gates'; the text gives an overall theoretical process efficiency of 1/8, independent of the dimension, because only two of the four Bell states can be distinguished unambiguously and the two multiports succeed with probability 1/4.
Published in Nature Photonics, researchers at TU Wien and Chinese institutions built a quantum logic gate that operates on photons encoded in four states simultaneously — not by using polarization but by using the spatial waveform of the photon itself. Different orbital angular momenta correspond to different quantum states, and the same photon that previously held one bit of information now holds two. The gate entangles pairs of four-dimensional photons, creating joint states in a sixteen-dimensional Hilbert space.
The structural insight is about where the dimensionality limit actually lives. Photons always had infinitely many spatial modes available — orbital angular momentum has no upper bound. The two-state restriction was never a property of the photon. It was a property of the gate. The photon was already high-dimensional; the operations imposed on it were not. Lifting the operational constraint does not change the information carrier — it changes how much of the carrier's capacity is accessed. The bottleneck was not in the physics but in the engineering that addressed the physics.