Gerçek zamanlı geri bildirim kuantum kaosunu denetim altına alıyor

Fault-tolerant quantum computers must detect errors while calculations continue, even though measuring qubits can disturb their fragile states. Rutgers, IBM and partner institutions demonstrated large-scale real-time feedback on a 156-qubit IBM Quantum Heron processor, using a connected chain of up to 100 qubits. Their program alternated between scrambling operations that spread information among neighbouring qubits and checks that measured and reset selected qubits. The team completed nearly 5,000 paired-qubit operations and roughly the same number of checks and resets. When scrambling dominated, the system remained chaotic; when interventions became more frequent, researchers steered it towards an orderly state. Near an even balance, a small shift caused an abrupt phase transition between the two regimes. Results also matched theoretical predictions more closely as the tested system grew. The experiment did not create a fault-tolerant computer, but it showed current hardware can coordinate repeated operations, measurements and resets during an active computation—a capability future error correction will require.