Quantum Error Correction Below the Surface Code Threshold

Key points

• ▸Google's 105-qubit Willow processor ran surface codes at distance 3, 5, and 7.
• ▸Each increase in code distance roughly halved the logical error rate.
• ▸The logical qubit outlived its best physical qubit, confirming error correction was a net win.

In plain language

Error correction only helps if your hardware is good enough to begin with. Below that quality bar, the extra qubits you add to detect errors introduce more noise than they remove, so bigger codes make things worse — and until this experiment, no machine had convincingly crossed the bar. Google built successively larger surface codes on its Willow chip and watched the logical error rate fall by roughly half at each step, with the protected qubit lasting longer than any individual physical qubit on the chip. That is the direction fault tolerance requires, and it is the strongest evidence so far that useful error-corrected quantum computers are an engineering problem rather than an open question.

Why it matters

Every fault-tolerance roadmap assumes that scaling up a code suppresses errors. This is the experiment that showed the assumption holds on real hardware, converting large-scale quantum computing from an open physics question into a scaling and engineering challenge.