E. :
I agree, China’s photonic qubits are clearly better than IBM’s transmon qubits. For transmons, they need to be cooled to nearly absolute zero, almost minus 273 degrees Celsius, to eliminate noise and interference. Even then, they still need between 1,000 to 10,000 physical qubits just to create one stable logical qubit because most of them are used for error correction. Photonic qubits, on the other hand, can operate at room temperature and don’t have that extreme cooling requirement.
For photonic qubits, their biggest hurdle is photon loss. They're fighting photon loss in a few main ways right now: using better materials like silicon nitride waveguides that lose way less light, making components super efficient with better detectors, improved photon sources, and low-loss connections, smarter designs like hybrid spatial-temporal encoding that Jiuzhang 4.0 uses to reduce how many components photons have to travel through, and advanced techniques like GKP encoding and error correction codes that are specifically good at handling missing photons. It's still their biggest headache, but they're making steady progress on it.
But topological qubits have the highest potential of all three. They store information in the braiding of quasiparticles, which makes them naturally resistant to errors instead of constantly fighting noise. Topological qubits have natural immunity to local noise because information is stored non-locally in the braiding of quasiparticles, built-in error protection, much longer coherence times, drastically fewer qubits needed, and precise gates through braiding. That means drastically fewer qubits are needed. You only need a handful instead of thousands. Also, you should look into something called the Thermodynamic Sampling Unit, or TSU. It's a game changer.
2026-05-21 05:27:10