FPGA Chip Revolutionizes Superconducting Qubit Stability in Quantum Computing

Scientists have developed a faster way to stabilise superconducting qubits using an on-FPGA workflow. This new method tackles the persistent issue of parameter drift, which can disrupt quantum computing performance. By moving key processes onto a chip, the team achieved millisecond-scale calibration and benchmarking—far quicker than traditional CPU-based approaches.

The breakthrough centres on shifting critical tasks—pulse generation, data collection, analysis, and adjustments—directly onto an FPGA chip. This eliminates the delays caused by sending data to a CPU for processing. As a result, the system can recalibrate continuously, keeping qubit performance stable even as conditions change.

Qubit coherence proved particularly volatile, with T1 values shifting within just two seconds. The new workflow counters this instability by enabling over 74,000 consecutive recalibrations, all while maintaining high gate fidelity. Tests showed that Clifford-randomized gate benchmarking, a standard performance check, took only 107 milliseconds—far faster than previous methods.

The research also explored the balance between estimation accuracy and decision speed when sampling data sparsely. Optimal settings were identified to streamline qubit and pulse parameter adjustments. However, the approach relies on simplified signal models, which may not fully capture the complexity of real quantum systems.

Beyond calibration, the workflow supports rapid readout tuning, spectroscopy, pulse-amplitude optimisation, and coherence tracking. By decoupling gate errors from control-parameter drift, it preserves the performance gains tied to qubit coherence.

The on-FPGA method marks a significant step toward real-time quantum system stabilisation. It replaces slow, CPU-dependent recalibration with near-instantaneous chip-based adjustments. This could lead to more reliable quantum computations, though further work may be needed to address its model-based limitations.