Silicon Quantum Bits Hit 99.99% Accuracy While Running Five at Once
Researchers at RIKEN in Japan have demonstrated five silicon spin qubits operating simultaneously with fidelity above 99.9%, and three of those qubits exceeding 99.99%, by identifying and correcting a quantum interference effect that previously degraded parallel performance.
A team at the RIKEN Center for Emergent Matter Science in Japan has cleared one of the biggest practical hurdles in scaling up silicon-based quantum computers. Silicon spin qubits are attractive because they can, in principle, be manufactured using standard semiconductor processes, but their accuracy has historically dropped sharply when multiple qubits run in parallel rather than one at a time. The RIKEN researchers traced the main culprit to the AC Stark shift: microwave pulses used to control one qubit change the effective energy levels of neighbouring qubits, causing unwanted phase errors. Instead of trying to calibrate every possible combination of simultaneous operations — an exponentially growing task — the team measured only pairwise interactions between qubits and used those measurements to predict and cancel multi-qubit interference. Using a five-qubit silicon array controlled by a single microwave line, three qubits achieved gate fidelity above 99.99% when operated simultaneously, while all five together maintained approximately 99.9% fidelity. As researcher Leon Camenzind stated, the result shows high-fidelity control can be maintained even when several qubits are operated at the same time. The findings were published in Nature Physics and reported by Phys.org and Interesting Engineering in October 2026, and are considered a significant step toward fault-tolerant silicon quantum processors compatible with existing chip-fabrication infrastructure.
